IRC logs for #opendune on OFTC at 2009-11-03
⏴ go to previous day
00:00:00 <TrueBrain> sometimes they include the offset in the place of the array ... very tricky stuff :(
00:00:11 <TrueBrain> most functions use an internal buffer ..
00:00:19 <TrueBrain> even 'strchr' or what it appears to represent
00:00:30 <TrueBrain> first time it is called with a buffer, next time without
00:05:06 <glx> hmm there are 3 functions looking like sprintf
00:06:29 <glx> maybe sprintf is not sprintf :)
00:07:13 <TrueBrain> the one you did identify does what sprintf should do
00:08:11 <TrueBrain> (for SCEN%c%03d that is :)
00:08:38 <TrueBrain> don't be suprised if there are functions close to it, which do things different .. :p
00:08:41 <TrueBrain> either way, I am off to bed
04:21:24 <DorpsGek> SVN: glx (r445) -Add: figured out a few variables
08:36:02 *** BoekaBart has joined #openDune
10:24:48 <BoekaBart> oh my, our new office fiber is 4 hops away from openttd.org, 5 ms latency, 6 megaBYTE/sec download...
10:27:22 <BoekaBart> And everything is open again, so I can svn:// again - REALLY fast
11:19:50 <TrueBrain> 4 hops .. 2 hops you always need in leaseweb, and most likely 2 in your own network, which means they have a peering with leaseweb, most likely in Evo
11:20:14 <TrueBrain> and 5ms latency puts you in 100km around Haarlem ;)
11:22:44 <Xaroth> which is 70% of holland :P
11:24:50 <Xaroth> I removed norton from this laptop the second I got it (few years ago)
11:24:55 <Xaroth> i just tried to run a vbs file.. norton popup
11:31:13 <TrueBrain> oh, sorry, 100km roundtrip, so 50km :p
11:31:27 <TrueBrain> Rotterdam for example is 13ms latency to AIX over fiber
11:31:42 <TrueBrain> s/latency/roundtrip/, to keep things clear :p
11:32:39 <Xaroth> and then to imagine, that in a few years or so people with 5ms latency will be considered 'slow'..
11:33:12 <TrueBrain> you will not get much faster
11:33:25 <TrueBrain> to the USA the fastest connection is around 43ms (average)
11:34:02 <TrueBrain> but the latter is the speed of light acting up :)
11:35:49 <TrueBrain> Xaroth: I think the fact you own all the posts scares people away :p
11:40:09 <TrueBrain> I already did that on fed2k yesterday
11:40:12 <TrueBrain> I had so much fun doing it :)
11:40:56 <TrueBrain> for everything is a day
11:41:09 <Xaroth> yes, but that one actually has a use! :P
12:31:55 *** TinoDidriksen has joined #openDune
12:48:56 <BoekaBart> TrueBrain: I lied before: 6 hosts, 5 hops, 6 a 7 ms avg.
12:49:53 <BoekaBart> 'router'.eventis.nl, signet.nl, signet.nl, leaseweb.net
12:50:18 <BoekaBart> leaseweb.net openttd.org
12:50:50 <TrueBrain> I love to be right :)
12:51:17 <BoekaBart> anyway the distance is > 200 km roundtrip
12:51:34 <TrueBrain> as I said, the fiber from Rotterdam to AIX is slower
12:52:19 <BoekaBart> what fiber runs from Eindhoven to AIX?
12:52:56 <TrueBrain> it also depends on the optical components used I guess ;)
12:53:36 <TrueBrain> but seriously, it is impressive, a 7ms roundtrip to Eindhoven and back
12:54:28 <TrueBrain> from Evo to Telecity2 already takes 2ms :p
12:55:48 <TrueBrain> Evo -> Sara is just 0.8ms .. lol :)
12:56:09 <BoekaBart> Ha, tracert to www.signet.nl (based and served in Eindhoven) goes via AMS
12:56:33 <Xaroth> TrueBrain: not sure if you want to be proud of that :P
12:56:44 <BoekaBart> funnily enough, the 1st hop outside our office is ams...
12:56:51 <Xaroth> SARA has been having issues the past year
12:57:01 <TrueBrain> gls .. I don't know gls
12:58:00 <TrueBrain> BoekaBart: signet blocks traces from this end .. sad :)
12:58:35 <TrueBrain> BoekaBart: if your first hop is ams, you have a good fiber connection :) But don't make mistakes, it is easy to hide a router from a traceroute
12:58:56 <TrueBrain> but it at least means you have a sane ISP, who knows what they are doing :)
13:00:23 <TrueBrain> 4 transits, nice :)
13:01:15 <TrueBrain> haha, BoekaBart, you have a direct peering with xsnews .. try using ipv6.xsnews.nl for very fast usenet binary access :p
13:01:21 <glx> TrueBrain: you saw my diff ?
13:01:30 <TrueBrain> glx: yes, and I applied it, not? Hmm ..
13:02:00 <DorpsGek> SVN: truebrain (r446) -Add: named a few more functions (glx)
13:02:02 <TrueBrain> it helps to press enter
13:02:48 <TrueBrain> accept ANY AND NOT {0.0.0.0/0} <- lol :)
13:04:09 <TrueBrain> k .. going to make a test install of openqrm .. lets see ..
13:08:59 <glx> hmm emu_String_sprintf_params is maybe emu_String_vsprintf indeed :)
13:11:32 <TrueBrain> glx: exact names are not important, as long as we have an idea what it does
13:21:04 <TrueBrain> lol, I love having a good CPU and enough memory
13:21:18 <TrueBrain> running XenServer, OpenFiler, Windows 7 and a debian install all on one machine
13:24:49 <TrueBrain> I believe it is now starting to run low on memory, as it is acting weird :p
13:31:24 <Xaroth> almost half way with the bottom half of the room
13:31:29 <Xaroth> about fooooking time as well
13:31:30 <glx> TrueBrain: you included one of your local changes it seems
13:32:47 <TrueBrain> in 445 I believe is a bit of extra code for the scenario load, but nothing fancy
13:33:28 <glx> removed extern void emu_Scenario_Load(); added #include "../src/scenario.h"
13:34:04 <DorpsGek> SVN: truebrain (r447) -Fix (r446): I fucked up ;)
13:34:44 <TrueBrain> too strong words for a commit message? :p
13:43:34 <Xaroth> hm, this grey is really.. grey o_O
13:43:43 <Xaroth> hope it dries a tiny bit lighter :P
13:43:56 <Xaroth> then again, carpet is going to be much lighter than this semi-dark-blue
13:50:28 <TrueBrain> wow, openqrm needs a lot of shit ...
13:50:39 <TrueBrain> but if it works .. it works ;)
14:42:11 <TrueBrain> okay .. OpenQRM fail because of lack of documentation
14:46:03 <Xaroth> TrueBrain: i take it binaries. is using normal directory listing?
14:46:47 <Xaroth> analytics :P but come to think of it, we're better off with the httpd log for those stats :P
14:47:42 <TrueBrain> really, to do things in OpenQRM you have to follow a list of 12 entries
14:47:45 <TrueBrain> which are .. vague, at best
14:48:00 <Xaroth> openqrm is nice, if it'd not be so damned bloated :P
14:48:17 <TrueBrain> better suggestions? :p
14:48:27 <TrueBrain> I now have to make an export of something ... NO idea how
14:51:27 <TrueBrain> we are in big need of some thing to manage our VPSes
14:51:36 <TrueBrain> and the paid products of Citrix and/or VMWare suck
14:51:42 <TrueBrain> (mostly as they require Windows :p)
14:51:59 <Xaroth> i'd say vmware infrastructure thingie
14:52:06 <TrueBrain> (mostly as they require Windows :p)
14:52:19 <Xaroth> but you already said, mostly as they require windows...
14:53:07 <TrueBrain> also, it cost A LOT for what you get
14:53:59 <TrueBrain> but OpenQRM is too poorly documented
14:54:05 <TrueBrain> that would takes hours and hours of testing and failing
14:54:27 <BoekaBart> TrueBrain: the only thing blocked now are NNPT and SMPT so .. xsnews.nl is no option :(
14:54:45 <TrueBrain> why would they block NNTP?
14:54:55 <TrueBrain> SMTP is logic, but NNTP?
14:55:34 <Xaroth> i'd be more tempted to block ssh than nttp...
14:55:49 <TrueBrain> why? SSH is always nice and good :)
14:56:47 <TrueBrain> BoekaBart: but there is always SSL for NNTP :)
14:59:03 <TrueBrain> I know XsNews support NNTP over a SSL port :)
14:59:54 <TrueBrain> as I said, a direct peering
15:07:21 <TrueBrain> oh, you also had crash-reports, not glx? Let me run them too ..
15:07:49 <glx> yes I had crash reports :)
15:09:45 *** BoekaBart has left #openDune
15:10:16 <TrueBrain> JIT keeps segfaulting, but okay, I believe it generates just enough data
15:13:46 <DorpsGek> SVN: truebrain (r448) [JIT] -Add: mapped another 10 functions (crash-bins provided by glx)
15:13:56 <DorpsGek> SVN: truebrain (r449) -Update (r448): update decompiled code
15:23:18 <TrueBrain> there is a max of 16 reinforcements you can get in a scenario
15:23:36 <TrueBrain> there only is no check if you go over that vlaue
15:23:57 <TrueBrain> Xaroth: maybe we should start a simple blog where we can post such 'knowledge' .. some fed2k might find that useful? Dunno ..
15:32:59 <TrueBrain> this is weird .. the '+' at the end of REINFORCEMENT is there in more than one scenario, but it seems the loading of a scenario never reads it due a bug in the code
15:35:31 * SmatZ totally doesn't know what you are talking about :(
15:37:02 <TrueBrain> lol, due to this bug, the game is MUCH easier :)
15:37:11 <TrueBrain> when ther eis a +, it should keep repeating (the reinforcements)
15:37:14 <TrueBrain> instead .. it never does :p
15:38:27 <TrueBrain> hehe, segra noticed the same problem, so it is not me :)
15:38:48 <SmatZ> those random soldiers appearing at borders of map :)
15:38:55 <TrueBrain> at your base, most of the time
15:40:06 <SmatZ> segra is sister in czech :)
15:40:28 <TrueBrain> to fix or not to fix this bug .... hmm ....
15:41:05 <SmatZ> fix and fix scenarios ;)
15:42:58 <TrueBrain> Xaroth / glx: what do you think? It is a big bug, and clearly a bug. Should it be fixed or should we keep it for now, and fix that later on or something? I wonder how close we should stay to Dune2, and how we want to do such fixes ...
15:43:29 <TrueBrain> I wonder if it is a good idea to like a TTDp like system, where with switches you can enable fixes :)
15:44:49 <TrueBrain> better suggestions / ideas?
15:45:49 <TrueBrain> arent you useful today ;) :)
15:45:56 <SmatZ> if you want to negate some idea, ask me ;-)
15:46:19 <TrueBrain> I really wonder how to approach this .. as all versions of Dune2 ignore this +, so it changes the gameplay a lot ..
15:47:28 <SmatZ> maybe detect "old-format" scenario and ignore the + there?
15:47:38 <TrueBrain> the scenario format won't change
15:47:42 <TrueBrain> it is a bug in dune2.exe
15:47:49 <TrueBrain> it checks at the wrong place for the +
15:48:15 <SmatZ> exceptions from rules are bad :(
15:48:46 <SmatZ> who will want to play scenarios in "old" dune2 (to compare behaviour)?
15:49:06 <SmatZ> even less people than who play ttdp now...
15:49:10 <TrueBrain> well, it is a big difference in difficulty if you get enemy attacks all the time, or just once
15:49:22 <SmatZ> if there was no ttdp, and only ottd, how many people would still play ttd
15:49:59 <SmatZ> or do you want to load scenarios for dune2 without any modifications?
15:50:25 <SmatZ> then there is no other choice than ignoring the +, or modifying scenario format (like, using differect command or so)
15:50:50 <Xaroth> TrueBrain: it should be fixed, though clearly documented in some form
15:51:27 <TrueBrain> SmatZ: well, fixing it is very simple, as the + only appears at the end of the line
15:51:34 <TrueBrain> so a minor modification to the code makes current scenarios work just fine
15:54:02 <TrueBrain> this is exactly what I have been worrying about for a while .. this clearly is a critical bug, but there will be less obvious bugs .. we can just fix them, but that might not always be the best thing
15:54:33 <TrueBrain> for example, as MrFlibble already tried to convince of us, the AI tends to build over your units to replace a building. Of course this is easy 'fixed', but the reason it does that, is to keep the game a tiny bit more fair
15:54:38 <TrueBrain> so fixing that would be really bad ..
15:55:12 <Xaroth> at some point it should be fixed, but only when the AI can do without it
15:55:22 <TrueBrain> which is never with the current AI
15:55:44 <TrueBrain> but where do you balance such fixes? And how much do you stay at the original?
15:56:00 <TrueBrain> what bug is critical enough to fix it .. like this '+'. It will change the game with a big factor
15:56:10 <Xaroth> if it was clearly part of the original design (as you can see by the fact the + are alreadyt there, but not used), it should be fixed
15:56:11 <TrueBrain> now you know if you have 2 waves of enemies, you can remove all your units from your base
15:56:22 <Xaroth> if it's been obviously placed there with a reason, it should be decided upon
15:56:29 <TrueBrain> well, I have seen a few entries in Dune2 already, where they did such thing for a reason
15:56:42 <TrueBrain> for example, the hp of a building in the scenario data is ignored
15:56:45 <TrueBrain> all buildings are always given 100%
15:56:53 <TrueBrain> is this intentional? Or is it too a bug?
15:57:01 <Xaroth> the code should be able to tell us that
15:57:23 <TrueBrain> in this last case, the variable is ignroed, and the building-info hp is loaded
15:57:40 <Xaroth> maybe it was planned, but never put into release
15:57:48 <TrueBrain> in the case of this '+', maybe someone did a quick-fix to make it non-working
15:58:01 <TrueBrain> for building-hp, all code is there; it is just not enabled
15:58:13 <TrueBrain> (the variable is ignroed, and another is used)
16:01:28 <Xaroth> then maybe we should have a good look at it
16:03:22 <TrueBrain> so I was thinking if we just make a global variable: _fix_dune2_bugs :p
16:03:42 <Xaroth> I was thinking of something similar
16:03:49 <TrueBrain> but SmatZ just said: no
16:04:06 <TrueBrain> and he is right .. how many peopple in TTDp play without mamoet active?
16:04:08 <Xaroth> i was thinking of defined constants to enable/disable it
16:04:39 <TrueBrain> or who would want to play OpenTTD where long tunnels gives you tons of money
16:04:55 <Xaroth> so not by patch features, but by compile options
16:05:34 <TrueBrain> dunno what is better .. in MP you need to check options anyway :p
16:05:52 <Xaroth> yeh, but we need to decide what is standard before we put in MP anyhow
16:06:17 <TrueBrain> but on the other side, it clearly is a bug, and so fuck original Dune2, we should fix it ..
16:06:22 <Xaroth> but MP will probably be a release or three away from now, and it's harder to re-trace all those bugs, than it is to permanently fix those switches
16:06:22 <TrueBrain> same with structure hp ..
16:07:00 <Xaroth> but seriously, if you lose due to reinforcement drops, you're doing something wrong ! :P
16:07:15 <TrueBrain> true, but this is just one example of many to follow
16:07:21 <TrueBrain> we need to create some guideline in this
16:09:08 <TrueBrain> and my problem is exactly this: on one hand I love the original game, including his bugs and trouble and stupidness (like the pathfinder)
16:09:19 <TrueBrain> on the other hand I want to make Dune2 better, more fun, ...
16:10:25 <TrueBrain> so I guess the question is here: do we want to offer people the same gameplay as Dune2 did, or do we want to give them our version of it only? Or let them decide (runtime)?
16:11:02 <TrueBrain> OpenTTD for example is following something in between all that :p Sometimes they just remove things, or fix things, or what ever. Other times they make switches for it. Not always in the most consistent way ;)
16:11:33 <SmatZ> if there wasn't TTDP, OTTD would have less switches
16:11:34 <TrueBrain> on the other hand, when we started we said: OpenDUNE will be 100% a dune2 copy/clone, with the option to make it better
16:11:40 <TrueBrain> SmatZ: I think you are right
16:11:55 <TrueBrain> SmatZ: I am for example clueless why the old pathfinders are still there; like anyone uses them?
16:12:27 <TrueBrain> also only 1 variant
16:12:32 <TrueBrain> at some point there were 4 types of PF
16:12:37 <SmatZ> but yes, in many scenarios NPF is better, and in many YAPF is better
16:13:24 <SmatZ> I was wondering the same as you ;)
16:13:46 <SmatZ> even removal of NPF was considered, but nobody did anything
16:14:01 <SmatZ> maybe because of lowlevel platforms that can't handle YAPF
16:14:12 <TrueBrain> so I really dislike switches ... what I had in mind when I started this project, is that after we made Dune2 readable in C, to make a framework, put Dune2 on it with all his bugs and problems, and make another 'mod' which is better and improved version of it
16:15:10 <TrueBrain> (and in both, no switches)
16:15:11 <Xaroth> TrueBrain: you do like your 'werkverschaffing' :P
16:15:31 <TrueBrain> if I would care about how much time it would take, I wouldn't have started this project in the first place :)
16:17:53 <TrueBrain> so I guess the easiest for now is to leave all the bugs and shit
16:17:57 <TrueBrain> and decide later? :)
16:18:12 <Xaroth> would be useful to mark them though
16:18:25 <TrueBrain> but no Xaroth, it is no solution :p As in 2 months, the question is still the same :)
16:18:48 <Xaroth> fully aware of that :)
16:19:04 <Xaroth> as you said, about caring about how much time etc :P
16:19:09 <Xaroth> all the time in the world :)
16:23:17 <DorpsGek> SVN: truebrain (r450) -Add: figured out the reinforcement struct in Scenario
16:23:26 <DorpsGek> SVN: truebrain (r451) -Add: figured out 2 more function names
16:28:03 <TrueBrain> @calc 0x97E7 - 0x8DE3
16:30:09 <Xaroth> TrueBrain: did you see the post fib made wrt icon?
16:31:02 <Xaroth> not sure if it's 'his creation' or a copy
16:37:51 <TrueBrain> @calc 999 * 2 / 1000
16:38:25 <TrueBrain> @calc 999 * 2 / 500
16:46:27 <TrueBrain> by the looks, a copy
16:48:15 <TrueBrain> testing my new scenario code ..
16:50:44 <TrueBrain> worm ate my harvester
16:51:20 <DorpsGek> SVN: truebrain (r452) -Add: found one more variable (an array)
16:52:39 <DorpsGek> SVN: truebrain (r453) -Add: first version of Scenario_Load (still lots to do)
17:19:19 <Xaroth> <@TrueBrain> worm ate my harvester << I see a trend :P
18:01:01 <SmatZ> I found directory with many my dune2 savegames
18:01:09 <SmatZ> from all levels of A/H/O
18:01:21 <SmatZ> hmm no, some ordos levels are missing
18:01:26 <SmatZ> can it be useful to you?
18:01:43 <Xaroth> i think TrueBrain made a dir with something similar
18:06:38 <TrueBrain> and of course it can be useful, even more if they are made with the original dune2 :)
18:18:42 <Xaroth> thought I could work on the plugin for mantis, but then realised i had all the info i needed on my real pc...
18:34:40 <TrueBrain> impressive SmatZ :)
18:35:38 <SmatZ> TrueBrain: are you decompiling version 1.07 or some older?
18:39:15 <TrueBrain> mostly because 1.0 doesn't work :p Ghehe :)
18:40:15 <TrueBrain> something related to the ticker
18:40:22 <TrueBrain> dunno what .. couldn't care much, as 1.07 worked :p
19:01:53 <SmatZ> what is that about? is going 160km/h that bad? :-x
19:02:10 <SmatZ> so there is a document about that :)
19:02:49 <TrueBrain> when you drive 50 km/h over the speed limit, you have to turn in your driver license
19:02:58 <TrueBrain> then this person totally flips .. and TOTALLY flips
19:03:11 <TrueBrain> going on a rampage it is about his origins and shit
19:03:24 <SmatZ> well if the limit wasn't 130 :)
19:03:51 <TrueBrain> the most fun part is that he says he didn't drive that fast
19:03:57 <TrueBrain> over 2 km they followed him with 160+ km/h
19:04:16 <TrueBrain> I love how stupid people can be :)
19:04:21 <SmatZ> police going 160km/h without any lights
19:04:39 <SmatZ> then either the limit is too low there, or they were behaving dangerously :-p
19:04:43 <TrueBrain> here emergency services don't have to carry lights nor sound
19:06:24 <SmatZ> still better than out 80km/h limit on highways inside city limits, no matter how low traffic is and how far from actual city it is :-p
20:19:51 <TrueBrain> /* 6E88(145) */ PACK char string_6E88[145]; /*!< " etainosrlhcdupmtasio wb rnsdalmh ieorasnrtlc synstcloer dtgesionr ufmsw tep.icae "
20:19:53 <TrueBrain> * "oiadur laeiyodeia otruetoakhlr eiu,.oansrctlaileoiratpeaoip bm" NULL terminated.
20:19:54 <TrueBrain> * Used to uncompress lang files (.ENG, ...). Also accessed via 0x6E98. */
20:19:57 <TrueBrain> glx: were you drunk? :p
20:20:26 <glx> no, it uses this to decompress lang files
20:20:56 <TrueBrain> decompress btw, not uncompress ;)
20:21:03 <TrueBrain> but it is very funny to read :)
20:21:50 <TrueBrain> ./*!< "\nInsert disk %s containing '%s' into any drive." NULL terminated. */
20:21:58 <TrueBrain> it always gives me the need to insert the disk in anohter computer
20:22:01 <TrueBrain> and call the helpdesk complaining :p
20:24:43 <TrueBrain> I Wonder why there is that unorderd list of dune0 .. dune20 in strings
20:27:05 <glx> but no driver, so no code
20:27:10 <TrueBrain> I meant more that: unordered part ;)
20:28:14 <glx> I don't try to understand the logic behind some stuff they did ;)
20:30:18 <SmatZ> [21:19:57] <TrueBrain> glx: were you drunk? :p <=== hehe :-)
20:30:23 <glx> like they set some global variables but never read them
20:30:37 <TrueBrain> glx: well, that can simply be in undecompiled parts
20:30:42 <TrueBrain> I have that with repeating reinforcement
20:30:56 <TrueBrain> I was wondeirng why 1 entry was never read .. but .. as it is never active, it is never read yet :)
20:31:08 <glx> but the weird thing is the buffers
20:31:24 <glx> disseminated everywhere in global memory
20:31:41 <TrueBrain> if the function is in an overlay, I can understand that (overlays can't have writable static memroy)
20:31:47 <TrueBrain> but for non-overlay pieces it is stupid
20:33:49 <glx> or global string duplicates with only one instance used
20:36:21 <TrueBrain> and now I have to wait for reinforcement! :)
20:52:10 <DorpsGek> SVN: truebrain (r454) [JIT] -Add: mapped one more function (recurring reinforcements)
20:52:18 <DorpsGek> SVN: truebrain (r455) -Update (r454): update decompiled code
20:55:19 <TrueBrain> also immediatly shows where reinforcements are brought in ;)
21:00:18 <DorpsGek> SVN: glx (r456) -Add: figured out a few variables
21:16:39 <Xaroth> iRobot is funny as hell
21:16:48 <Xaroth> U.S.Robotics... I had routers made by them o-O
21:28:22 <TrueBrain> would Nyerguds know how fucking annoying he is?
21:30:09 <Xaroth> either that or he doesn't care, at all
21:43:41 <TrueBrain> bah, tricky to make Scenario_Load even more readable, because subfunctions require 16bit memory addressing ...
21:48:09 <Xaroth> this'll be interesting.. trying to svn co opendune on my laptop...
21:48:16 <Xaroth> not got msvc.. but still :P
21:59:39 <DorpsGek> SVN: truebrain (r457) -Change: renamed 'building' to 'structure'
22:14:25 <TrueBrain> hmmm .... I am now looking at a few functions, and I get very weird strings for structure names :p
22:14:52 <TrueBrain> Structure_StringToId
22:15:16 <Xaroth> which file is that in, for those without the joys of grep
22:16:53 <TrueBrain> oh, that is more because there is no UTF-8 support
22:17:19 <TrueBrain> it reads one line too many
22:18:54 <TrueBrain> dunno from where exactly
22:21:16 <SmatZ> [23:16:47] <Xaroth> 65536 == 0xFFFF? 65536 == 0x10000
22:21:28 <Xaroth> er, good point SmatZ :P
22:25:18 <TrueBrain> k .. something is very fishy .. either with this routine, or my conversion
22:28:55 <TrueBrain> when I start a mission with a completely clean (no src/*) version, things are VERY fishy
22:30:21 <TrueBrain> no ... it is just a bit weird
22:30:24 <TrueBrain> and also a bit wrong
22:30:27 <TrueBrain> but I like to say weird
22:30:31 <TrueBrain> but you know .. also a bit wrong
22:31:49 <TrueBrain> somehow something is corrupted in my clean version ..
22:34:35 <TrueBrain> graphics are all weird
22:35:15 <glx> there is garbage at the end of some scenarios
22:35:50 <TrueBrain> this just doesn't make sense ... why would a src/* less copy give me weird graphics ..
22:35:54 <TrueBrain> the gameplay and stuff seems in order
22:37:28 <SmatZ> maybe you are overworked today :)
22:37:41 <SmatZ> when things like that happen to me, it's best time to go to bed ;)
22:37:55 <Xaroth> you got a good point, SmatZ
22:37:56 <TrueBrain> can you guys check on a clean checkout if a newgame looks okay?
22:39:14 <SmatZ> TrueBrain: still doesn't work on amd64,right?
22:39:27 <TrueBrain> which you don't need
22:39:58 <SmatZ> it's giving me some link errors
22:42:06 <TrueBrain> it made the CY 3x3 instead of 2x2 :s
22:42:43 <SmatZ> TrueBrain: oh great, I had problem with linking 32bit libemu to 64bit opendune ;)
22:43:04 <SmatZ> TrueBrain: seems to work :)
22:45:09 <TrueBrain> and the svn trunk doesn't have them?
22:45:43 <TrueBrain> paint me yellow and call me a banana
22:46:32 <SmatZ> hg shows strange behaviour
22:46:45 <SmatZ> I have rafinery around my construction yard
22:46:53 <TrueBrain> yes, it thinks it is 3x3
22:47:15 <SmatZ> so that's the only difference
22:47:29 <TrueBrain> a decompiler is not THAT selective
22:47:57 <TrueBrain> well, how can 2 identical versions onlyhave THAT difference
22:47:59 <SmatZ> and stupid gimp fails to start, AGAIN! grrrr
22:48:01 <TrueBrain> the rest seems to work perfectly
22:49:13 <TrueBrain> even if I load the trunk src files, the difference is still there
22:49:35 <TrueBrain> - emu_ax = emu_get_memory16(emu_ds, emu_bx, 0x2D52);
22:49:36 <TrueBrain> + emu_ax = emu_get_memory16(emu_ds, emu_bx, 0x2D40);
22:52:29 <TrueBrain> order of files .. how cute
22:52:37 <TrueBrain> Xaroth: you were right, JIT corruption
22:53:03 <glx> though both address are in unknown_2C5E
22:54:08 <TrueBrain> lucky I have a fairly good idea which of the two is corrupted
22:54:14 <TrueBrain> mostly as one shows weird stuff, the other not :p
22:54:29 <TrueBrain> okay, this really was very weird :)
22:57:54 <SmatZ> valgrind ./opendune seems to hang :(
22:58:08 <TrueBrain> valgrind can't handle it :p
22:58:13 <DorpsGek> SVN: truebrain (r458) [JIT] -Remove: 3 stat-files which are corrupted files
22:59:58 <TrueBrain> okay, it seems that I introduced the weirdness of scenario loading :(
22:59:58 <SmatZ> would reducing timer (IRQ9) frequency help?
23:00:11 <TrueBrain> _pic_speed to higher value
23:01:26 <TrueBrain> hmmm ... Dune2 also does something weird ..
23:01:31 <TrueBrain> can't put my finger on it, but it is not right
23:02:31 <TrueBrain> well, it seems to repeat the last line a few times till it happens to hit a \0
23:04:16 <TrueBrain> but in the 'clean' version it happens in a slightly different way
23:19:43 <DorpsGek> SVN: truebrain (r459) -Fix (r453): nasty mistake in a copy/paste
23:40:49 <TrueBrain> glx: does 'strcasecmp' exist on MSVC?
23:43:50 <DorpsGek> SVN: truebrain (r460) -Change: move HouseInfo to house.h, despite the fact it is located in g_global. Avoid using the g_global access by renaming it to 'donotuse_houseInfo'
23:43:50 <DorpsGek> SVN: truebrain (r460) -Add: C-ified (House|Structure|Unit)_StringToType
23:43:56 <TrueBrain> k, then r460 breaks :)
23:44:33 <TrueBrain> euh .. would a include/function.h make sense?
23:44:41 <TrueBrain> or functions, to keep it the same as types
23:45:00 <TrueBrain> and glx, is it defined in strings.h?
23:45:38 <TrueBrain> or how would you suggest to fix the upcoming mess of MSVC naming vs POSIX naming?
23:49:10 <glx> stricmp is declared too, but deprecated
23:49:40 <TrueBrain> strcasecmp is in strings.h (plural)
23:50:15 <glx> implicit declaration of strcasecmp with mingw
23:50:24 <TrueBrain> so ... what shall we do ..
23:50:30 <TrueBrain> src/os/strcasecmp.h
23:50:35 <TrueBrain> which takes care of it for all OSes?
23:50:39 <TrueBrain> or src/core/strcasecmp.h
23:50:40 <glx> but links without problem :)
23:50:52 <TrueBrain> or src/core/strings.h
23:53:08 <glx> weird, strings.h is a wrapper aroung string.h and strcasecmp is in string.h
23:53:09 <TrueBrain> glx: any opinions/ideas?
23:56:12 <TrueBrain> as I currently really care how mingw does stuff .. we know we will hit this problem with tons of functions, so we need some kind of common ground in some kind of structure
23:56:22 <glx> ok it's not declared because -ansi :)
23:58:02 <TrueBrain> so yes/no? Ideas? no?
23:58:07 <TrueBrain> well .. then we will see tmorrow :)
23:58:12 *** Guest677 has joined #openDune
23:59:29 <glx> well dunno if we should have an header for each function
continue to next day ⏵
