IRC logs for #opendune on OFTC at 2009-12-08
⏴ go to previous day
00:01:40 <DorpsGek> SVN: truebrain (r696) [JIT] -Add: mapped another 26 functions (dune.log)
00:01:55 <DorpsGek> SVN: truebrain (r697) -Update (r696): update decompiled code
00:02:05 <TrueBrain> and I am still fucking annoyed by my ex-labpartner
00:02:28 <TrueBrain> he just likes to flame, because he feels attacked
00:02:41 <TrueBrain> and he tried to push everything on me, instead of just recognizing he was in error
00:03:06 <Xaroth> heh, sounds like one of my colleagues
00:03:33 <TrueBrain> yeah; it is hard to keep standing above that
00:03:37 <TrueBrain> and not go with the flaming
00:03:43 <TrueBrain> still let the right persons know that you are not to blame
00:03:45 <Xaroth> why do you think I got his phone ringer on remote control? :)
00:04:17 <Xaroth> my boss knows that I take responsibility if I fucked up, so by now he can look at me and know if it was my doing, or his doing
00:04:37 <Xaroth> people who tend to shove off their faults on others usually try to look too perfect
00:05:07 <glx> TrueBrain: didn't you forget to add a file ?
00:05:09 <Xaroth> by taking responsibility on a (small, ofcourse) error, you can point the finger back at him, without looking like an arse, or degrading yourself to his level
00:05:31 <DorpsGek> SVN: truebrain (r698) -Fix (r697): I forgot to add a file
00:06:04 <TrueBrain> Xaroth: well, I took a much easier approach: "I will refrain myself from defending those accusations, and will leave it by the fact that cooperation is no longer possible"
00:06:26 <Xaroth> but I usually leave that until after my idea :P
00:06:36 <TrueBrain> glx: I think dune.log is a 'demo' thingy
00:06:40 <Xaroth> since with the first you build credibility, and with the second you force them to pick sides
00:07:13 <TrueBrain> all I need to do is make sure the teacher knows what is going on
00:07:16 <TrueBrain> I don't care about the rest
00:07:28 <Xaroth> I have a career to keep in mind ^^
00:08:05 <DorpsGek> SVN: glx (r699) -Fix (r698): forgot MSVC project ;)
00:08:06 <Xaroth> (and a pay raise, hopefully at the end of the month)
00:08:15 <TrueBrain> forgot more I think
00:08:36 <DorpsGek> SVN: truebrain (r700) -Fix (r698): also forgot svn propset
00:08:56 <Xaroth> heh, i keep updating like this :P
00:10:13 <TrueBrain> 4 revisions to do 1 revision of work
00:11:35 <Xaroth> f__1FB5_0005_000F_62CC increases global 0x986C, calls emu_Drive_Get_Default_Wrapper, then decreases it again
00:14:15 <Xaroth> happening quite frequently in cs__1FB5
00:14:46 <TrueBrain> it is pretty normal :)
00:14:50 <TrueBrain> there is minor 'threading'
00:16:07 <Xaroth> 0x6C91 is read as 16bit
00:17:16 <Xaroth> f__2598_0000_0017_EB80 swaps global->variable_6C91 with arg1, and returns the old value of global->variable6C91
00:17:23 <Xaroth> ... with correct spelling and stuff
00:17:38 <TrueBrain> a lot of core functions tend to do that
00:17:44 <TrueBrain> the idea is that you store the old value
00:17:46 <TrueBrain> so you can reset it later on
00:18:10 <TrueBrain> glx: make sure to check the 379A for other values then 0 and 1
00:18:19 <Xaroth> yeh, i got interested by the 108 different Called From: entries
00:18:25 <Xaroth> for something so simple
00:18:42 <TrueBrain> that is about layering and simplicity :)
00:19:56 <TrueBrain> yes, dune.log seems to be the 'record' file for the demo
00:20:49 <glx> and this stupid anticopy keeps asking me for speeds
00:22:11 <Xaroth> f__257A_000D_001A_3B75 does some interesting things :o
00:22:13 <glx> well 0x379A only tests for == 0x0
00:22:43 <TrueBrain> I was reading things on dune.log
00:22:54 <TrueBrain> there should be a variable which switches it to playback
00:23:07 <glx> ok 3 questions, 3 speeds :/
00:23:32 <Xaroth> so basically, there's a built-in replay record feature? :P
00:23:49 <TrueBrain> from what I read: yes
00:25:45 <glx> hmm emu_push(emu_get_memory16(emu_ds, 0x00, 0x379A));
00:25:59 <glx> maybe there are different values :)
00:26:20 <TrueBrain> but 2 should work too I guess
00:26:36 <glx> f__257A_000D_001A_3B75 <-- yes 0, 1 or 2
00:27:29 <Xaroth> anyways, off to bed, before it really gets half past two :P
00:28:22 <TrueBrain> ha, the demo recording things works nice :)
00:28:41 <glx> could be used to reproduce crashes :)
00:29:06 <TrueBrain> it just doesn't seem to work flly :p
00:29:16 <TrueBrain> playback differs :p
00:33:36 <TrueBrain> no, you need to load the EXACT same map :p
00:35:51 <DorpsGek> SVN: truebrain (r701) [JIT] -Add: mapped another 28 functions (dune.log playback)
00:36:07 <DorpsGek> SVN: truebrain (r702) -Update (r701): update decompiled code
00:41:17 <glx> it's fun to watch the replay :)
00:42:00 <TrueBrain> and it worksin the way I suggested for OpenTTD
00:42:11 <TrueBrain> record mouse movement and commands, and you will be fine
00:42:23 <TrueBrain> (instead of the much more complex methods I have seen in the years :p)
00:42:28 <glx> for openttd the problem is resolution
00:42:41 <TrueBrain> I resolved that very simple, in one of my demo patches :)
00:43:16 <TrueBrain> (relativate everything to a 640x400 resolution :p)
00:45:50 <glx> and it closes once finished :)
00:52:09 <glx> indeed the replay is not exactly the same
00:52:27 <TrueBrain> but it does record 'load savegame' and stuff :p
00:52:45 <glx> yes I just recorded A1 playing :)
00:53:14 <glx> but 1 of the ordos unit is killed differently each time :)
00:53:57 <glx> oh and in this replay the total killed is 4
00:54:04 <glx> it was 3 when I recorded it
00:54:30 <glx> the game is not deterministic ;)
00:57:46 <TrueBrain> this time for real ;)
00:57:52 <TrueBrain> for now: good night :)
09:29:42 <TrueBrain> I am still so fucking annoyed by my labpartner
09:29:48 <TrueBrain> fucking hell ... what an ass ....
09:30:52 <TrueBrain> My complain: you are never on time, and you don't keep your promises. His comeback: I think this is unfair, I haven't complained about your low participation, in this and this and this instance, have I? .... eeuuh ... I think ... you missed my point .. by like ... a mile?
09:31:18 <TrueBrain> participation != being on time ...
09:31:32 <TrueBrain> Xaroth: if your college is anything remotely like him, I feel bad for you
09:32:27 <Xaroth> I.. don't do college :)
09:32:51 <Xaroth> he does try to counter in the most odd form
09:33:01 <Xaroth> and as such, missing the point by miles.
09:33:43 <TrueBrain> sometimes I don't get people ... they are intelligent beings, are they? My sister always made me laugh: I don't like what you just said to me, but I don't want to talk about it
09:34:09 <TrueBrain> one way to hold off any discusion :p
09:34:59 <TrueBrain> she in fact is a genious in it :) "It sucks that I can't borrow the car, I think it is really stupid and I think I should be able to borrow it. But I don't want to talk about it if you keep on saying I can't"
09:35:20 <TrueBrain> I guess at least you know where you are at :p
09:40:05 <TrueBrain> either way .... how were the Team functions going Xaroth? :)
09:40:46 <Xaroth> cleaned it up a bit, needs testing still
09:41:47 <TrueBrain> testing is a bit tricky in this case ;)
09:42:06 <Xaroth> well, running a full mission or two without crashing should indicate it's not fucked up :)
09:42:14 <TrueBrain> and see team attacks :p
09:42:38 <Xaroth> the scenario fix sorted that, right
09:42:58 <TrueBrain> even the old ones has team attacks, but only by foot folks :p
09:57:05 <TrueBrain> BigBangTheory is VERY good, the last episode :)
10:03:36 <TrueBrain> k .. time to go to the uni, see if I can somehow savage this project, or that it is lost ...
10:04:48 <TrueBrain> ah; and it wasn't intended :p
12:01:51 <TrueBrain> I love to talk with people who understand you
12:38:22 <Xaroth> TrueBrain: what kind of dns package do you use? :o
12:38:36 <TrueBrain> server? client? relay?
12:39:04 <Xaroth> I'm getting annoyed by the staticness of bind
12:39:14 <TrueBrain> Our frontends are PowerDNS
12:39:50 <Xaroth> that's what I was aiming at
12:39:53 <TrueBrain> and the secondary over SQLite
12:41:14 <Xaroth> I looked at mydns, as I had fiddled with it on our ispconfig test rig.. but it looks a tad limited
12:41:34 <TrueBrain> PowerDNS has shown itself to be very stable under very heavy loads
12:41:39 <TrueBrain> and doesn't have the bugs/problems bind has
12:41:50 <TrueBrain> no other DNS server software comes close to it, as far as I know
12:41:58 <TrueBrain> PowerDNS doesn't have cache polution
12:42:05 <Xaroth> going to suggest setting up a test case for bind then :)
12:42:25 <TrueBrain> our backends run a mix of software
12:42:33 <TrueBrain> bind, custom software, ...
12:42:40 <TrueBrain> but they all publish (AFXR) to the frontends
12:50:15 <Xaroth> now to find a managing tool/frontend/whatever for it
12:50:40 <Xaroth> I somehow think that out of 4 people in this room, 2 know how to manually run sql queries :P
12:53:50 <TrueBrain> we use phpmyadmin :p haha :)
12:53:52 <TrueBrain> there is a tool for it
12:54:05 <TrueBrain> but okay, we here use other tools around bind
12:54:08 <TrueBrain> which publish to PowerDNS :p
12:54:41 <Xaroth> well the thing is, you can make really nice tools around bind
12:55:00 <Xaroth> but that doesn't take away that some of the design features of bind, are flawed
12:57:03 <Xaroth> not that bind doesn't work, i mean, it's a good performing piece of kit, but i'd rather have top notch with all the features we need, than the #2 with limited stuff
12:58:24 <Xaroth> for a half working frontend, that works MUCH better than most of the bind frontends i've tried :P
12:59:27 <Xaroth> (account test // test )
13:00:02 <TrueBrain> writing something for PowerDNS will also be very easy, so ..
13:01:31 <Xaroth> easier than for bind ^^
13:27:39 <TrueBrain> so what will we be doing today ...
13:31:29 <Xaroth> fiddling with powerdns :P
13:35:06 <glx> I'm looking at f__2502_00FA_0009_9EB2() :)
13:39:14 <TrueBrain> I am trying random 1A34 functions
13:39:33 <TrueBrain> a weird one .. a sandworm which requests a structure by index ... weird ...
13:46:29 <glx> ok time to add 3 new files I guess, strings.h, strings.c and emu_strings.c :)
13:47:26 <glx> String_GetString and related
13:47:43 <TrueBrain> euuh ... string.c, I guess?
13:48:00 <glx> more consistent indeed :)
13:48:01 <TrueBrain> String_Get_ByIndex is btw a better name I guess :p
13:48:30 <glx> true but I'll start with String_Decompress :)
14:00:28 <Xaroth> trying netinst cd for debian lenny
14:00:37 <Xaroth> can't find the vmware virtual hd?
14:00:59 <TrueBrain> never had that problem
14:01:05 <TrueBrain> both in SATA as IDE mode
14:13:43 <DorpsGek> SVN: truebrain (r703) -Add: the understanding of one more flag
14:14:24 <DorpsGek> SVN: truebrain (r704) -Add: named one more function
14:24:11 <Xaroth> maybe a bare bone install is a bit TOO barebone
14:41:47 <TrueBrain> apt-get install less
14:45:12 <glx> hmm dune2 uses a strange encoding for accents
14:50:31 <glx> another smart solution :)
14:50:55 <glx> even if that means custom encoding
14:54:09 <Xaroth> ugh, setting up ldap is much easier on ubuntu :P
14:54:23 <Xaroth> manually editing pam files *shudder*
14:54:34 <Xaroth> and even still, errors -_-
14:54:39 <TrueBrain> install the right ldap package :p
14:54:55 <TrueBrain> there are 2 .. one sucks, other is very good
14:55:30 <Xaroth> on ubuntu it's libnss-ldap , there's a built-in auth-client-something that does the pam editing for you.
14:55:36 <Xaroth> worst part is, getent works
15:12:44 <Xaroth> i just spent 20 minutes trying to trace down a bug
15:12:59 <Xaroth> when it was another service causing it
15:24:23 <glx> TrueBrain: when you name a function, do it for all references ;)
15:24:46 <glx> or compile before commit :p
15:24:58 <TrueBrain> sorry, decompiler doesn't do src/ :p
15:25:01 <TrueBrain> can you fix that? :)
15:26:51 <DorpsGek> SVN: glx (r705) -Fix (r704): forgot 2 references
15:27:35 <DorpsGek> SVN: glx (r706) -Add: named and C-ified String_Decompress()
16:48:23 <glx> oh the stuff set by f__01F7_20B5_0076_64ED() is used indeed
16:48:54 <glx> it stores command line somewhere, then emu_GameLoop_Main() parses it
16:49:23 <glx> of course it's unused in emulated version :)
17:15:37 <DorpsGek> SVN: glx (r707) -Add: figured out a few variables
17:55:52 <Xaroth> g_global->string_31D8 + g_global->language * 4 << string + (language * 4) ?
17:56:26 <Xaroth> oh wait, nevermind, you can do tricks like that in C :P
18:47:50 <TrueBrain> glx: only comment I have that '
18:47:58 <TrueBrain> glx: only comment I have that '= 0' might be better to be '= 0x0'
18:48:10 <TrueBrain> as we talk pointers .. somehow I like that :p
18:48:16 <TrueBrain> at least other files are like that
18:48:31 <TrueBrain> and there is a function to get the IP of a g_global entry
18:48:38 <TrueBrain> to remove the 0x8282 :p
19:00:47 <glx> refreshed again, I decided to use snprintf
19:06:50 <DorpsGek> SVN: glx (r708) -Add: C-ified String_GenerateFilename()
19:07:33 *** Guest773 has joined #openDune
20:48:08 *** Guest773 is now known as planetmaker
20:48:45 *** planetmaker is now known as Guest782
20:51:02 *** Guest782 is now known as planetmaker
21:35:41 <Xaroth> ooo LoL store going to be opened tonight :o
21:52:45 <Xaroth> I take it you don't mind me to swap arguments around a bit when c-ifying
21:54:48 <Xaroth> I find it more meaningful that if you AddUnitToTeam, you specify the unit you want to add first, then the team to add it to
22:02:26 <Xaroth> TrueBrain: What is the correct way if a pointer is actually pointing to something..
22:03:19 <Xaroth> or is there no sanity checks required in the team functions
22:07:20 *** Xaroth sets mode: +vvvv tneo SmatZ planetmaker nsz
22:09:43 <glx> usually I don't put sanity check when dune2 doesn't
23:16:33 <TrueBrain> Xaroth: the only 'sanity' check you need to do is != NULL
23:16:46 <TrueBrain> we assume the rest of the system is sane
23:17:09 <TrueBrain> Xaroth: and the general thumb-rule is that your first parameter is what it is about
23:17:16 <TrueBrain> so having Team first makes much more sense
23:17:20 <TrueBrain> to a team, you add a unit
23:17:24 <TrueBrain> hence: Team_AddUnitToTeam
23:17:32 <TrueBrain> you don't add a team to a unit
23:17:44 <TrueBrain> parameters should be like a book, which you read from left to right
23:18:01 <TrueBrain> Unit_AppendToTeam, in that case 'Unit' would come first
23:18:27 <TrueBrain> and that is something one could debate, to name the functions: Unit_AddToTeam and Unit_RemoveFromTeam
23:18:36 <TrueBrain> but if you name it Team_AddUnitToTeam, Team should come first
23:25:57 <Xaroth> I'll stick with NULL checks then
23:26:18 <glx> these functions look more like Unit_AppendToTeam(unit, team) and Unit_RemoveFromTeam(unit)
23:26:19 <TrueBrain> well, what else? I mean .. you can't validate if something is a Unit * :p
23:26:27 <TrueBrain> glx: that is why I suggested it ;)
23:26:49 <TrueBrain> lets leave it up to Xaroth if he agrees with our perspective on that :)
23:26:58 <Xaroth> glx: hence why I was confused yesterday at the order of the args
23:28:12 <TrueBrain> the original code really had a name like Team_AddUnitToTeam
23:28:20 <TrueBrain> one can just debate if that name would have made sense :p
23:28:59 <TrueBrain> glx: do you think there is any way to remove the fact it now reads either intro.eng or dune.eng?
23:29:15 <TrueBrain> I mean, not now, but that somehow you have an extra parameter to Get_ByIndex which is set to a value which indicates that?
23:29:17 <Xaroth> glx: I'm tempted to give that one a tl;dr
23:29:28 <TrueBrain> which is -1 in the emu_ for now
23:29:35 <TrueBrain> and when converting, you have to pick the right one or something?
23:30:16 <Xaroth> hm, I agree on the Unit_AddToTeam bit
23:30:41 <glx> ha right I forgot a space againe
23:30:54 <TrueBrain> don't forge tto remove the function from function_names.txt
23:31:06 <TrueBrain> spacing in emu_String_GetFromBuffer_ByIndex is off
23:31:16 <glx> right, and I need to add some comments too
23:31:49 <TrueBrain> nevermind the last comment :p
23:31:54 <TrueBrain> I am just being blind
23:32:19 <TrueBrain> 'maybe' cast String_GetFromBuffer_ByIndex(buffer, index) - buffer; to 'uint8'? Not sure ...
23:33:05 <Xaroth> msvC#e has a mass rename quickoption if you rename a function
23:33:10 <glx> or uint16 (uint8 is definitely too small)
23:33:28 <TrueBrain> instead of the 'char *'
23:33:36 <Xaroth> yeh, that's the normal f&r
23:33:48 <Xaroth> the .net one was intellisense-powered
23:34:04 <TrueBrain> glx: any ideas on making String_Get_ByIndex more robust for the future? Like that we can give it STR_ blabla shit?
23:34:12 <TrueBrain> or shall we do that some time later?
23:34:33 <glx> something for later I think, because for now it still uses global buffers
23:34:54 <TrueBrain> well, the non-emu in fact don't directly ;)
23:35:03 <TrueBrain> but using non-emus is very hard with most of the other shit not yet translated :p
23:35:20 <Xaroth> TrueBrain: the @name should reflect the function name right? (aka, not used by decompiler)
23:35:21 <TrueBrain> clever pointer solutions
23:35:36 <TrueBrain> Xaroth: @name should be the name used, yes, for the emu_ variant, which should always be a void ()
23:35:36 <glx> was that or returning csip
23:35:43 <TrueBrain> Decompiler reads it, and uses it for the @implements
23:35:49 <glx> but for future it makes sense to return char*
23:35:55 <TrueBrain> glx: yes, it should return char *
23:35:59 <TrueBrain> and I like what you have done with the rest
23:36:06 <TrueBrain> csip.s.ip += String_Get_ByIndex(index) - emu_get_memorycsip(csip);
23:36:08 <TrueBrain> made my head spin ;)
23:36:50 <TrueBrain> So I give it a A-, because of the 'uint16*' mistake :p Ghehehehehe :)
23:40:15 <glx> first version was return buffer + buffer[index * 2];
23:41:26 <TrueBrain> oh well, off to bed
23:41:30 <Xaroth> syntax error : missing ')' before identifier 't'
23:41:42 <glx> btw string file format is smart
23:42:10 <glx> but you already knew that ;)
23:42:46 <Xaroth> extern uint16 Unit_AddToTeam(Unit *u, Team *t);
23:42:55 <Xaroth> somehow the Team *t part is wrong? :o
23:44:24 <glx> better not add another include just for 1 thing
23:45:55 <Xaroth> ok, now you confused me :o
23:47:07 <Xaroth> that doesn't make it stop bitching about it tho
23:47:19 <Xaroth> need to edit externals
23:48:07 <glx> most Structure_* functions need to know House, but we don't include house.h
23:48:32 <Xaroth> ..\src\unit.c(666) << diabolical
23:54:09 <Xaroth> bound to have a fair few coding style issues, but at least it's a start
continue to next day ⏵