IRC logs for #opendune on OFTC at 2009-09-12
⏴ go to previous day
10:16:35 <TrueBrain> it seems MarkJ was suprised by my tool :p
10:39:47 <Xaroth> He's not the only one surprised by yer tool :)
10:44:27 <Xaroth> I see you figured out Overlays last night as well :o
10:45:13 <TrueBrain> now I need to find a solution too :p
10:45:23 <TrueBrain> which is just a tiny bit more tricky :)
10:57:19 <Xaroth> tricky or not, you've figured out the rest without frying your brain so I'm pretty confident you can figure out this one too :)
10:59:57 <TrueBrain> je weet hoe je iemand moet paaien he :p (sorry, can't translate)
11:00:16 <Xaroth> Compliments are in order from time to time :)
11:00:23 <Xaroth> else I'll turn into petern.
11:00:33 <TrueBrain> enough references to him ;)
11:01:10 <Xaroth> he's just a too good of an example on how not to treat people you work with :)
11:04:41 <TrueBrain> you clearly never worked with Tron :p
11:17:25 <TrueBrain> okay, I can't get the CRC to work to have as unique id for a cs-frame
11:17:34 <TrueBrain> the cs is also not good enough
11:18:38 <Xaroth> is the cs somewhat-uniqu?
11:18:49 <TrueBrain> well, not for overlays
11:18:58 <Xaroth> combination of the two?
11:19:01 <TrueBrain> as 1 overlay can be loaded on cs A one time, and on cs B the other time
11:20:13 <TrueBrain> hmm .. combining doesn't give the easiest result, but it might just do the trick
11:23:13 <Xaroth> see i have smart ideas from time to time ^^
11:23:29 <TrueBrain> for sure you are not stupid .. but I hope you didn't doubt that :p
11:23:39 <Xaroth> especially for somebody who doesn't know half of what you're doing ^^
11:25:44 <TrueBrain> k, made the alterations .. now lets see what the result is ..
11:25:52 * Xaroth crosses fingers and toes
11:26:44 <TrueBrain> slow ... much more files which are written
11:27:28 <TrueBrain> scrap: too many files :p
11:29:23 <TrueBrain> I know overlays are loaded between 0x3FF1 and 0x4448
11:29:30 <TrueBrain> so .. I only have to do this between those values :)
11:32:12 <TrueBrain> okay, the values so far look promising
11:40:44 <TrueBrain> nevertheless, the results look good
11:40:51 <TrueBrain> just where does it go wrong, is now the question ;)
11:45:42 <TrueBrain> Insert disk 0 containing 'ut]#un.mpdipl.ENG
11:52:12 <Xaroth> I think something went wrong there
11:54:42 <TrueBrain> it happens every time when I try to detect overlays ... how annoying :p
11:54:53 <Xaroth> they are trying to hide from you :P
11:58:24 <TrueBrain> euh, yes ... a runtime-validation-check caused the problems :)
11:58:51 <TrueBrain> yes, now it continues :)
11:59:01 <TrueBrain> okay ... I seperated the overlays now I think ..
11:59:31 <TrueBrain> intro screen runs ... moon moving for planet ....
11:59:37 <TrueBrain> now lets try to start a level again
12:00:54 <TrueBrain> Building Windtrap :)
12:01:43 <TrueBrain> graphic glitches .. hmm ..
12:01:53 <TrueBrain> something is not 100% right in my decompiler, not sure what yet :p
12:01:57 <TrueBrain> but okay :) Game works ;)
12:03:04 <TrueBrain> Harvester arrived! :)
12:03:23 <TrueBrain> the game only really does not like when I move the mouse in or out the screen :p
12:03:38 <Xaroth> gotta keep playing then :)
12:04:22 <TrueBrain> lol, I just GAIN money by canceling a building :p
12:04:30 <TrueBrain> Build something till you are out of money
12:04:33 <TrueBrain> build something else
12:04:41 <TrueBrain> and you are left with more money then you started :p
12:04:45 <TrueBrain> (like 3 or 4 credits, but still)
12:08:08 <TrueBrain> and overlays are perfectly seperated! :)
12:08:24 <Xaroth> so, another step closer?
12:08:44 <TrueBrain> now I only need to make the return-detector aware of the -2 jumps
12:08:48 <TrueBrain> and correct for them
12:09:17 <Xaroth> then to find out what new wall hides behind the one torn down
12:10:29 <TrueBrain> stupid enough, the changes are 6 lines:
12:10:30 <TrueBrain> if (emu_last_cs > 0x3FF0 && emu_last_cs < 0x5000) {
12:10:32 <TrueBrain> jump_from.cs = emu_get_memory_crc(emu_last_cs, 0x0, 0x80) | 0x8000;
12:10:35 <TrueBrain> if (emu_cs > 0x3FF0 && emu_cs < 0x5000) {
12:10:36 <TrueBrain> work_cs = emu_get_memory_crc(emu_cs, 0x0, 0x80) | 0x8000;
12:10:43 <TrueBrain> just to point out how simple a solution can be, but how hard the road to the solution :p
12:23:48 <TrueBrain> hmm ... this is weird ... I thought some returns were just an overlay which was moved
12:23:54 <TrueBrain> but now the CRC turned out different for such overlays
12:24:03 <TrueBrain> so I am no longer sure if it is the same overlay ....
12:24:08 <TrueBrain> would be strange if it was not, but ... hmm ..
12:24:22 <TrueBrain> Fuzzy return at depth 7. Expected E3AB:0223, but got B6A4:0223 from 421F:0A23 <- stuff like this
12:24:35 <TrueBrain> the values before :0223 is the CRC of the overlay
12:25:07 <TrueBrain> taking a checksum at the wrong time is bad! :p
12:25:25 <Xaroth> time to get some food for the next few days
12:27:51 <TrueBrain> okay, another problem bites the dust
12:35:30 <TrueBrain> there, even less fuzzy errors
12:35:34 <TrueBrain> this only leaves a few to deal with
12:53:53 <TrueBrain> lets ignore those few cases in the static version, and just see what happens ;)
12:57:00 <Xaroth> technically, what -would- that cause?
12:57:04 <Xaroth> missing functions of some sorts
12:57:08 <Xaroth> or missing code in a function
12:57:19 <TrueBrain> no, it is the overlay which hooks itself via a return into an overlay
12:57:23 <TrueBrain> but I am unsure what it does
12:57:32 <TrueBrain> I THINK it just cleans up overlay memory
12:57:42 <TrueBrain> as it happens randomly, not really in any consistant way
12:58:37 <Xaroth> Verder veel succes met decompilen en developen van OpenDUNE. Ik kijk er zeer naar uit. :)
12:59:09 <TrueBrain> it can also be like this: if the overlay unloaded an overlay, it makes sure that on all returns from an overlay to another overlay it is called, to make sure the overlay you return to still exists
12:59:19 <TrueBrain> hehe :) Nice of him ;)
12:59:38 <Xaroth> ah, MarkJ replies once more :)
13:00:47 <TrueBrain> I will make a thread special for him to explain the things I know about the overlay :p Although once you read it, it is pretty straight forward
13:03:33 <Xaroth> think it'll be an interesting read, even for me
13:03:44 <TrueBrain> but for that I will first gather more details ..
13:03:52 <TrueBrain> but far before that, I want this to work :p Ghehe :)
13:05:25 <TrueBrain> there is ony strange return which worries me .. and which I can't figure out 1-2-3 .. but okay, first the rest :p
13:05:29 <TrueBrain> (stop stalling already! :p)
13:17:42 <TrueBrain> Xaroth: you use Windows, not?
13:18:48 <TrueBrain> sucks :p As I am very unsure if my emulator layer compiles for Windows ... and I don't have a way to test :p
13:19:06 <TrueBrain> and I use the LD_PRELOAD unix way to attach the JIT ..
13:19:08 <Xaroth> aka, you want me to test? :P
13:19:11 <TrueBrain> oh well, something to toy with later :)
13:19:54 <TrueBrain> "It is indeed my pleasure to write you this letter, which I believe will be a surprise to you as we have never met before, and I am deeply sorry if I have in any manner disturbed your privacy. Please forgive this unusual manner of contacting you, but this particular letter is of exceptional and very private nature. There is absolutely going to be a great doubt and distrust in your heart in respect of this email, coupled with the fact that, so many
13:19:55 <TrueBrain> individuals have taken possession of the Internet to facilitate their nefarious deeds, thereby making it extremely difficult for genuine and legitimate business class persons to get attention and recognition."
13:32:57 <TrueBrain> I suck in: connect the dots :p
13:51:47 <TrueBrain> somewhere .. there seems to be a bug :) Hehe :)
14:01:32 <TrueBrain> omg ... I made such a terrible mistake :s
14:02:12 <TrueBrain> somehow I cant get into this brain of mine that you FIRST need to push cs, THEN ip ...
14:02:15 <TrueBrain> not the other way around :'(
14:02:31 <TrueBrain> I made this mistake like 100 times now :p
14:02:33 <TrueBrain> it gets annoying :)
14:02:45 <Xaroth> same as forgetting the ; after things :P
14:02:51 <TrueBrain> that is just stupid :p
14:09:23 <TrueBrain> wow .... it built a graphical display
14:09:27 <TrueBrain> I am getting there ;)
14:10:22 <TrueBrain> even more stupid mistakes .. lol :)
14:10:31 <TrueBrain> lets see if you can figure it out :p
14:10:34 <TrueBrain> emu_cs = emu_get_memory16(emu_cs, emu_si, 0xA);
14:10:37 <TrueBrain> switch ((emu_cs << 16) + emu_get_memory16(emu_cs, emu_si, 0x8)) {
14:10:52 <TrueBrain> hmm, nevermind, it is not wrong
14:16:17 <TrueBrain> the error btw: using emu_cs after it is set, assuming it is not set :p
14:20:01 <TrueBrain> whoho, stack corruption! I love stack corruption :)
14:23:21 <TrueBrain> (infinite loop waiting for the timer to tick .. modern computers are just too fast for those 'tricks' :p)
14:25:04 <TrueBrain> wopla, now I am at the point of overlay loading :)
14:38:11 <TrueBrain> bah, it seems the return thingy I thought I could ignore, is important :p
14:42:37 <TrueBrain> hmm .. a 'quick' patch failed :p
14:44:45 <TrueBrain> ah, hmm ... yes ... sometimes it accessess data from the overlay, but of course I don't know the real segment of the overlay, so it reads some random data :p
14:45:14 <TrueBrain> (for those who don't know me: yes, I always talk this much :p)
14:51:26 <TrueBrain> okay, I now kind of need two things. On one hand I want to be able to tell my decompiler I rewrote a few functions and he should use them (and that of course in a clear way). On the other I need to fix this problem I am having now ;)
14:51:50 <TrueBrain> I need the one to test the other, and the other to make the one easier :p
14:55:01 <TrueBrain> okay, time for a good break :) Back in a while
17:06:57 <TrueBrain> what about them? :p
18:12:20 <TrueBrain> I justed watched "Frequently Asked Questions About Time Travel" .. it is suprisingly good for a BCC movie
18:59:36 <TrueBrain> ik heb net denk ik de meeste lelijke routine ooit geschreven
18:59:38 <TrueBrain> but it might work :)
19:05:55 <TrueBrain> I am already patching the resulting C files, isn't that awesome? :p
19:06:14 <Xaroth> nice, so the resulting stuff works? or still issues that need to be addressed?
19:06:22 <TrueBrain> compiling right now
19:07:04 <TrueBrain> darn, still somewhere a problem ... let me try to trace that ..
19:08:27 <TrueBrain> sadly, I didn't write any clever things yet which makes recompiles not take very very long :p
19:08:41 <TrueBrain> just 1809 files to compile :p
19:10:32 <TrueBrain> now that part I wanted to skip gives me a problem
19:10:34 <TrueBrain> of course .. hmm ..
19:16:17 <TrueBrain> hmm .. now I am just getting annoyed ...
19:26:04 <TrueBrain> somehow 'cs' is damaged, and indicates a wrong value
19:53:42 <TrueBrain> hmm .. yes yes .. this is a tricky problem .. hmm ..
19:54:02 <TrueBrain> as always, solve one problem, gain 2 :)
19:56:57 <nsz> oh so you made some progress, nice :)
19:57:15 <TrueBrain> I even made a lot of progress :)
20:03:41 <TrueBrain> hmm .. I somehow dislike it when the dynamic version didnt took a jump the static version wants to take :p
20:07:07 <TrueBrain> mostly because that feels very wrong .. for some reason :p
20:08:53 <Xaroth> oo survivorman in 2 hours on discovery
20:11:31 <TrueBrain> hahaha, an usleep ever 0x1000 memory access makes block-fades VERY slow :p
20:11:45 <TrueBrain> (those images where one image goes to another image via small blocks :p)
20:17:21 <TrueBrain> I don't really understand why in the dynamic mode it does A, and in static mode it does B ..
20:19:17 <TrueBrain> on various of places, and that is the only thing I can think off that can make a difference
20:19:19 <nsz> if it uses any world entropy source then it can behave differently
20:19:51 <TrueBrain> random is every time the same if you don't do any mouse and/or keyboard input :p
20:20:33 <TrueBrain> but okay, even if so, I just have to run the application a few time in dynamic mode, then it should harvast all that too ...
20:21:24 <TrueBrain> the only difference I can think off is that in dynamic the ticker goes faster from the application point of view (because the application goes slower, ghehe :p)
20:24:42 <nsz> well if it can have different pointers for the same data and then uses the pointers for something..
20:24:53 <TrueBrain> within an overlay everything is relative
20:25:13 <TrueBrain> (else it will tell me ;))
20:25:51 <TrueBrain> maybe I should move my usleep to a place it should be at :p
20:26:02 <TrueBrain> maybe my hack is what is causing the problem
20:29:10 <TrueBrain> it doesn't make sense! :'(
20:30:18 <TrueBrain> lol, my stack is 180 functions deep at a random moment :) But that was to be expected :)
20:35:39 <TrueBrain> haha, cool, in valgrind my ticker ticks sooner then my main is executed :)
20:36:30 <TrueBrain> downside of using timers ... code becomes bitchy :)
20:39:53 <TrueBrain> valgrind makes applications slow :p
20:44:54 <TrueBrain> internal revision 232 .. lol
20:45:15 <TrueBrain> my internal mercurial repos, I am at revision 232 already :p
20:45:17 <TrueBrain> (for my decompiler)
20:47:32 <nsz> then you commint often :)
20:48:18 <TrueBrain> and I started with this repo 15 Jul 2009
20:48:58 <DorpsGek> Xaroth: 3.59677419355
20:49:54 <TrueBrain> why oh why does my static version want to jump ... the dynamic doesn't, neither does DosBox
20:53:06 <TrueBrain> okay, I was wrong ... DOSBox does run it .. then why doesn't my dynamic variant do it :p
20:59:49 <TrueBrain> woepsiedoesieo! I saw the Westwood logo! :)
21:00:09 <TrueBrain> then it segfaulted :p Ghehe :)
21:00:22 <TrueBrain> stack corruption, what a suprise!
21:01:22 <TrueBrain> Xaroth: we need to start thinking about a way of releasing and a release cycle
21:01:41 <TrueBrain> I know that might be a bit early to start thinking about that at this point, but I know a lot of projects get problems with that later on :)
21:02:04 <TrueBrain> and I don't like the OpenTTD: release once a year a new version :p
21:02:57 <nsz> publish a hg repository somewhere
21:03:05 <nsz> and occasionally add tags
21:03:23 <Xaroth> I hate hg's way of making each revision some random tag
21:03:37 <TrueBrain> Xaroth: Mercurial also has a incremental number
21:03:47 <nsz> well that's the problem with distributed systems
21:03:54 <nsz> there are no global count
21:03:58 <TrueBrain> but I agree, svn is easier for central versioning; but it doesn't really matter, as the question is more: when to tag ;)
21:04:05 <nsz> or any global state for that matter
21:04:16 <Xaroth> TrueBrain: based on milestones ?
21:04:23 <Xaroth> aka, NoAI would be it's own release
21:04:30 <Xaroth> instead of part of a bigger release
21:04:30 <TrueBrain> Xaroth: yes, but you see how that goes for OpenTTD .. it can take ages for a milestone to finish :p
21:04:47 <nsz> the incremental number in mercurial is only usable locally
21:05:17 <TrueBrain> I will have a local hg, and push to a svn when I find that piece ready ;)
21:05:25 <TrueBrain> hg works so much faster ....
21:05:32 <TrueBrain> you want to try something? locally branch
21:05:40 <TrueBrain> you want to make a patch based on a patch? start a queue
21:05:51 <nsz> also no unnecessary networking
21:06:12 <TrueBrain> so yes, for local development, hg is perfect :)
21:06:31 <TrueBrain> as main source for a project, I think it is better to have svn (mostly because then you can point to a revision :p)
21:06:39 <TrueBrain> issue tracking becomes easier that way :)
21:07:03 <TrueBrain> Xaroth: I was thikning along the lines: every 3 months a release, 2 weeks before a release no new features, only bug-fixes
21:07:18 <Xaroth> hm, 3 months might be a bit fast from time to time tho
21:07:42 <TrueBrain> does it really matter? When I have this working (and I see the westwood logo more and more), any time is as good as the other
21:07:59 <nsz> i guess if you tag releases then issue tracking is just as easy
21:08:16 <TrueBrain> nsz: tagging every night might not be so nice :p
21:09:08 <TrueBrain> that is my main problem with hg, in a centralized project it is hard to track which version is where, and what users refer to
21:10:26 <Xaroth> Tb: 3 months good enough
21:11:31 <TrueBrain> version numbering .. we start with 0.1, and keep on increasing to 0.2, 0.3, .. 0.11, 0.99 every 3 months?
21:11:55 <TrueBrain> and we keep 0.2.1 or what ever in case anything really bad happens :p
21:12:11 <TrueBrain> (with a 3 month cycle I doubt :p)
21:12:39 <TrueBrain> WOPLA! Westwood logo does what I expect it to do :)
21:13:07 <TrueBrain> need to add a few usleep()s to fix infinite loops :p
21:16:03 <TrueBrain> nsz: I do plan to make, like OpenTTD, a 'nightly' every night, where people can test the state-of-the-art
21:16:17 <TrueBrain> it has shown to give a nice amount of input on bugs, before a release happens
21:16:42 <TrueBrain> on the other hand, with a 3 month cycle that might not be so much used as with OpenTTD, as with OpenTTD there are SO MANY new things in a nightly compared to latest stable, that is just insane :p
21:17:01 <nsz> i'm not so keen on project management issues
21:17:24 <nsz> but i guess it's important if you are working on a popular project
21:17:27 <TrueBrain> whoho, the 'and' shows :)
21:17:37 <nsz> opendune might be popular i guess
21:17:54 <TrueBrain> but at least I want to be prepared, to avoid pitfalls I have seen around me :)
21:19:04 <nsz> when we did wowmapview/wowmodelview we didn't care anything about management
21:19:21 <nsz> no community, no forums, no mailing list,..
21:19:34 <nsz> and still it was quite hyped..
21:19:59 <nsz> but that was because of the popularity of wow
21:20:10 <TrueBrain> it depends on your userbase I guess ;)
21:23:18 <TrueBrain> for the rest I can only hope OpenDUNE becomes a hype ;) But I am already happy if it keeps me busy for the next 2 years :p
21:24:23 <nsz> if you treat the dune code like blackbox then how do you plan to extend it?
21:24:44 <TrueBrain> how do you mean? The idea is to produce C code, and extend that.
21:25:16 <nsz> i thought the generated c code is gibberish :)
21:25:24 <TrueBrain> I hope to clean that up a lot :)
21:25:29 <TrueBrain> and the rest will be hand-work ;)
21:25:53 <TrueBrain> the resulting code is good readable if you know assembly .. of course a lot can be made in C-structures
21:26:02 <TrueBrain> the biggest challenge will be auto-detecting structs ;)
21:29:28 <TrueBrain> Xaroth: you don't have a linux machine with SDL (GUI) support at all?
21:29:49 <Xaroth> I have vmware workstation with unubto
21:31:44 <TrueBrain> hmmm .. there are minor errors ... somehow the cs is somewhere trashed again ..
21:32:36 <TrueBrain> but okay, it is a nice example :)
21:32:40 <TrueBrain> Xaroth: you have dune2 1.00 I guess?
21:33:29 <Xaroth> 1.07 on this machine iirc
21:33:33 <TrueBrain> oh, fuck it, I just add the files :p
21:33:36 <Xaroth> but i should be able to find 1.00 somewhere
21:34:46 <TrueBrain> bah, uploaded a not-working version
21:35:11 <Xaroth> An error occurred on the page you were trying to access.
21:35:11 <Xaroth> If you continue to experience problems please contact your redMine administrator for assistance.
21:35:15 <Xaroth> when going to repository
21:36:18 <TrueBrain> Xaroth: a restart of the lighttpd magicly fixed it
21:36:22 <TrueBrain> it happens more often, I hear ..
21:36:59 <TrueBrain> oh, it was out of memory :s
21:37:09 <TrueBrain> why does everything tend to leak memory :'(
21:37:20 <TrueBrain> default setting: LEAK MEMORY! :p
21:38:01 <TrueBrain> nice proyvind :) I found that site before, as the links are coloured :p Ghehe :) /me bookmarks this time :p
21:38:08 <TrueBrain> proyvind: btw, do you know what this HitSquad stuff is?
21:38:29 <proyvind> TrueBrain: it's a cd release version
21:38:42 <proyvind> you'll find it on the site above :)
21:39:25 <TrueBrain> lol, stripping the binary made it go from 4MB to 1MB :p
21:41:15 * Xaroth fires up vmware workstation
21:41:30 <TrueBrain> tomorrow I will run it over a windows compiler :p
21:41:44 <Xaroth> I love vmware's quick restore thingie
21:41:51 <TrueBrain> don't bug me with error-reports, there are plenty :p All it does: show Westwood logo, and Virgin logo
21:42:07 <TrueBrain> but I just want to show it isn't only talk, it really does something :)
21:42:08 <Xaroth> ah, real teaser then :)
21:44:12 <Xaroth> also, TrueBrain, I want a different approach to community-made features getting into trunk
21:44:26 <Xaroth> cuz the way OpenTTD does it is complete crap
21:44:43 <TrueBrain> (yes, we learn from problems of others :p)
21:45:33 <Xaroth> As you saw on the teams page I split up the team in developers and managers, it's not much to do with what a person does for the project, but more where his fields of expertise lie
21:45:41 <Xaroth> A manager can develop, and a developer can manage
21:45:51 <Xaroth> but managers are better in managing things, and developers better in developing
21:46:11 <Xaroth> as such the manager(s) should be in charge of the decision, in good discussion with the developers
21:46:42 <Xaroth> I really hated to see some good ideas got shot down by the ottd dev team because the devs couldn't be arsed fixing the patch to trunk
21:46:51 <nsz> ok just dont overengineer this
21:47:05 <Xaroth> I'm not overengineering
21:47:16 <Xaroth> I just don't like vetoing of stuff
21:47:47 <TrueBrain> Xaroth: well, nsz has a point, as we kind of are ;) But okay, lets at least think about it, if we ever execute it, is a second ;)
21:47:48 <Xaroth> the managers will pick up the request for things to put in trunk, they run it by the devs, summarize pros/cons, and make a balanced decision
21:49:35 <Xaroth> as a result, we should get an objective view of features to be put in trunk, or at least some form of objectivity
21:50:27 <TrueBrain> I personally don't think that will work, mostly as still a few select will be doing the decision making. But okay, we can at least try it ;)
21:50:57 <TrueBrain> I do suggest a few other things: dev-mailing list, where people can post patches/ask questions. I suggest we try to answer any and all mails within 7 days, if not alone a reaction of delay
21:51:21 <TrueBrain> on long (long) term (if the project ever becomes something): a website where people can ask opinions of other users
21:51:30 <TrueBrain> I always wanted to do that for OpenTTD ;)
21:51:56 <Xaroth> can't we use the development forum for that?
21:52:14 <Xaroth> authors posting their patch ideas for others to give their oppinions?
21:52:19 <TrueBrain> well, not with what I once had in mind, but I realise that it is something for much much later :p
21:52:28 <TrueBrain> either way: did you try it already?! :)
21:52:36 <Xaroth> no, have to restart instance
21:53:30 <TrueBrain> okay, I don't understand the overlay manager .. that is .. it sometimes puts itself in between, but I don't get why and when ..
21:53:38 <TrueBrain> seems random to me ...
21:55:22 <Xaroth> Program Termination: jumped to 2B6C:00E5, which is not decompiled. The jump was triggered at src/decompiled/f__2B6C_00C5_0045_415A.c:16 The jump appears to originate from 2B6C:00E1.
21:55:31 <TrueBrain> don't press any keys :p
21:55:55 <Xaroth> that looks FOOKING AWESOME
21:56:31 <TrueBrain> including the tiny stars on the letters :)
21:57:04 <Xaroth> clicking into screen is also a nono i see
21:57:13 <TrueBrain> fixing that right now :)
21:57:22 <TrueBrain> as that should work :p But I think I found the bug :)
21:57:43 <TrueBrain> finally .. PROGRESS! After 3 (or is it 4?) weeks of being annoyed with those overlays, I now finally have something that WORKS! :)
21:59:23 * Xaroth screenshots for sentimental value
22:00:07 <TrueBrain> what was it called ... chronical-blog or something?
22:00:16 <TrueBrain> a very simple blog that fits in your user-dir :p (as it puts out html files)
22:00:33 <Xaroth> as it runs on xaroth.nl
22:00:46 <TrueBrain> wopliedoeh! The menu works .. now the 'exit' part ..
22:01:07 <TrueBrain> Xaroth: for all I care you install a blog (simple blog!) on opendune, I will post once in a while, no problem :p
22:01:14 <nsz> nice, publish the code :)
22:01:28 <TrueBrain> nsz: will do :) Tomorrow or so :p
22:01:43 <Xaroth> Joomla isn't -that- bad.
22:02:18 <nsz> from code quality it is (at least it has bad security record)
22:02:25 <TrueBrain> you do know why they changed the name of the software, right?
22:02:52 <TrueBrain> yes, bad rep for a VERY good reason
22:02:56 <TrueBrain> the software is just crap (security-wise)
22:03:06 <Xaroth> they actually improved on security quite a bit
22:03:16 <Xaroth> that doesn't mean it's not crap, but.. less
22:04:43 <TrueBrain> tomorrow I will be testing my resuming .. that is: if the static version aborts with the nice code, I can attach the JIT, run it, recompile the static, and then that problem should be fixed :)
22:06:37 <TrueBrain> it is going to be legand
22:09:43 <TrueBrain> new version online, it 'fixes' the menu
22:09:48 <TrueBrain> it still crashes a lot, but at least you can see it now :)
22:09:57 <TrueBrain> and sometimes even make it to pressing Exit game :p
22:10:10 <TrueBrain> let me pack the source for nsz ;)
22:11:08 <TrueBrain> oh, I see the names are wrong ..
22:11:22 <Xaroth> still crashes at same place for me :P
22:11:38 <TrueBrain> Xaroth: press the mouse when the Westwood logo just appears
22:11:47 <TrueBrain> and 'crash' as segfault, or crash as a nice error?
22:12:03 <Xaroth> no click == nice error
22:12:09 <Xaroth> click multiple times == segfault
22:12:22 <TrueBrain> segfaults are harder to trace
22:12:24 <Xaroth> click once == nice error
22:12:28 <TrueBrain> they are mostly caused by stack overflow
22:12:48 <Xaroth> but once i left the bounds of the window, segfault
22:12:56 <TrueBrain> yeah .. as I said ;)
22:14:44 <TrueBrain> what the decompiler does at this stage, is to create a new file every jump
22:14:51 <TrueBrain> this makes the logic MUCH easier
22:15:04 <TrueBrain> but of course is useless for this :) Therefor a next stage will collect it, and put them in one file
22:15:17 <TrueBrain> but as that next stage is not important for the result, I didn't do that yet ;)
22:15:40 <Xaroth> I mean, f_0070_*_*_*.c has 8 files, if that is anywhere near an average you end up with just 300 fiels rather than 2500 :P
22:15:44 <TrueBrain> I now manually patches 21 files, of which a few should be done automaticly, so ...
22:16:09 <TrueBrain> well, if you look through the files, you will see that you can't in fact group 0070
22:16:13 <TrueBrain> but they are also kind of special :)
22:16:27 <TrueBrain> mostly you can group things, namely the ones that just jump to eachother (not calls or what ever)
22:16:39 <TrueBrain> they are mostly one function, and have some C-logic in them
22:16:49 <TrueBrain> but .. that logic I need to implement first ;)
22:16:49 <Xaroth> well yeh, but now there's 1 file per function
22:17:01 <TrueBrain> for now, it is sufficient :)
22:17:10 <TrueBrain> (as you can see, as it works :p)
22:17:14 <TrueBrain> just compiling takes for ever :p
22:17:41 <TrueBrain> k, going to remove the 'data' dir from the linux package and put it up as a seperate file
22:18:40 <nsz> restarted it with make -j4
22:18:42 <TrueBrain> nsz: lovely slow, don't you agree ;)
22:19:02 <nsz> there are lots of warnings
22:19:26 <TrueBrain> at least the amount of files
22:19:29 <nsz> you should declare those functions in a central .h...
22:19:50 <TrueBrain> as I explained above, I should put a lot of functions in one functions :)
22:20:00 <TrueBrain> then generate .h files based on the 'cs' value
22:20:02 <nsz> duh make: *** [main] Error 1
22:20:11 <TrueBrain> huh? It should compile ...
22:20:14 <TrueBrain> make 2>&1 | grep -v warning
22:21:53 <TrueBrain> lol, this already is a $1.2M project :p (sloccount)
22:22:23 <nsz> ah i should have placed libemu.so there
22:22:35 <Xaroth> it should say "9 person-years.... or 4 weeks kicking overlays"
22:23:06 <Xaroth> imagine that, for every week you were kicking the overlays, you did 2 person-years worth of work O_O
22:30:32 <TrueBrain> Xaroth: please enable [url] for phpBB :p
22:30:58 <nsz> hm i don't see the logo (yet)
22:31:15 <TrueBrain> nsz: starting takes a while; I see the black screen for like 10seconds or so
22:31:54 <TrueBrain> I believe Dune2 times the CPU, which takes most time :)
22:32:07 <TrueBrain> (and which is sadly enough a nearly infinite loop :p Despite my usleep(0))
22:32:26 <Xaroth> first to figure out where that is
22:32:49 <TrueBrain> nsz: but 'main' itself never eats much CPU. Only X11 does
22:32:58 <TrueBrain> thank you Xaroth :)
22:33:17 <nsz> btw when you package a tar it's nice when it contains only one top level dir
22:33:33 <TrueBrain> nsz: darn, they ar ein the root now? :(
22:34:09 <nsz> ah and the colors are right
22:34:32 <TrueBrain> of course ;) That took me a bit of time, but it is pretty simple :)
22:35:27 <nsz> mind if i show the screenshot around? :)
22:35:40 <TrueBrain> I posted the source and binaries on the forum too, so
22:38:59 <TrueBrain> and there should now be a parent dir in all packages ;)
22:40:47 <TrueBrain> Xaroth: clear enough dir-structure you think?
22:48:28 <TrueBrain> nsz: let me/us know the responses on it ;)
22:51:03 <nsz> ok well, it's on low traffic chan and everyone is sleeping
22:51:18 <nsz> but igor2 will like it when he gets up ;)
22:54:49 <TrueBrain> Although I am also sure he will have something to complain about :p
22:55:46 <nsz> he won't be able to compile it on his machine less than an hour :))
22:56:17 <TrueBrain> (igor2 uses p1 and shit :p)
22:56:36 <Xaroth> sounds like my linux server at home
22:56:41 <TrueBrain> nsz: but at least it is C, so he won't hate it :p
22:57:23 <nsz> not a fan of many things.. :)
22:57:47 <TrueBrain> hence my: he will have something to complain about :) He always does ;)
22:57:54 <Xaroth> he'll hate me then, being a C# coder n stuff
22:58:09 <Xaroth> i'll be the devil incarnate
22:59:34 <TrueBrain> but igor2 is one hell of a programmer :)
23:00:40 <Xaroth> I'll look forward to his request to join the dev team then ;)
23:01:07 <nsz> i doubt he has much time nowdays
23:01:11 <TrueBrain> he has little time lately
23:01:26 <TrueBrain> but he always talked about a dune2 clone ;) So we will see ...
23:01:30 <Xaroth> an hour of dune a day, keeps the doctor away :)
23:02:00 <TrueBrain> added a few legal notes on our main page :)
23:02:06 <TrueBrain> should solve any future problems :p
23:02:36 <nsz> just don't mention web forums to him
23:02:49 <TrueBrain> he liked Redmine! It has a TXT output :p
23:03:33 <TrueBrain> k, there .. enough for one day, I think :)
23:03:38 <TrueBrain> I have been on OpenDUNE for 12 hours now ...
23:03:42 <TrueBrain> it is time to get some rest :)
23:03:47 <TrueBrain> let me see .. plan for tomorrow ...
23:04:13 <TrueBrain> 1) Windows compile. 2) Resuming, to ease up development. 3) Better working teaser, maybe even include the first level.
23:04:39 <TrueBrain> oh, and 4) Fix up all stack overflows. 5) Figure out wtf the overlay manager is trying to do from time to time :p
23:04:46 <TrueBrain> k .. plan set :) Good night all :)
23:04:54 <Xaroth> well once 1 is done I might be able to help out with testing :)
23:05:09 <TrueBrain> the fact that nsz could compile, makes me happy too :)
23:05:11 <Xaroth> means I can actually do something useful :)
23:05:16 <TrueBrain> and I should release the source of the libemu :)
23:05:17 <Xaroth> what's nsz using then?
23:05:27 <TrueBrain> but it means I didn't use anything funky :p
23:05:37 <TrueBrain> I need a project page for libemu ... hmm ...
23:06:39 <TrueBrain> something to find tomorrow :)
23:06:44 <TrueBrain> nsz: tnx for your input and testing :)
continue to next day ⏵
