GLQuake Bug: Nvidia Extension Limit must be off

[JasonX]

I couldn't possibly ask for more, mh. Thank you so much for the help!

[JasonX]

Quick question: are there any problems with GLQuake's netcode? Or it's just bugged on the graphics part? =p

[mh]

Aside from not working across firewalls or routers? And being pretty crap to begin with?

The sound code is also f-cked by the way. Writing to the primary buffer with DirectSound is not a nice thing to do, and no well-behaved app should do it.

The netcode fix is something that's already in ProQuake and a few others. There's a partial fix in many other engines too, but that doesn't work on some servers. I put a full fix in DirectQ some time back but never really investigated exactly what was required (must do that sometime!) As soon as info on the full fix emerges I'm sure it'll be posted here and tutorialized.

For the sound code you need to hack around in SNDDMA_InitDirect. All you really need is this block (it should be obvious what it replaces):

Code: Select all

#ifndef MHGLQFIX
    if (!COM_CheckParm ("-snoforceformat"))
    {
        if (DS_OK == pDS->lpVtbl->CreateSoundBuffer(pDS, &dsbuf, &pDSPBuf, NULL))
        {
            pformat = format;

            if (DS_OK != pDSPBuf->lpVtbl->SetFormat (pDSPBuf, &pformat))
            {
                if (snd_firsttime)
                    Con_SafePrintf ("Set primary sound buffer format: no\n");
            }
            else
            {
                if (snd_firsttime)
                    Con_SafePrintf ("Set primary sound buffer format: yes\n");

                primary_format_set = true;
            }
        }
    }
#endif

#ifndef MHGLQFIX
    if (!primary_format_set || !COM_CheckParm ("-primarysound"))
    {
#endif
    // create the secondary buffer we'll actually work with
        memset (&dsbuf, 0, sizeof(dsbuf));
        dsbuf.dwSize = sizeof(DSBUFFERDESC);
        dsbuf.dwFlags = DSBCAPS_CTRLFREQUENCY | DSBCAPS_LOCSOFTWARE;
        dsbuf.dwBufferBytes = SECONDARY_BUFFER_SIZE;
        dsbuf.lpwfxFormat = &format;

        memset(&dsbcaps, 0, sizeof(dsbcaps));
        dsbcaps.dwSize = sizeof(dsbcaps);

        if (DS_OK != pDS->lpVtbl->CreateSoundBuffer(pDS, &dsbuf, &pDSBuf, NULL))
        {
            Con_SafePrintf ("DS:CreateSoundBuffer Failed");
            FreeSound ();
            return SIS_FAILURE;
        }

        shm->channels = format.nChannels;
        shm->samplebits = format.wBitsPerSample;
        shm->speed = format.nSamplesPerSec;

        if (DS_OK != pDSBuf->lpVtbl->GetCaps (pDSBuf, &dsbcaps))
        {
            Con_SafePrintf ("DS:GetCaps failed\n");
            FreeSound ();
            return SIS_FAILURE;
        }

        if (snd_firsttime)
            Con_SafePrintf ("Using secondary sound buffer\n");
#ifndef MHGLQFIX
    }
    else
    {
        if (DS_OK != pDS->lpVtbl->SetCooperativeLevel (pDS, mainwindow, DSSCL_WRITEPRIMARY))
        {
            Con_SafePrintf ("Set coop level failed\n");
            FreeSound ();
            return SIS_FAILURE;
        }

        if (DS_OK != pDSPBuf->lpVtbl->GetCaps (pDSPBuf, &dsbcaps))
        {
            Con_Printf ("DS:GetCaps failed\n");
            return SIS_FAILURE;
        }

        pDSBuf = pDSPBuf;
        Con_SafePrintf ("Using primary sound buffer\n");
    }
#endif

Input is f-cked too. As is physics, and the filesystem (it's case-sensitive!), the memory allocator is an unholy mess, the chase cam is f-cked, etc etc etc.

These are all problems that have been worked over throughout the years, and I'd seriously say that I don't envy you one bit starting fresh at them.

[frag.machine]

Input is f-cked too. As is physics, and the filesystem (it's case-sensitive!)

Hmm, case-sensitive filesystem rocks. Too bad Windows can't handle this properly

[mh]

case-sensitive filesystem rocks.

Not if you're an end-user it doesn't.

[frag.machine]

case-sensitive filesystem rocks.

Not if you're a Windows-only end-user it doesn't.

Here, I fixed that for ya

[mh]

case-sensitive filesystem rocks.

Not if you're a Windows-only end-user it doesn't.

Here, I fixed that for ya

No, I've managed Unix servers with big production Oracle databases on them, and believe me, this does not rock. Waking up at 4 in the morning screaming is not a nice experience.

[dreadlorde]

No, I've managed Unix servers with big production Oracle databases on them, and believe me, this does not rock. Waking up at 4 in the morning screaming is not a nice experience.

But why doesn't a case-sensitive filesystem not rock?

[Feared]

I noticed the link went down.
Had it sitting in my downloads folder for a while now.

http://bfeared.com/library/index.php?di ... GLQFIX.zip

[leileilol]

Here, I fixed that for ya

No, you have to also deal with case-sensitivity when it comes to pak files.