Blog Posts

Hurricane heads for tim32...

Posted On Tuesday 6 October 2015 by Murray Colpman (Muzer)

IPv6 is now enabled on most tim32 services (with the notable exception of IRC, which I intend to look into) thanks to Hurricane Electric who have an excellent tunnelbroking and dynamic DNS service.

If you notice anything horribly broken, let me know.



Timlan meeting #8 minutes - 2014-02-14

Posted On Friday 14 February 2014 by Murray Colpman (Muzer)

We have made up for our massive lack of Timlan meetings, by making a load of important changes to Timlan that fix EVERYTHING, all to do with sets. See the recent blog posts from today for details. BEAST

sup means "is a superset of"

uq means to the same extent as, or of equal set cardinality

Groups make sense as concept and it's a useful term, but they can be thought of as just a set inside another set. This means that some verbs simply operator on the "top set", but others will operate on the members of the top set.

From this train of thought, we are modifying the syntax of these 'set operators' take 'sets' (and 'groups') (and adjectives, sometimes) and the concept of a 'nounoid' in syntax means the elements that comprise a set.

We need to work out what Adjectives are... This has gone on too long now.


compar is gone.

Posted On Friday 14 February 2014 by Murray Colpman (Muzer)


It's been replaced quite thoroughly with erom and foben. These are now used as verbs, and their new definitions may be found in the dictionary. This can be used in conjunction with the awesome rila/rilo words to make something that makes sense. For example

eromelem riloiqeht kel rilanahaeqeht b-ah ta cute riloiqeht kel eiami cuteOpen TCP Editor

Oh, and with erom, you can also use it with sets to compare their cardinality:

eromelem suiami cute eiami worapotOpen TCP Editor


sets, groups = lists, sets

Posted On Friday 14 February 2014 by Murray Colpman (Muzer)

So, sets now means what lists used to mean, and groups now means what sets used to mean. We've also changed the words sit~s and lit~s to git~s and sit~s respectively, and changed any terminology in the dictionary and speedy intro.


We've also changed the default meaning of a noun on its own, to mean "the set of all things of this noun", and changed the meaning of aq to now mean "subset of". This is because types are basically sets. This means we've removed ayik because it is now useless and we don't like saying it.

We've also changed the meaning of terit so that it actually makes sense now. It is now used with rila or rilo, and means "representable as". So, you effectively say "the group that is representable as this given set" (or "the group that this given set is representable as", they're essentially equivalent)


Rail Journeys - Steam on the Met and exploring diversions.

Posted On Saturday 14 December 2013 by Murray Colpman (Muzer)

This, the second post in this series, is the first post about a rail trip for its own sake - ie, for rail enthusiast reasons rather than to actually go somewhere. As such it's a lot more complicated and a lot more oddities occur.

I'll just say that, though it seems like there are loads of diversions, this is only because we were actively seeking them out because they're interesting to rail enthusiasts. And besides, diversions are good - the works have to be done anyway, so better diversions than no choice but a replacement bus service. Just think of the diversion opportunities possible before the Beeching axe fell...

So, last Sunday, a friend and I went out with a mission:

* To explore strange diversions
* To seek out new life and new... er, I mean, Steam on the Met
* To boldly risk missing our connection home by going a really circuitous route on the tube

So yeah. I had previously booked tickets for Steam on the Met for the both of us, and we decided to make a day of it. Not a very long day like we have done in the past, but a day nonetheless. And the most interesting thing for a rail enthusiast to do in a day is explore diversions, of course!

So, Reading to Basingstoke was closed, as was Woking to Guildford, so it made for some interesting diversions down our end. There were also some things of interest going on in London as well. My friend came up with the plan for the day.

Our ticketing situation: the same ticketing anomaly as mentioned in the first post to get us to Guildford, then from there a Guildford to London Zones 1-6 route CLANDON/WOKING super off-peak day travelcard.

So, we woke up to get the first train out. Now, this is a Sunday, and one of the biggest things I've learnt as a rail enthusiast is that Sunday service sucks quite a bit. So the first train out was 0820 (actually quite early). We got a Unilink bus to Southampton Central - I bought my Guildford travelcard online on the way there, and picked up it along with buying the anomalous ticket from the TVM. Unfortunately we had both misread the bus timetable because the Unilink website doesn't work properly (I could start a very large rant about buses here, but I won't), but fortunately we'd left enough padding to still arrive in plenty of time for the train. I even had time to buy a panini for breakfast.

We headed to platform 1 and waited for our train. We caught the 0820 diverted CrossCountry (XC) service from Southampton Central (SOU) towards Manchester Piccadilly, arriving 0934 Guildford (GLD)*, formed of 220010, headcode 1M34. In order to avoid the Reading to Basingstoke line closure, this train departed from SOU as normal on the South Western Main line, but at St Denys, diverted onto the West Coastway line. From there, it went on to Havant and joined the Portsmouth Direct line to GLD, which was the first call after SOU. Finally, it joined the North Downs line on to Reading, and resumed its ordinary calling pattern from there.

The train itself was very quiet, at that time in the morning, and so we had a quick chat to the guard about the diversion before departing. We were debating whether to ask the driver for a cab ride, but never did ;)

After leaving the train at Guildford (where we arrived at platform 6 a minute early), we hung around the station for half an hour, my friend (who is into Photting) and I taking various photographs, before heading to platform 3 and catching a diverted Portsmouth Direct service. The Portsmouth Direct line between Woking and Guildford was also closed, so this particular train was diverted fast via Effingham Junction and additionally called at Surbiton, presumably for Woking passengers to change and double back. This was the 1006 South West Trains (SW) service towards London Waterloo (WAT), calling at Surbiton, Clapham Junction, and London Waterloo at 1047. It was formed of 444029, headcode 2P28, and was moderately busy, but not too busy to stop us getting seats. We left on time, but unfortunately there was probably some congestion on the busy junction to the main line, because we lost a few minutes there, arriving at platform 15 at 1053.

We headed for the Underground, but this was our mistake - we stupidly went to the Northern and Bakerloo entrance instead of the Jubilee one, which wasted a lot of time walking along travelators. When we eventually got to the Jubilee line platforms (platform 6 - Eastbound), we caught the first train that arrived towards Stratford, train number 302/trip 4, arriving at 1100**, with end vehicles 96023 and 96078, to West Ham, timetabled to arrive at 1116½. Calling at all stations, of course - Southwark, London Bridge, Bermondsey, Canada Water, Canary Wharf, North Greenwich, Canning Town and West Ham. This train was rather busy when we got on, but quietened down enough for us to find a seat before too long.

Unfortunately, due to a combination of our late arrival into Waterloo and (more importantly) walking the wrong route to the Jubilee line Underground platforms, we missed our planned train upon arrival at West Ham platform 6, so had to cut out a small portion of our route, visiting Pudding Mill Lane. A bit of a shame, but no big deal - this was a part of our route we only grafted on on the train to London anyway, so removing it again was no problem.

So, we continued on our original route by heading over to platform 1, and catching the 1129½ (public timetable 1129) c2c (CC) service towards Stanford-le-Hope, alighting at Barking, platform 4 - the first stop after West Ham. This train was formed of 357027, with headcode 2D16. It was again moderately busy, but we didn't bother trying to find a seat since we were only going one stop. It departed West Ham and arrived at Barking perfectly on time.

We then had a half-hour wait at Barking for our third diversion of the day - the 1157½ (public timetable 1157) CC service to London Fenchurch Street, arriving 1221 (public timetable 1223), but diverted via Stratford for the benefit of Christmas shoppers. Not all services were diverted this way because it was purely for the benefit of the public rather than due to engineering works. We crossed over to platform 8 - this platform due to it being one of only two through platforms that can access the junction, and watched as the train, formed of 357018, headcode 2B23, arrived.

The train was very quiet, so we found a table seat quite easily. We departed two minutes late, at 1159, and soon diverged at Barking, Tilbury Line Junction West onto the Gospel Oak to Barking Line, fondly known as the GOBLIN. We passed through Woodgrange Park, before diverging at Woodgrange Park Junction onto a chord to the Great Eastern Main Line, which we joined at Forest Gate Junction. We proceeded down the amusingly-named Up Electric line - amusing because all the lines have been electric for quite some time! We passed through Forest Gate and Maryland stations before calling at Stratford, platform 5. We ran parallel to the Docklands Light Railway, seeing plenty of Crossrail developments as well as the work in progress on Pudding Mill Lane DLR station, until we diverged at Bow Junction onto a short line back to the London, Tilbury and Southend line, rejoining at Gas Factory Junction. We called at Limehouse, and finally terminated at London Fenchurch Street, platform 1 at 1222, one minute late by the working timetable and on time by the public one.

From there, we walked to Tower Hill, platform 3 (inner rail/eastbound), in order to catch a Circle Line train to King's Cross St. Pancras. After a short wait, a Circle Line service with end vehicles 5557 and 5521 (C stock), forming train number 206 on trip number 8, arrived. We had a minor wish to travel on some C stock, so this was lucky. It was timetabled to leave at 1232, and arrive at King's Cross St. Pancras at 1245, platform 1. On the journey there (in which we managed to find seats without too much hassle), we commented on all the development around that area, mostly relating to Thameslink, and how it was hard to keep up with it all! We called at Aldgate, Liverpool Street, Moorgate, Barbican, Farringdon and King's Cross St. Pancras.

Upon arrival at King's Cross St. Pancras, we headed to London St. Pancras station, and (after missing the escalator the first time whilst trying to navigate the frankly ridiculous shopping complex), up onto the East Midlands Trains, high-level platforms. Part of the Thameslink core was closed, and so Thameslink trains were running from these platforms rather than the usual low-level ones.

When we made it up the steps, we noticed they had a rather unusual arrangement - a few ticket barriers had been claimed and segregated from the rest, and opened up (presumably normal ticket acceptance rules would have to be out the window!). They had a few people managing the queue from the barriers to platform 1, where the train stood, while we waited for a massive flood of people to depart. A small trickle, including us, boarded the train - the 1303 First Capital Connect service towards Bedford, scheduled to arrive at West Hampstead Thameslink at 1309½ (public timetable 1310), formed of 377522, headcode 1G24. The train, again, was very empty, so we boarded and it departed on time, after getting photos of it in this unusual location. The train passed through Kentish Town without stopping, and arrived on time at platform 4 of West Hampstead Thameslink. We walked to West Hampstead (tube), platform 1 (Northbound), and boarded the first Jubilee line train - end vehicles 96020 and 96037, forming train number 325 on trip number 7, the 1316½ towards Stanmore. We called at Kilburn, Willesden Green, Dollis Hill, Neasden, and Wembley Park, platform 3 at 1326¾ where we alighted.

We crossed the island to platform 2 and caught the 1329½ Metropolitan Line service towards Uxbridge, formed of the train with end vehicle 21084, forming train 440, trip 8. We called at Preston Road and Northwick Park before alighting at 1337 at Harrow-on-the-Hill, platform 4.

Our roundabout journey complete, we were now here in good time for the 1350 "Steam on the Met" service run by the London Transport Museum which we were booked on. Being a steam charter (a cheap one, incidentally!) with a rather tight timetable, naturally it was rather late, but this is only to be expected. It eventually arrived on platform 6, with locomotives L150 (GWR 4575 class, renumbered from 5521) and Metropolitan Railway Electric Locomotive 12 "Sarah Siddons" on the Uxbridge end, and Class 20 20142 on the rear. Metropolitan 1 was suffering technical problems so was replaced by L150. They also weren't sure whether or not Sarah Siddons would be ready in time, but it was.

We entered one of the 1950s ex-BR 4TC carriages (coach B, the one we were booked into), in a compartment. We had an enjoyable but all too short trip up the line to Uxbridge when we eventually left at 1400. When we arrived at Uxbridge (platform 2/3) at 1421 (compared to the scheduled time of 1408), we headed back without being allowed to alight, departing at 1426 (compared to scheduled 1412) and arriving back at Harrow-on-the-Hill platform 6 at 1444 (compared to scheduled 1430). The trip both ways was enjoyable, and more comfortable than you come to expect on the tube...

We had discovered that as a result of Sarah Siddons being ready there was one extra trip than had been advertised originally (presumably "buy on the day" only), so we decided to chase it down the branch on its next trip and take photos. To this end, we crosed to platform 4 once more, and caught the 1452½ Metropolitan Line service towards Uxbridge, calling at West Harrow and Rayners Lane on the way, formed of front vehicle 21115, forming train 422/trip 10. We alighted at Rayners Lane platform 1 at 1457, and got photos and videos of the train coming through on its next Westbound working.

To catch the return, we then stayed on platform 1 and travelled up to Eastcote, on the 1514 Metropolitan Line service towards Uxbridge, Eastcote (platform 1) being the first stop at 1516½. The train number was 425, with trip number 9, and the front vehicle 21068. We caught the train returning as the light faded. Finally, again staying on the platform, we caught the 1551 Metropolitan Line service towards Uxbridge, calling at Ruislip Manor, Ruislip, Ickenham, Hillingdon and Uxbridge platform 4, where we alighted at 1556½. It was formed of front vehicle 21108, forming train number 433, trip number 10.

We found a Burger King in Uxbridge where we had lunch/dinner and stocked up on chocolate in the M&S. We returned to Uxbridge station to take photos of the steam locomotive when it arrived, using the light inside the station as it was already dark outside.

Once we got our photos and the train had left, we embarked on the trip back. From platform 4 we caught the 1653 Metropolitan Line service towards Aldgate, formed of front vehicle 21097, forming train 444 with trip 11, and calling at Hillingdon, Ickenham, Ruislip, Ruislip Manor and Eastcote before we alighted at 1706½ at Rayners Lane platform 2 to change onto a Piccadilly Line service.

So we waited for a while on the same platform, until at last we caught the 1719½ Piccadilly Line service towards Cockfosters, formed of front vehicle 126, forming train 240 with trip 9. It called at South Harrow, Sudbury Hill, Sudbury Town, Alperton, Park Royal, North Ealing, Ealing Common, Acton Town, Hammersmith, Barons Court, and, at 1752, Earl's Court platform 5 where we alighted. We briskly walked up the escalator, aware that we were running out of time.

At the Earl's Court subsurface platforms, we looked at the tube times map and the "next train" diagram, did the maths and realised that we would not arrive in time. We started to form a contingency plan - there was another train we could pick up that would provide a cross-platform interchange with ours at Surbiton, where ours waits for some time.

However, this turned out to be unnecessary, for the train we caught on platform 4, the 1757 District Line service towards Wimbledon, formed of front vehicle 5596, train number 74, trip number 19, ran early. We called at West Brompton, Fulham Broadway, Parsons Green, Putney Bridge, East Putney, Southfields and Wimbledon Park before arriving at Wimbledon, timetabled at 1815. We arrived at (from memory) platform 2, and ran over the footbridge to platform 8, and caught the 1816½ (public timetable 1816) SW service towards Guildford with a few minutes to spare. It was formed of 455710, with headcode 2G53, and called at Surbiton, Hinchley Wood, Claygate, Oxshott, Cobham & Stoke d'Abernon, Effingham Junction, Horsley, Clandon, London Road (Guildford) and Guildford at a timetabled time of 2042½ where we departed.

This train was quite quiet, and departed a minute late at 1817, but unfortunately lost a few minutes throughout the journey, eventually arriving at 1913. It seemed to be accelerating very slowly out of stations, which implied to us it was probably leaf fall/wheel slip issues, which is only fair enough.

We arrived at Guildford platform 1 (the bay platform) and switched to platform 6 to catch the 1920 XC service towards Bournemouth, formed of 221133, headcode 1O24. It followed the route of the diversion from the morning, arriving into Southampton Central platform 4 2042½ (public timetable 2043). We departed at 1923 and arrived at 2044, making up a minute of delay along the way.

Finally, at the end of a long day, we changed onto the 2059 SW "Salisbury 6" service towards Romsey on platform 2, formed of 158888, headcode 2R76. We called at St Denys, before alighting at Swaythling platform 1 at 2106½. We were just a couple of minutes late, departing at 2101 and arriving in the end at 2108.

So ended our journey - we went our separate ways. All in all, it was rather good - a few minor delays throughout the day but nothing particularly game changing. In fact, an early train helped us quite a bit!

*In the first blog post I listed the calling pattern for the whole service - I think this is probably not worth it unless it is odd in some way, so in this post onwards I'll only list the stations called at between the ones we boarded and alighted from.

**London Underground usage is rather spontaneous and generally rushed, so I rarely record actual arrival/departure times, instead just recording the train number and later looking up the trip number and timetabled arrival/departure times. Unless otherwise stated, all London Underground timings are timetabled only.


Rail Journeys - Southampton Airport (Parkway) to Guildford and back

Posted On Sunday 8 December 2013 by Murray Colpman (Muzer)

Yesterday was the first ever Student Robotics Guildford tech day, and a few of us from Southampton wanted to help out, so what better way to visit than by train?

So, the four of us met at Swaythling [SWG] Station at 0705 - despite not visiting this station by train today, it's a useful meeting place where our walking routes home fork, and also allows us to buy tickets from SWG from the Ticket Vending Machine [TVM] in case we get more lucky on the way back.

Speaking of tickets, I was using a particularly good anomaly to get to Guildford [GLD]. I won't reveal it here, but people who know me probably know which one it is. Three of us got this with Groupsave-3 [GPS-3], and the final person buying a full-price ticket because he wasn't coming back with us and needed to use the ticket to go further along the route.

In order to avoid hassle at the ticket barriers at GLD, we also bought GPS-4 London Road (Guildford) to GLD returns. We did not actually use these tickets for travel, but just to avoid explaining why tickets are valid to barrier staff.

Before leaving SWG we spotted a large Freightliner train pass, but were unable to identify the loco class.

So, we walked to Southampton Airport (Parkway) [SOA] in good time to catch the 0738 South West Trains [SW] service towards London Waterloo - calling at Winchester, Woking [WOK] and London Waterloo; headcode 1W92. After watching the anticlockwise SW "Salisbury 6" local service (formed of a 158) depart and a CrossCountry Empty Coaching Stock [ECS] move (formed of a 220), our train, formed of 444005 and 444016, arrived on platform 1.

The train was empty enough for us to get a table for the four of us, with plenty of free seats, but not exactly quiet either - surprisingly busy for me, but then again, I never understand train usage at weekends ;). The train left on time, and the journey was comfortable - exactly what I expect from a 444. After a brief stop at Winchester, we continued on non-stop to WOK, with a hassle-free ticket check shortly before the stop, even gaining a minute somewhere along the line - arriving into WOK platform 2 one minute early, at 0820 compared to the scheduled 0821.

We crossed via the footbridge from Platform 2 to 5, in time to catch the invalid connection of the 0825½ (public timetable 0825) SW service towards Portsmouth Harbour - calling at GLD, Haslemere, Petersfield, Havant, Fratton, Portsmouth & Southsea and Portsmouth Harbour; headcode 1P17. The train arrived into the platform mere minutes after we did, formed of 450105 and 450101.

This train was significantly more empty than the previous one - we easily found a group of six seats facing each other to sit in, with enough room to be comfortable despite the 3+2 seating layout of the 450s due to the quietness of the train. We left on time, and though the journey was short, I can say once again that it was comfortable, as I can expect with a 450. We arrived at GLD platform 4 at 0832½, on time (public timetable 0833).

We exited the station via the West exit, it being closest to the university. The barriers accepted (and ate) our GPS-4 London Road (Guildford) to GLD tickets without hassle.

The tech day itself went rather well.

After finishing the tech day, three of us made our way back to the West entrance of the station, and entered through the barriers with the London Road (Guildford) to GLD tickets, again without fuss. After spending a while getting realtime trains to load (my phone's fault, not the site's!), we made our way to platform 5 for our train home, the 1934½ (public timetable 1934) SW service to London Waterloo, calling at Worplesdon, WOK, Clapham Junction and London Waterloo; headcode 2P66. After watching many 450s and a 455 pass through the station, our train, formed of 450544 and 450039, arrived.

The train was even more empty than the previous one, and again we found six seats facing each other to sit in. We left one minute late, and only just making up that one minute, arriving into WOK platform 2 at 1944, on time. The train was once again comfortable but the journey really too brief to appreciate.

We remained on the island platform but moved over to platform 4, watching a SW service towards Exeter St. David's and Bristol Temple Meads, formed of 3×159, leave. After once again watching multiple trains arrive and depart, including a rather fast 444 from Weymouth passing through without stopping, our train, formed of 444031, arrived. It was the 2000 SW service to Weymouth, calling at Winchester, SOA, Southampton Central, Brockenhurst, Bournemouth, Branksome, Parkstone, Poole, Hamworthy, Wareham, Wool, Moreton, Dorchester South, Upwey and Weymouth; headcode 1W81. We found a block of four seats in the front coach, unfortunately not facing each other, and sat down. After leaving on time, the guard came around almost instantly and again checked tickets without hassle. The train was again comfortable, though unusually rattly for a Desiro (but not nearly as rattly as the average Turbostar, for example). The journey was comfortable, although the Rail Gourmet-run trolley service had run out of Coke which I was planning on buying.

We gained a minute in the run to Winchester, but departed on time so arrived at SOA on time exactly at 2040½ (public timetable 2041), on platform 2. Unfortunately this did not time nicely with a local service to Swaythling, so we walked back.

Shortly after leaving SOA, we spotted a light loco running from West Ealing Plasser Sdgs to Totton Yard, again being unable to identify the class.

So, as can usually be expected from SW, an almost perfect journey. This is a good start to this blog...

EDITED 20131214 - added platform numbers where not explicit; added public timetable.



Posted On Tuesday 3 December 2013 by Murray Colpman (Muzer)

Because I'm bored/crazy/something, I decided that I'll blog about every train journey I make - short or long, eventful or uneventful. This is because Tom Cairns still hasn't fixed RailMiles, and also because I want to show that most train journeys are perfectly normal and when there are delays, they're probably for a good reason (I'll try to find out what that reason is for anything more than about ten minutes).

I have train logs that I've been keeping on my phone to various levels of detail from late July onwards. I might upload these at some point. This logs the journeys I made to a level of detail that will allow me to determine which route the train took (possibly with the help of external documents that I hold), but aren't really intended for public consumption. They also contain the unit numbers on which I travelled (if I remembered to note them down - most of the time), headcodes or train/trip codes or any equivalent that I can find easily, and scheduled (always WTT) and actual timings.

To start us off, next weekend I'll be going to Surrey on the Saturday followed by London on the Sunday, so I should have some interesting things to talk about by then.

I'm a rail enthusiast but have many other interests too, so don't expect anything like a post every week or something.


Timlan meeting #7 minutes - 2013-11-29

Posted On Friday 29 November 2013 by Murray Colpman (Muzer)

Timlan, this is, Timlan.

Last week we had a meeting, but we didn't post minutes for it. In that meeting, we made some scripts that are useful for TTS. They can be found in our github organisation (see below).

We added a word for "start" so that we can use time phrases by saying "5 hours after the start of today", for example, for 5am. It should be relatively self-explanatory - just remember that you use afrax/whatever before is with tion.

We also fixed VOSFace to do UTF-8-ing properly, as well as the timlan lookup tool.

Oh, we also fixed input.php because it was broken, again.

Finally, we decided that Soviet Timlan would be cool and would only have one mood, the imperative.

бинашеьт тимлан суиOpen TCP Editor

We also started using git to store timlanny things. It's not all been placed in git yet, but the stuff that has is here


More potentially breaking things

Posted On Monday 18 November 2013 by Murray Colpman (Muzer)

We have done some database/PHP-related maintenance.

Please report anything broken that looks unintentionally broken.


Bye bye y... hello y

Posted On Saturday 16 November 2013 by Murray Colpman (Muzer)

OK. Very shortly after the last post, we decided to rectify this once and for all. First of all, a restatement of the problem.

* W isn't a problem because the "semi-vowel" used by timlan is far enough away from the closest actual vowel in timlan (our u) to be distinguishable relatively easily (compared to y) - see ua and uwa. Therefore, w gets to stay for now. However, remain vigilant when pronouncing ua and uwa.
* Y is a problem because the "semi-vowel" is the same as timlan i, so iya (for example) sounds the same as ia unless a glottal stop is inserted into the latter, which we didn't want to make mandatory for all words.

Therefore, we have replaced y with the OLD pronunciation of x - that is, the "the" th, or the "with" th.

This is easy to say for English speakers and removes all ambiguity of pronunciation (hopefully). Hooray!

Two words have been changed because of this:

hyorax --> horax
binahyo --> binayo


Uh-oh, glottal stops on the attack again.

Posted On Saturday 16 November 2013 by Murray Colpman (Muzer)

After getting confused with the difference between unogiy? (? is the fonti representation of the normally unwritten and optional schwa, remember?) and unogi? (obviously not a word but there\'s no reason there couldn\'t be any similar confusion between actual words), we decided we need to pull our old friend, the glottal stop out of the cupboard again.

They can now go between any two vowels. They still MUST go between two identical vowels, but you can now optionally place them between any two vowels within a word, if it makes pronunciation easier. This includes between the last vowel and the optional schwa.

In this case, you can easily distinguish, therefore, between unogi? and unogiy? - pronounce the former as unogi^? and the latter as it stands.

We might try to come up with a better method of doing this, since it\'s mildly horrific. In the mean time, this probably works fine.


Timlan meeting #5 minutes - 2013-11-15

Posted On Friday 15 November 2013 by Freddie Nash (FredFace)

Minutes. Five of them. Equal to 300 seconds. And now you can say that:

eqeht c-d-r-eta unis h-eta karOpen TCP Editor

So, that's a thing. We also have derived units that comprise latin characters only. Finally, we have the ability to derive arbitrary units, using /!!/ blocks.

/!unimac-isuc-!/Open TCP Editor

== metres per second

or, for short

/!unimaisu!/Open TCP Editor

I'm sure you can work out how these work, so I don't need to bother telling you.

Ordinary units are defined like elements, but since capital letters actually matter, they are instead denoted by replacing the "i" with an "o", thus: unopiaOpen TCP Editor

This works for base units and named derived units. These are all in the dictionary for ease of use - therefore, they are proper words, unlike elements.

However, unnamed/arbitrary derived units are obviously not words, but they are still timlan so need to be placed in /!!/ blocks. They are constructed by appending an "a" after each unit for a positive exponent or a "u" for a negative one. If the modulus of the exponent is 1, the next unit (preceded by an i or an o as appropriate) may optionally follow immediately - otherwise, a number must be provided, this being the modulus of the exponent.


/!unikigaimaisud-!/ /*kilogram metres per second squared, also known as newtons, or unon*/Open TCP Editor


We haven't really done much else, aside from reminding ourselves about how time phrases must work, and looking at the potential for a timlan TTS.


Timlan meeting #4 minutes - 2013-11-01

Posted On Friday 1 November 2013 by Murray Colpman (Muzer)


Yes. We did a rather large thing involving aspects that Freddie has blogged about.

We also hacked on VOSFace but we won't bore you with the boring details.

Finally, we made this

Based on a much earlier thing, that we managed to find using the keywords "it's an image", "it has lines", "frame of reference".

By the way, get on with backslashes, Freddie



Posted On Friday 1 November 2013 by Freddie Nash (FredFace)

We have added the Perfect Progressive Aspect to Timlan. HUZZAH!!!

It has only taken us 2 hours to work out why it is useful. Work is also a noun, bear that in mind.

This has brought along a fairly important change also. You may be wondering why we have the Perfect aspect, the Perfect Progressive Aspect, but not Progressive Aspect? We DO have the Progressive Aspect, and it is now the default for all tenses, except the generic tense.

This does work nicely, as generally when you use what we might have called the "Simple" Aspect, you probably meant to use the Progressive Tense (I'm eating) or the Perfect Tense (I have eaten).

Naturally, using the Perfect Progressive Aspect in Timlan is as easy and using the Perfect Aspect in Timlan. In this case you just stick "pete-" at the start of the word.

For example:
ceveabus unafelem eiami lapopis eateht hat~s ei onafelem eiami tidela peteeateht hat~s eiOpen TCP Editor
Please note that TCP does not support pete at this time

The example uses the rather unused "ceve-" - hello ceve! Ceve is the verb identifier for conjunctions, as most of you will not recall. The Perfect Progressive appears to be useful for distinguishing between a current state of operation, and a former state of operation, but where you are performing the same operation.

This is a pretty extreme example, because we can't think of many uses for this new Aspect, but we know it has uses, and that it seems to work.

This new Aspect is also useful when combined with Time Phrases, something Timlan has sadly little support for. You could, for example, say that you have been working for PI hours, but that you are working for 5 hours, meaning you have an irrational amount of time left to get on with it.

Because of the confusing nature of Aspects, and our lack of support for examples using them, this is a bit of an experimental thing (don't tell Muzer I said that), but we'll see how it goes when we are thinking about it in real life, doing real things, not trying to make up use cases, something we don't seem to be very good at describing to each other.

Final note, the short-name for this Aspect is "perfprog", in case you see that about the place.


Timlan meeting #3 minutes - 2013-10-11

Posted On Friday 11 October 2013 by Murray Colpman (Muzer)


Timlan meeting #2 minutes - 2013-10-04

Posted On Friday 4 October 2013 by Murray Colpman (Muzer)


Timlan meeting #1 minutes - 2013-09-28

Posted On Monday 30 September 2013 by Murray Colpman (Muzer)

We have decided to post the results of our weekly meetings on the blog.

  1. We decided that titan (to) is fine as it is. Definition added from Wiktionary at a later date.

  2. We decided that the syntax of ton is fine, but the English shorthand has been changed to "self-determining", which is less misleading.

  3. The Tring Timlan logo was adopted despite its offcentreness. An alternative version for Timlan Lookup was developed and put in place.

  4. Documentation on the Timlan wiki commenced with the Introduction page.

  5. Created a mediawiki plugin for inline TCP, based on Tom's TAIOS plugin.

  6. Identified the use case of a new affix, called erek, for identifying the type of contents of a /!!/ block. This is so humans know what to expect, and so that parsers have a way of knowing whether or not they should try to parse the inside of the block without having to try to parse it first. This should be added and documented at some point in the future.

Next meeting: Friday 2013-10-04


Tim36 intro

Posted On Monday 5 August 2013 by Murray Colpman (Muzer)

Welcome to tim36!

First of all, if you notice anything broken, let us know. The only thing that could break things is probably the hostname change.

Here are some comparisons of the specifications

Attribute       | Tim32                           | Tim36
Laptop model    | Fujitsu Siemens Lifebook S-4546 | HP Pavilion dv6000
RAM             | 256MB                           | 1GB
CPU             | Pentium III Coppermine          | Pentium Dual-Core T2060
CPU speed       | 600MHz                          | 1.6GHz
HDD capacity    | 20GB                            | 250GB
Everything else | Irrelevant                      | Irrelevant


Unexpected additional change

Posted On Monday 5 August 2013 by Murray Colpman (Muzer)

I also found a spare computer, and now points to it. However, this new computer is not tim32.

Since it was brought into service 4 years later than tim32, we give you: tim36.

Tim32 is still alive and will be coming into service in the near future, for which purpose is still not known for certain.


memcpy(murrays_house+free_space, tim32_location, sizeof(tim32));

Posted On Thursday 1 August 2013 by Murray Colpman (Muzer)

Tim32 is moving to my house this Sunday.

It'll probably be down for quite a bit of the day, but I'll try to make the migration as fast as is humanly possible. We shouldn't need to do any DNS reconfig, so aside from dodgy caches it should not take any additional migration time.

We might try migrating the HDD across to a bigger one (which is a much-needed upgrade).


PHP for you and me and everybody because it's not actually a very nice language but it's an industry standard so oh well

Posted On Thursday 1 August 2013 by Murray Colpman (Muzer)

PHP has been reconfigured, so if you see anything broken, please inform a member of railway staff or a police officer. Er, I mean, me.


SE32 - security-enhanced Tim32

Posted On Friday 1 March 2013 by Murray Colpman (Muzer)

Today we've performed a security-related task we've been putting off for a while that involves databases.

If you notice any database errors on anything user-facing, do let us know. We did leave a few things intentionally broken but hopefully nothing anybody uses.


PS: Sorry for the downtime today. It was due to firewall issues. Well, we said we'd have downtime, it just came a couple of days late ;)



Posted On Tuesday 26 February 2013 by Freddie Nash (FredFace)

Components or all of Tim32 may be down for chunks of tomorrow (27th Feb), from 13:00 (UTC) onwards.

This isn't downtime for maintenance, but if we have to take it down, then we may make use of said downtime to perform some long due maintenance (and depending on how many engineers we have about to make decisions and stop me getting things wrong).

tokouhaht setuh sui


Inline TCP

Posted On Saturday 19 January 2013 by Freddie Nash (FredFace)

Having thought about a couple of nice ways we might make the timlan blog less scary, we had Tom write some nice JavaScript to provide inline TCP support! You can now see the TCP parsing of timlan in blog posts, and change it to see how that changes the output.

iqeht kel tiOpen TCP Editor

Just click the little icon on the right of the timlan text, and you will see the mini-inline TCP parser appear. Click it again to close it. It works really rather well, and we are all very pleased with it (thanks Tom!)

It uses tcp.exe (same as this), which is the timlan parser written in C++. It is much nicer code than the C# parser (which is always up to date) but sometimes falls behind the times a bit. For now it is pretty good, we updated it just for this.

The only real issue with with the system is that TCP is very forgiving, something a timlan parser shouldn't be! It will much rather spit out rubbish than tell you the sentence is nonsense. There are couple of issue that need resolving in TCP anyway, so that will hopefully be improved along the way.

tokouhaht setuh suiOpen TCP Editor


Relative clause overhaul

Posted On Saturday 19 January 2013 by Murray Colpman (Muzer)


We've been discussing this for a while but only just got down to putting it properly formally today.

We have overhauled the relative clause system. Rela and relo are joined by two new prefixes, rila and rilo, which have finally been added to the fixionary DB.

Basically, we've always had the issue with timlan, as to whether:

taropeht relaiqeht luc hat~s eiOpen TCP Editor

...means "I like pasta (which is lush)", or "I like lush pasta". Now, this issue is solved!

taropeht relaiqeht luc hat~s eiOpen TCP Editor means "I like pasta (which is lush)", and...

taropeht rilaiqeht luc hat~s eiOpen TCP Editor

...means "I like lush pasta".

Rila/rilo are used when the relative clause is part of what determines the noun in that context. If removing the relative clause would completely change the meaning of the main clause, then this is probably the case and rila/rilo should be used. In other cases, rela/relo should be used. Note that if you're using tif, which is used for determiners (don't get these confused with determining relative clauses!), it only makes sense to use rila/rilo.

Supplements are now all defined as expanding to rila/rilo, to make supplements non-optional.



Posted On Tuesday 20 November 2012 by Murray Colpman (Muzer)

After a little more talk, we've decided to in addition to the previously mentioned changes, add a determiner for proper nouns. Not really much to say here, it's a determiner, and it works on proper nouns ;). It must be used on proper nouns both of the Timlan and foreign import variety.

This determiner is ton. For instance:

taropeht toneta timlan eiOpen TCP Editor

"I like timlan"

In addition, we've decided to make /::/ default to English, for brevity reasons when using it for proper nouns. We chose English because it is the language of code and the internet, not at all because we happen to speak it ;).

eqeht toneta /:Murray:/ ei

But compare:

haiseht /:Murray:/ ei

Because there Murray is just being used as a word rather than a proper noun.


/ mini-overhaul

Posted On Monday 19 November 2012 by Murray Colpman (Muzer)

We have forever been noticing constantly issues with /, but never getting around to fixing them. Now, we finally have. In order from easiest to hardest:

* Nomenclature. The two characters together of an opening thing, for example /*, is called a marker. The / is called the mark, and the second character is called the er. Note that */, for instance, is called an ermark!

* Pronunciation. We realised that there isn't actually a nice place to find the pronunciation for the ers. Initially I tossed around the idea of changing the ers to be the actual character pronounced (c, p, k, m, etc.), but neither of us liked this idea, so we decided to keep it as it is and just add the er characters to fonti. This will be done shortly.

* Nesting and escaping. Nesting is now allowed, but it'll probably take an age for it to be added to VOSFace and even longer for TCP. Escaping has always been a problem we've ignored, but not for longer! Escaping in text and in speech works in a similar way but may be done at different points due to the obvious differences between pronunciation and writing in this special case. Only the mark character or sound, if it appears in text, needs to be escaped; the ers never do. To escape the mark, you simply write or say it twice - for instance, /*This is a website*/, or (in speaking) /"I want a bi^it of toast"/ (where ^ is the glottal stop, of course, as in usual fonti). Of course, this means if you do say a mark sound and want it actually to be interpreted as a mark, the next sound you say cannot also be a mark - this shouldn't be a problem, as this would rarely occur, and if it does, in the last post you saw that a schwa can always be added to avoid ambiguity!

* We've overhauled !. It now doesn't mean a proper noun, and in fact can never be used for foreign text. You'll see what the replacement is later. ! is now used for Timlan text that is not a Timlan word. There may be some ambiguity as to what this means, so I'll spell it out:
** Timlan text is anything that is pronounced with fonti.
** Not a timlan word means that it is either a special case construction (like eli-) or nothing to do with a word at all. This does not extend to numbers as they are so vitally important. Examples of things that are timlan text but not timlan words could include spelling something out with the Fontibet (/!m'uz'er'!/ for instance), and elements and whatever we later decide to use for compounds.
This is so that parsers won't have to get out of hand coding for special cases so that parsing doesn't break if they come across a special case they don't even care about, but they can still parse the special case if they do. Elements are a prime example of this - now a parser can simply ignore /!elih!/ if it doesn't care about elements, and still be able to parse the timlan properly. Finally, timlan proper nouns *not* in the dictionary would also fall into this category, but we're not sure whether those should or shouldn't exist yet. If we decide they do, they'll be covered by this.

* We've added a new er. The new er is :, and is finally an er for specifying foreign text! It is pronounced as timlan "x", for "eXternal". "But wait", you might ask, "how do I specify the language in which the foreign text is written?". Well, the answer is simple - use /!!/ with the ISO-639-3 code! ISO-639-3 code is a standard that aims to define a code for every significant known language ever in existence, so you shouldn't have too many problems with missing languages, with one obvious exception - Timlan isn't there! If we ever have a need for a code, we'll figure out something else, but for now, this should never be needed for Timlan phrases anyway. You specify the code by spelling it out inside the /!!/ marker/ermark - for instance, /!eng!/ (pronounced /!en'g'!/). You attach this to the /::/ marker/ermark by using the inilan verb or the alan suffix (these will be added to the dictionary/fixionary in due course), thus: relainafelem /!eng!/alan /:France:/ eateht hat~s ei. If you do want to specify a very exotic language that doesn't have a code, but does have a name in a language that does have a code, you can always chain suffixes together thus: haiseht /!eng!/alan /:Na'vi:/alan /:Palki:/ eiami cute



Posted On Monday 19 November 2012 by Murray Colpman (Muzer)

After some deliberation, we've decided to repurpose the recently (by Timlan standards!) introduced schwa vowel for a more useful use. This does not conflict with its previous use, but it does conflict with the explanation in the previous blog post.

Imagine we're living in a parallel universe where the Timlan vocab actually has many many words. Imagine that "ha" and "t~sei" are two of those words. When speaking, how would you distinguish the following two sentences:

eateht hat~s ei

eateht ha t~sei

This has been a problem largely ignored since the start of Timlan, but one that now has an answer:

Schwas can be optionally used where clarity is of the utmost importance to signify the end of a word. For information on what a schwa actually is, see the previous post. For learning purposes only, schwas are written ? (but this does not occur in normal Timlan). So, if you knew confusion could occur, you could optionally say:

eateht hat~s? ei

or even

eateht? hat~s? ei?

Fonti will be updated shortly.



Posted On Saturday 20 October 2012 by Murray Colpman (Muzer)

(Sorry for this being the first timlan post in ages)

It has come to our attention recently that there are a few word-ending consonants that can be rather hard to say in Timlan without making it hard to distinguish between the consonants or inadvertently adding a vowel onto the end. Freddie has particularly noticed this with "r"-ending words like "lir".

So, we've added a new rule (probably will be added to fonti.txt at some point) - schwas that occur anywhere in words can safely be ignored. This is because Timlan has no schwas, and its vowels cannot be replaced by schwas unlike in English. Schwas, for the unaware, are basically that indistinct vowel sound you hear a lot in (mostly unstressed) English syllables, often transcribed literally by people as "uh" (but doesn't actually sound much like "uh"). It can be found, for instance, in "the" when spoken by a person who doesn't pronounce it as "thee" (I told you it was common), and as the first syllable in the word "above" for instance. So in this case, you could prononce lir as "lir(schwa)", which would probably be transcribed by an English person as "leeruh". Schwas are represented in IPA by the character 'ə'.

Schwas should be avoided when it is possible to say a word without one, but it is perfectly permissible to have them where it would decrease phonetic ambiguity or increase ease of speech.


Something about something

Posted On Wednesday 22 August 2012 by Freddie Nash (FredFace)

Sorry about the Tim32 website downtime (Tim32 hasn't been down, except very briefly for another router change).

We pulled it down to allow us to (slowly) make important security modifications.

If anything seems to be !working, let us know.


Going, going, gone, come back up again, maybe going again in the future

Posted On Friday 20 July 2012 by Murray Colpman (Muzer)

We had a little hitch with Tim32 today. Nothing a quick reboot didn't fix, but we're not really sure what went wrong and are generally slightly worried. Apologies for the downtime.

Freddie is going to Spain for a couple of weeks soon. Tom and I will have a backup ready by then and one of us will run a minimal mirror should Tim32 go down when Freddie is away.

Data will remain safe.


New DNS Provider

Posted On Wednesday 4 July 2012 by Freddie Nash (FredFace)

Turns out DynDNS have turned, and we have had to flee in fear for our uptime to No-IP, an alternate provider. HUZZAR

Our thanks go to 123-reg for not being useless, and everything ought to be fine now.

If you notice anything not working a few days after today (4th July 2012) then let us know.


Sets and lists and matching

Posted On Wednesday 30 May 2012 by Timlan Team (timlanami_ludef)

After reading a mostly-incomprehensible document about Lojban, we've realised that we needed some way to specify exactly how groups of entities interact with other single entities or groups of entities. We have come up with two different systems which should be used together to make a fully useful system.

Firstly, we need to define two new (to timlan) terms:

* A LIST is a group of entities in which every entity is taken individually; the sentence can be split into multiple separate sentences, one for each entity, and effectively mean the same thing.

* A SET is a group of entities taken as one entity grammatically.

*** Simple relationships ***

Many entities to one entity.

Three people carried the seat:

This sentence is an example of a many to one relationship. In this sense, we mean that there is one seat, and three different people carried it, but not at the same time. In other words, it means the same as:

Person1 carried the seat. Person2 carried the seat. Person3 carried the seat.

So, in Timlan, we therefore want the subject to be taken as a list. How do we do that? Well, we do nothing, because we've decided for multiple reasons to have multiple entities be a list by default:

lugicoht seluxeta kesit f-eta rek

= Three people carried this seat, translates also to "person1 carried this seat. Person2 carried this seat. Person3 carried this seat."

One entity to one entity

We will use the same example sentence, "three people carried the seat". In this case, the three people would have been acting together, as one, to carry the seat. Therefore, we want to convert the list that is the three people into a set. So, we use a new "verb":


This verb means "treated as". You use it with sit~s or (rarely) lit~s (set and list respectively) as the object in order to specify a conversion. Lists can be converted to sets, and we haven't decided about the rest yet:

teritelem sit~s lot

= "treat (thing) as a set".

So, the sentence would be:

lugicoht seluxeta kesit teritelem sit~s f-eta rek

= "Three people (treated as one entity) carried the seat" - in other words, three people carried the seat together.

Similar things can be worked out for one person and three seats in both interpretations of "three seats", this is left as an exercise for the reader.

*** Complex relationships ***

If you're intelligent like us, you might at this point be thinking, "but what about many to many relationships? Aren't they illegal or something?". Yes, we did have some issues when coming up with how to handle this, but we think we have a solution - you specify the matching with a verb attached to the clause. These clauses should usually be relative so that words like puc can still be used without horrible confusion ensuing.

Our example sentence: Three people carried three seats.

So, we have another lot of words:

mac - matched as (verb). The subject is the clause to be matched, the object is the matching type (see below).
fev - one-to-one bipartite. Each item in the subject list is matched with its corresponding item in the object list. For obvious reasons, the lists have to be the same size.
kev - complete bipartite. Each item in the subject list is matched with every item in the object list. These lists can, of course, be different sizes.

We have left it open so that, in the future, we can specify user-defined matchings, but we haven't yet decided on a decent syntax for that so we can leave it for another day/year.

If a mac clause is not given, it is left vague - you have no idea who carried which chair in our example.

So, without further ado, here are the examples and what they mean:

lugicoht f-eta rek f-eta kesit

Since f-eta rek and f-eta kesit are both lists, this sentence is intentionally vague in who exactly carried what - the important information it conveys is just that, by whatever means, three chairs were carried, and it was three people who were involved in doing it.

macelem fev lugicoht f-eta rek f-eta kesit

fev is the one-to-one matching specifier. So, this sentence means that each person carried a single chair.

macelem kev lugicoht f-eta rek f-eta kesit

kev is the complete matching specifier. So, this sentence means that of the three people, each one at some point carried all the seats (but not at the same time necessarily).

*** ANDS ***

And, finally, we realised that having "and" for and in every English usage of the word "and" was a failure of an idea. So, we made another word for and:


This one is used just as a generic verb, indicating that 2 entities (or lists of entities) are to be handled as one list.

For instance

taropeht danelem ohk tok ei

= "I like apples and oranges (as a list)"
= "I like apples. I like oranges."



BETRAYAAAAAAAAL! cokoiqeht luxenal sui

Posted On Tuesday 15 May 2012 by Murray Colpman (Muzer)

On Friday, the 20th of April, a person with IP address using a Vodafone contract and Android 4.0.4 with the default browser on an HTC Legend (Ice Cream Sandwich Build IMM76D) with root database access modified the definitions of some Timlan words when he should have been concentrating on Brownian Motion!

The changes have been undone. Any further transgressions will result in more severe DISCIPLINARY ACTIONS!!!

This kind of nonsense is what would be expected from creationists, not rational Tim32 engineers...

ON A LIGHTER note, we're working on a new communication-themed Vocab O'Clock, but it'll probably take a while because we're actually being competent in how many words we add this time...


Multi-digit basification

Posted On Monday 14 May 2012 by Murray Colpman (Muzer)

So, we realised that having |bases| limited to <=16 isn't particularly a good plan, so we needed a SOLUTION. And then one arrived into someone's mind, called multi-digit bases.

Basically, these are bases in which each "logical" digit is actually composed of two "physical" digits (ie, numerals). Think of the two (or more!) digits combining to make a SUPER-DIGIT or something...

Basically, it's quite easy. You just use a higher number after y- for multi-digit bases. The number of digits in a given base where |base| > 1 is given by the integer part of:

N = (log(base - 1)/log(16)) + 1

(Where log is a logarithm to any base (NOT sln...))

(For base = 1 or base = 0 or numbers in between, number of digits is obviously 1...)

So, remember that when you have a base that uses two (or more!) physical digits, each logical digit MUST contain that number of physical digits, so insert a leading 0 if not...

For instance

c-b-b-d-y-c-c- = c-c-d- = 274 in Arabic


Some new rules.

Posted On Wednesday 2 May 2012 by Freddie Nash (FredFace)

As of today, Timlan no longer recognises Prime numbers greater than 15, here is why:


With any other number, you can nicefully set up divisors by changing to a suitable base (eg. 10, for base 100)

Actually, the above is lies, but it seemed a fun/scary way to introduce Multi-digit bases which we will make a post about when we've verified everything, and blah blah blah.



Posted On Monday 16 April 2012 by Freddie Nash (FredFace)

No, not pirates, just y-

I've now convinced myself that Muzer was right all along, and that y- should be a Timlan base modifier! HURRAY!

It is now, meaning Timlan fully supports all bases between 0 and 16 (and maybe some negative ones, but I can't work out the bounds)


In Timlan, it used to be hard to say numbers like 5.8, because (8 / 10) * 16 is not a natural number (its 12.8).

Conversly, it is hard in decimal to say c-z-m- (0x1.9) because (9 / 16) * 10 is not natural either (its 5.625)

(If you don't follow the above, don't worry about it - its just me explaining why a flexible base is good)

As you notice, in English we kind of have a way to say hex numbers, with the 0xblah notation, but that is just stolen straight from the computer scientists.

In Timlan, using base 10 is as easy as sticking y-n- at the end of any number. No really.

y- means that the following changed the base
n- is 10
y-n- means the base is 10

So in Timlan, if you want to say 5.8 (horrid in hex) you just say h-z-l-y-n-!

Note: TCP is yet to be updated, but will be soon


DB Fail

Posted On Monday 16 April 2012 by Freddie Nash (FredFace)

Ok, so, from the start, I have done an AWFUL job of the Timlan word database. Now that I have less free time, I've been thinking that maybe I should expend valuable time sorting out word catagories (yes, catagories) because they are really unhelpful.

As mentioned before, prepositions have always been distinct, and there originally (and currently (and from now onwards)) this is pointless.

The biggest issue is that adjectives have always been 'nouns'.

Only recently did this strike me a stupid, all for the same reasons that we added 'aq' (typeof). Here is why

Adjective in use:

iqeht ripeli timlan

Noun in use:

aqeht c-eta langer timlan


So, if the other Timlan engineer(s) agree, I will whitle away hours (minutes?) going through the nouns and changing those that are adjectives into nouns.

In Developing Timlan, we make a point of not being wary of ripping it apart and re-build whole slices of the lush cake like metaphor for it. However, this is as good an instance of forward planning as any. Had I not been mislead by my desire to remove all the forms of is, I would have had an adjective catagory, and all would have been good in every way, as if we decied to change them all to nouns (hypothetically) we could do so with a simple script.

Enough of me typing, just really wanted to let you (valued probably non-existant readers) know that a bigish (slow) change might incur, and to remind you that with Timlan you are never in doubt about what a word in a real sentence is, even if the dictionary lies.


Vocab o-Clock

Posted On Saturday 14 April 2012 by Freddie Nash (FredFace)

This Vocab o-Clock (sorry, no good puns) is on the theme of Police (why not?)

susan - suspect
efqsot - police force
efqsaset - police officer
evimon - criminal
ofalak - break (verb) (eg. rules)
rul - rule
rulen - law (lots of Laws (British Law is lots of distinct Rules)
oxanen - apprehend (verb)
pecus - pursue
nutezal - neutralise
kontan - confine
foregat - investigate

tokouhaht setuh sui



Posted On Tuesday 10 April 2012 by Freddie Nash (FredFace)

Indeed not, however for some reason Tim32's IP address keeps changing and the DNS not working, so I have to tell the nice interface at 123-reg our new address so that can work again.

Sorry about this.

We'll probably sort out a proper solution in the near future.



Posted On Monday 26 March 2012 by Freddie Nash (FredFace)

(We [have]thought &null;)

peceheekaht setuh /!peve!/

'peve', that nice prefix that isn't nice and made sense, turns out to be effectivly the same as that even nicer prefix that definatly makes sense, 'rela/relo'! (The subject/object returning relative clauses prefixs for ye unenlightend).

Here is how to use 'prepositions' (for some not very good reason I decided they should have their own catagory (yes, catagory) in the DB):

OK, SO - a quick intro to relativeness

Think of a BINARY TREE (like a Timlan Sentence). IT IS A TIMLAN SENTENCE.

The thing that is at the very top of the tree is the main phrase. When you use relative clauses, it is like passing the sbj/obj up the tree instead of the phrase itself, so...

iqeht ripeli relataropeht hat~s seluxeta cute

(this cat (that likes pasta) is epic)

this looks like this:

         |                |
       relataropeht     ripeli
    |             |
seluxeta cute    hat~s

(Sorry, no FTP client here to upload pretty pictures)

The main clause would have to be at the very top, and it is, but what is exactly?

Spolier: iqeht ripeli seluxeta cute

The sub clause relataropeht hat~s seluxeta cute retursn the sbj, due to 'rela'. This means that you can essentially omit all the clause ness and replace it with just the sbj (would be obj with relo): seluxeta cute! (Remember supplements (determiners (selux (eta))))

Hopefully you understand now! If not, learn to code, and it will ALL make sense (and we will not have had to put any effort into said revelation).

Now, for when the relative clause is ABOVE the main clause. This is bascilly the same, it just returns the main clause.

Consider: relaiqeht kel eatoht hat~s sui

((it is good that) you ate pasta)

         |                |
       eatoht            kel
    |             |
   sui          hat~s

Here, the eatoht hat~s sui is the main clause, and it is passed up to the top by the relative clause.

See? How much fun is that!

Anyway, prepositions are just scary verbs (which arn't scary), so you can just use rela to state the something occurs, and that it takes place somewhere.

Regarding tenses... prepositions are funny and tend to take their tense from another verb. When outside the main clause (passes it up) the main clause provides this tense, and the preposition (verb) will take the generic tense, effectivly inheriting the frame of reference.

Now, for when not to use rela.

I am in France.

This is only one clause, and it needs a tense. 'am' is a non word, 'in' is a preposition (verb).

inafeht /!France!/ ei

Job done!

This is the same as ever, but without 'peve' glued on the front.

None of this is nearly as hard as it seems, its just lush binary logic goodness.

Some sample sentences:

inafelem /!France!/ eateht hat~s ei
((I eat pasta) in "France")

taropeht relainafeht /!Bristol!/ ayiketa lapopis ei
(I like (all houses in "Bristol"))

iqeht ripeli relaaqeht ripeli timlan
(tautology :D)



Posted On Wednesday 21 March 2012 by Murray Colpman (Muzer)

We're just thinking about the following:

* The difference between "I kill humans who are stupid" and "I kill humans who happen to be stupid"

* Other things.

* Whether or not we should replace soho- and peve-


Beware Docs

Posted On Sunday 11 March 2012 by Freddie Nash (FredFace)


(We will soon make the fixionary front end, and then everything will be better).


tokouhaht setuh sui



Posted On Thursday 1 March 2012 by Murray Colpman (Muzer)

We've added a way to accurately reference any previously-mentioned noun or clause in Timlan, as long as the speaker has a good memory ;)

To that end, we've added three new words:

to - refers to the object of n sentences ago (default: 1)
ta - refers to the subject of n sentences ago (default: 1)
tq - refers to the clause n sentences ago (default: 1)


iqeht nrib eiami cute. iqeht pali ta

= "My cat is blue. It is cute."

iqeht nrib eiami cute. iqeht value to

= "My cat is blue. The blue is shiny."

iqeht nrib eiami cute. taropeht tq ei

= "My cat is blue. I like that."

These can be stacked, in two different ways. The trees always start from the root clause (indicated by the first non-supplement word). First, as one might expect, they can use possessives:

taropeht relaeateht sohohat~s eiami cute ei. iqeht pali toami ta

= "I like my cat that is eating pasta. It is cute."

taropeht relaeateht sohohat~s eiami cute ei. iqeht luc toami to

= "I like my cat that is eating pasta. The pasta is lush."

If the tree ends on a relative clause, the subject or object of the clause (depening on whether rela or relo is used) are inherited by default. So, the first sentence above can be phrased simply:

taropeht relaeateht sohohat~s eiami cute ei. iqeht pali to

= "I like my cat that is eating pasta. It is cute."

If you wish to refer to a clause, you must always end the tree with "tq". This will stop parsers (people, programs) from inheriting nouns from relative clauses, but must also be used for referring to any clause, just for clarity purposes:

taropeht relaeateht sohohat~s eiami cute ei. iqeht kel toami tq

= "I like my cat that is eating pasta. The fact that it is eating pasta is good."

Trees have a short form, as using so many "ami"s could get tiresome. The short form is simply concatenating all of the "to"s, "ta"s and "tq"s into one word:

taropeht relaeateht sohohat~s eiami cute ei. iqeht pali tota

= "I like my cat that is eating pasta. It is cute."

taropeht relaeateht sohohat~s eiami cute ei. iqeht kel totq

= "I like my cat that is eating pasta. The fact that it is eating pasta is good."

Finally, an ordinal number can be prepended to change the number of sentences ago you are referring to. The previous sentence (the default) is c-ah, the sentence before that (for instance) is d-ah, and the current sentence (not easy to use in speech, but possible in written things!) is b-ah:

taropeht relaeateht sohohat~s eiami cute ei. aqeht sihelol cute. iqeht pali d-ah tota

= "I like my cat that is eating pasta. Cats are animals. It [my cat] is cute."


iq or nahaiq?

Posted On Tuesday 28 February 2012 by Murray Colpman (Muzer)

(Sorry the title is not in timlan, but I've realised we have no word for "or"!)

We've made a couple of changes recently that we haven't yet documented, so I've taken the liberty to write this post about one of them.

For a while, we've had an issue, that iq is too vague. You could use it for all sorts of things, and it didn't really have a clear defintion.

Now, however, that has changed!

iq now means "subject has the PROPERTY of the object". So:

iqeht nrib eiami cute

My cat is blue

But, what if you want to say "this food is pasta"? After all, this food doesn't have the property of pasta (it doesn't make much sense). What you want to say, is that this food has the type of pasta. So:

aqeht sohohat~s seluxeta main

= "This food is pasta(s)"

With the latter, you use determiners (or plurals) on both the subject and the object when referring to any specific instance of the subject. If you're referring to things in general, both subject and object should have no determiner. EG:

aqeht sihelol cute

= "Cats are animals" (in general)


Timlan Team account

Posted On Wednesday 22 February 2012 by Timlan Team (timlanami_ludef)

We might use it. We might not. We'll decide. So there.


Timlan Rolls Onward

Posted On Wednesday 22 February 2012 by Freddie Nash (FredFace)

Howdy, Folks! All is fine and dandy where we live and work on Timlan, so, here is a nice update on what we are doing ATM and all that Jazz:

Clarity Clarification Project
So far, we've clarified about 1/3 of our words (ignoring elements), and any new words we add are being added properly. (its really tedious work).

Many thanks to Wiktionary for providing generally viable definitions. The CCP has helped me to realise how dreadful the English dictionary is, its truley horrible.

The Great Verb Project (a subset of the CCP) is working out fine, our simple syntax seems to work. Don't forget that this lushness works alongside Timlan's perfectly inferable sentence structure, you don't need to know any words syntax to work out a sentence structure, though you may be unable to derive any meaning.

Say, that word reminds me of the word Dictionary! This is intentional, it is a dictionary of affixes, you guessed right (hopefully): we have FINALLY got around to making the afix list an online system, and soon will be modifying software to utilise this. Rejoice!

{{Insert Link To Fixionary Listing Page (Build Fixionary Listing Page)}}

Alongside this grand progression, we will be writing individual blog posts for each and every affix, with detailed information about them and extencive examples. This will take a while, as we want to make these posts both coherent && meaningful, which will be first for us (ish).

Having decided that Timlan is totally epic, we intend to add full chemical compound support (like with elements). This is currently !(quite working) but isn't far off (we want it to support structure, which may take some thinking).

Random Facts
Known Timlan Speakers: 2
Current Word Count: 381 words (+elements)
Our rough accessment of Lojban: tacky
Theoretically you can learn Timlan with just 3 documents:
- Dictionary
- Fixionary
- Fonti.txt

tokouhaht setuh sui


Affix changes again? YES, TIS TURE

Posted On Tuesday 7 February 2012 by Murray Colpman (Muzer)

Two more affix-related changes have occurred due to our making the fixionary and noticing silly things:

* -ah has been split off into two, -ah and -ame. Don't worry, we haven't done a U-turn on ordinal numbers; this is something else. Words that apply to whole clauses are split off into -ame, and take the old "as" long form (with the peve preposition marker) that the always did; -ah still has the rest with the "iq" long form. It all makes more sense this way.
* Modal verbs now have a long form! It's using mqdq. You don't make it a preposition as it isn't; it affects the meaning of the sentence. You just say: mqdqeht carun eatelem hat~s ei (I can eat pasta). This way, you can say mqdqeht ceveand carun musisen heekaht setuh ei (I can and must die).



Posted On Monday 6 February 2012 by Murray Colpman (Muzer)

Despite entitling the other blog post "Comparatives/Superlatives", we never actually touched on superlatives on the web (we did a bit on IRC I believe).

This is how you do them:

iqeht relacompareht ayiketa cute erometa kel eiami cute

(="My cat is more good compared to all cats")



peceheekeht setuh /"ota"/

Posted On Monday 6 February 2012 by Murray Colpman (Muzer)

We've decided that, since we've generalised -eta to mean determiners, we've also decided that -ota needs sorting out. And it's going to be sorted out by merging it with -ah. Here are the reasons:

* Like -ah, ordinal numbers are attributes of the noun (the first cat will forever remain the first cat, no matter which weird sentences he may be in that may or may not specify this)
* number-ah thing was previously undefined.

So, -ota is now permanently dead. FOREVER. All previous uses will now be served by -ah.



Posted On Tuesday 31 January 2012 by Freddie Nash (FredFace)

As you probably know, the only Timlan consonants not used in our number system are x and y. BUT, this has changed.

Now, x- means "times 0x10 to the power of".

For example:

is 0x120000

but you can now say




So, this works.

Some day y- may be put to good use, BUT THAT DAY IS NOT TODAY.


Timlan Word Count += 100 (and a bit)

Posted On Sunday 29 January 2012 by Freddie Nash (FredFace)

Not sure how many elements there are, BUT WE HAVE THEM NOW.

In Timlan, we like to make life easy, like, you know, making everything make sense, and not be silly.

Now, we recognise that Scientists like this also, so we are using the Standard Symbols for elements so that we don't have to come up with our own names, I mean, be consistent and all.

To make an element, stick 'eli' at the start, and another 'i' between any remaining chars in the symbol.


Hydrogen, H, elih
Silicon, Si, elisii (glotal stop, remember)
Potassium, K, elik

If you don't like the symbols, use the elements atomic numbers:

c-ota eli (1st element, Hydrogen)

That better not need explaining.

tokouhaht setuh sui

Reminder: DO NOT diphthongise 'iw' (in eliw (tungsten)) AND SUCH


SOPA/PIPA blackout.

Posted On Tuesday 17 January 2012 by Murray Colpman (Muzer)

You know that thing that Wikipedia is doing, to protest against SOPA/PIPA?

We are too, though I assume not many will care, it's still FUN :D

The only thing that will stay up is the machine-readable (plain) Timlan dictionary. Only the website will be affected; everything else hosted on the server will stay up (so IRC, mail, and some external sites, etc.).



Posted On Sunday 8 January 2012 by Murray Colpman (Muzer)


OK. Comparing things to other things. We've sort of had vague ideas on having this construct for years but none of us have actually sat down and worked it out...

We've added a new verb, compar, which is always used in a relative clause. We have also tightened definitions of the two mores (which are both now determiners):

erom = to a greater extent:

iqeht relacompareht rapot erometa kel cute

"Cat is more good compared to dog"
(Cats and dogs in general due to lack of determiner)

vintitia = additional; more in quantity:

eveeht relacompareht sui vintitiaeta cute ei

"I have additional/more cats compared to you"

So you can sort of see the syntax from that... ANY QUESTIONS NON-EXISTANT READERS????



Posted On Sunday 8 January 2012 by Murray Colpman (Muzer)

We have realised that determiners exist, and aren't the same as adjectives (which are nouns) :)

And we've also realised LEese ARe NoT In MaLtA Now.

And we've also realised that some of our existing words are determiners, and so are cardinal numbers.

So it stands to reason to use the cardinal postfix for all determiners. These are words like all (ayik), this (selux), etc. that are like adjectives that instead of describing attributes of a noun, describe information about a noun that might be valid only in that context.


seluxeta hat~s

= this pasta




Posted On Saturday 7 January 2012 by Freddie Nash (FredFace)

Its Change Vocab with respect to Time again, and todays topic is Mathematics.

To clarify, that is Mathematics, not Programming, so no distinct assignment and comparision operators.

This also comes with a lush syntax for handeling maths, which we've spent most of the last 2 days working on. Hopefully the time has paid off, in a !commercial sense.

Now some vocab... This first list of verbs will all always take the 'elem' (general) tense, as maths doesn't change of time like the Timlan vocab.

pad - Add
sub - Subtract
mul - Multiply
div - Divide
mod - Modulate
var - Variable
cost - Constant
ras - Raise to the Power
eq - equals

These words are used for making maths work in-line.

solute - Solve
eval - Evaluate

Now the fun parts that have NOTHING to do with vocab.

How to build expressions

Timlan having a VOS syntax, this is fairly evident.

In common SVO notation nonsense, you might write:

(4 + 2) / 2 = 3

In Timlan, using symbols this would be:

= 3 / 2 + 2 4


eqelem f- divelem d- padelem d- g-

Which is probably a bit hard to follow, if you don't know Timlan numbers (see

And to compliment this...

A new Marker!!


The # is pronounced as the Timlan 'm', for maths, or something else that starts with 'm', doesn't not matter.

Inside these Lush expression markers, symbols can be used inplace of lengthy verbs with postfixes (but verbs can be used also, and must be used properly). Here are the symbols we allow:

add +
sub -
mul *
div /
mod %

Expression markers also allow decimal numbers (0 to 9), which means that this is valid:

= 3 / 2 + 2 4

If it seems familiar, it should be (its the example above all this).

Note that the clear spacing is ESSENTIAL (especialy between numbers when they are written).

Do not say decimal numbers; this is all a bit hap-hazard ATM, so some kind of meaningful information regarding this may be released at some point in the future.

Following on...


Variables must be followed by $ (pronouned as the 'u' in 'put').

This applies to text inside and out of expression markers.

Nor much more to say about these...

Some Samples

tokosoluteaht /# = ^ 2 x -1 #/ sui
iqelem wuhilot padelem c- l-
evaleht divelem b- c- ei

tokouhaht setuh sui



Posted On Friday 6 January 2012 by Murray Colpman (Muzer)

Timlan alphabet!

We decided we needed an alphabet for (for instance) spelling out (non-Timlan) words, having single-letter variable names, etc.

So, we made one!

To this end, we have introduced another vowel. It is represented by '. The IPA is ɜ and it is pronounced like the "i" in British English "bird" (I'll add it to fonti.txt shortly). A few dialects of British English and most of American English that are "rhotic" (ie, actually pronounce the "r" in "car", "bird", etc.) must remember not to pronounce the "r" - if you're moving your tongue whilst saying it, or your tongue is in (nearly) the same position it would be to say "r", you're doing it wrong.

As for the order, as usual, we've been logical! Number letters come first, followed by the remaining consonants, then followed by the affricates (consonants with ~), then the normal vowels and finally the special vowels - / and '.

So, the full alphabet is written thus:

b' c' d' f' g' h' j' k' l' m' n' p' r' s' t' v' w' x' y' z' t~s' d~z' t~c' d~j' a e i o q u - / '

Note that when writing out letters, you don't write the ' unless you want to specifically emphasise that the letters should be pronounced individually - just like in English, you write "B" but you pronounce it "bee".

This will come in handy when we finalise the mathematics implementation...


Lush Goodness of the Future

Posted On Thursday 5 January 2012 by Freddie Nash (FredFace)

Here are some of the highlights of inprogress and upcoming Timlan projects:

- More Vocab o'Clock (including a planed mathsy one)
- The Timlan Webpages will be finalised/created/reverted for the sake of consitance and existance.
- The Verb Project, where we will be adding syntax definations to ALL Timlan Verb-oids (Conjunctives, Prepositions, etc.)
- The Clarity Clarification Project, which will involve added concise definitions for all Timlan words (will take place along side the Verb works, and a whole load of Dictionary capability improvments)

Because these last two projects will be fairly-long term, some currently functional Timlan orientated software may fail. If you find that something stops working, please inform us of this matter via comments, as there is the chance we won't notice and it will go unfixed for longer than it should.

We will probably make some sort of Project Page (other than the Tim32 one) so that you can see how far we are getting with changes, and see detailed descriptions of what is actually going on.




Posted On Thursday 5 January 2012 by Freddie Nash (FredFace)

Not sure why they are called markers, but hay.

Markers (/**/, /!!/, /""/) now have official definitions and pronunciations! Isn't that marvellous.

/!proper noun!/

The forward slash ('/') in each case is pronounced as the 'i' in 'in'.

The * is pronounced as the Timlan 'c'
The ! is pronounced as the Timlan 'p'
The " is pronounced as the Timlan 'k'

For example, saying /!France!/, one says '/p France p/' (where '/' is the i vowel thingy). The reasoning and logic behind this should be evident.


BURN (and other verbs)

Posted On Thursday 5 January 2012 by Murray Colpman (Muzer)

TRANSITIVE. All verbs that make sense to be transitive are transitive (and only transitive). Basically, if something can be used transitively in any other language and make sense, it is transitive (and only transitive) in Timlan. This isn't a concrete definition; we might decide to modify the dictionary to clarify on a word-by-word basis.


*kadameht setuh ei mean ("I burn") is INCORRECT. It really means "I burn an undefined object". Instead, to mean "I burn", you would write:

kadameht ei setuh

("Undefined subject burns me")

You can also, therefore, do sentences like:

kadameht sui ei

("I burn you").



Posted On Thursday 15 December 2011 by Freddie Nash (FredFace)


Yes, Vocab-o-Clock has DIED,and returned, but more fun!

Here is why:
- There is no longer a specific time on which the post will be made (and it doesn't have to be daily, how dramatic is that?)
- There are an arbitrary number of words defined

On with the vocab (not all nouns this time!)

Monitor - siati
Key - kej
Physical Array - dasik
Array [programming] - dasin
Keyboard [physical array of keys] - kejasik
Compute - komput [VERRRRRRRRRRRRRRB]
Null - nul
Compile - asim [ VERB}
ConCatEnate - miaw (V.b.*/
Processing unit - roce
CPU - rocec
GPU - roceg
FPU - rocef
PPU - rocep (phyics)
KPU - rocek (kookie processing unit; oven)
DSP - roced (digital signal processor)
Web (of worldwide) - lir
Internet - lira
Mouse - kilik




Timlan Signing

Posted On Thursday 15 December 2011 by Freddie Nash (FredFace)


We are now signing copies of our NEW BOOK!

Timlan for Toons

Its just £29.99 or €3 in the continent! (Sorry, we can't ship to the US or Australia)


Signing Timlan numbers just became possible!

Just prepend w- to the number:

c- = 1
w-c- = -1
v-v- = 255
w-v-v- = -255

We also decided that z- will be the decimal point (well, hexadecimal point really)

c-z-c- = 1.0625 (17/16)
v-v-z-l- = 255.5
w-v-v-z-l- = -255.5


Anything we write will of course be online, for future refrencness (;))


Brown Brownady Brown Brown!

Posted On Thursday 15 December 2011 by Freddie Nash (FredFace)

How does one say BROWN in Timlan!?!?1

Step Zero:
- Find RGB Value for BROWN of your Choice, here are a selection: (102, 51, 0), (61, 31, 0), (86, 57, 29)

Step One:
- Convert the RGB Dec values to Hex Representations: #663300, #3D1F00, #56391D

Step Two:
- Note that 'nri' is the word for 'Colour' in Timlan - that is important, so don't forget!

Step Three:
- Turn Hex Colours into Timlan! (This is the easy bit): j-j-f-f-b-b-, f-s-c-v-b-b-, h-j-f-m-c-s-

Step Four:
- Colours are ordinal! Duh!: j-j-f-f-b-b-ota nri, f-s-c-v-b-b-ota nri, h-j-f-m-c-s-ota nri

Step Five:
- Post a comment: IS THIS A GOOD IDEA?!?1?



IF you want a shade of Red/Green/Blue... we just HAPPEN to have words for these!


Find the byte index of your lovelyful shade (64) and make that Timlan number (64 == #40 == g-b-). Then... IT'S AN ORDINAL BLUE!

g-b-ota nrib




Posted On Sunday 4 December 2011 by Murray Colpman (Muzer)

We've sort of decided to make prepositions as verby as possible, albeit still keeping their identity in the form of the peve prefix. However, this is not 100% agreed yet, so take with a pinch of salt; it may quickly change.

So, gone is the differentiation between usage in clause-applying and noun-applying forms:

peveinafeht /!France!/ eateht hat~s ei

= "In France, I'm eating pasta"

But then, you may ask, what does changing the tense mean? Well, this:

peveinafoht /!France!/ eateht hat~s ei

...would mean something like "In a place that was France, I'm eating pasta"

Adding of tenses solves the problem that, if:

peveaseht ceveand torixi efiferogqj eatelem setuh ei

...means "I eat in an effective and complex manner", then how do you say "I eat in a manner that WAS effective and complex"? The answer is simply now changing the tense on the preposition:

peveasoht ceveand torixi efiferogqj eatelem setuh ei

This increased verb-like-ness also means that relative clause markers are used when appropriate. There are two main categories of its use:

* In the form when a preposition applies to a noun. With no exceptions.
* In the form when a preposition applies to a clause, when the clause is effectively acting as a relative clause that happens to have a preposition.

An example of the first is easy to come up with, and to see the merit: Not only can you say "I'm eating pasta which is in France" as before:

eateht relapeveinafeht /!France!/ hat~s ei

...but you can also say "I'm eating France, in which there is pasta" simply by changing one character:

eateht relopeveinafeht /!France!/ hat~s ei

An example of the second is as follows:

eateht relapeveinafeht /!France!/ relakadamoht ei rek ei

(note the double use of the relative there, to bring the human (rek) two layers out): "I am eating the person who burnt me, and the burning occurred in the place that is France"


ARTICLES! Or lack thereof!

Posted On Tuesday 29 November 2011 by Murray Colpman (Muzer)

The latter is the case!

This (selux) is used when referring to a specific instance of a noun. Lack thereof indicates things in general.

When referring to objects in general (eg "I like cats"), plurals are NOT used.

taropeht cute ei

Plurals are only used when saying (for instance) "[some] cats are running":

risiseht setuh sohocute


Double Vowels and Glotal Stops

Posted On Tuesday 22 November 2011 by Freddie Nash (FredFace)

You may, or may not, be familiar with the ^ glotal stop in Timlan... IT IS NO MORE.

As of now, we will be ridding the dictionary (which we will be adding outstanding Vocab-o-Clock words to at some point) of the carets, as they are taken out of the phonology, in that sense. Glotal stops will instead be an 'invisible' phone, used to separate  double vowels (eg. 'heek', 'kecaaht').

While written, double voewls will be just that, two of the same vowel next to each other, but when spoken, a glotal stop should be slung between them, making them distinct.

So, heek will no longer be a hesitant 'hek' (double vowels used to just be held longer, which has obvious issues) but he^ek, with the caret being used unofficially (don't use the caret ever again) to mark the Glotal Stop.

I hope that was clear, the DB should be purged of ^ soon, and one of us will post the changes made.



Posted On Wednesday 26 October 2011 by Freddie Nash (FredFace)

Sorry about the lack of posts the last two days, but now, it is time for Vegetables

Carrot - karoten
Potato - pataten
Parsnip - pasisnen
Pea - peaten
Bean - binen

Again... all nouns...

...and I'll try and re-write the Mega Input soon, such that it works...



Posted On Sunday 23 October 2011 by Freddie Nash (FredFace)

Todays topic is Pirates, and sailing in general.

Sailor = peviaset
Navy = pevilud
Pirate = pirate
Sail = kavas
(the white blowy in the wind things)
Map = rutel
Dock = pevioy
Sea/Ocean = sufeloy
(Ship = cipevi)

All nouns as ever



Posted On Saturday 22 October 2011 by Freddie Nash (FredFace)

Today's slightly late post is on Stationary

Pen - inekepi
Pencil - kapi
Eraser - nakapi
Ink - ineke
Paper - papie

All nouns, as ever...

And some time soon I'll find the time to actaully add some of these words to the DB...

The Mega Input doesn't work, and the code is really not very nice, and I am NOT adding each word one by one



Posted On Friday 21 October 2011 by Freddie Nash (FredFace)

Today's topic is Clothing

Shirt - celofel
Tie - kavatie
Hat - hedik
Trousers - tawes
Sock - sok
Jumper - wamefel
Vest - tinofel
Coat - kuta



Posted On Thursday 20 October 2011 by Freddie Nash (FredFace)

Vocab-o-Clock again, and this time, the topic is Transport!

transport - tanevek
vehicle - tavin
train - tarin
aeroplane - ahevin
helicopter - helivin
ship - cipevin
submarine - subevin
car - kavin
taxi - kabevin
bus - buvin
lorry - lqvin
bicycle - biwivin
chariot - qsevin
hovercraft - supevin

All nouns... again... (avoiding verbs as we may need to modify DB and ALL verb definitions some time in the future...)



Posted On Wednesday 19 October 2011 by Freddie Nash (FredFace)


Which means... WE MUST ADD EXTRA WORDS today to make up for it.


Job - dulet
Scientist - peragaset
Doctor - mediaset
Engineer - bobaset
Factory - makoy
Factory Worker - makoyaset
Shop - rital
Retailer - ritalaset
School - lenre
Teacher - lenreaset
Fireman - hefijaset
Police Officer - efqsaset
Soldier - defenaset
Business - magel
Business Person - magelaset
Chef - mainaset

All nouns... will be added to the DB shortley



Posted On Monday 17 October 2011 by Murray Colpman (Muzer)

New feature. Post vocab at 22:22 each day.

Today's topic is: Holidays

*** BEACH - ORAN ***
*** HOTEL - PACE ***
*** MUSEUM - YESER ***

All nouns.

Will be added to DB shortly.


Timlan: tenses, aspects, moods.

Posted On Monday 17 October 2011 by Murray Colpman (Muzer)

We've decided to change a few pretty major things:

* We have added aspects and moods. In Timlan they are sort of the same thing, but are visually differentiated by the prefixes being in a different form (they are both prefixes). An aspect and a mood can be applied to one verb by stacking them in any order.
* Questions and imperativeness are the "moods". coko remains the question "mood", and toko becomes the imperative mood (this replaces elet, the imperative 'tense'). It should be obvious when to use these (if not, look at past documentation).
* We have one aspect so far - the perfect. This is defined rigidly as the following:
the verb has happened prior to the time that the tense refers to, has changed a state, and that state remains changed at the time the tense refers to

It CANNOT be used for ANY other purpose.

For example:

pevemituh gahek pecerecoveht /!Ripley!/ setuh

Ripley is covered in blood.

(Luther referance (as is the verb ripeli (epic)))

We are considering the best way to define and add a progressive aspect, if required. Currently all tenses still exist apart from imperative, but if a progressive aspect is created, it will replace the generic tense (elem).


New Number System

Posted On Monday 3 October 2011 by Freddie Nash (FredFace)

I say NEW, we didn't really have one before...

BUT, in a moment on simplicity, we have devised a nice number system. It works like this:

Each digit (0-F (Hexadecimal)) is allocated a consonant or fricative (B through V), and numbers are spoken per-digit as this phone with the 'a' from 'father' vowel following it.

You represent this vowel in written Timlan as '-'. When you see '-', you must think of it as a vowel, rather than a marker or anything.

The digit/phones are as follows:

b - 0
c - 1
d - 2
f - 3
g - 4
h - 5
j - 6
k - 7
l - 8
m - 9
n - A
p - B
r - C
s - D
t - E
v - F

This means that the number 125 (7D in hex) is: k-s-, pronounced something like "karsar" (Southern non-rhotic British English)

As mentioned above, the written notation k-s- is used, making the relevant phone clear, and indicating it is a number directly. Any place a dash is, the 'a' sound will be. It follows the final digit (ie. s-c-d-eta)

That is about all there is...

Changes to the BD and VOS Parsers... I have to update the parser a bit to actually support numbers. The digits in the DB will be removed, and possibly replaced by a phonetics system, based on the new phones (so that numbers can be stated extra clearly (ie. saying phone numbers down a bad line)).



NEW CONCEPT - multiple objects

Posted On Wednesday 28 September 2011 by Murray Colpman (Muzer)

We have finally come up with a solution for words with multiple objects. These include English words such as "give" (I give you the cat), "grant" (I grant you a wish), "introduce" (I introduce the cat to you), and "tell" (I tell you the requirements). There are even ones with three objects, such as "trade" (I trade my cat and your pasta with you).

In Timlan, the concept we've come up with is that you can have a "stack" of objects. As with any good processor architecture, it's the least significant object you see first ;)

To push an object onto the stack, you use the "puc" semi-verboid*. This semi-verboid is special in that it takes no prefixes, and that the subject must be another clause (possibly another puc clause).

For purposes of backwards-compatibility, and because we see no reason to change it, in any instance in which using "to" as a preposition would make equal sense to using two objects, either form can be used.

Now, for some examples!

I give him the pasta (multiple object form):

puc ti pieselem hat~s ei

As with all languages I know of, the order of importance is essentially decided by how much the objects are directly affected - the object that is most directly affected by the verb (in this case the pasta, because it is physically being given to the person, who is not having an action applied to him) becomes the main object, and the person becomes the secondary object.

I trade my apple and her orange with her: (we don't yet have a word for "trade")

puc ti puc tiami ohk /!trade!/eht eiami tok ei

We will soon start to redo the database, describing accurately how many objects each verb takes and what exactly each one means.

(Please note that this is a serious job, and we are all VERY busy, so don't expect much progress on this front any time soon)

*we've also come up with some new terminology.
verboid = any word that can be used as a verb
nounoid = any word that can be used as a noun
semi-verboid = any word that can be used in some situations as a verb
semi-nounoid = any word that can be used in some situations as a noun


Minor change to use of prepositions

Posted On Wednesday 28 September 2011 by Murray Colpman (Muzer)

We have decided that to ease the differentiation between prepositions and verbs, they should always have the preposition prefix peve-. This is in addition to existing rules, not instead of, so you still use tense suffixes in the short form and no tense suffixes in the long form.


All sorted

Posted On Saturday 13 August 2011 by Murray Colpman (Muzer)

All issues involving Freddie's change of router should now be sorted. If you find a problem, let us know via comments.



Posted On Saturday 6 August 2011 by Freddie Nash (FredFace)

Hello All,

The recent downtime (from about 18:00 UK GMT till 00:00) was due to the changing of the Tim32 router. This requried a cople of reboots, and a fair bit of filling out data as well as alot of general waiting.

Unfortunatly, this new router doesn't seem to support Dynamic DNS, don't ask us why. This means that the domain '' no longer works, and we have to manually adjust '', which takes time to sort itself out.

Since at this time hasn't yet been updated (I'm assuming no one needs the service now), I can't say whether or not subdomains will be working in the near future... I'd assume they won't without further adjustments, but we shall see.



Timlan Demo

Posted On Friday 22 July 2011 by Freddie Nash (FredFace)

Here are some demo's I wrote, for no real reason

adiliadelet setuh sui
(V O S)

taropeht /!Industrial!/ah /!Lighting!/ ei
(V Os O S)

bobelet heekaht setuh sohoselug sui
(V (V O S) S)

iqehteht miselak heekaht setuh sui
(V O (S O S)

peveinaf /!France!/ eateht cevemituh felolez sohorisic ei
(P O (V (C O S) S))

peveinaf eiami /!Office!/ linekeht setuh haiseht ceveand /!Holly!/ /!Rhowan!/ zeieta rapot
(P Os O (V O (V (C O S) S)))


Site Change

Posted On Saturday 11 June 2011 by Freddie Nash (FredFace)

In order to step back a few version numbers, Timlan is now in verson 0.9! Hurray!

This is part of the move to the new site location, and site redecoration, to make it seem less grim, and more inviting, as well as being more useful, hopefully.

The new site link


Ordinals, and some clarifications to relative clauses

Posted On Thursday 9 June 2011 by Murray Colpman (Muzer)

Right, first of all, we needed ordinals (first, second, third, etc.). These are made in exactly the same way as cardinals (normal numbers), but with -ota instead of -eta:

iqeht einota cute ti

= it is the first cat.

If you want to say "the cat is first", you remove the suffix:

iqeht ein cute

And now, the clarification to relative clauses.

The general rule is that relative clauses add information or clarification, rather than changing the whole meaning. Therefore, an adjective/ordinal applied to the subject of a relative clause applies only to the subject, not to the relative clause as a whole. For example:

eateht hat~s relaiqeht nrib einota cute

= The first cat, who happens to be blue, is eating pasta. NOT "The first blue-coloured cat is eating pasta".

If you want to say the latter, you must use the long form:

eateht hat~s relaiqeht ein relaiqeht nrib cute

= "The first blue-coloured cat is eating pasta"

FWIW, you can also use the long form to produce the other meaning:

eateht hat~s relaiqeht nrib relaiqeht ein cute

= "The first cat, who happens to be blue, is eating pasta."


A clarification and a minor change and a major change

Posted On Wednesday 8 June 2011 by Murray Colpman (Muzer)

A clarification about modals: The one way in which they do not act like adverbs is that they can only be applied to things that act as verbs - so, verbs, prepositions and conjuctions.

And a minor change - because of this distinction, we've decided to make modals use a different postfix to adverbs to more easily distinguish them. So, the new modal postfix is -amo, rather than -ah.

And a major change - they can no longer be used as nouns, eg in an "as" clause, so this is no longer valid:

*peveas ceveand musisen carun eataht setuh ei

Instead you must say:

ceveand musisenamo eataht setuh ei carunamo eataht setuh ei

This is because "-ah" for a verb is just the short form of an "as" construct, and "-ah" for a noun or another "-ah" word is just the short for of an "iq" construct. "-amo" isn't short for any construct, we've decided, and wouldn't make much sense to lump it with adverbs.


Modal verbs and adverbs

Posted On Wednesday 8 June 2011 by Murray Colpman (Muzer)

We've been thinking about modal verbs, and also about how to use multiple adverbs, and we've come up with a solution we all like:

First of all, I'll talk about a new construct for adverbs. As everyone knows, the standard "short form" construct is:

efiferogqyah eatelem setuh ei

= I eat effectively

As you saw from the previous post, these always apply to the term immediately to the right, so:

torixyah efiferogqyah eatelem setuh ei

= "I eat in a way which is complexly effective."

If you want to say "I eat in a way which is effective and complex", you can now use the "long form" construct, which is:

peveas ceveand torixy efiferogqy eatelem setuh ei

= "I eat in a complex and effective way" - we translate it as "I eat as complex and effective".

So, this brings us onto modals. These work EXACTLY like adverbs:

musisenah eataht setuh ei

= "I must eat"

This means you can stack them like adverbs:

musisenah carunah eataht setuh ei

= "I must be able to eat" ("I must can eat")

And you can use the new long form like adverbs:

peveas ceveand musisen carun eataht setuh ei

= "I can and must eat".

Now here's where it gets complicated, but still makes sense - using them WITH adverbs.

If you want to say "I quickly must eat" - that is, I must eat in a short time from now, but it doesn't matter how I eat - you can say:

rapidah musisenah eataht setuh ei

If you want to say "I must quickly eat" - in other words, "I must eat in a quick way, but it doesn't matter when I eat" - you can say:

musisenah peveas rapid eataht setuh ei

And if you want to say "I quickly must quickly eat" - in other words, "I must eat in a quick way in a short time from now" - you can say:

musisenah peveas rapid musisenah eataht setuh ei

I hope it's all clear!



Posted On Wednesday 8 June 2011 by Murray Colpman (Muzer)

We've just come up with an immensely easy way of doing gerunds - that is, using a verb as a noun to mean "the act of doing that verb". Basically, you just use the verb as if it were a noun, ie without a tense:

iqeht nehe yelok

= "Walking is fun"

taropeht risis ei

= "I like running"

Much more sensible than English!


Descriptive words

Posted On Wednesday 8 June 2011 by Murray Colpman (Muzer)

Just to clarify - describing words only apply to the word immediately following them:

taropeht kempotah nribah cute ei

= "I like the awesome-blue cat" - you're saying that the BLUE is awesome, not necessarily the cat.

If you wanted to say "I like the awesome, blue cat", you would have to say:

taropeht relaiqeht ceveand kempot nrib cute ei

= "I like the cat that is blue and awesome".

I don't particularly like it, but it's the only way we could think of without producing unnecessary limitations. For example, with this method, you can say:

taropeht relaiqeht relaiqeht ceveand kempot value nrib cute ei

= "I like the cat that is blue that is shiny and awesome"

or even:

taropeht relaiqeht ceveand relaiqeht ceveand kempot value nrib nrig cute ei

= "I like the cat that is green, and blue that is shiny and awesome"



Posted On Wednesday 8 June 2011 by Murray Colpman (Muzer)

We decided we needed vocatives in Timlan, last night, and after a heated debate, we decided we will have two vocative markers. Like most other things in Timlan, these act as verbs.


This roughly translates as "hailing" - it is used at the beginning of a conversation, or rarely, during one, when you need to attract somebody's attention:

voceht /!FredFace!/ dieeht sui ei

"Hailing FredFace: I greet you."


This roughly translates as "regarding" - it is used during a conversation with multiple people, when you already have the attention of the other person, but want to directly address a comment or question to them:

vokeht /!FredFace!/ iqeht wuhihyorax seluxah hyorax

"Regarding FredFace, what's the time?"

In other news, we may attempt to start translating Star Trek episodes.


Relative Clauses

Posted On Wednesday 8 June 2011 by Freddie Nash (FredFace)

Those things which add more information that can't be considered as valuable to the bulk of the sentence.

This morning we added-relato the army of pre-fixes. Nice and easy to remember, it indicates that a verb is starting a relative block/clause. For Example:

eataht relaiqeht ceveand nrib nrig hat~s ei

Roughy translates as 'I am going to eat pasta that is blue and green'

When you start a new relative section of a sentence, that is like a new sentence all together - inside that the same rule applies, such as in this rather senseless example:

eataht relaiqeht relaiqeht nrib eineta nri hat~s ei

Meaning I will eat pasta that is  a colour, that is blue.

This may lead you to the logical conclusion that the first verb in any sentence should follow suit - as long as you know that it doesn't and don't think about it too hard you should be OK (its just a inferable and clear with and without).

The Timlan team



Posted On Tuesday 7 June 2011 by Freddie Nash (FredFace)

Snigger snigger (like drowning? no? OK)

We are having a great night! Lots of good changes to Timlan!

We have changed all the nasty post and pre fixes, so that they conform to fonti (always).

The new list is as follows.



The new conjunction and prepostiion pre-fixes conserve clarity of syntax without knowledge of vocab. If a preposition happens to have a tense denoting postfix, this represents a short hand of the verbiq(to be), eg:

peveinaf /!France!/ iqeht setuh ei
P        O          V     O     S


peveinafeht /!France!/ ei
P (V)       O          S

The introduction of these new prefixes make the usage ofelemfor convaying prepositions and conjunctions redundant, and flawed, so don't use it unless as explained a bit above. elemnow does just indicate the generic tense.


The Timlan team



Posted On Tuesday 7 June 2011 by Murray Colpman (Muzer)

In a similar, but certainly not the same, topic to the previous post, I'll now talk about conjunctions. This has had much less thought go into it, but it's quite a simple topic. Just remember one thing:


Right - conjunctions are most used to connect two clauses together, or two nouns together. Although the meanings in each case are not completely identical, they are certainly very closely related, so unlike some languages (eg Na'vi), Timlan uses the same words to connect clauses as it does to connect nouns.

In fact, the syntax is pretty much the same between both:

andelem eateht setuh ei yelokeht setuh ei

= I am walking and I am eating.

eateht andelem hat~s felolez ei

= I am eating pasta and cheese

So, as you can see, conjuctions are ALWAYS used in the generic tense. NO EXCEPTIONS.

EDIT: I missed this:

Some languages allow you to use conjunctions to join verbs if the subjects are the same:

You will suffer and die.

This is not possible in Timlan, at least for the moment, so you must say:

and katamaht setuh sui heekaht sehtuh sui

...which is a bit verbose, but it's curently the only way.


VOS prepositions!

Posted On Tuesday 7 June 2011 by Murray Colpman (Muzer)

Freddie and I have finally sat down and worked out prepositions FOR GOOD!

BASICALLY, IF the preposition applies to the clause as a whole, and CANNOT be attributed directly to one of the nouns, it goes outside the clause, and is given the GENERIC tense:

inafelem /!France!/ titanelem /!Person!/ pieseht sohkusi ei

= (((( I  give [present]  cookie [plural] )  to [generic]  Person )  in [generic]  France ) )

As you can see, this applies to indirect objects (which are created in VOS using prepositions), and general locations. This particular sentence means that a cookie is being given to the person, and the giving is taking place in France.

If the preposition only applies to one of the nouns, however, you essentially turn that noun into its own clause, and you give it a tense:

titanelem inafeht /!France!/ /!Person!/ pieseht sohkusi ei

= ((( I  give [present]  cookie [plural] )  to [generic] ( Person  in [present]  France ) ) )

In this case, the PERSON is CURRENTLY in France, but the giving does not necessarily take place in France. For instance, the cookie could have been posted from Britain to the person in France.

This basically means you can go completely crazy with much more detail than is practical in English, like this:

inafelem /!Belgium!/ titanelem inafeht /!France!/ /!Person!/ pieseht inafoht /!Germany!/ sohkusi ei

= (((( I  give [present] ( cookie [plural]  in [past]  Germany ) )  to [generic] ( Person  in [present]  France ) )  in [generic]  Belgium ) )

In this one, the cookie was in Germany, the person is in France, and the giving is done in Belgium. This is quite hard to make sense of, since it describes a situation which wouldn't really happen in real life, but, for example, the cookie could have been bought in Germany, then moved to Belgium and posted to the person in France.

/me cheers, for completing this blog post

EDIT: That above line should have been posted to IRC, but I'll keep it because it correctly describes my attitude :p

EDIT2: So, we've been thinking about this, and have come to the main conclusion of this post:

* If the preposition can be applied equally well (or poorly as the case may be) to both the subject and the object, it should be nested outside the main clause and given a generic tense
* If the preposition cannot sensibly be given a tense, it should be nested outside the main clause and given a generic tense
* If neither of the above is true, it should be nested inside the clause along with the noun it applies to, in the tense that makes most sense.

I cannot think of any occasions in which the first bullet point would be true but the second wouldn't, or vice versa, and I expect linguistically wise there aren't any, so I believe we now have a very solid basis for this.

-- The Timlan team.


More Prepositions

Posted On Sunday 5 June 2011 by Freddie Nash (FredFace)

Snigger snigger...

Prepositions are Verbs again! Exclusivly!

You don't be somewhere, (you be -) somewhere!

inafelem /!France!/ iqeht setuh sui
V (P)    O          V     O     S

Stay tuned for more changes, they are sure to be the same as the ones before, for different reasons.

Note that some parsers have-ent, some don't, just wait till we make a dicision on this one!

Timlan team



Posted On Sunday 5 June 2011 by Freddie Nash (FredFace)

In case you are for some reason keeping an eye on this blog, updates are far and few at the moment becuase the Tim32 members are caught up in Exams and such, the likes of which will be over soon.

Hopefully once we have more time, we'll be able to provide a greater presents, and do something interesting!



Posted On Monday 9 May 2011 by Freddie Nash (FredFace)

After getting confused, I've decided to post this, as a reminder to me as to anyone else who happened to psychic.

When stating your location, YOU USE THE VERB 'iq'!!

For example:

iqeht inafent /!Italy!/ ei
V     OS (P)  O         S

If you need to state something happens somewhere, use prepositions as verbs, like before.

inafelem /!France!/ eateht hat~s ei
V (P)    O          V      O     S

I eat pasta in France
S V   O     P  O

Proably destined to add a special post-fix for verb based Prepositions, Something similar to -ent, no doubt.


LFP Examiner

Posted On Friday 22 April 2011 by Freddie Nash (FredFace)

Hello! After getting confused about SOMETHING (can't remember what) this morning, a friendly text led me to ammend the LFP VOS Examiner for Timlan 1.1 - hurray!

Usual place, here


Grammar Stuff

Posted On Saturday 16 April 2011 by Freddie Nash (FredFace)

Been revising some grammar and such, here are some revisions.

!WARNING! These changes will (officially) be part of 1.1, but we suggest you learn them instead of 1.0, as they are significant. Online resources currently don't support the new '-ent' (read on) postfix.

Prepositions now have a proper postfix, so arn't really verbs anymore. Enjoy this example:

iqeht inafent eimai lapopis ei
V     P (OS)  OS    O       S

In English:

I am in my house
S V  P  OS O

This means Prepositions are form of supliment.

Modal Verbs
These seem to make sense... when in doubt, try jamming sel in as a subject (make it reflexive), eg:

musiseneht eataht hat~s sel ei
V          V      O     S   S

Putting ei in again would be weird...

Modal verbs may end up as Adverbs.

Remember kim? Well, you probably won't need to soon. It would apppear that it acts as an overcomplicated work around to the main issues resolved by the resolves above. Expect a post soon declaring its removale.

I'm VERY happy about this (though, it means ripping quite a bit of code out of parser, which takes time), no doubt anyone who thought they uderstood it (I didn't) is too.


GIMP Single-Windowed?

Posted On Thursday 14 April 2011 by Tom Leese (TomMan)

Although GIMP 2.8 is not out yet, its Single-Window mode is available for testing!

I have been using GIMP 2.7 for some time now and I much prefer the GIMP Single-Window mode.

BBCode-included image

It is still a little bit buggy in some aspects, for example, if you open a new dockable window it does not appear within the single-window mode, but I cannot wait till GIMP 2.8 is released!

You can follow the development of GIMP here: GIMP Progress Tracking


For(;;) a While(true)

Posted On Saturday 26 February 2011 by Freddie Nash (FredFace)

Incase someone is wondering, for(;;) and while(true) compile identically in CPP and C#. C# is my forte, so I'll demonstrate the intermediate code produced (the same) for each. Consider the following C# Code:

for (;;)
    if (threadBreaker)
while (true)
    if (threadBreaker)

Clearly, it represents two loops which will be broken out of when a variable (bool threadBreaker) is set to true (this is just give something to look at).

After moving each loop to a separate function and compiling (without optimizatins) the IL for the functions is identical, and thus:

  .maxstack  2
  .locals init ([0] bool CS$4$0000)
  IL_0000:  nop
  IL_0001:  br.s       IL_0013
  IL_0003:  nop
  IL_0004:  ldsfld     bool ConsoleTestGround.Program::threadBreaker
  IL_0009:  ldc.i4.0
  IL_000a:  ceq
  IL_000c:  stloc.0
  IL_000d:  ldloc.0
  IL_000e:  brtrue.s   IL_0012
  IL_0010:  br.s       IL_0017
  IL_0012:  nop
  IL_0013:  ldc.i4.1
  IL_0014:  stloc.0
  IL_0015:  br.s       IL_0003
  IL_0017:  ret

A quick skim will show up how very inefficiant this is - the need for a stored local variable alone is uncalled for, but it does show that the said variable is assigned the value true (IL_0013 onwards), but it is never compared to anything.

With oprtimizations, the code size is reduced dramatically, but oddly the maxstack is quadrupled in size. Both functions are again, identical:

  .maxstack  8
  IL_0000:  ldsfld     bool ConsoleTestGround.Program::threadBreaker
  IL_0005:  brfalse.s  IL_0000
  IL_0007:  ret

Well, nothing more to say...



Battle of the Programming Languages

Posted On Saturday 26 February 2011 by Tom Leese (TomMan)

I recently put together a series of tests to find out which is faster between: Python, Lua, and PHP.

The tests consisted of:

Here are the results:

python3  11.29s
lua      8.46s
php      15.01s

As you can see, Lua is the fastest language, python is the next and PHP is the slowest.

I did all these tests on an Intel Atom processor under Arch Linux. With all the tests I had: Chromimum, Kate and XChat running (and KDE).

Python Code · PHP Code · Lua Code


New RSS Feed!

Posted On Sunday 9 January 2011 by Tom Leese (TomMan)

Just a quick post to say that the Tim32 Blog now has an RSS Feed!!! You can see it here: RSS Feed

(Edit: Photos now also have RSS feeds.)



Posted On Saturday 8 January 2011 by Tom Leese (TomMan)

Here at Tim32, FredFace and I have started creating some fun comics including some of the members of Tim32. You can find them here:

We hope you enjoy them! :)


Tim32 Website Update!

Posted On Saturday 8 January 2011 by Tim32 Admin (admin)

Welcome to the new updated Tim32 website! It has had a complete design re-think to make it simpler and easier to use! I am currently in the process of moving the old posts to this new blog.


VOS -> Timlan

Posted On Thursday 9 December 2010 by Freddie Nash (FredFace)

Yes, we've finally given it a name (mainly because it turs out VOS and SVO etc. are infact proper terms).

Timlan (Tim32 + Language) shuold be pronounced as in Fonti, thus: Team-Lan

Most things will remain a 'VOS *' and what-not, its just that if we ever end up talking to anyone who does this sort of thing for a living, we don't want to end up confused and looked down apon.

Oh, just to say - all the docs are out of date... I've started re-writing Comprehensive VOS for VOS 1.0, it's in the VOS 1.0 docs dir.
Enjoy. Lots spelling/grammar errors fixed in the content also. Unclarified sections have been removed for the time being.



Posted On Thursday 21 October 2010 by Freddie Nash (FredFace)

Yes, it does exist.

Planar is a fun platformer, where you run about and are vapourised regularly inorder to reach the next level.

Its heavily under development at the moment and the game can be downloaded here:

It uses some slightly hacky Line Collsion detection I wrote almost a year ago, and currently adopts Glut and OpenGL for the rendering. There is a GDI+ rendering 'option', but this sequence is no longer updated.

Planar uses Tao and FreeGlut, where before it employed the 'SimpleGLControl'. This change happened only today and so it all looks a bit haphazard (2 forms, one accepts input, one render).

If you like moding games, you might as well mod Planar, is could hardly be easier. Here are some reasons why:
- Nothing is compressed
- 'Animations' comprise of string referenced PNGs
- It has a nice set of commands (the docs for which I can't be bothered to update, just read the code if you need to find something)
- Runtime console helps debugging (levels etc.) and can result in excessive amounts of fun
- As implied, I provide the source
- I'll probably write a tutoruial for moding at some point

Have fun.


VOS 1.0 is here!

Posted On Sunday 10 October 2010 by Freddie Nash (FredFace)

And it is more shiney than ever!

Oh, yeah - no blending of post/pre fixes and root words, it just spoils all the fun.

So you know, as long it supports ~, ^, ., and , (heh) VOS can be written on any keyboard, capitals are not requried to be distinct.


fonti - New VOS Phonology

Posted On Thursday 7 October 2010 by Freddie Nash (FredFace)

VOS has a new Phonology!

Its not scary, don't worry, just stricter and more flexible than the old one.

Now, Blending is ILLEGAL! NO BLENDING!

Why? Because blending reduces clarity, and thus ruins everything that VOS is designed for.

Other than that, to stop people blending characters, and make words nicer to say, we are no longer allowing any non-Fricative, non-Nasal constonants to be next to each other.

We have also added new charcters which are blends, but only those which are usful and have an IPA representation. The most notable of these is 't~s'. The format is basiclly a tilda joining 2 chars which when blending make the blended noise. I this case, the 'ts' in cats, hats, bats. This is so that we can have the word 'hats' (pasta), but it will no doubt come in handy in the future.

Unfortuanalty (as we hope you will have worked out) the new rules make some words in VOS ilegal. 'seth', for example. The online DB is currently down, so some services may fail... Muzer is hard at work scripting a system to automatically verify each VOS word, and make ammends if needed.

The upcoming release of the 1.0 VWV has the ability to check the spelling of words now, identifying invalid words by their spelling.

We are hoping to have fonti fully introduced, and the online Dictionary repaired by Monday, when we will almost certainly be declaring VOS functional, and starting work on the 1.1 special vocab aditions.


As ever, the Phonolgy can be found here under the new name of 'fonti'.

We felt a new name for this shiney phonology was neccessary to make it more appealing, and seemingly more 'portable' to other langs that may which to adopt it. If you do (for some good reason, I've given you one) then feel free.



Posted On Wednesday 1 September 2010 by Freddie Nash (FredFace)

Hello! Its another pointless test of the .NET Framework!

I ran a quick test just now, to see which is faster (without csc optimizations):

I won't go into detail, and shall just give you my results.

The times in Milliseconds are:

string.Concat(): 44
string.Format(): 186
a + b:           25

No suprises there, as String.Format does a lot more then the other 2 methods, but this does show how much slower it is. I never use it now :(


VOS Overview

Posted On Monday 16 August 2010 by Freddie Nash (FredFace)

Hello everyone,

This is a joint post by Murray and me as an overview of VOS.

For starters, we think VOS is fun. The most recent version is 0.9, which can be found here: VOS 0.9 Docs. Note that VOS is infact an acronym for Verb Object Subject - the basic word order that defines the structure of any sentence written in VOS on any level.

VOS has been shown to work. Perfectly. In every case we have considered (we've spent alot of time when we should be sleeping verifying the grammar, it seems solid (thanks to Sh4rk for intense unbias scrutiny)). English, on the other hand, makes no sense, and should be illegal (thankfully it isn't yet else we souldn't be able to spread VOS effectivly).

The ambiguity in English (and many other langauges) is terrifying, and please note the differnce between 'ambiguity' and 'abstraction' - it is a BIG and IMPORTANT differnce.

This is ambiguous:

> I'm going to the bank

Bank could be a river bank, and bank (stash) of something, or a Bank (eg. the building). If we assume this means the store of something (eg. a Box Bank) then this is NOT ambiguous:

> I put the item in the bank

It is abstract, because bank encompasses selection of similar items.

Oen of the worst crime commited by the English language is the torment of having different endings for different words which have the same effect. For example, with plurals:

Cats - Pretty much standard, OK
Sheep - Ambiguous - Not OK
Mice - Complete restructure - Not OK

In VOS, plurals are indicated in one way only, with the prefix 'soh':


This is the same in Hungarian, though this is merely coincidence.

The most important thing to learn about VOS is LFP (Lisp For People). We recommened reading through Expantion.txt in the VOS 0.9 docs if you are strugling to figure this out - it may seem weird at first, but it is infact perfect.

The LFP Examiner will help you check your VOS skills, and acts as a translator. LFP could technically be turned to make SVO (almost making sense English word order) work with some minor additions.

If you think that having pre-set prefixes etc. makes VOS open to error, you are right! We just verify everything using the VWV - its quick, easy, and is the only way to make the language work properly without any performance hit.

We hope you atleast consider the chance to learn VOS, or atleast the grammar structure - we've learned alot from making it, and it is really easy to learn - probably because it is so simple.

A VOS 'dictionary' can be found here: VOSLookup

This is the only place where any record of current VOS words is maintained, and any software we provide which appears to contain a long list of words is probably just downloading this.

Some things (like Boolean operators) can be found in 'Vocab.txt' in the docs - anything in there is either special, or being considered for value - chances are it will be kept, unless we comment that it may die (be removed).

There is a good chance that we will create a Language category, just for people to discuss the horror that is modern English, and the poor usage of it.

iqelet hilf sui


VOS 0.8

Posted On Sunday 8 August 2010 by Freddie Nash (FredFace)


VOS 0.8 is out, with a new Parser (this one actully handels LFP) and VOS Phonology.

Muzer has made the Phonology as easy for English people to learn as possible while retaining some degree of sense (English phonology is a complete mess, I'm just very confused about everything though).

The only particularly obscure characters are:

j, i, x, u, q, c

You can find it in the usual place:

Evil Lying VOS Place

Have fun!


What is the best way of making games in C++ in Open Source toolkits?

Posted On Thursday 5 August 2010 by Tom Leese (TomMan)

In C++ there are many ways in which you could create games. Here are your options:

There are probably some more that I've missed but these are the toolkits I am most familiar with.

Firstly, we have SDL. SDL is designed to be used for making games from the beginning. SDL is very fast and very lightweight. Unfortunately, it is low-level so it can be a hard to start programming with SDL. In general, SDL is an excellent toolkit for writing games.

Next, OpenGL. This is the best option for people who want awesome looking and extremely fast 2D games, or excellent 3D Games. OpenGL can't be used on it's own, it has to be used with a library for managing windows, etc. From what I know you can use OpenGL with Qt, SDL, Windows Forms, GTK, and GLUT.

Next, Qt! Now, Qt is by far my favorite toolkit of them all. It is fast, easy to use and has amazing documentation. Unfortunately though, Qt is not the best toolkit for making games. It has a lot of useless technologies behind it when you write games in it which can make the games slow compared to SDL or OpenGL.

Finallly, GTK. I have not really used GTK must, but I used to use PyGTK loads because I prefered it to PyQt. From what I can see GTK is not really a very good toolkit for making games, but it is an excellent toolkit for making applications.

In conclusion, I have decided that if I were writing a game I would choose SDL + OpenGL because it is fastest and although it requires some learning it is an excellent solution for creating games.

Further Reading


Qt and its Awesomeness!

Posted On Thursday 29 July 2010 by Tom Leese (TomMan)

I would just like to say a few things about Qt and its awesomeness :p

Ok, so everything in Qt is based around the concept of events. This means that everything from the Window Manager wanting the window to be redrawn, to a mouse button being clicked, to the window being resized is done in events.

Here is some example code of drawing a circle in the middle of the window (200x200) when the window manager wants it.

void MainWindow::paintEvent(QPaintEvent *event)
    QPainter painter(this);
    painter.setRenderHint(QPainter::Antialiasing, true);
    QPen pen(Qt::blue, 2);
    QBrush brush(Qt::green);
    painter.drawEllipse(20, 20, 160, 160);