Back to Home / #openttd / 2009 / 07 / Prev Day | Next Day
#openttd IRC Logs for 2009-07-24

---Logopened Fri Jul 24 00:00:47 2009
00:08-!-kkb110 [] has joined #openttd
00:35-!-Pygma [~quassel@] has quit [Remote host closed the connection]
00:52-!-el[cube] [] has joined #openttd
00:55-!-eleusis [] has quit [Ping timeout: 480 seconds]
01:08-!-FloSoft [] has quit [Remote host closed the connection]
01:11-!-FloSoft [] has joined #openttd
01:40-!-Splex [~splex@] has quit [Ping timeout: 480 seconds]
01:48-!-fonsinchen [] has joined #openttd
02:01-!-kkb110 [] has quit [Remote host closed the connection]
02:06-!-stuffcorpse [~rick@] has quit [Remote host closed the connection]
02:06-!-stuffcorpse [~rick@] has joined #openttd
02:12-!-Zorn [] has joined #openttd
02:12-!-ecke [~ecke@] has quit [Quit: ecke]
02:13-!-fonsinchen [] has quit [Remote host closed the connection]
02:13-!-reldred [] has joined #openttd
02:26-!-reldred [] has quit [Quit: Leaving.]
02:36-!-LadyHawk [] has quit [Quit: Never underestimate the power of stupidity.]
02:45-!-tux_mark_5 [] has joined #openttd
02:59-!-Splex [~splex@] has joined #openttd
03:05-!-HerzogDeXtEr [~Flex@] has quit [Read error: Connection reset by peer]
03:05-!-HerzogDeXtEr [~Flex@] has joined #openttd
03:24<TrueBrain>Xaroth: a clone, by Dutch law, is possible, just not from reversed engineered code ;)
03:24<TrueBrain>but of course there is always the Clean Room method ;)
03:26-!-Polygon [] has joined #openttd
03:38<CIA-1>OpenTTD: rubidium * r16934 /trunk/src/ (6 files): -Codechange: introduce a simple helper function to check whether a station is pending deletion or not
03:52-!-Yrol [] has joined #openttd
04:06-!-Alberth [] has joined #openttd
04:08-!-Polygon [] has quit [Quit: Verlassend]
04:25-!-fjb_ is now known as fjb
04:29-!-reldred [~reldred@] has joined #openttd
04:50-!-Azrael- [] has joined #openttd
04:51-!-OwenS [] has joined #openttd
05:05-!-Yrol [] has quit [Quit: MOOOOOOOOOO?????]
05:14<CIA-1>OpenTTD: alberth * r16935 /trunk/src/strings.cpp: -Codechange: Improve wording of .lng file error message (found by cmoiromain).
05:19-!-ecke [~ecke@] has joined #openttd
05:23<planetmaker>@define setdef
05:23<planetmaker>meh.. wrong channel :-P
05:23<TrueBrain>lucky for you there is no glx :p
05:24<planetmaker>hm... I guess no one tried those commands before :-)
05:24-!-DephNet[Paul] [] has joined #openttd
05:27-!-DephNet[Paul] [] has quit []
05:27<planetmaker>hm... Rubidium maybe it's getting time to change the pf.wait... settings to 255 by default?
05:27<planetmaker>that question is asked more and more frequently it seems to me.
05:28<@Rubidium>people shouldn't mix signal types if they don't know what they're doing
05:29<@Rubidium>and trains turning around in PBS blocks when it's 100% certain there's no route in that direction is the 'main' cause, not the signal settings
05:29<planetmaker>Nah, how could you not mix them? And it's not like they don't know what they're doing. Even with correct signaling it's annoying.
05:30<planetmaker>it breaks a perfect network when you have a small jam.
05:30<@Rubidium>but... the train turning around isn't annoying, it's the not turning around back when it can't go into that direction
05:30-!-DephNet[Paul] [] has joined #openttd
05:31<planetmaker>but in most cases, nowadays one-way signals are used. So... turning around is useless in most cases anyway
05:31<planetmaker>the turn-around setting was mainly useful when you only had two-way signals.
05:35<FloSoft>planetmaker: yes that makes more problems then solve them
05:35<FloSoft>if a train's stuck, you always have to manually let him drive over the red signal
05:35<FloSoft>that turning around annoys if you have only some kind of high traffic
05:38-!-Tekky [] has joined #openttd
05:44<Noldo>hi Tekky
05:45<CIA-1>OpenTTD: alberth * r16936 /trunk/src/rail_gui.cpp: -Fix [FS#3053]: Rail station build window with NewGRFs could crash due to uninitialized string parameter.
05:47<Tekky>hi Noldo :)
05:48<TrueBrain>I have the Dune2 intro on my screen, lalalalalalalalalalalaa
05:48*TrueBrain is so happy! :)
05:49<Tekky>TrueBrain: Hehe, are you using DosBox as an MS-DOS Emulator?
05:49<TrueBrain>C code ...
05:49<TrueBrain>really sick, not important :)
05:49<TrueBrain>and terrible slow ... :p
05:49<@Rubidium>only incanity ;)
05:49<Alberth>I shall need to dig up my Dune2 files soon :)
05:50<Noldo>dune2 has quite bad ui
05:50<TrueBrain>it was the first in its kind ... I give them that, no problem :)
05:51<Alberth>I still remember playing level 5 with turrets for the first time, what fun that was!
05:51<TrueBrain>I am watching the Intro now .. :)
05:51<TrueBrain>lalala :)
05:51<Alberth>building turrets so the enemy gets slaughtered automatically :)
05:52<TrueBrain>(well, mostly because I can't continue to the menu as mouse and keyboard are not functional :p :p
05:52<@Rubidium>mind control?
05:52-!-keoz [] has joined #openttd
05:52<TrueBrain>still not implemented
05:53<Alberth>so all the implemented functions work ;)
05:53<TrueBrain>the 55Hz timer fails :p
05:54<TrueBrain>euh, 55ms timer
05:54<TrueBrain>it runs more at like ... 1000ms :p
05:55<Noldo>what are running and with what?
05:56<CIA-1>OpenTTD: alberth * r16937 /trunk/src/rail_gui.cpp: -Codechange: Make the station build window look good when resized.
05:57<Alberth>computers, electricity
06:00*OwenS burns OpenSolaris CD
06:04<Tekky>I am currently working on a patch which I will call "custom signal routes" or something like that. This patch will allow you to define all possible routes from one PBS signal to the next signal and modify the pathfinder cost for these individual routes. This patch will hopefully be very useful for making trains select the best route when entering/leaving a station, thereby hindering the...
06:04<Tekky>...other trains less who are entering/leaving the station. The reason I decided to make this patch is because it happens often that trains make pathfinding decisions which severely hinder other trains.
06:04<Tekky>However, it is a lot of work to understand all the pathfinding code of OpenTTD :(
06:05<TrueBrain>OwenS: why?
06:06<OwenS>TrueBrain: For the server I'm building
06:06<OwenS>My primary attractions are ZFS and Zones
06:06<OwenS>Any more questions? I'm gonna have to turn this machine off to steal it's disk drive ;-)
06:07<TrueBrain>enjoy :)
06:07-!-OwenS [] has quit [Remote host closed the connection]
06:11<Noldo>Tekky: what do you need those custom routes for?
06:14<planetmaker><Alberth> I shall need to dig up my Dune2 files soon :) <-- I shall do that then, too :-)
06:17<Tekky>Noldo: I have a ro-ro station with 12 platforms and trains enter the station from the main line using 6 tracks. However, it often happens that trains take the shortest route into the station, instead of taking the route into the station which hinder the trains on the other 5 tracks least.
06:17-!-LadyHawk [] has joined #openttd
06:18<Tekky>Noldo: Therefore, I see the need to be able to influence the pathfinder cost of all possible routes myself, so that trains prefer taking routes which hinder other trains least.
06:19<Noldo>how many inputs?
06:19<Noldo>ah 6
06:19<LadyHawk>i missed the rest of the conversation but i 'trick' trains on crossings
06:20<Noldo>make the pathfinder look at all 6 trains trying to enter and solve the problem so that as many as possible can proceed
06:20<LadyHawk>a train has 1 tile to go either left or right, and it'll choose left.. if at that same junction, there's a 90 degree slab on the tile, it'll choose right.. makes no sense but i trick trains that way
06:22<LadyHawk>trick em all to choose paths that have the least chance of hindering other trains that way.. yeah it still happens, but not often
06:24<LadyHawk>i also limit their choices.. 1 huge junction where 20 tracks merge into 18 tracks.. each of the 20 tracks has 2 exit options only, not 18
06:24<LadyHawk>so if a train decides to hinder another.. the other has a second chance to go free
06:24<LadyHawk>i feel like i make no sense
06:25<Tekky>Noldo: Yes, it would be the ideal solution if the pathfinder would detect all possible conflicts with other trains and make pathfinding decisions accordingly. I have looked into such a system myself, however it would be very complex. Therefore, I think the best solution for now would be to implement these custom signal routes, which does not automatically take other trains into account....
06:25-!-OwenS [] has joined #openttd
06:25<Tekky>...Instead, it only allows the user to influence the pathfinding cost of individual routes.
06:26<LadyHawk>i'd love a pathfinding system like that though lol, maybe if someone gets bored enough for 2 weeks straight and works on it..
06:26<LadyHawk>if it ever does get made, it's prolly not gonna get touched again ever XD
06:27-!-Exl [] has joined #openttd
06:30<Tekky>LadyHawk: I have been thinking a lot about pathfinding in which trains influence each other's pathfinding decisions. This would allow cool things such as bi-directional double track, as I have described in the following article:
06:32<Noldo>when a train does pathfinding will it plot the whole route all the way to the destination tile by tile?
06:34<Tekky>Noldo: Yes, it will, but not necessarily "tile by tile", as track segments are cached in YAPF.
06:36<Noldo>I might be good idea to make rough signal to signal route first and then more detailed routing only on shorter scale
06:37<Tekky>LadyHawk: Yes, there are many possibilities to influence trains' pathfinding decisions using the methods you described, for example with 90 degree turns. However, I consider such methods as "ugly hacks". It would be much nicer and also more reliable if you could simply define all permissible routes and their pathfinding cost explicitly. That is what I am currently working on.
06:39<Tekky>Noldo: Depends what you mean by "rough". The pathfinder must know exactly whether it is able to reach a certain location if it takes a certain route. For this, the pathfinder must examine every individual tile. Only after it has done this at least once, can the performance be improved by caching the results.
06:41<Noldo>make a cache of routes from signal to signal
06:43<Tekky>YAPF actually caches straight track segments, i.e. track with no switches in between. Even if there are several signals on a straight piece of track, it caches the whole straight piece of track as one segment. This is slightly more efficient.
06:44<LadyHawk>very true Tekky
06:45<LadyHawk>i tried to make a bidirection passthrough station a month ago or so but the only way i could manage that was to have entrance track, exit track, entrance track, exit track and somehow make them crossover later on.. it was a mess so i just gave up
06:46<Noldo>Tekky: will it count trains or red signals on those segments?
06:47<LadyHawk>only image that confuses me is the last track image before the first station image
06:47<OwenS>Yaay! Solaris has entered X11 :-D
06:47<LadyHawk>regular track train makes strong reservation.. i can understand that.. but surely the one on the other track needs a strong reservation to get out of deadlock's way
06:48<LadyHawk>thinkin bout it.. hmm.. it doesn't have to on that image cuz there's no trains from opposite side coming
06:48<LadyHawk>nm ;p
06:48-!-KenjiE20 [~KenjiE20@] has joined #openttd
06:48<Tekky>Noldo: When using cached data, the pathfinder does not take red signals into account. That is why the pathfinder only uses cached data for distant track segments. It does not use cached data in the immediate vicinity, so it can take red signals and PBS reservations into account.
06:49<TrueBrain>playing Dune2 with keyboard is no fun :(
06:50<LadyHawk>dune 2 <3
06:50<LadyHawk>i haven't played that in so long cuz i know of a horrible bug that just takes all the fun away ;(
06:51<TrueBrain>did you know that when yous tart a level, you start at negative credits
06:51<TrueBrain>but that they don't have an icon for this? :p
06:51<LadyHawk>were you the one that was messing with dune2?
06:51<LadyHawk>maybe you can 'mess' in a bugfix? XD
06:52<Tekky>LadyHawk: A "weak" reservation tells other trains: You may use this track as long as you promise not to get stuck on it, because I will be needing this track myself very soon, otherwise I risk getting stuck on the opposite track (which would cause a deadlock).
06:54<LadyHawk>Tekky for what i read on the page you linked it sounds very promising
06:54<LadyHawk>but there's so many things to keep into account it's just... heck, must be confusing to code... long long days of work
06:55<Tekky>LadyHawk: Bi-directional passthrough stations are now possible with the new PBS signals in OpenTTD. The article I gave you was made at a time before these new PBS signals existed. This link explains how to make a bi-directional pass-through station with the new PBS signals:
06:55<LadyHawk>yeah those are the ones i used
06:55<LadyHawk>but i dont stick any signals at the station though
06:56<LadyHawk>the junction is sitting right next to the station at both sides
06:56<LadyHawk>it reserves a path to safe stopping point so there's no need for those
06:56<LadyHawk>just try duplicating that into a 10 track station with so many trains on it that it needs at least 2 entrance and 2 exit tracks on both sides
06:57<LadyHawk>trains will horribly start to block eachother off
06:58<LadyHawk>pf.path_backoff_interval 20 Ticks between checks for a free path. Increase if you have a slow computer.
06:58<LadyHawk>i might have a look at reducing that
06:58<Tekky><LadyHawk> it reserves a path to safe stopping point so there's no need for those <--- I disagree. A train entering a platform will reserve a route outside the station, thereby blocking a junction, if you don't have the platforms surrounded by signals.
06:59<LadyHawk>i haven't noticed that
06:59<LadyHawk>the route gets unreserved behind them
06:59<LadyHawk>leaving the way clear
07:01<Tekky>LadyHawk: Did you activate the option to display PBS reservations?
07:01<LadyHawk>i like knowing what my trains are doing
07:01<Tekky>LadyHawk: Yes, the path behind the train gets unreserved, I am talking about the path in front of the train.
07:01<LadyHawk>oh, no it wont try to reserve that till it wants to leave the station
07:02<LadyHawk>cuz the signals reserve the path to a safe stopping point.. nothing after it
07:02<Tekky>LadyHawk: What version of OpenTTD are you using? It does with me, if the station platforms are not surrounded by signals.
07:02<LadyHawk>dunno, whatever latest version was
07:02<LadyHawk>not the nightlies
07:04<LadyHawk>my hd got wiped recently but it was uh.. 0.7.3 i think
07:05<LadyHawk>i think i still have screenshots
07:05*LadyHawk digs
07:07<LadyHawk>you're right
07:08<Tekky>LadyHawk: I am using a recent nightly build and it behaves differently with me. However, I am quite sure that this (possibly controversial) feature to reserve routes past a station platform was introduced by michi_cc during the development of YAPP, before YAPP was included into trunk. Therefore, it should exist both in trunk and in the 0.7 branch.
07:08<LadyHawk>i dont have the signals at the entrance but i do have them at the exit
07:08<LadyHawk>so i suppose you're right with the bi-whatever it was stations
07:09<Alberth>LadyHawk: 0.7.3 does not exist, maybe 0.6.3 ?
07:09<LadyHawk>that screeny's a month old
07:10<LadyHawk>i musta remembered wrong
07:10<Tekky>LadyHawk: Hmmm, platform entrances and exits? I don't understand? In a bi-directional pass-through station, platform entrances are also platform exits, since the platforms are bi-directional.
07:10<LadyHawk>yeah... i said you were right with those
07:11<LadyHawk>the screeny i linked only has trains entering from 1 side
07:12<Tekky>Alberth: The new YAPP PBS signals were first introduced in version 0.7.0.
07:12<LadyHawk>so if that station turned into bi-directional.. you'd be right
07:12<Alberth>it say 0.7.1 at the top :)
07:12<LadyHawk>signals both ends
07:13-!-Singaporekid [] has joined #openttd
07:15<CIA-1>OpenTTD: rubidium * r16938 /trunk/src/ (12 files in 3 dirs): -Codechange: introduce helper function to tell whether a tile is either a rail station or rail waypoint tile
07:16<Tekky>I don't quite understand why michi_cc changed the behavior of YAPP to reserve routes past a station platform, since a station platform should always be a safe waiting location, in my opinion.
07:18<OwenS>Tekky: Not necessarily
07:18<Ammler>Tekky: to support that: ;-)
07:19<Ammler>what stops you from placing a pbs signal after the plattform?
07:20<OwenS>Whew! Nearly typed root password on wrong keyboard!
07:21<LadyHawk>dummy trains o_O
07:21<LadyHawk>i dont like those
07:21<LadyHawk>expenses you dont need
07:22<Tekky>Ammler: Yes, I always place signals on the exit of a platform, also because it looks better.
07:22<Ammler>well, that station uses some complex things but still looks much better then your, imo. :-P
07:22<OwenS>There are also situations, such as when using stations as penalties, that stoppiung at them is a really bad idea
07:23<LadyHawk>late in game stopping a train for only a second could cost you lots of money
07:24<LadyHawk>thats why instead of merging trains at a station, i had to make up junctions that would merge tracks into other tracks that lead to the station
07:24<Ammler>you think, it isn't true?
07:24<LadyHawk>divide the traffic over the whole station
07:24<LadyHawk>instead of trying to merge 5 into the corner of it
07:25-!-eleusis [] has joined #openttd
07:25<LadyHawk>what do dummy trains do?
07:25-!-Polygon [] has joined #openttd
07:25<LadyHawk>surely they're full at some point.. they still keep ratio up or something cuz of the 3 day thing on the station rating?
07:25<Ammler>they full load a train without full load order.
07:26<LadyHawk>yeah i get that part
07:26<Ammler>nothing about rating
07:26<Ammler>we play with one company only, so we don't need to cheat the rating ;-)
07:27<LadyHawk>the rating thing is the only thing i can think of that would benefit a dummy train
07:27-!-el[cube] [] has quit [Ping timeout: 480 seconds]
07:28<Ammler>[13:26] <Ammler> they full load a train without full load order.
07:29<Alberth>Ammler: From "Your e-mail will never be published." <-- shouldn't that be 'email-address'?
07:30-!-sdafsdf [] has joined #openttd
07:30-!-LadyHawk [] has quit [Read error: Connection reset by peer]
07:30<Ammler>if you comment?
07:30-!-sdafsdf is now known as LadyHawk
07:30<Ammler>that is wordpress, I am not english pro ;-)
07:31<Ammler>KenjiE20: ^
07:31<KenjiE20>blame wordpress
07:31*Alberth blames wordpress
07:31<KenjiE20>I'm pretty sure mine says the same thing
07:32<KenjiE20>it means essentially the same thing to most people
07:32<Ammler>but I agree to Alberth :-)
07:32<Alberth>no problem, I just noticed it
07:34<KenjiE20>ah, it's the template
07:34<KenjiE20>what's on mine --> "Mail (will not be published) (required)"
07:34<Ammler>LadyHawk: the trains which do the effective transport to the drop-off station don't have any load order, so they need to be sure, there is something to load.
07:35<Ammler>so the dummy train loads full and transfer the whole to the "main" train.
07:35<Ammler>more about srnw is on our wiki.
07:36<LadyHawk>thanks for that, made it a bit clearer
07:36<@Rubidium>so "Dear sir/madam, I'm writing to inform you of a fire. Looking forward to hearing from you. All the best, Maurice" would be valid?
07:38<TrueBrain>who knows how you can place structures with the keyboard only in Dune2? :(
07:40<LadyHawk>lmao Rubidium
07:41<Ammler>I changed it, thanks Alberth :-)
07:41<OwenS>"Transfering Contents" =)
07:41<LadyHawk>i never played dune2 with just keyboard.. why dont you use a mouse with it TrueBrain?
07:41<TrueBrain>because I didn't implement that yet, mostly :p
07:41<LadyHawk>fair enough
07:42<Ammler>it was in the skin, so not sure, if wordpress is really the right address to blame.
07:42<LadyHawk>i'd assume arrow keys or nummeric keypad
07:42<TrueBrain>but placing itself I can't find
07:42<TrueBrain>like moving a unit ...
07:42<LadyHawk>enter, space?
07:42<LadyHawk>hit keys till it places
07:42<TrueBrain>believe me, I tried
07:42<TrueBrain>so if you don't know, don't try to guess
07:43*TrueBrain !== stupid ;)
07:43<LadyHawk>in that case i wont be able to say anything
07:43-!-LadyHawk [] has quit [Quit: Never underestimate the power of stupidity.]
07:43<TrueBrain>that has to be a suprise ;)
07:44<TrueBrain>well .. that is overreacting I guess?
07:45-!-nfc [] has quit [Read error: Connection reset by peer]
07:45<@Rubidium>or 'p'
07:45-!-nfc [] has joined #openttd
07:45<TrueBrain>Rubidium: nope .. I tried all keys on my keyboard :(
07:45<@Rubidium>then the manual I found on the internet is incorrect ;)
07:46<TrueBrain>Rubidium: 'p' allows you to place it, but the really placing itself fails :p
07:47<CIA-1>OpenTTD: rubidium * r16939 /trunk/src/ (18 files in 4 dirs): -Codechange: s/RailwayStation/RailStation/ to unify the way it's written.
07:53<TrueBrain>Rubidium: which manual btw?
07:54<Ammler>TrueBrain: continue webdune?
07:56<TrueBrain>k, it really needs a mous eI guess :p
07:56<TrueBrain>tnx Rubidium
08:02-!-nfc_ [] has joined #openttd
08:04-!-nfc [] has quit [Ping timeout: 480 seconds]
08:10-!-LadyHawk [] has joined #openttd
08:10<Tekky>TrueBrain: In Dune, I think you must place a foundation for buildings before you can actually build the buildings.
08:10*roboboy may reconnect soon
08:11<TrueBrain>Tekky: nope
08:11<TrueBrain>but in fact, I tried placing a slab
08:11<LadyHawk>you don't have to place foundations... placing a building on no foundations gives it 50% health
08:11<LadyHawk>which means sticking it on slabs is cheaper
08:12<TrueBrain>but takes longer to build
08:12<LadyHawk>finished trying to multitask html/css bugfixing and talking here
08:12<LadyHawk>wasn't working together lol
08:12<TrueBrain>it starts to rain here ... and I mean really RAIN
08:13*Alberth is going bankrupt very slowly :s
08:13<@Rubidium>TrueBrain: that only happens *just* before I want to go to sleep
08:14<@Rubidium>and if it just rains a bit it's sleep inducing, but 'really RAIN' ain't... too noisy
08:14<TrueBrain>emulating the mouse is easier then I expected :)
08:14<LadyHawk>a forum post i find here says the hotkey could be the d key
08:14<LadyHawk>to place
08:16-!-roboboy [] has quit [Quit: ajax IRC Client]
08:16<@Rubidium>hmm, the precipitation surplus in Enschede is way of the Dutch average... and it was always so dry :(
08:17<TrueBrain>LadyHawk: nope, fails too ...
08:18<LadyHawk>The message on the Production button has now changed to a flashing
08:18<LadyHawk>"Place It." To place the structure, click on the Production button ot type
08:18<LadyHawk>"P." Your cursor will now turn into a glowing rectangle in the shape of
08:18<LadyHawk>your completed structure. Move the cursor over a portion of rock or
08:18<LadyHawk>concrete foundation adjacent to any of your structures and click the left
08:18<LadyHawk>mouse button or press the space bar.
08:18<LadyHawk>from the dune2 manual
08:18-!-roboboy [] has joined #openttd
08:19<TrueBrain>not in my emu
08:19<TrueBrain>not in DosBox
08:19<LadyHawk>you make me want to try it now =P
08:20<TrueBrain>for me it is easier to get the mouse to work :p
08:20<TrueBrain>movement is almost correct ...
08:21<TrueBrain>correction: IS correct
08:21<TrueBrain>now clickertheclick
08:21<LadyHawk>mouse makes life so much easier
08:26<TrueBrain>YES! Mouse works :) That was easy ;)
08:28<Alberth>that was the last we hear from TB today :)
08:30<TrueBrain>Alberth: ghehehe :)
08:30<TrueBrain>well .. still a lot to do to make it playable :)
08:30<TrueBrain>details ;)
08:30<@Rubidium>,%2017th%20Sep%201970.png <- oh it is such a hack ;)
08:30<planetmaker>TrueBrain, is it a binary to c de-compiler or... what exactly is it what works now?
08:31<TrueBrain>that, yes
08:31<planetmaker>TrueBrain, I'll be very interested to give it a try on another thing than Dune2 :-)
08:31-!-Chruker [] has joined #openttd
08:31<TrueBrain>send it over :p
08:31<planetmaker>Rubidium, cool!
08:32<planetmaker> <-- TrueBrain
08:32<TrueBrain>planetmaker: doesn't run in dosbox
08:33<planetmaker>it does
08:33-!-Zahl [] has joined #openttd
08:33<LadyHawk>don't suppose anyone wants to know the spoiler bug for dune2? lol
08:33<LadyHawk>i wish they'd fix that
08:33<TrueBrain>I only want to know why moving vehicles fails ...
08:33<LadyHawk>but the game's too old :'(
08:33<TrueBrain>planetmaker: it tells me it misses some ovl file
08:34<planetmaker>oh? Hm... might be you need some additional file. Let me find it...
08:36<planetmaker> <-- unpack this and than unpack the first zip into the same dir
08:36<planetmaker>if that still doesn't work, I screwed up somewhere :-)
08:36<planetmaker>in my own description and installation instructions
08:37-!-Tekky [] has quit [Read error: Connection reset by peer]
08:37<Chruker>Has anybody here ever encountered a computer (with winxp) that just loses the clock, but never the date?
08:37-!-Tekky [] has joined #openttd
08:37<TrueBrain>Stub Error (2003): can't find dpmi16bi.ovl
08:37<TrueBrain>(why does it want to load dpmi anyway)
08:37<planetmaker>hm... it might need... ship files.... err... hm?
08:38<planetmaker>never seen that error, honestly :S
08:38<planetmaker>what's the dpmi?
08:38<TrueBrain>planetmaker: get dosbox
08:38<TrueBrain>get it to towk
08:38<LadyHawk>interesting problem Chruker.. only thing that comes to mind is the motherboard battery, but you'd think it would also lose the date
08:38<TrueBrain>pack it in one zip
08:38<TrueBrain>give that to me
08:38<planetmaker>I'll do that
08:38<OwenS>TrueBrain: Perhaps it's a DPMI app? Doesn't explain why it doesn't detect DosBox DPMI
08:38<TrueBrain>DosBox has DPMI disabled
08:39<OwenS>That may be why :P
08:39<TrueBrain>how the comment says: never used, and mostly broken
08:39<planetmaker>always worked with dosemu - which is faster :-)
08:39<Chruker>ladyhawk, yes exactly, but over the last month it has repeatedly had the wrong clock (both hour and minutes) but never the date.
08:39<TrueBrain>planetmaker: either way, I read opcode 0x66 in the executable
08:39<TrueBrain>I only decode 16bit real mode application
08:39<TrueBrain>0x66 means 32bit
08:39<OwenS>0x66 means what exactly?
08:39<OwenS>MOV cr0, X?
08:39<TrueBrain>Operand Size Prefix :)
08:40<OwenS>16-bit apps used 32-bit operands
08:40<TrueBrain>so all instructions become 32bit
08:40<LadyHawk>Chruker have you looked at the synchronisation settings in winxp? it might be trying to synchronize with a wrong timezone or whatever
08:40<TrueBrain>planetmaker: so can't help you out there .. well .. not at this stage :)
08:40<TrueBrain>adding dword is kind of nasty :p
08:40<TrueBrain>howdie Belugas :)
08:41<LadyHawk>Chruker or you can try a different time sync server
08:41<@Belugas>howdie tired, TrueBrain
08:41<@Belugas>and ya?
08:41<LadyHawk>i've had issues with it in the past, none recently though
08:41<TrueBrain>dune2 now keeps on failing when trying to move units .. sucks ...
08:41<LadyHawk>or you can try to disable synchronization altogether and see how that works out
08:42<TrueBrain>time to do some shopping .. I am all out of .. well .. everything :p
08:42*LadyHawk shivers at the thought of shopping
08:42<OwenS>OK somethings up... not detecting my SATA drive =/
08:44<Chruker>I've had it running a few days with both timesync servers and also with it disabled and it still lost the clock. WHen I manually ask it to update the clock it does fail the timesync, but since it still loses clock with the sync disabled I'm not sure it is it.
08:45<Chruker>Although I guess there could be something else timesync and failing silently.
08:45<Chruker>*else timesync'ing
08:45<LadyHawk>how does it 'lose' the clock... does it fall behind every time?
08:45<LadyHawk>or is it random jumps hours ahead, hours behind etc
08:46<Chruker>After the computer have been off for a while (mostly nights) and it boot up, the clock is just behind.
08:46<planetmaker>TrueBrain, hm... here starting it in dosbox it works without any modification
08:46<Chruker>When I unplug the power and do shorter reboots it doesnt loose time.
08:46<planetmaker>hm... but the ovl is in the dir :-P
08:46<LadyHawk>typically a clock on battery will start falling behind if the battery is getting low.. the timesyncs failing might explain why it's never set back to the proper time
08:48<LadyHawk>if it is the battery it'll keep falling behind further and further, until you get the dreaded "settings reset to default, press f1 to continue" thing and you know for sure you have to get a new battery lol
08:48<@Rubidium>Chruker: is the clock going too fast or too slow?
08:48<OwenS>I was gonna say whats an "OVL" anyway, the nit hit me... OVerLay
08:49<planetmaker>:-) | OwenS
08:49<Alberth>you know, in the days that not everybody had 1G mem :)
08:50<Chruker>rubidium, sometimes it just have the wrong clock (hour + minute) when the computer starts. It never loses the date.
08:51<TrueBrain>OwenS: stupid enough, the files are called OVL, but Borland names the functions OVR :p
08:51<TrueBrain>go figure ;)
08:51<TrueBrain>oeh, I was gong to do some shopping .. bbl :)
08:51-!-nfc_ [] has quit [Ping timeout: 480 seconds]
08:51<OwenS>OK, I need to relabel this disk because Solaris is looking on the wrong one for it's root partiton
08:52<planetmaker> <-- @ TrueBrain
08:52-!-Yrol [] has joined #openttd
08:52-!-glx [glx@2a01:e35:2f59:c7c0:d532:8cee:15af:7043] has joined #openttd
08:52-!-mode/#openttd [+v glx] by ChanServ
08:52<planetmaker>TrueBrain, best playing experience, if started with "vpa 9 ."
08:52<planetmaker>(as other player files are not present)
08:52<Yrol>hello every body and mind :o)
08:53<LadyHawk>i've got a clock on battery here and it falls behind by 5 minutes every week, but it's been doing it for the past 3 months and still hasn't drained the battery completely
08:54<LadyHawk>i know it's different but when a pc is turned off.. it's the same kinda principle
08:55<LadyHawk>hey, Chruker:
09:05<Eddi|zuHause><Tekky> However, it is a lot of work to understand all the pathfinding code of OpenTTD :( <- afair YAPF has code that checks the first 10 signals for their state (red/green), and modify the penalty based on that, you should probably hook into that code
09:13-!-OwenS [] has quit [Remote host closed the connection]
09:23<Tekky>Eddi|zuHause: Well, that would work somewhat, but it wouldn't be ideal, as the pathfinding information in the distance would be incorrect, then.
09:24<Tekky>Eddi|zuHause: I would prefer the distant pathfinding information to also be correct.
09:27<Eddi|zuHause>hm, i think that's a minor problem or even a desirable side effect
09:27<TrueBrain>planetmaker: if I ever add wide-mode, I will run your app :p
09:27<Eddi|zuHause>you primarily will want to avoid blockings on THIS junction, and worry about the next junction later
09:34<Yrol>hm, planetmaker, im just reading about that alain2007-trying-to-do-a-megapatchpack. from what i read of your post there, i can just "checkout" my local trunk against any version available in the repository/remote trunk (?) and then have that version here locally available to patch against?
09:41-!-reldred [~reldred@] has quit [Read error: Connection reset by peer]
09:45<planetmaker>TrueBrain, "wide mode"? Oh, you mean the memory management it has?
09:45<planetmaker>Yrol, you can always check out any version you like, yes
09:45<TrueBrain>planetmaker: I mean 32bit access
09:46<planetmaker>Using mercurial, you only need to do it once as you have everything then locally
09:46<planetmaker>TrueBrain, yeah, hm... ok.
09:46<planetmaker>a pity it doesn't work then now :-)
09:46<TrueBrain>too new ;)
09:46<TrueBrain>to new
09:46<TrueBrain>what ever
09:46<planetmaker>too :-P
09:47<planetmaker>maybe I can find an old version... that game is at least as old as Dune2
09:47<TrueBrain>in the time of dune2 there was already wide mode
09:47<TrueBrain>ever since 386
09:47<planetmaker>But that binary is way newer, so maybe the current maintainer or a previous one added that support
09:47<Yrol>planetmaker, thank you
09:48<planetmaker>TrueBrain, that game ran on anything which had a VGA graphics card. And I only mean VGA.
09:48<planetmaker>like 286 was fine
09:48<TrueBrain>hehe :)
09:48<planetmaker>I played it back when I had one such system :-)
09:48<TrueBrain>if there is a 286 version, it should work :p
09:49-!-reldred [~reldred@] has joined #openttd
09:49<planetmaker>Naturally I never tested the newer binaries for 286 compatibility...
09:49<TrueBrain>0x66 is not a 286 spec, so it shouldn't ;)
09:49<TrueBrain>or my decompiler fucks up :p
09:50<TrueBrain>always a posibility :)
09:50<planetmaker>Might be that that particular binary is compiled for 386. Dunno really
09:58<Yrol>dad a chack?
09:59-!-Tekky [] has quit [Read error: Connection reset by peer]
09:59-!-Tekky [] has joined #openttd
10:00<Yrol>its just a sound that crabs make
10:00<Yrol>whats pompiedom(piedom) ?
10:00<TrueBrain>if I say: pompiedom
10:00<TrueBrain>you say: lalala
10:00<TrueBrain>got it? Good! Now remember that :)
10:01<Yrol>°salutes° yassir!
10:01<TrueBrain>so lets try:
10:01<@Belugas>tsoin tsoin!
10:01<TrueBrain>.... sigh ....
10:01<TrueBrain>people :'(
10:01<@Belugas>naaa... just bots
10:02<planetmaker>lalala.... Narrow streets of cobble stone...
10:02<Yrol>°goes and plays kraftwerk - boing bumm tschack°
10:05-!-HerzogDeXtEr1 [~Flex@] has joined #openttd
10:12-!-HerzogDeXtEr [~Flex@] has quit [Ping timeout: 480 seconds]
10:13-!-Polygon [] has quit [Quit: Verlassend]
10:21-!-Rexxars [~rexxars@] has quit [Ping timeout: 480 seconds]
10:30-!-Pygma [~quassel@] has joined #openttd
10:33-!-DephNet[Paul] [] has quit [Read error: Connection reset by peer]
10:33-!-DephNet[Paul] [] has joined #openttd
10:35-!-Default_ [] has joined #openttd
10:40-!-Default__ [] has quit [Ping timeout: 480 seconds]
10:42-!-goodger_ [] has joined #openttd
10:44-!-|Jeroen| [] has joined #openttd
10:44-!-michi_cc [] has quit [Ping timeout: 480 seconds]
10:45<Yrol>°waves° a good day to all of you :o)
10:45-!-Yrol [] has quit [Quit: MOOOOOOOOOO?????]
10:48-!-HerzogDeXtEr1 [~Flex@] has quit [Quit: Leaving.]
10:48-!-goodger [] has quit [Ping timeout: 480 seconds]
10:50-!-thisismynick [~chatzilla@] has joined #openttd
10:51-!-thisismynick [~chatzilla@] has quit []
10:56-!-michi_cc [] has joined #openttd
10:56-!-mode/#openttd [+v michi_cc] by ChanServ
11:07-!-Polygon [] has joined #openttd
11:14-!-Dartteri [] has quit [Ping timeout: 480 seconds]
11:18<CIA-1>OpenTTD: rubidium * r16940 /trunk/src/ (10 files in 2 dirs): -Codechange: make the pathfinders behave the same when finding waypoints or stations, i.e. don't force exactly one destination tile for a waypoint
11:22-!-DephNet[Paul] [] has quit [Ping timeout: 480 seconds]
11:23<TrueBrain>you don't say such things to people
11:23<Eddi|zuHause># lalala, text vergessen, scheiß egal
11:24<Eddi|zuHause>if i only remembered which song that was...
11:24<TrueBrain># And so we start to sing: LA LA LAAA!!! LALALAAAAAAAaaaa
11:24-!-DephNet[Paul] [] has joined #openttd
11:24<Eddi|zuHause>possibly "Katzenklo"
11:27<CIA-1>OpenTTD: rubidium * r16941 /trunk/src/station_map.h: -Document: some map accessors
11:42<Ammler>he, multi wps is comming :-)
11:44-!-Zorn [] has quit [Quit: ( :: NoNameScript 4.2 :: )]
11:48<TrueBrain>Belugas: again: IEUW!
11:48<TrueBrain>go back to your work or something :p
11:50-!-P0lygon [] has joined #openttd
11:53<Ammler>yeah, go back to your something.
11:54<TrueBrain>Ammler: don't be rude! :p
11:54<Ammler>oh, I forgot a bit smilie
11:54<TrueBrain>a bit smile
11:55<TrueBrain>how do they look?
11:56<Ammler>the smilies?
11:56<TrueBrain>a smiley constructed from bits
11:56<TrueBrain>nah, looks silly :p
11:56<@Rubidium>flapping your (large) ears?
11:56-!-Polygon [] has quit [Ping timeout: 480 seconds]
11:56<Eddi|zuHause>what's that supposed to be?
11:57<TrueBrain>Eddi|zuHause: you tell me?
11:58<@Rubidium>a reason to ignore him?
11:59<@Rubidium># what's wrong with you?
12:02-!-DephNet[Paul] [] has quit [Quit: Leaving]
12:04-!-KenjiE20 [~KenjiE20@] has quit [Ping timeout: 480 seconds]
12:17-!-KenjiE20 [~KenjiE20@] has joined #openttd
12:22-!-frosch123 [] has joined #openttd
12:24-!-DephNet[Paul] [] has joined #openttd
12:31<Tekky><Eddi|zuHause> hm, i think that's a minor problem or even a desirable side effect ... you primarily will want to avoid blockings on THIS junction, and worry about the next junction later <---- Yes, but if the user changes the pathfinder weights significantly on the custom signals, there will be a big difference between the results of the local pathfinder and the distant pathfinder. For...
12:31<Tekky>...example, as long as the custom signal is outside the range of the local pathfinder, the pathfinder will consider the junction with the custom signal to be the shortest route and therefore select that route. However, as soon as the train approaches the junction with the custom signals and the custom signals thereby getting in range of the local pathfinder, the pathfinder may find this...
12:31<Tekky>...route to now be significantly longer and therefore decide to turn around and select a completely different route.
12:34<Eddi|zuHause>well, then make the number of signals to be checked configurable (and have a setting for "infinity")
12:34<Eddi|zuHause>just the number of signals previously was limited because the signal states can't be cached
12:35<Eddi|zuHause>so it could be a huge performance hit if you send trains across the entire map
12:37<Tekky>Eddi|zuHause: I see no reason why the weight of the user-defined routes should not be included in the YAPF cache. The YAPF cache can simply be invalidated whenever the user makes changes to the routes defined by the custom signal.
12:38<Eddi|zuHause>yes, but that assumes that the weights can't be made dynamic (e.g. for possible load balancing)
12:40<@Rubidium>you can't cache it in YAPF anyway as it doesn't cache signal-signal
12:41<Eddi|zuHause>Rubidium: well, that can be easily solved by treating "active" signals like junction tiles, i.e. have segments end there
12:42<Eddi|zuHause>the existing "passive" signals would not need to be changed
12:43<Tekky>Eddi|zuHause: Currently, I have no plans to make them dynamic. I think it would be nice to have programmable signals that can be scripted, such as in the programming language Squirrel, since that language is already used by the AI. However, my current project is far less ambitious.
12:43<Eddi|zuHause>Tekky: it doesn't need to be in the plan, but it should be considered for future expandability
12:43<Eddi|zuHause>it's a design decision that should be made carefully
12:46<Tekky>Rubidium: Yes, one of my custom signals would have to trigger and end of segment in the YAPF cache and start a new segment.
12:46<Tekky>Rubidium: Yes, one of my custom signals would have to trigger an end of segment in the YAPF cache and start a new segment.
12:53-!-fjb [] has quit [Ping timeout: 480 seconds]
12:54<Tekky>Eddi|zuHause: Yes, I agree that it is important to make such design decisions carefully. Actually, I am thinking whether I should simply make my patch implement "player-defined routes with custom pathfinder weight", so that these custom routes can be linked to every track piece instead of only to signals. However, when using PBS signals, all pathfinding decisions are made at signals, so it...
12:54<Tekky>...seems most logical to link the routes to signals, at least when using PBS signals.
12:56<Tekky>I am not sure whether to implement this as a PBS feature or as a general feature.
12:57-!-ecke [~ecke@] has quit [Ping timeout: 480 seconds]
12:57<_ln>how much disk space should i leave for PS3 itself?
12:57<Sacro> <- ARRGH DX
13:05-!-Eddi|zuHause [] has quit [Ping timeout: 480 seconds]
13:05<CIA-1>OpenTTD: rubidium * r16942 /trunk/src/station_cmd.cpp: -Codechange: document CanExpandRailStation and remove the 'magic' array of numbers for a named struct
13:07-!-Eddi|zuHause [] has joined #openttd
13:08-!-fjb [] has joined #openttd
13:09<Eddi|zuHause>that was weird...
13:12<Eddi|zuHause>S.M.A.R.T. says the hard drive that i just got from warranty replacement has uncorrectable sectors...
13:13<TrueBrain>Eddi|zuHause: send it back, and call them names
13:22<fjb>You can not fully trust S.M.A.R.T.. And even uncorrectable sectors are not always a problem. They should be remapped.
13:24<TrueBrain>but when you get a new disk
13:24<TrueBrain>it should work
13:26<fjb>It was a warranty replacement. That are usually repaired disks.
13:27<fjb>And even uncorrectable sectors are no sign that the drive does not work.
13:27<fjb>Just write every sector once and the problem should be gone.
13:29-!-krushia [] has quit [Remote host closed the connection]
13:29<@Rubidium>I've heard that shooting it with a shotgun resolves uncorectable sectors as reported by SMART ;)
13:29<TrueBrain>oeh, I like to try that :)
13:31<KenjiE20> <-- well it's not a HD, but it IS being hit by buckshot
13:31<TrueBrain>why oh why does stupid unit movement not work :'(
13:31<fjb>Shotgun? Sounds like the french method of solving problems.
13:31<@Rubidium>fjb: it's American!
13:32<@Rubidium>the french would use a guillotine
13:32<@Belugas>or a baril or wine
13:33<@Belugas>GIVE ME ONE!
13:33-!-krushia [] has joined #openttd
13:33<@Rubidium>as I suspected, is of an American
13:36<TrueBrain>video removed by user
13:36<TrueBrain>hmm ..
13:38<fjb>Americans are shooting with everything. But the french are known for their love to shotguns.
13:40<Sacro>and white flags
13:40<Sacro>and cheese/wine/frogs/armpit hair
13:42-!-FR^2 [] has joined #openttd
13:43-!-|Jeroen| [] has quit [Quit: oO]
13:45<CIA-1>OpenTTD: translators * r16943 /trunk/src/lang/ (8 files): (log message trimmed)
13:45<CIA-1>OpenTTD: -Update from WebTranslator v3.0:
13:45<CIA-1>OpenTTD: simplified_chinese - 5 changes by Gavin
13:45<CIA-1>OpenTTD: dutch - 1 changes by webfreakz
13:45<CIA-1>OpenTTD: german - 1 changes by Roujin
13:45<CIA-1>OpenTTD: indonesian - 1 changes by prof
13:45<CIA-1>OpenTTD: korean - 5 changes by darkttd
13:47<TrueBrain>sjoep sjoep
13:51-!-OwenS [] has joined #openttd
13:51<TrueBrain>wb OwenS
13:52<OwenS>Bloody hell that took some doing.
13:53<OwenS>One of the disks and the motherboard misnegotiated at SATA2 speeds (They're in SATA-1 only enclosures)... Had to boot a CD to reconfigure trhe disk
13:53<OwenS>Naturally the first two (DVD+RW) burns the system refused to boot
13:53<OwenS>And the whole process was made more irritating by the need to continually swap CD drive
13:54<fjb>Sounds like a lot of fun.
13:55<TrueBrain>OwenS: but now it works? :)
13:55<TrueBrain>concratz :)
13:55<OwenS>Well, I'm having permissions trouble doing the cross-network mounting, but it mostly works :P
13:56-!-KUDr [] has joined #openttd
14:00<Ammler>does someone have an idea how/where to check, why the load average was for a short time excessed?
14:00-!-Singaporekid [] has quit [Quit: Leaving]
14:01<Ammler>has been*?
14:01<+glx>AI ?
14:01<Ammler>no, on a linux server :-)
14:01-!-tux_mark_5 [] has quit [Quit: KVIrc Insomnia 4.0.0, revision: , sources date: 20090115, built on: 2009/03/07 00:45:02 UTC]
14:01<+glx>DOS attack ?
14:02<Ammler>nothing in messages and warn logs
14:02<TrueBrain>processes start and stop
14:02<TrueBrain>what is in the past, is in the past
14:03<Ammler>it was just strange, some even complained the web wasn't reachable
14:04<TrueBrain>if it doesn't happen again .. who cares :p
14:04<Ammler>well, IF :-)
14:04<TrueBrain>yes, I didn't use: when
14:05<Alberth>you could look in the web access log
14:05<Ammler>nothing really special there either.
14:05<Ammler>just no call for around 5 mins.
14:05<TrueBrain>I once notice on a server (production server, several M hits per minute), that the HD did a spin down .. caused a nice stall for the next 15 seconds :p
14:06<TrueBrain>and the backlog that comes with that :p
14:06<TrueBrain>shit happens :)
14:06<TrueBrain>computers are supposed to be deterministic .. I wonder if they are :p
14:06<Alberth>yes, if only the network with all those independent users were :p
14:07<Ammler>to determine dos attacks, I would need to make tcp dumps or so, else I wouldn't see then?
14:07<TrueBrain>a DDOS of less than 5 minutes?
14:07<TrueBrain>you should count yourself lucky ..
14:07<TrueBrain>really Ammler, if it doesn't repeat, don't worry about it
14:08<Ammler>well, brians server ahs that all the time.
14:08<TrueBrain>when things repeat, you can trace it
14:08<TrueBrain>like we do with
14:08<TrueBrain>we have ... 'attacks' happening from time to time :p
14:09<OwenS>As I have just discovered, Solaris can suffer from the same "wrong drivers in initrd" problems as Linux
14:09<Ammler>how do you "trace" it?
14:09<OwenS>Fortunately, fixing them is very simple
14:10<TrueBrain>Ammler: depending what is going on .. first of, check the network activity
14:10<TrueBrain>next: check IO
14:10<TrueBrain>and if the load increased, check which process caused it
14:10<TrueBrain>but al that can only be done WHEN it happens .. or you need to enable extended logging ;)
14:10<TrueBrain>for example, the nodes I manage are normally running default logging .. network, IO, processes, etc etc
14:11<Ammler>well, on brians server, a restart of apache solved it mostly everytime
14:11<TrueBrain>when things start to spike, extended logging kicks in
14:11<TrueBrain>next time, check the amount of memory it consumes
14:11<TrueBrain>for we recycle the httpds every 24hour
14:11<TrueBrain>lighttpd currently is under daemontools control, being recycled faster I suppose
14:11<TrueBrain>(simply because after a while it consumes so much memory, it starts to fail)
14:12<TrueBrain>svnserve needs a reset every week, as it leaves stall connection all over the place
14:12<Ammler>he restarts the server at least 3-5 times per day
14:12<OwenS>TrueBrain: Try nginx
14:12<TrueBrain>then someone should look into it in more detail :)
14:12<OwenS>nginx > lighthttpd
14:12<TrueBrain>OwenS: considered it .. even have Cherokee running next to it
14:12<TrueBrain>but lighttpd does exactly those things we need .. and Cherokee is not mature enough :p
14:13<TrueBrain>currently things are 'stable', so it is fine for now
14:13<TrueBrain>(and more stable then those times we used apache :s)
14:13<OwenS>Resarting every 24h is not my idea of stable :p
14:13<TrueBrain>well, it has to do with proxies and giving files of several megs big :p
14:14<OwenS>nginx is fine with that kind of load without bloat
14:14<TrueBrain>(like requesting certain SVN revisions via trac)
14:14<TrueBrain>they said that about lighttpd too
14:14<TrueBrain>it turns out that what we have, is very ... special :)
14:14-!-tux_mark_5 [] has joined #openttd
14:14<OwenS>TrueBrain: No, lighthttpd has a memory leak bug which has been reported which they classify as no bug
14:15<OwenS>nginx doesn't
14:15<TrueBrain>it doesn't have a memory leak bug
14:15<TrueBrain>it just consumes massive amounts of memory in certain scenario
14:15<TrueBrain>so please stop the commercial for nginx
14:15<TrueBrain>I am sure when we install that, it will give problems too :)
14:15<TrueBrain>(like all httpds did so far)
14:15<TrueBrain>"if it aint broken, dont fix it", is what we currently say :)
14:15<OwenS>It's not a memory leak in lighthttpd as such. It's more that it allocates lots of memory to store the file, mallocs a bit, frees big chunk, and malloc is left with big lump it can't return to the system
14:16-!-Nite_Owl [] has joined #openttd
14:16<TrueBrain>OwenS: no, it is much more simple
14:16<TrueBrain>when you request a 100 MiB file from a proxy
14:16<TrueBrain>it allocates that
14:16<TrueBrain>there things go wrong
14:16<Nite_Owl>Hello all
14:16<TrueBrain>morning Nite_Owl
14:16<Nite_Owl>Hello TrueBrain
14:16<TrueBrain>OwenS: we had people downloading such files at 2 kbit/sec
14:16<TrueBrain>you do the math on that ;)
14:17<TrueBrain>files and stuff are loaded from one side of the kernel to the other .. takes no real memory :)
14:17<OwenS>TrueBrain: (Last nginx advert) nginx doesn't allocate such huge buffers - it reads as fast as it can write out the socket and no more
14:17<TrueBrain>Cherokee too
14:17<TrueBrain>but then there were other problems .... ;)
14:17<TrueBrain>no httpd is a saint
14:17<TrueBrain>this works, so why replace it? When it starts to fail, we will replace it
14:17<OwenS>No, but nginx is pretty good (Unless you need old-style CGI)
14:18<TrueBrain>oh! Yet another advert!
14:18<Ammler>again :-)
14:18<TrueBrain>how often can you say nginx?
14:18<TrueBrain>please .. stop it .. this setup works, how insane it might be. The future might hold other things .. for now, it is enough :)
14:18-!-ecke [~ecke@] has joined #openttd
14:18<TrueBrain>(you do not want to know the time spend on services so far by me and Rubidium)
14:19<OwenS>Now, if I could figure out how to configure Solaris' nfsd...
14:19<TrueBrain>the 'hidden' time wasted on open source projects :p
14:21<TrueBrain>our configuration is 2500 lines long :p
14:21<TrueBrain>(for lighttpd)
14:21<TrueBrain>euh, 608 for lighttpd, 2000 for cherokee
14:21<TrueBrain>and the apache I can't count .. crazy ass format :(
14:22<OwenS>Haha. Mine shrunk noticably when I dumped Apache also
14:22<TrueBrain>managing 26 domains
14:22<TrueBrain>of which 12 have 'special' attention
14:22<OwenS>I do "include sites.d/*" and have a config file for each domain in there.. I prefer it that way
14:22<TrueBrain>500 lines of scripts for git/hg
14:23<TrueBrain>1000 lines of other scripts
14:23<TrueBrain>(ranging from gentoo ebuilds to wt3 import, to corrections (Dates and stuff))
14:24-!-ecke_ [~ecke@] has joined #openttd
14:25<Ammler>don't forget the awesome compile farm :P
14:25<planetmaker>hehe :)
14:25-!-Dred_furst [] has joined #openttd
14:25<planetmaker>apropos... compile farm... what about adding grfcodec and nforenum to it?
14:25<TrueBrain>haha, lets not count the lines there :p
14:26<TrueBrain>and let's not count the patches we have for various of tools
14:26-!-ecke [~ecke@] has quit [Ping timeout: 480 seconds]
14:26-!-Yexo [] has joined #openttd
14:26<DaleStan>ISTR someone saying something about me needing to change the build process for that. But not what would need to change.
14:26<TrueBrain>howdie Yexo :)
14:27<Yexo>good evening
14:27<Yexo>hello TrueBrain
14:27<OwenS>Oh screw it I'll just install Samba. Configuring Samba is evil but it's evil I know
14:27<TrueBrain>OwenS: hehe :) I know that feeling :)
14:27<TrueBrain>planetmaker: I remember vaguely that Rubidium tried it, and failed at some level .. dunno what the problem was .. maybe he still remembers :)
14:27<TrueBrain>it was something silly
14:28<@Rubidium>something with the PPC IIRC
14:28<TrueBrain>(and not something we could fix)
14:28<planetmaker>TrueBrain: yeah. Well, the binary for, iirc renum, didn't run on my mac
14:28<@Rubidium>might have or might not have been fixed by now
14:29<TrueBrain>ah, yes, the one having asm, which is not supported by most of our targets
14:29<TrueBrain>the other not working on MAC
14:34<planetmaker>TrueBrain: well. It works, if compiled natively.
14:34<planetmaker>So that's a problem basically of the CF :-)
14:34<TrueBrain>if I remember correctly, Rubidium rolled out an OSX update last time
14:34<TrueBrain>so maybe now it works ..
14:34<planetmaker>we could try :-)
14:35*planetmaker can easily say so... little to do for me :-P
14:35<@Rubidium>DaleStan: for the compile farm we basically need a way to determine the version, named, and a way to build bundles easily e.g. make bundle_zip. See
14:36<TrueBrain>blegh, and it needs Boost?
14:36<@Rubidium>TrueBrain: yes, it needs boost
14:37*TrueBrain grumbles and goes back porting Dune2 ..
14:37<TrueBrain>planetmaker: some other day :p
14:37-!-krushia [] has quit [Remote host closed the connection]
14:37<planetmaker>he @ TrueBrain
14:37<TrueBrain>Rubidium: btw, the findversion stuff we might just copy from the CF to all projects .. as they all tend to need it :)
14:37<TrueBrain>(and we are copying over the same file :p)
14:37<@Rubidium>TrueBrain: most need special needs
14:37<TrueBrain>only the SRC_DIR, not?
14:38<@Rubidium>e.g. is there or isn't there a src/ directory
14:38<@Rubidium>furthermore grfcodec/nforenum seem to use ${REV}E instead of ${REV}M for modified builds
14:38<TrueBrain>hehe, something to improve, I say ;)
14:39<TrueBrain>but okay, the CF normally only compiles non-modified versions
14:39<TrueBrain>so that isn't a real issue from the CF point of view
14:39-!-krushia [] has joined #openttd
14:40<TrueBrain>either way ... CF rewrite .. still in the pipeline ...
14:40<TrueBrain>first WT3.1 I guess ...
14:40<TrueBrain>next week :p
14:40<TrueBrain>BaNaNaS rewrite .. sigh ..
14:40<Ammler>rewrite? not extendable?
14:41<@Rubidium>caught exception: std::overflow at TrueBrain
14:41<TrueBrain>Rubidium: yeah ... I feel like that :)
14:41<TrueBrain>still haven't done anything on my paid job :p
14:41<frosch123>well, as long as you are paid :)
14:41<TrueBrain>per assignment
14:41<TrueBrain>so not really helping :s
14:43<TrueBrain>stupid dosbox uses -exit instead of the normal --exit
14:43<TrueBrain>go figure
14:44<Alberth>java style long options
14:45<TrueBrain>in that case: FUCK JAVA
14:45<TrueBrain>why if the whole world goes left, there is always some big thing who wants to go right?
14:45<Alberth>you can say that of GNU too :)
14:46<TrueBrain>personally, I couldn't give a rats ass which method is used .. as long as they all would do the same
14:46<Eddi|zuHause>it's always a question of who is "the other side"
14:46<Eddi|zuHause>TrueBrain: that is just not going to happen :p
14:46<TrueBrain>as now I was trying --exit over and over, thinking: HUH?
14:48<_ln>who was it that planned porting ottd to C#?
14:49<planetmaker>haha :-) Long forgotten that is ^
14:49<@Rubidium>planetmaker: does work on intel apple?
14:50<Eddi|zuHause>_ln: lots of people wanted to port to lots of languages...
14:50<TrueBrain>and in the end, it is all useless
14:51<Eddi|zuHause>hm... kaffeine crashes on opening a 5MB png
14:51-!-GioGio [] has joined #openttd
14:52*OwenS starts the data copy to new server =)
14:52<TrueBrain>OwenS: concratz
14:52<TrueBrain>I am trying to figure out how to write nasm assembly :p
14:52<OwenS>I'm personally an AT&T syntax person
14:52<Eddi|zuHause>with a text editor :p
14:53<OwenS>I love confusing people with leal 8(%rax, %rbx, 32), %rcx type stuff :p
14:54<@Rubidium>that doesn't confuse me
14:54<TrueBrain>I just confused debug of DosBox :)
14:54<TrueBrain>I am just trying to print a message to the screen :p
14:54<@Rubidium>although I just read "I love confusing people with <some asm I can't be bothered to interpret> type stuff"
14:54<Eddi|zuHause>ah... TV is shit...
14:54<Eddi|zuHause>90% of it is rubbish
14:55<Eddi|zuHause>and the other 10% i have already seen
14:55<TrueBrain>Eddi|zuHause: join the club
14:55<Eddi|zuHause>the internet totally spoils the TV :p
14:55<@Rubidium>90% is rubbish?
14:55<@Rubidium>that can't be true
14:55<Eddi|zuHause>Rubidium: conservative estimate ;)
14:56<@Rubidium>then 10% must be good
14:56<OwenS>Note to self: Get gigabit network
14:56<@Rubidium>which equates to... given say 10 channels... a full day of good TV
14:56<TrueBrain>OwenS: check
14:56<planetmaker>Rubidium: seems to work like a charm
14:57<Eddi|zuHause>well, i did not specify that i was talking about the prime time programming
14:57<OwenS>Theres no point having high speed RAID arrays and 100mbit ethernet :p
14:57*TrueBrain enjoys his gbit every day :)
14:57<Eddi|zuHause>of course i totally ignore daytime programming
14:57<DaleStan>OwenS: If 8 is the scale and 32 is the base, I know what's going on there. Except that I thought the l indicated a 32-bit width, and the registers all imply 64-bit width.
14:57-!-GioGio [] has left #openttd []
14:58<Eddi|zuHause>so, given 10 channels, one channel has an interesting prime time each day, and this one i have already seen
14:58<OwenS>DaleStan: Woops with the l/q mixup. And other way round; %rcx = %rax + %rbx * 32 + 8
14:58<@Rubidium>planetmaker: your OSX still has rosetta, right?
14:59<OwenS>And on the "90% of TV is rubish line":
14:59<TrueBrain>oeh, segfaulting DosBox, cool :)
14:59<planetmaker>uhm... the ppc emulator/simulator? Yes
14:59<@Rubidium>in other words, does work too?
15:00<planetmaker>hm. no. It segfaults
15:01<@Rubidium>planetmaker: so big endian nforenum doesn't work
15:02<planetmaker>I'm not 100% confident to what degree that emulator works correctly here, though
15:02<planetmaker>I'd want to test a different big endian ppc binary before I'm 100% sure :-)
15:02<planetmaker>though I *think* I did that ages ago with some ... openoffice or so.
15:03<OwenS>I'd say shame PPC did away with little endian mode.. but actually little endian mode wasn't ever real little endian :p
15:03<TrueBrain>grrr, I forgot how tricky COM-files work ...
15:03<@Rubidium>planetmaker: is bash still big endian with your version?
15:04<@Rubidium>it was big endian in (at least some versions of) 10.4
15:04<OwenS>I thought PPC-LE was done away with on the PP64 (G5) and newer processors?
15:04<DaleStan>OwenS: rbx*32 !? 32bit only allows 0,1,2,4,8 as the scale. And I see no indication in the Intel programmer's manual that the REX prefixes extend the scale field of SIB.
15:05<OwenS>DaleStan: You're reading too far into ASM I've written in an IRC conversation. Particualrly when I haven't been writing x86 ASM for a while
15:06<DaleStan>Well, that's why I made 8 the scale and 32 the base. 32's not a valid scale, but 8 is.
15:06<@Rubidium>planetmaker: could also be that my boost isn't endian aware or so
15:07<OwenS>I always forget the little silly restrictions of x86 like that (Only 2 bits for scale?!)
15:07<planetmaker>Rubidium: bash is running native
15:08<planetmaker>how do I check that actually?
15:08<planetmaker>IIRC, though, I'd get a kind of hint, that now the emulation is started and blah blah blah
15:08<planetmaker>and it usually is SLOW
15:09<@Rubidium>planetmaker: try bash --version
15:10<CIA-1>OpenTTD: rubidium * r16944 /trunk/src/saveload/ (industry_sl.cpp saveload.h station_sl.cpp vehicle_sl.cpp): -Codechange: remove needlessly complex way of getting the offset of items within a struct that are within a struct
15:10<planetmaker>GNU bash, version 2.05b.0(1)-release (powerpc-apple-darwin8.0)
15:10<planetmaker>Copyright (C) 2002 Free Software Foundation, Inc.
15:10<planetmaker>hm... that ... is... powerpc
15:11<planetmaker>and I wonder actually why they ship it with an ancient bash, too. But well
15:12<@Rubidium>because they can
15:12<Eddi|zuHause>mine is 3.2, so that 2.05 is really ancient ;)
15:13<@Rubidium>planetmaker: at least it's as up-to-date as Apple ships their bash
15:14<planetmaker>back then when 10.4 was current ... maybe :-)
15:14<@Rubidium>although they've backported a few bugfixes from 3.2
15:14-!-nfc [] has joined #openttd
15:14<@Rubidium>planetmaker: 10.5.7 still ships with bash 2.05b (if is a credible source)
15:15<planetmaker>I'm actually looking forward to 10.6 :-)
15:15<planetmaker>oh, do they?
15:15<planetmaker>Then it's a bit... stupid
15:15<Eddi|zuHause> <- Fox "News" tells taliban to execute a capured soldier, to save the us army some legal hassle :p
15:17<CIA-1>OpenTTD: rubidium * r16945 /trunk/src/base_station_base.h: -Fix (r16940): comments got swapped
15:18<Nite_Owl>also to avoid Darth Cheney from being prosecuted for war crimes
15:25<Nite_Owl>technically the Taliban could water board the poor soldier and simply claim that they were trying to get important information
15:27<TrueBrain>yeah, wrote my own DOS program ! (first time this year :p)
15:28<Eddi|zuHause>some DOS versions distributed ".com" files that were actually ".exe" files renamed
15:28<OwenS>Doing what?
15:28<TrueBrain>haha, you really want to know? :p
15:29<OwenS>Kinda :P
15:29<Eddi|zuHause>OwenS: not much more than a "Hello World" :p
15:29<OwenS>Eddi|zuHause: Thats nothing. Visual Studio includes a .com which is a win32 app
15:29<TrueBrain>it prepares the registers, runs a test (one single instruction over and over), dumps all register entries in the memory, opens file, dumps all data there
15:33-!-Mucht [] has quit [Remote host closed the connection]
15:41<TrueBrain>if I now take a picture of my room view, and I make a jigsaw out of it, I can sell it for a good price
15:45<fjb>I would take some kind of steel to make a jigsaw out of.
15:45<@Belugas>naked girls dancing under a rainbow?
15:45<TrueBrain>fjb: yeah, good luck with that
15:45<TrueBrain>Belugas: only in your dreams
15:45<TrueBrain>(or mine .. hmm ...)
15:46<@Belugas>in OURS!
15:46-!-lime [] has quit [Read error: Connection reset by peer]
15:47<fjb>Belugas: You are married. :-)
15:47<@Belugas>yeah. so?
15:47-!-lime [] has joined #openttd
15:47<TrueBrain>I also miss your point fjb
15:47<@Belugas>Does not mean that we cannot use our eyes :S
15:47<@Belugas>use our hands is a little bit more.. troublesome
15:48<@Belugas>using ... well... you know... is even worse
15:48<@Belugas>but eyes.... come on...
15:48*OwenS wishes this file copy were faster
15:50*TrueBrain helps OwenS wishing
15:50<TrueBrain>okay .. my compare test works, if not that flags are stored differently :)
15:50<TrueBrain>I write 0x0206, DosBox writes 0x7212 ..
15:51<@Belugas>sonds like my wife... i earn $206, she spends $7212...
15:52<OwenS>Speeding it up would require buying a new network card (For the server... Atoms don't come with gigabit =() and router (Cause I hate this one and may as well kill two birds with one stone)... and probably relaying my Ethernet cable
15:52-!-PeterT [~Peter@] has joined #openttd
15:57<OwenS>I do love over-network X11
15:59<TrueBrain>okay .. first bug in my emulation layer .. flags! WHOHO! :(
16:02-!-CIA-1 [~CIA@] has quit [Ping timeout: 480 seconds]
16:03<TrueBrain>euh .. who can help me with this: mov ax, 0x1; shl ax, 0x1 .. in both cases PF should be 1, not?
16:03<Eddi|zuHause>if you earn $206, you should maybe consider switching jobs ;)
16:04<OwenS>I'd presume so
16:04<TrueBrain>Eddi|zuHause: he didn't define a timespan :p
16:04<TrueBrain>OwenS: DosBox says not :(
16:04<@Rubidium>Eddi|zuHause: $206 an hour is acceptable for me ;)
16:04<OwenS>Actually no
16:04<TrueBrain>oh .. even .. lol
16:04<Spoons>TrueBrain: Ooh, you're writing an x86 interpreter / translator? What host language?
16:04<OwenS>Isn't PF 1 when even parity?
16:04<TrueBrain>oopsie :)
16:04<@Belugas>Eddi|zuHause, never spoke about the frequency of the events...
16:05-!-PeterT [~Peter@] has quit [Quit: Leaving]
16:05<TrueBrain>Spoons: C (and small parts C++)
16:05<TrueBrain>emulator layer is pure C
16:05<TrueBrain>tnx OwenS
16:05<Spoons>That's not interesting, there's plenty of those. :(
16:05<OwenS>I was thinking SHR for a moment :p
16:06<TrueBrain>when is it 'interesting' in your book?
16:06<@Belugas>something flashy, full of buttons, that rumble and shake all over...
16:06<OwenS>w00t! Music copy finished!
16:06<blathijs>TrueBrain: Why are you building it?
16:07<Spoons>Interesting is not NIHing something, or a novel host platform, or a translator of any kind.
16:07<TrueBrain>blathijs: it started with me wanted to look inside Dune2 executable .. finding not many tools who can do it (although I now found IDA) .. got a little bit bigger with me wondering how hard it could be to decompile a 16bit application with modern computers .. turning out to be a big project :p
16:07<Eddi|zuHause>why do i think Civ4:Colonization is boring?
16:07<OwenS>And now the video copy... I think this may take all night, something I'd prefer it not to, as my machine doesn't make a good sleeping partner
16:07<Yexo>the original colonizations is way more fun
16:07<Eddi|zuHause>it does have neither the feel of Civ4 nor of Colonization
16:07<@Rubidium>Eddi|zuHause: because 90% of stuff is crap?
16:08<blathijs>TrueBrain: ida is pretty nice, yeah :-)
16:08<TrueBrain>OwenS: ghehehehehe :p
16:08<@Rubidium>and because Civ II was good, there must be 10 crappy versions
16:08<TrueBrain>blathijs: yup, it is :) But I have to say ... I like my decompiler more :) Somehow I can do more in C than I can with asm :p
16:08<blathijs>TrueBrain: Hehe, decompiling is nasty, especially when you want to do it statically (ie, without actually running the code)
16:08-!-Spoons is now known as FauxFaux
16:08<TrueBrain>blathijs: that turned out to be impossible
16:08<TrueBrain>so now it is a JIT decompiler :p
16:08<TrueBrain>it stores what it has seen
16:08<Eddi|zuHause>i believe there are way more than 10 versions (including alpha centauri and colonization)
16:09<TrueBrain>and makes a static binary of that later on :)
16:09<blathijs>TrueBrain: Cool. You really have too much time, haven't you? :-p
16:09<TrueBrain>well, I wasn't planning on spending so much time on it ..
16:09<TrueBrain>although .. much time .. hitting the 60 hours or so now
16:09<TrueBrain>so still within reason
16:09<TrueBrain>(well, 60 programming hours :p)
16:10<@Rubidium>TrueBrain: I reckon blathijs can use some of your free time for maintaining the debian package ;)
16:11<TrueBrain>free time is such an ugly word
16:11<OwenS>Aww hell this copy IS gonna take all night
16:11-!-Alberth [] has left #openttd []
16:12<TrueBrain>this is odd .. shl should not affect 'adjust' flag .. yet it does ..
16:12-!-Fuco [] has joined #openttd
16:12<TrueBrain>well, it is not specified what it should do, but DosBox suggest it shouldn't do anything ..
16:13<blathijs>Did you check the intel manuals? I think those are really good
16:14<TrueBrain>they suck, but yeah, I have them here
16:14<TrueBrain>doesn't change the fact it is not specified ;)
16:14<OwenS>AMD manuals say OF, SF, ZF, PF and CF. AF is undefined
16:14<OwenS>AKA AF may be anything
16:14<TrueBrain>but I wonder why DosBox makes it 1
16:14<TrueBrain>makes it hard to compare with my results you see :p
16:15<OwenS>Perhaps the processors do that and some app relies on it?
16:15<TrueBrain>okay, maybe I should word this differently:
16:15<TrueBrain>DosBox source code doesn't redefine AF
16:15<TrueBrain>yet it changes
16:15<Ammler>anyone here, who has blender installed?
16:16<Ammler>and might be able to make a 8bpp pcx from a .blend
16:21-!-Tekky_ [] has joined #openttd
16:22<TrueBrain>OwenS: do you by any chance know when the 'nested task' flag is set and unset?
16:23<Aali>it's set whenever you nest a task, and unset when you unnest it \o/
16:23<TrueBrain>why do I get this urge to kick someone ... :p
16:24<OwenS>You're emulating real mode, correct? In that case, it should never be touched IIRC; thats all to do with protected mode and hardware task switching and is royally ugly and completely unused
16:24<Aali>what he said
16:24<TrueBrain>tnx OwenS
16:24<TrueBrain>yeah, no diff! :)
16:25<OwenS>I believe it's set when a task does a call far task_segment:any_random_ignored_address and unset when a task does a retf. Or something like that
16:25<Aali>well, actually, hardware task switching is okay, however it's less flexible and not any faster than doing it yourself
16:25<OwenS>Aali: And nobody uses it. And AMD64 doesn't have it.
16:26<OwenS>And it's also slower but this is completely irrelevant when most of the time is spent doing other stuff anyway
16:27-!-Tekky [] has quit [Ping timeout: 480 seconds]
16:27-!-CIA-2 [~CIA@] has joined #openttd
16:27-!-Tekky_ is now known as Tekky
16:29<TrueBrain>I still can't find a single trace of SHL setting AF in DosBox .. yet it does ..
16:30<TrueBrain>maybe I am looking at the wrong source version :p
16:32<TrueBrain>unless it defined 'false' as 1
16:32-!-Dred_furst [] has quit [Read error: Connection reset by peer]
16:36<@Belugas>bye bye
16:36<TrueBrain>nightynight Belugas
16:36<@Belugas>night sweety sugar
16:36<planetmaker>night Belugas
16:36<planetmaker>and have a nice WE
16:37<frosch123>"For a non-zero count, the AF flag is undefined", what what does it contain for zero count :o
16:37<TrueBrain>frosch123: hehe :) Shifts with zero-count are a big no-no
16:39<TrueBrain>haha, slowly I find NASTY errors ... SHLw for example still takes only 1 byte as param, despite being wide :p
16:39<TrueBrain>(the documentation is NOT clear on this subject :p
16:41<planetmaker>Sounds like "Please write a 100 page document. I'll read the introduction to judge your analysis"
16:42<frosch123>mine shl only takes imm8
16:42-!-tokai [] has quit [Ping timeout: 480 seconds]
16:42<TrueBrain>oh, you are right, it says imm8 data
16:43<TrueBrain>where normally it says immediate data
16:43<TrueBrain>(without 8 or 16, meaning it can be either one :p)
16:44<TrueBrain>SET_FLAG(AF,(lf_var2b&0x1f)); <- introduced in Dosbox 0.71 or 0.72 ...
16:44<TrueBrain>how lovely :p
16:45-!-tokai [] has joined #openttd
16:45-!-mode/#openttd [+v tokai] by ChanServ
16:48<TrueBrain>basicly, if AF is set, the shift did anything :p
16:56<TrueBrain>now Dune2 no longer wants to start :'(
16:59<Eddi|zuHause>then you did either something wrong or you did not enough things right
17:00<TrueBrain>you know
17:00<TrueBrain>I think you are right about that :p
17:07-!-Azrael- [] has quit [Ping timeout: 480 seconds]
17:15-!-Nite_Owl [] has quit [Quit: Read You Soon]
17:17<TrueBrain>yeah, fixed it .. lalala :)
17:21-!-tux_mark_5 [] has quit [Quit: KVIrc Insomnia 4.0.0, revision: , sources date: 20090115, built on: 2009/03/07 00:45:02 UTC]
17:31<Chruker>Somebody should kick that guy until he shortens his damn exit message
17:32-!-Pygma [~quassel@] has quit [Read error: Connection reset by peer]
17:32-!-Pygma [~quassel@] has joined #openttd
17:36-!-KritiK [] has joined #openttd
17:38<TrueBrain>it is hard to kick someone who is gone ...
17:38<TrueBrain>I can kick you, on the other hand?
17:38<TrueBrain>oeh, more interest in WT3 :)
17:38<TrueBrain>Monday I will start with the new version :p
17:41<Chruker>you could, but I would just come back and back and back
17:41<TrueBrain>you could, but I can fix that ;)
17:42*Chruker wonders if TrueBrain could/would block connections via the tt-forums irc chat
17:43<TrueBrain>don't challenge me :)
17:43<TrueBrain>I think my hand reached futher then you want to know :p
17:46<TrueBrain>my english sucks .. my dutch too
17:47<TrueBrain>I keep mixing both ... in ways which are not always called: okay
17:49<OwenS>TrueBrain: I would disagree on the suckiness point
17:50<OwenS>Interesting question: Do Solaris zones get hostnames at the same level or one below their host machine?
17:51<TrueBrain>no clue :)
17:51<OwenS>(As in, which do I assign them :-P )
17:52<OwenS>(A zone can be considered to be a really lightweight virtual machine)
18:07<TrueBrain>hmm .. a good movie before going to sleep .. let me see ...
18:11-!-frosch123 [] has quit [Remote host closed the connection]
18:11<Chruker>"Eddie, does Dallas" ?
18:12<Ammler>oh, kde has a bookmark manager which allows sync between all browsers.
18:20<fjb>One the same level afaik.
18:21-!-Nazeradom [] has joined #openttd
18:21<fjb>OwenS: They are almost like FreeBSDs jails. So I guess on the same level.
18:21<fjb>Hi Nazeradom
18:24<Nazeradom>Well I just lost 6 hours refining a junction
18:24<Nazeradom>Only realised the time when I was suddenly hungry
18:25<Eddi|zuHause>i only realise the time when it's getting light outside
18:25<TrueBrain>we all have been there :p
18:27<Nazeradom>When I started drawing network diagrams I decided it was time for a break...
18:33-!-Lakie [~Lakie@] has joined #openttd
18:34-!-Exl [] has quit [Quit: Bitches.]
18:43-!-tokai [] has quit [Ping timeout: 480 seconds]
18:44<TrueBrain>hmm .. found a bug in DosBox :p
18:45-!-Ridayah [] has quit [Ping timeout: 480 seconds]
18:46-!-tokai [] has joined #openttd
18:46-!-mode/#openttd [+v tokai] by ChanServ
18:50<Eddi|zuHause>i want to play something, but i really should be going to bed, because i need to get up early...
18:50<TrueBrain>oh no
18:50<TrueBrain>WHAT TO DO!
18:50*TrueBrain start running around naked
18:52<Eddi|zuHause>i'm disturbed by that image
18:52*glx put the film on youtube :)
18:54-!-Ridayah [] has joined #openttd
19:05<TrueBrain>glx: really?
19:05<TrueBrain>you are that desperate?
19:06<OwenS>80 of 346GB copied... It's gonna be noisy in here tonight
19:06<TrueBrain>hihi, that was mean, sorry glx :)
19:06<TrueBrain>OwenS: shit happens ;)
19:07<TrueBrain>I fixed tons of flags-errors related to shifts ... documentation can be so nasty to read correctly
19:07<OwenS>It's also going to be brighter than normal due to silly illuminated fans
19:07<TrueBrain>hahahaha :)
19:07<TrueBrain>stupid DosBox btw made a bug because of their own unclear codingstyle
19:07<TrueBrain>val2w and val2b are the 2 values
19:07<TrueBrain>but either one is assigned a value when sub-functions are called
19:08<TrueBrain>then things tend to go wrong I guess ...
19:08<Eddi|zuHause>yay for functions with side effects :p
19:08<TrueBrain>but still my vehicles in Dune2 do not move :(
19:09-!-unambiguous [] has joined #openttd
19:09-!-unambiguous [] has left #openttd []
19:09-!-unambiguous is "crossposting" on (unknown)
19:09<TrueBrain>thank you for your visit
19:09<TrueBrain>drop by any time
19:14-!-sdafsdf [] has joined #openttd
19:14-!-LadyHawk [] has quit [Read error: Connection reset by peer]
19:14-!-sdafsdf is now known as LadyHawk
19:14<TrueBrain>0x11 << 0x1
19:15<TrueBrain>my emu says: oh, that is 0x10
19:15<TrueBrain>oh, it helps if I look at the right statement
19:18<TrueBrain>OwenS: the Overflow Flag on a SHL
19:18<Eddi|zuHause>conceptual question: if you want to make a static binary from the results of a JIT compiler, doesn't that require that every part of the code is actually run at least once?
19:18<TrueBrain>what do you think it should be?
19:18<TrueBrain>Eddi|zuHause: yup! Which is the tricky part ..
19:18<TrueBrain>but the good part: it doesn't have to happen at once
19:18<TrueBrain>if I give you the static binary, and you come at a point you miss something, it will exit, tell you about it
19:18<TrueBrain>you give me the info it prints you
19:18<TrueBrain>and I can make you a new binary
19:19<TrueBrain>of course it is not ideal, but the best that can be done :)
19:19<TrueBrain>OwenS: the reason I ask is because I followed the docs I have here, but DosBox does something different :s
19:19<Eddi|zuHause>you could include the original binary and the JIT compiler in the resulting binary for such cases ;)
19:20<TrueBrain>Eddi|zuHause: that is also possible, yes
19:20<TrueBrain>original binary is already there
19:20<TrueBrain>as it needs to be loaded in the memory ;)
19:20<TrueBrain>well .. when converted enough to C, I guess it will be needed less and less
19:20<OwenS>TrueBrain: I don't know actually what the OF should be All I know is it's modified; AMD's docs aren't particularly specific :P
19:21<TrueBrain>OF := high-order bit of r/m <> (CF);
19:21<TrueBrain>reads my docs
19:21<TrueBrain>so I made that: (result >> 7) ^ CF
19:21<TrueBrain>DosBox reads: (lf_resb ^ lf_var1b) & 0x80
19:22<OwenS>Yeah, OF = CF xor MSB of result
19:22<OwenS>Had to dig into the text :p
19:22<TrueBrain>but that is not what DosBox does :s
19:22<OwenS>Aah. Of the result only for 1 bit shifts. For longer shifts it's undefined
19:23<TrueBrain>of course :(
19:23<TrueBrain>in which case DosBox is right
19:23<OwenS>Make it work for one bit shifts and assume the processor does the same thing everywhere? :P
19:23<TrueBrain>I think my implementation is more clean .. but it makes my test fail when comparing to DosBox
19:23<TrueBrain>oh well ..
19:24<OwenS>If you want a more difinitive reference, try looking at Bochs
19:24<Eddi|zuHause>you need to exclude the "don't care" parts from the test case ;)
19:24<TrueBrain>Eddi|zuHause: hard to exclude a single bit ;)
19:24<OwenS>&~(1<<bit) - or make a macro of it :p
19:25<TrueBrain>file to file compare :p
19:25<TrueBrain>-00000c0 7212 7213 7213 7216 7212 7a13 7a13 7212
19:25<TrueBrain>-00000d0 7212 7a13 7a13 7212 7212 7213 7213 7216
19:25<TrueBrain>+00000c0 7212 7213 7a13 7216 7212 7a13 7a13 7212
19:25<TrueBrain>+00000d0 7212 7a13 7a13 7212 7212 7213 7a13 7216
19:25<TrueBrain>in that way ;)
19:25<Eddi|zuHause>difficult ;)
19:26<OwenS>Do the masking in the test program? :P
19:26<Eddi|zuHause>make a one-bit-per-line output :p
19:29-!-KritiK [] has quit [Quit: Leaving]
19:32<OwenS>Trivia: Britain is the only country in the world who's postage stamps don't have it's name on
19:32<TrueBrain>weird people :p
19:33-!-Eddi|zuHause [] has quit []
19:33<TrueBrain>okay, final try to see if this all fixed my dune2 problem ...
19:34<TrueBrain>nope ...
19:34<TrueBrain>although I can send it to attack my own base ..
19:34<TrueBrain>just moving by itself ..
19:35<TrueBrain>oh well .. tomorrow I have 256 test-cases to write :p
19:35-!-Eddi|zuHause [] has joined #openttd
19:35<TrueBrain>wait .. today I finishes 26 of them
19:35<TrueBrain>230 test cases to write
19:35<TrueBrain>wait .. 20 or so are not valid commands (or I don't uspport it)
19:35<TrueBrain>210 test cases to write
19:35<TrueBrain>wait .. I got to bed
19:35<TrueBrain>night all :)
19:35<TrueBrain>tnx for your help OwenS
19:35<TrueBrain>good luck with the noise
19:35<TrueBrain>and lights
19:36-!-Eddi|zuHause [] has quit [Remote host closed the connection]
19:36-!-Eddi|zuHause [] has joined #openttd
19:36-!-Eddi|zuHause [] has quit [Remote host closed the connection]
19:36-!-Eddi|zuHause [] has joined #openttd
19:36-!-Eddi|zuHause [] has left #openttd []
19:37-!-Eddi|zuHause [] has joined #openttd
19:37<Eddi|zuHause>something's odd with my reconnect script...
19:44<OwenS> I presume that domain is out of date? :P
19:44-!-OwenS [] has quit [Quit: Konversation terminated!]
19:44-!-OwenSX [] has joined #openttd
19:45-!-OwenSX is now known as OwenS
19:52<Eddi|zuHause>what do you mean?
19:52<Eddi|zuHause>i've had that hostmask for years...
19:54<fjb>What's wrong with
19:54<OwenS>"dial in"
19:55<OwenS>As in the reference to dial up networking
19:55<fjb>I still don't see the problem.
19:56<Eddi|zuHause>well, i guess that meaning shifted to generic non-static login
19:56<Eddi|zuHause>but that doesn't make it wrong...
19:59-!-KenjiE20|LT [] has joined #openttd
20:01-!-KenjiE20 [~KenjiE20@] has quit [Quit: WeeChat 0.3.0-rc2]
20:19-!-OwenS [] has quit [Remote host closed the connection]
20:19-!-Mucht [] has joined #openttd
20:23-!-Zorn [] has joined #openttd
20:38-!-Zahl [] has quit [Quit: *schiel*]
20:41-!-FR^2 [] has quit [Quit: Der Worte sind genug gewechselt, lasst mich auch endlich Taten sehn!]
20:41<Tekky>I need two additional bits in the map array for storing whether a PBS signal is custom or not, for my new project. I need two bits per tile because one tile may contain two signals. Judging from the documentation, it would be appropriate for me to use bits 12 and 13 in m2. That would be ok, wouldn't it? This is the first time I am using the map array, so I guess it is best to ask to make...
20:41<Tekky>...sure I am not doing something stupid. :)
20:52-!-Zorni [] has joined #openttd
20:59-!-Zorn [] has quit [Ping timeout: 480 seconds]
20:59-!-Zorn [] has joined #openttd
21:03-!-Zorni [] has quit [Ping timeout: 480 seconds]
21:04<Aali>Tekky: rest assured that *atleast* one of the devs think you're doing something stupid, especially if it involves signals
21:11-!-P0lygon [] has quit [Quit: Verlassend]
21:15-!-ecke_ [~ecke@] has quit [Quit: ecke_]
21:51-!-George [~George@] has joined #openttd
21:57-!-George3 [~George@] has quit [Ping timeout: 480 seconds]
21:58-!-reldred [~reldred@] has quit [Remote host closed the connection]
22:03-!-reldred [~reldred@] has joined #openttd
22:06-!-Chruker [] has quit [Read error: Connection reset by peer]
22:10-!-goodger_ [] has quit [Quit: +++ Out Of Cheese Error +++]
22:27-!-KenjiE20|LT [] has quit [Quit: Leaving]
22:50-!-Nazeradom [] has quit []
22:56-!-Lakie [~Lakie@] has quit [Quit: Leaving]
23:00-!-glx [glx@2a01:e35:2f59:c7c0:d532:8cee:15af:7043] has quit [Quit: bye]
23:08-!-TinoDidriksen [] has quit [Ping timeout: 480 seconds]
23:12-!-TinoDidriksen [] has joined #openttd
23:17-!-Fuco [] has quit [Ping timeout: 480 seconds]
23:39-!-TinoDidriksen [] has quit [Ping timeout: 480 seconds]
23:39-!-stuffcorpse [~rick@] has quit [Read error: Connection reset by peer]
23:40-!-reldred1 [~reldred@] has joined #openttd
23:42-!-Singaporekid [] has joined #openttd
23:42-!-stuffcorpse [~rick@] has joined #openttd
23:43-!-reldred [~reldred@] has quit [Ping timeout: 480 seconds]
23:43-!-TinoDidriksen [] has joined #openttd
23:58-!-reldred [~reldred@] has joined #openttd
---Logclosed Sat Jul 25 00:00:55 2009