#exult@irc.freenode.net logs for 6 Feb 2003 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage

[00:18:36] <-- DarkeZzz has left IRC (adams.freenode.net irc.freenode.net)
[00:18:36] <-- jerbear has left IRC (adams.freenode.net irc.freenode.net)
[00:18:36] <-- royalsexy has left IRC (adams.freenode.net irc.freenode.net)
[00:18:36] <-- Servus has left IRC (adams.freenode.net irc.freenode.net)
[00:18:36] <-- Matt_O has left IRC (adams.freenode.net irc.freenode.net)
[00:18:36] <-- Yuv422 has left IRC (adams.freenode.net irc.freenode.net)
[00:18:36] <-- Dark-Star has left IRC (adams.freenode.net irc.freenode.net)
[00:18:36] <-- jer|w0rk has left IRC (adams.freenode.net irc.freenode.net)
[00:18:54] --> Yuv422 has joined #exult
[00:18:54] --> DarkeZzz has joined #exult
[00:18:54] --> Servus has joined #exult
[00:18:54] --> Dark-Star has joined #exult
[00:18:54] --> jerbear has joined #exult
[00:18:54] --> Matt_O has joined #exult
[00:18:54] --> royalsexy has joined #exult
[00:18:54] --> jer|w0rk has joined #exult
[00:18:54] <-- Dark-Star has left #exult ()
[00:26:55] <-- Matt_O has left IRC (adams.freenode.net irc.freenode.net)
[00:26:55] <-- Servus has left IRC (adams.freenode.net irc.freenode.net)
[00:27:16] --> Servus has joined #exult
[00:27:16] --> Matt_O has joined #exult
[00:28:00] --> Kirben has joined #exult
[00:28:00] --- ChanServ gives channel operator status to Kirben
[01:18:58] <-- Matt_O has left IRC ("Client exiting")
[01:27:24] <-- Servus has left IRC ()
[02:57:43] --> Matt_O has joined #exult
[03:13:54] <-- Kirben has left IRC ("System Meltdown")
[04:06:36] --> Cless|Away has joined #Exult
[04:07:26] --- Cless|Away is now known as Colourless
[04:07:28] --- ChanServ gives channel operator status to Colourless
[04:07:35] <Colourless> hi
[04:07:40] <Yuv422> hi colourless
[04:08:02] <Colourless> just came on to ask you, since you are not a windows coder, do you want me to do the initial porting of SDL to Dx8 ?
[04:08:31] <Yuv422> I'm having some issues with Dx8 and SDL maybe you could help me
[04:08:45] <Yuv422> How do I use the C API for dx8
[04:08:46] <Colourless> it will probably require more changes to work on xbox, but it would make it easier for you
[04:09:45] <Yuv422> The video seems straight forward enough.
[04:10:27] <Yuv422> I'm just having this problem with the fact that SDL is compiled in C and it looks like the direct X xbox libs like C++ style
[04:11:22] <Colourless> if it's anything like the windows version, you 'should' be able to use macros in c
[04:11:33] <Yuv422> is there a trick to getting dx8 to compile in plain C
[04:12:22] <Colourless> yes and no
[04:13:29] <Yuv422> :\Eric\Projects\XBox SDL\src\video\xbox\SDL_xboxvideo.c(274) : error C2037: left of 'CreateDevice' specifies undefined struct/union 'Direct3D'
[04:13:56] <Yuv422> when I try to do this
[04:14:33] <Yuv422> g_pD3D->CreateDevice(0, D3DDEVTYPE_HAL, NULL,
[04:14:35] <Yuv422> &d3dpp, /*&xbox_dx8.*/&g_pD3DDevice);
[04:15:11] <Colourless> that will not work in C
[04:15:34] <Yuv422> ok isn't it just a function pointer?
[04:15:47] <Colourless> doesn't quite work that way :-)
[04:15:56] <Yuv422> oh
[04:17:42] <Colourless> try using IDirect3D_CreateDevice(g_pD3D, 0, D3DDEVTYPE_HAL, NULL, D3DCREATE_HARDWARE_VERTEXPROCESSING, &d3dpp, /*&xbox_dx8.*/&g_pD3DDevice);
[04:18:10] <Yuv422> right. I was looking at those defs in the header. ;)
[04:18:35] <Yuv422> brb just ziping the headers
[04:19:06] <Colourless> you can also 'normally' use this way as well in c (but the code isn't portable to C++)
[04:19:30] <Colourless> g_pD3D->lpVtbl->CreateDevice(g_pD3D, ...);
[04:27:52] <Colourless> also, there 'is' 2 ways that 8 bit surfaces could be implmented btw.
[04:29:05] <Colourless> method one: create the buffer in normal memory, and save it's palette. on a buffer swap or rectupdate do all the palette look ups and copy them to the back buffer/another surface
[04:29:51] <Yuv422> do I need to provide an 8 bit surface?
[04:32:59] <Colourless> method two: make the buffer a 1024x512 texture of format D3DFMT_L8 or D3DFMT_A8. make a second 256x1 D3DFMT_A8R8G8B8 texture for the palette. Then when you need to do a buffer swap/rectupdate, draw the texture to the back buffer using direct3d and a pixel shader to do a dependant texture read. This is likely way more complex than required
[04:33:29] <Colourless> now, do you need to provide it. probably not, but it's not hard to add in support for it, since exult 'would' like it, but doesn't need it :-)
[04:34:13] <Yuv422> hmm sounds like you know yourstuff when it comes to directx.
[04:35:05] <Colourless> i know a 'bit'
[04:35:28] <Colourless> have never programmed anything substantial in d3d, but i know much of the theory behind it :-)
[04:35:52] <Colourless> the latest thing i've written in d3d is http://www.users.on.net/triforce/d3d_fsaaviewer/new9x.png which probably has no relevance to you :-)
[04:38:06] <Yuv422> I got it working with the direct function.
[04:38:19] <Yuv422> I just had to add an 8
[04:38:27] <Yuv422> IDirect3D8
[04:39:58] <Colourless> i'll probably add some in some macros for the SDL stuff to easily allow me to compile the Xbox Direct3D code as Windows Direct3D8 code since everything just needs an 8 added :-)
[04:52:52] <Yuv422> when do you think I should flip the surface?
[04:53:26] <Yuv422> does SDL only call the flip routine for double buffered vid modes?
[04:53:28] <Colourless> when ever an SDL_UpdateRects() is called
[04:53:42] <Colourless> no idea. i haven't looked into that yet
[04:54:42] <Yuv422> right
[05:00:02] <Colourless> if you want to wait a bit. i will probably work on this myself for a few hours and will get back to you
[05:00:32] <Yuv422> If you like
[05:00:38] <Yuv422> that's fine with me.
[05:01:27] <Yuv422> it's probably best to leave the event stuff for now. :)
[05:01:36] <Yuv422> just focus on the video first.
[05:02:32] <Colourless> well, for me it's easy enough since i'm on windows. the event code is already written for me :-)
[05:02:48] <Yuv422> :)
[05:04:01] <Colourless> you using sdl 1.2.5 for this?
[05:04:30] <Yuv422> yeah
[05:04:37] <Yuv422> and SDL_mixer 1.2.4
[05:19:41] <Colourless> got to go. bbl
[05:19:42] <-- Colourless has left IRC ("casts invisibility")
[06:02:02] --> Kirben has joined #exult
[06:02:02] --- ChanServ gives channel operator status to Kirben
[09:31:46] --> Servus has joined #exult
[09:46:45] <-- Servus has left IRC ()
[10:26:13] <-- Kirben has left IRC (Read error: 104 (Connection reset by peer))
[10:28:48] --> Kirben has joined #exult
[10:28:48] --- ChanServ gives channel operator status to Kirben
[12:31:18] <-- Kirben has left IRC ("System Meltdown")
[12:48:51] --> Cless|Away has joined #Exult
[12:49:02] <Cless|Away> hi
[12:49:06] <Yuv422> hey
[12:49:16] --- Cless|Away is now known as Colourless
[12:49:18] --- ChanServ gives channel operator status to Colourless
[12:49:36] <Colourless> Yuv422: i've got 'something' working, but it's not yet at an entirely usable level :-)
[12:50:07] <Yuv422> cool what state is it in?
[12:50:57] <Yuv422> I've been playing around with my copy on the xbox
[12:51:11] <Colourless> works displays everything fine, but isn't exactly fast, nor does work as I actually intend to do it (i.e. doesn't use hardware surfaces at all)
[12:51:37] <Colourless> it also has some code hacked in to get it to compile in windows with input and stuff :-)
[12:51:42] <Yuv422> ok how does it get at the video buffer then?
[12:52:19] <Colourless> in each call to XBOX_UpdateRects() i do the following
[12:52:22] <Colourless> IDirect3DDevice8_GetBackBuffer(g_pD3DDevice,0, D3DBACKBUFFER_TYPE_MONO,&lpBackBuffer);
[12:52:22] <Colourless> IDirect3DSurface8_LockRect(lpBackBuffer, &Locked, 0, 0);
[12:52:44] <Colourless> the manually copy the rect from the 'software' surface to the locked back buffer
[12:52:48] <Colourless> then do
[12:52:48] <Colourless> IDirect3DSurface8_UnlockRect(lpBackBuffer);
[12:52:48] <Colourless> IDirect3DSurface8_Release(lpBackBuffer);
[12:52:48] <Colourless> lpBackBuffer = 0;
[12:52:48] <Colourless> IDirect3DDevice8_Present(g_pD3DDevice, 0, 0, 0, 0);
[12:53:12] <Yuv422> right
[12:53:14] <Colourless> which is hardly the correct way of doing things :-)
[12:53:42] <Yuv422> yeah I was wondering how you would present the backbuffer while it was still locked to the SDL app
[12:54:05] <Yuv422> I'm guessing you can't flip a locked buffer onto screen ;)
[12:54:26] <Yuv422> does it handle listmodes?
[12:54:29] <Colourless> i'm betting you can, you really just might not have the results you expect :-)
[12:55:46] <Yuv422> would it be better to ask exult to use a HW surface and get it to lock / unlock the drawing sections? or would this be too big a code change?
[12:55:57] <Colourless> is there some sort of #define you can check for to detect if you are compiling for XBOX. There is code in there that is only for windows
[12:56:11] <Yuv422> yeah I'm using XBOX
[12:56:32] <Yuv422> and ENABLE_XBOXVIDEO for the video code.
[12:56:35] <Colourless> exult 'can' be forced to be double buffer mode friendly
[12:56:41] <Colourless> so there is a define XBOX ?
[12:56:51] <Yuv422> yes
[12:57:02] <Colourless> however, i should be able to get things faster
[12:57:50] <Colourless> i'll do some tests here myself at the speed difference between an emulated hardware single buffer mode and the double buffer mode.
[12:58:02] <Yuv422> when I first ran my test SDL app on the xbox I couldn't even get SDL_Init() to return :(
[12:58:06] <Yuv422> but that is fixed now
[12:58:25] <Yuv422> now it's crashing / spinning out in SDL_SetVideoMode
[12:58:40] <Colourless> you should wait for my code
[12:58:49] <Yuv422> but I've got very primative debugging support so it is taking along time to work through this
[12:59:04] <Yuv422> the first issue was with getenv
[12:59:17] <Yuv422> which the xbox obviously doesn't handle. ;)
[12:59:46] <Yuv422> yeah. I'm just iching to see something on the screen. ;)
[13:00:01] <Yuv422> I guess I should be patient. ;)
[13:01:33] <Yuv422> Is exult multi-threaded?
[13:02:38] <Colourless> not any more as far as I know. only the windows native midi driver uses a separate thread as far as I know
[13:02:55] <Yuv422> are you storing the dx8 vars in the SDL_PrivateVideoData struct?
[13:03:14] <Colourless> no, just using things global
[13:03:26] <Yuv422> ah k.
[13:03:47] <Colourless> not that i've added anything yet from your code :-)
[13:04:06] <Colourless> in regards to new vars
[13:04:14] <Yuv422> There's not much to add. ;)
[13:04:45] <Yuv422> I wish I had a debug xbox. :)
[13:05:22] <Yuv422> at the moment I send strings back to my computer via TCP/IP and that's all the feedback I get as to what the program is doing. ;)
[13:06:01] <Yuv422> and that is using a C++ class so I can't embed it into SDL :(
[13:06:31] <Colourless> well, you 'could' actually
[13:06:42] <Colourless> create wrapper functions
[13:07:17] <Yuv422> ah k
[13:07:26] <Colourless> you use extern "C" when declaring the functions so they can be linked to the C parts
[13:08:07] <Colourless> you'd basically do something like this:
[13:08:12] <Colourless> extern "C" int Class_Function (void *_this, char *an_arg)
[13:08:12] <Colourless> {
[13:08:38] <Yuv422> right
[13:08:45] <Colourless> return ((Class*)_this)->Function(an_arg);
[13:08:46] <Colourless> }
[13:09:20] <Yuv422> ah k. all this new C++ stuff to learn. ;)
[13:10:09] <Yuv422> so _this is a pointer to the instansiated object?
[13:10:12] <Colourless> yes
[13:10:28] <Yuv422> and that i availible in the C namespace?
[13:10:42] <Yuv422> how do I reference that?
[13:11:05] <Colourless> well, you can't really.
[13:11:34] <Colourless> you need wrapper functions for everything you would normally do with the objcet
[13:12:26] <Yuv422> can I just cast the object to void * and pass it out of my wrapper?
[13:12:33] <Colourless> yes
[13:12:51] <Colourless> as long as the object being passed from your wrapper is a pointer
[13:13:01] <Colourless> so you need to create things with new
[13:13:15] <Colourless> so you'd have something like
[13:13:16] <Yuv422> right
[13:13:34] <Colourless> void *Class_contruct(int a)
[13:13:35] <Colourless> {
[13:13:40] <Colourless> return new Class(a);
[13:13:41] <Colourless> }
[13:14:00] <Colourless> s/contruct/construct/
[13:14:34] <Yuv422> sed. :)
[13:15:14] <Colourless> then you need a Class_destruct too :-)
[13:15:19] <Yuv422> looks like I should make some wrappers for the xbox TCP/IP classes.
[13:15:22] <Colourless> \
[13:15:39] <Yuv422> either that or convert my xbox to a debug machine. ;)
[13:15:50] <Colourless> which would just do
[13:15:50] <Colourless> delete (Class*) _this;
[13:16:09] <Yuv422> right
[13:17:30] <Yuv422> do you think we should bother with a keyboard event handler.
[13:17:44] <Yuv422> can exult use the joystick interface?
[13:17:57] <Colourless> nope, exult doesn't understand joysticks
[13:18:18] <Colourless> exult only knows about keyboards and mouses
[13:19:10] <Yuv422> well I guess I should map the controller joystick to the mouse interface and then map some of the buttons to keyboard keycodes
[13:19:59] <Colourless> yeah. I would imagine mapping A,B,C,X,Y,Z,L,R to the keys of the same name might be a half decent idea :-)
[13:20:35] * Colourless assumes he has the correct key names/
[13:21:15] <Yuv422> well the xbox controller has two analog joysticks
[13:21:20] <Yuv422> one d-pad
[13:21:24] <Yuv422> start
[13:21:25] <Yuv422> back
[13:21:32] <Yuv422> a,b,x,y
[13:21:38] <Yuv422> white button
[13:21:41] <Yuv422> black button
[13:21:50] <Yuv422> left, right analog triggers
[13:22:27] <Colourless> Start would best be mapped to Enter and Back to escape
[13:22:28] <Yuv422> did keyboard conversation control ever make it into exult?
[13:22:39] <Colourless> yes it's been in for 'years' :-)
[13:22:57] <Yuv422> with the arrow keys. ;)
[13:23:06] <Colourless> you use keys 1 to 9 (possibly zero), but no, not arrow keys
[13:23:12] <Yuv422> I remember you could choose with the numbers
[13:23:20] <Yuv422> ah k
[13:23:56] <Yuv422> I think that could speed things up a lot when using an xbox controller
[13:24:40] <Colourless> well, i think one of the joysticks should be mapped to the mouse, and the other be mapped to the cursors
[13:24:50] <Colourless> exult can have movement using the cursors
[13:25:01] <Yuv422> yeah that sounds good. :)
[13:25:36] <Colourless> exult has reasonably flexible key mapping
[13:26:22] <Yuv422> are there any major game functions that can't be done with the mouse?
[13:26:44] <Colourless> name input and savegame names
[13:27:13] <Colourless> i 'was' planning on implementing an on screen keyboard for them, but i never got round to doing it so far
[13:27:44] <Yuv422> wouldn't it be easier to just allow up and down to cycle through the alphabet?
[13:27:54] <Yuv422> when in text input mode.
[13:28:36] <Yuv422> I must admit an on screen keyboard would be nice.
[13:28:46] <Colourless> that would probably be easiest to 'hack' in. the on screen keyboard was going to be for the pocketpc
[13:29:03] <Yuv422> right.
[13:29:21] <Yuv422> did you see the people talking about proting exult to the GBA! ;)
[13:29:25] <Yuv422> porting
[13:29:35] <Yuv422> heheh now that would be an impossible task
[13:31:18] <Colourless> we've spoken about it before
[13:31:30] <Colourless> it is impossible
[13:31:38] <Colourless> exults data set is too huge :-)
[13:31:49] <Yuv422> I think the GBA is a littleunder powered for that.
[13:32:25] <Yuv422> maybe ultima 1-5 ;)
[13:32:28] <Colourless> just doesn't have the memory
[13:32:37] <Yuv422> yeah
[13:32:42] <Colourless> at least, not with exults inefficent use of it
[13:32:49] <Yuv422> even the dreamcast would be a stretch. :)
[13:33:16] <Colourless> using memory mapped datafiles from the cart might make the gameboy slightly do-able, but it's still going to need a big cart
[13:33:40] <Colourless> dreamcast, could theoretically be doing, but it wouldn't be easy
[13:33:56] <Colourless> s/doing/done/
[13:33:58] <Yuv422> what's the U7 dataset size? about 20MB?
[13:34:19] <Colourless> just under
[13:34:23] <Colourless> SI is getting closer to 30
[13:34:44] <Colourless> or not
[13:34:47] <Colourless> SI is just over 20
[13:35:03] <Yuv422> just chop the dungeon plane. That would save some space.
[13:35:21] <Yuv422> ;)
[13:35:40] <royalsexy> hehe
[13:36:16] * royalsexy removes that damn mountains of freedom i always get stuck in
[13:37:33] <Yuv422> I wonder if controller state changes are queued
[13:37:42] <Yuv422> or if I have to poll it myself?
[13:38:02] <Colourless> no idea, that would depend on the API
[13:38:32] * Yuv422 looks at the input sample code.
[13:40:25] <Yuv422> it looks like we can control the ramble motors too. :)
[13:40:39] <Yuv422> that might make combat more interesting.
[13:56:47] <Yuv422> looks like a simple call to XBInput_GetInput()
[13:57:03] <Yuv422> will fill a struct for all for controller devices
[14:09:41] <Colourless> ok, i 'think' i should have something i could send to you
[14:10:38] <Yuv422> great :)
[14:11:05] <Colourless> not what I would call a final version, but it should work
[14:12:50] <Yuv422> we can polish it off once it is working.
[14:13:09] <Colourless> yes
[14:18:40] <Colourless> sent you an email
[14:18:46] <Colourless> so, where in .au do you live?
[14:21:14] <Yuv422> Sydney
[14:21:28] <Yuv422> I see you are from SA. :)
[14:21:40] <Colourless> yes, not really hard to work out where :-)
[14:22:13] <Colourless> it's note like the DNS for my ip doesn't say it or anything
[14:22:17] <Colourless> s/note/not/
[14:22:22] <Yuv422> ;)
[14:23:00] <Yuv422> I'll look at intergrating your code into my tree tomorrow
[14:23:16] <Colourless> i'll probably have more changes by then anyway :-)
[14:23:27] <Yuv422> heheh ;)
[14:26:13] <Yuv422> oh you can take out that CINTERFACE define
[14:26:21] <Colourless> yeah I know :-)
[14:27:46] <Yuv422> you might want to move your private variables into device->hidden
[14:27:54] <Yuv422> which is defined in the header.
[14:28:15] <Colourless> yeah I know. I'm just getting started :-)
[14:28:51] <Colourless> it is correct programming to do so, despite the fact that under the platform this is being done for, it shouldn't exactly matter one bit :-)
[14:29:15] <Yuv422> ;)
[14:33:02] <Yuv422> I see the SW surface in hidden. :)
[14:33:46] <Yuv422> should current->pixels be a pointer to that buffer?
[14:33:56] <Colourless> ah geez, that code isn't supposed to be there anymore :-)
[14:34:09] <Yuv422> ;)
[14:34:12] <Colourless> delete from line 425 to 431
[14:34:39] <Colourless> should just have
[14:34:39] <Colourless> if (g_pD3DDevice == NULL) return NULL;
[14:34:39] <Colourless> if ( ! SDL_ReallocFormat(current, bpp, Rmask, Gmask, Bmask, 0) ) {
[14:35:09] <Yuv422> so the buffer allocation is chopped?
[14:35:09] <Colourless> note that I added code so xbox will never attempt to run windowed :-)
[14:35:19] <Colourless> yes
[14:35:24] <Yuv422> yeah I saw that. :)
[14:37:41] <Yuv422> right well I'm off to bed
[14:37:48] <Colourless> the code is pretty crude at the moment. I get 'extreme' amounts of input lag :-)
[14:37:54] <Colourless> ok cya
[14:38:24] <Colourless> but it seems reasonably fast
[14:38:36] <Yuv422> can you mail me any changes you make. :)
[14:38:44] <Colourless> of course
[14:39:16] <Colourless> i would advise that for input, the analog stick should be the 'mouse'
[14:39:27] <Yuv422> yeah
[14:39:30] <Colourless> the harder you push it, the more the cursor moves
[14:39:41] <Colourless> s/harder/further/
[14:40:00] <Yuv422> I'm a bit worried about the cursor image though
[14:40:11] <Colourless> why?
[14:40:18] <Yuv422> are you sure it's not provided by the os layer under SDL?
[14:40:28] <Colourless> yes. it's all exult
[14:40:44] <Yuv422> device->ShowWMCursor = WIN_ShowWMCursor;
[14:40:51] <Yuv422> what is this line doing?
[14:40:54] <Colourless> that doesn't matter for exult
[14:41:05] <Yuv422> ah k great. :)
[14:43:17] <Yuv422> it seams to be hanging in SetVideoMode ATM with my video driver
[14:43:27] <Yuv422> hopefully this won;t happen with yours. ;)
[14:43:46] <Yuv422> I might spend some time fixing the debug code.
[14:44:14] <Colourless> ah shit
[14:44:17] <Colourless> made a mistake :-)
[14:44:23] <Colourless> #ifdef XBOX
[14:44:23] <Colourless> // Windowed only for XBOX
[14:44:23] <Colourless> flags &= ~SDL_FULLSCREEN;
[14:44:23] <Colourless> #endif
[14:44:32] <Colourless> isn't what it's supposed to be :-)
[14:44:46] <Colourless> should be this
[14:44:47] <Colourless> #ifdef XBOX
[14:44:47] <Colourless> // Fullscreen only for XBOX
[14:44:47] <Colourless> flags |= SDL_FULLSCREEN;
[14:44:47] <Colourless> #endif
[14:45:31] <Yuv422> righto changed
[14:46:06] <Yuv422> does this need changing?
[14:46:09] <Yuv422> current->pixels = 0;
[14:46:33] <Colourless> no. that gets set in XBOX_LockHWSurface
[14:46:42] <Yuv422> right
[14:47:19] <Yuv422> so does that mean that exult must pass the HWSurface flag?
[14:47:44] <Colourless> no, it works fine without it
[14:47:58] <Yuv422> ok
[14:48:00] <Colourless> SDL will still create it's own software surface
[14:48:30] <Yuv422> well I;m off now
[14:48:42] <Colourless> obviously exult should be updated to a hw surface
[14:48:44] <Colourless> cya
[14:49:01] <Yuv422> I'll probably talk to you tomorrow. :)
[14:49:02] <Yuv422> cya
[14:49:06] <-- Yuv422 has left #exult ()
[15:09:42] <Colourless> updating exult to hw surface... really really difficult...
[15:55:15] <Colourless> possibly nearly impossible actually, execpt for scaled surfaces...
[16:19:03] --> wjp has joined #exult
[16:19:03] --- ChanServ gives channel operator status to wjp
[16:19:05] <wjp> hi
[16:19:10] <Colourless> hi
[16:19:34] <wjp> actual activity on the exult ML, wow :-)
[16:30:58] <DarkeZzz> Hi!
[16:31:08] <DarkeZzz> Yeah, it's been like that for the last couple of days too. *grin*
[16:31:15] <Colourless> :-)
[16:34:09] * DarkeZzz nodes that kdevelop3 is rather pretty. Still lacking some of the features of the previous version though, but it is an alpha.
[16:35:11] <DarkeZzz> Apart from the aforementioned blight of a scrolling tabbed dialogue box. *shudder*
[16:35:55] <wjp> hmm... gnome 2.2 out
[16:36:10] * wjp wonders if he should unmask the required ebuilds
[16:36:38] <wjp> (apparently portage is in somewhat of a freeze so no new stable packages are being added)
[16:37:28] <DarkeZzz> I'd wait a couple of days. Whilst most of the packages are stable, a couple are a bit... touchy. *grin* I've complained about kde3.1 and it's mysterious crashing konqueror, yes?
[16:37:41] --- DarkeZzz is now known as Darke
[16:37:46] * Darke decides to awaken for a bit.
[16:38:47] * Darke ooohs, gcc3.2.2 is out too...
[16:38:57] --- ChanServ gives channel operator status to Darke
[16:45:01] <wjp> tetex 2 is finally out too, btw :-)
[16:45:14] <wjp> (since last saturday, actually)
[16:45:21] <wjp> hm, or was that monday
[16:48:52] <wjp> hm, I don't think I heard you about a crashing (krashing?) konqueror (conqueror?) yet, actually :-)
[16:49:11] * wjp wonders if that has anything to do with the fact that he's only been online after 5-6pm this last week
[16:50:52] <Darke> Hmm... probably that. You'll no doubt find my rantings in the logs though. *grin* As a general synopsis, in kde3.1 konqueror starts up ok, browses the web ok, but if you close the window it crashes with a sig11, which isn't inconvenient until you've opened up half a dozen windows from that one (say reading slashdot or something) then close one, only to have them all simultaneously crash and close. *grin*
[16:51:03] <wjp> <Darke> Want me to DCC a screencap? It's got to be the most hideous, I-don't-get-the-point-of-tabs dialog I've *seen*! <-- can I see? please? :-)
[16:52:23] <Darke> Just a sec and I'll locate it again. *grin*
[16:54:20] <wjp> --- Received a malformed DCC request from Darke.
[16:54:20] <wjp> --- Contents of packet: "DCC SEND hell_is_a_dialog1.png ,NK84 llPZe&$ 7^,NK84 llPZ 54960".
[16:54:45] <Darke> wjp: Erk. Weird. I've had that with another person too. Obviously my firewall setup is b0rk3n.
[16:57:47] * Darke pokes things.
[16:58:19] * royalsexy takes his things away from Darke for fear of having them poked
[17:02:03] * Darke gets a looooooong stick and pokepokepokes the things royalsexy has removed.
[17:27:13] --> Fingolfin has joined #exult
[17:27:14] --- ChanServ gives channel operator status to Fingolfin
[17:27:22] <wjp> hi
[17:27:28] <Fingolfin> yo
[17:28:19] <wjp> question on forum: "Do files on OSX still have the resource fork?" Answer by OSX user: "What is a 'resource fork'?"
[17:29:41] <wjp> btw, do you think it would help if someone from that thread submitted a new bug report about this to apple?
[17:29:54] <wjp> (this = fstream.read() bug)
[17:32:11] <Fingolfin> besides the one from me?
[17:32:18] <Fingolfin> and yes OS X still has resource forks
[17:32:41] <Fingolfin> but as under OS 9, resource forks are nothing mandatory, i.e. there are plenty of files w/o one, too :-)
[17:33:34] <wjp> out of curiosity; any way to access them with 'standard' unix tools?
[17:33:35] * Fingolfin goes to check Radar
[17:34:30] <wjp> radar?
[17:34:36] <Fingolfin> most standard unix tools ignore them, but there is a special notion, you can do something like "cp foo/rsrc bar" to copy the res fork of foo to the data fork of bar etc. (the syntax is a bit different I think, never use it)
[17:34:41] <Fingolfin> Radar is the Apple bug tracker
[17:35:30] <Fingolfin> not sure if I said it, but I confirmed yesterday that the fstream bug is still there
[17:35:50] <wjp> there == in radar or there == in the latest devtools?
[17:36:43] <Fingolfin> the bug still occurs
[17:36:50] <Fingolfin> oh wow, new web layout for Radar :-)
[17:36:54] <Fingolfin> the old one was nicer :-)
[17:36:59] <Fingolfin> let's see
[17:37:14] * wjp nods; I had someone on the forum run that readtest.cc program you sent me a while back, and it occured for him too
[17:37:17] <Fingolfin> it has status "closed/duplicate"
[17:37:51] <Fingolfin> and due to this shitty closed bug tracking system, I have no way to find out about its status unless I use my contacts (i.e. ask some friends working at Apple to look at it for me
[17:38:19] <wjp> you can't find out which bug it's a duplicate of?
[17:38:23] <wjp> bah
[17:38:34] <wjp> doesn't really encourage community involvement, then
[17:38:37] <Fingolfin> no
[17:38:44] <Fingolfin> and in fact, you can't find out anything about the bug
[17:38:53] <Fingolfin> e.g. any comments made to it are not visible to outsiders
[17:39:07] <Fingolfin> I can only see the text I used when I submitted it, its current status, and its id, nothing else
[17:39:13] <Fingolfin> it's a very crappy thing IMHO
[17:39:24] <wjp> maybe it's time to look for some security holes in the interface ;-)
[17:39:24] <Fingolfin> I mean, I understnad that certain bugs are confidential, but *all* ?
[17:40:06] <Colourless> some companies are known to keep in contact with the bug finder with the status of the bug
[17:40:08] <Fingolfin> there is a bug i submited over 1 year ago and it's still in "open/analyze" status. It was a very obscure memory leak, I have to admit, but it was one nonetheless :-)
[17:40:14] <Fingolfin> Colourless: sometimes they do
[17:40:38] <Fingolfin> i.e. sometimes I get mails from Apple on bug reports where they ask me further details, or tell me its fixed and I should please verify it, etc., but usually nothing at all
[17:41:02] <wjp> "Workarounds are considered to be Technical Support incidents. You must purchase Technical Support if you would like to receive a workaround." <-- hehe :-)
[17:41:59] <Fingolfin> yup
[17:42:29] <Fingolfin> my former boss used to pull all sorts of strings to get Apple to fix his stuff. It helps if your product is of some importance to the platform, or if you know the engineers involved
[17:42:37] <Fingolfin> but sometims you have to wait ages for things to be fixed
[17:42:40] * Darke blinkblinks and thinks he's coming to understand exactly why it took so long to get pch out of apple's compiler and into the standard gcc one.
[17:42:56] <Fingolfin> like, we are still using a buggy version of Perl 5.6.0! didn't changhe since the original release of the OS!
[17:43:10] <Fingolfin> reason: the guy responsible for it was fired back then, and nobody dared to touch the code afterwards
[17:43:13] <wjp> oh, speaking of pch, in which version of gcc where those added?
[17:43:20] <Darke> 3.4.
[17:43:21] <wjp> (in the 'main' gcc, that is)
[17:43:42] <Colourless> OMFG :-)
[17:44:00] <wjp> ...and the long wait continues... :-)
[17:44:03] <Fingolfin> that goes so far that Apple has an official website which tells you how to manually install perl 5.8.0 (from sources), because the guys working on the pages promoting Apple has a platform for web services needed a newer Perl, but the system engineers couldn't be bothered to update it in the system for 2 years now
[17:44:08] <Darke> It's also got a Brand New Parser(tm) that happens to fix lots of parser bugs that have been piling up over the last few months... and introduce more of it's own too. *grin*
[17:44:15] <Fingolfin> well, I do have pch of course <g>
[17:44:29] <wjp> we're well aware of that *sigh* ;-)
[17:44:32] <Darke> wjp: http://gcc.gnu.org/gcc-3.4/changes.html
[17:44:34] <Fingolfin> since Apple included it in their GCC for 2-3 years now :-)
[17:44:45] <wjp> and that... *bigger sigh* ;-)
[17:45:05] <wjp> OTOH, we have perl 5.8 and a working fstream::read() ;-P
[17:46:01] * Darke will happily take a less buggy and more standards comforming compiler over a faster one anyday. *grin*
[17:46:33] <Colourless> you know, implementing PCH into pentagram with gcc3.4 should be fairly simple. pent_include.h is included at the top of every file for a reason :-)
[17:46:50] <wjp> yeah, I was pretty sure that was one of the reasons for that :-)
[17:47:43] <Colourless> the other reason is it simplies compiling on systems with hacks of compilers *cough* mine *cough*
[17:47:53] <Colourless> s/simplies/simplifies/
[17:47:58] <wjp> :-)
[17:48:05] <Fingolfin> wjp: bah who needs fstream and perl 5.8 :-)
[17:48:11] <Fingolfin> I can install perl 5.8, too, mind you
[17:48:22] <Fingolfin> it's just that it's lame to ship 5.6.0, not even 5.6.1 !
[17:48:53] <Fingolfin> and in fact there is a bug in the dlopen interface code, basically teh perl lib exports dlerror and dlclose but not dlopen/dlstatus, which leads to a big mess (it should export none of them)
[17:49:03] <Fingolfin> I can fix it manually, but ugh, Apple should fix it, not me :-)
[17:49:39] <Fingolfin> ahah
[17:49:44] <Fingolfin> I asked my contact <g>
[17:49:58] <wjp> got the ID of the duplicate?
[17:50:12] <Fingolfin> the bug is known, was there in gcc 3.1, too, but is now fixed in FSF gcc, and they merged it into their tree -> next GCC update should ahve the fix
[17:50:19] <wjp> great :-)
[17:50:33] <Fingolfin> the ID of the duplicate would be useless: I can only view my own bugs... and in those, i see ntohing besides its ID, status, and what I wrote
[17:50:42] <wjp> ah
[17:50:59] <wjp> (well, if you could see the status of the duplicate, you'd at least be able to see the 'fixed' status)
[17:51:22] <wjp> pretty stupid, btw... 'duplicate' tells you exactly nothing about the status of the bug
[17:51:36] <wjp> 'fixed' or 'rejected' or whatever at least gives you some idea of what's going on
[17:53:50] <Fingolfin> yeah
[17:54:23] <Fingolfin> they really should have to kinds of bugs... or at least allow the bug submitter to see a little more info on the bugs
[18:13:34] * wjp grins; I 'only' need to upgrade 65 package to get gnome 2.2
[18:13:42] <Fingolfin> hehe
[18:13:53] <wjp> s/package/packages/
[18:26:47] <jer|w0rk> :|
[18:33:48] <jer|w0rk> unf
[18:36:05] <wjp> ?
[18:39:37] * Darke yawns and decides it's sleeptime for him. Night!
[18:39:42] <wjp> g'night
[18:39:45] --- Darke is now known as DarkeZzz
[18:39:49] <Colourless> cya
[18:40:29] <Fingolfin> wjp: "schenk water bij dat net van de kook af is" = "add boiling water" ?
[18:40:42] <wjp> basically, yes
[18:40:51] <wjp> 'water that just stopped boiling'
[18:41:41] <Fingolfin> yeah
[18:44:09] <Fingolfin> hah, just run it through an automated translator for the fun of it: "water pours that of boil down exactly is"
[18:44:18] <wjp> lol
[18:44:20] <Fingolfin> reminds me of an asian motherboard manual I once had =)
[18:45:04] <wjp> hm, the translator actually changed the word order, so it wasn't a word-by-word translation
[18:45:19] <wjp> not sure if the result is much better than a word-by-word one would've been :-)
[18:45:55] <Colourless> what would it be word by word?
[18:46:41] <wjp> pour water <bij is part of 'bijschenken'> that just of the boil <af... of? from?> is
[18:53:13] <Colourless> hmm, i guess that if it were "pour water that is just from the boil" it would make some sense
[19:09:23] <Colourless> i think i should go
[19:09:24] <-- Colourless has left IRC ("casts invisibility")
[21:16:48] <-- wjp has left IRC ("brb")
[21:21:09] --> wjp has joined #exult
[21:21:09] --- ChanServ gives channel operator status to wjp
[22:24:06] <wjp> time to go; g'night
[22:24:09] <-- wjp has left IRC ("Zzzz...")