Requiem MAP21 Den of the Skull demos

[vdgg]

There are 3 UV Max demos available on DSDA. If I playback rq21-838.lmp it goes out of sync right away.

I tried to follow these instructions , but to no avail. I downloaded ZenNode 1.2.1, typed zennode -rz rr21.wad (the file I created with DoomBuilder 1.68 using "Save Map Into"). Now the demo plays back correctly until about 3'20" (I use Prboom-plus and add 3 files to the command line: reqmus, requiem and the newly created MAP21 file).

Also, the link to the file created by Rudy Jurjako is broken. So, what to do in order to watch these demos? And perhaps, to record anything?

[Andy Olivera]

Redownload the WAD from the DSDA. It contains both the Compet-N and DSDA versions of MAP21. And the reason the version you rebuilt didn't work is because it has to be the same version of Zennode(v0.98, I believe).

[vdgg]

As for "where to get the right files?" problem, many thanks! It's not the first time when I complicate matters whereas the thing I'm looking for is right there at the DSDA :)

But I still do something wrong, I guess.


R21.lmp (Rudy, 96% kills, 8:10) indeed works with req21_competn
req21-uv.lmp (Kai-Uwe, 103% kills, 11:21) works with req21_dsda
rq21-838.lmp (Xit Vono, UV Max, 8:38) goes out of sync with req21_dsda around 3:20, when the player is fighting in the area with no floor nor ceiling (skies everywhere)
rn211050.lmp (Xit Vono, nightmare!) is out of sync with req21_dsda around 3:55 (arch-vile fight)
rq21-636.lmp (Xit Vono, UV Speed, 6:36) goes out of sync with req21_dsda around 3:30

I tested this with Prboomplus 2.5.0.2 (no -complevel options), Doom2.exe (I changed the names req21_dsda etc. so that they have <= 8 characters). My requiem.wad file has 11464486 Bytes, of course. Any ideas?

[Andy Olivera]

Hmm. Xit's max demo should play back correctly with the Compet-N version(the text file says it was used), however it didn't sync for me, either. The only explanation I can come up with is that the fixed WAD wasn't used for any of Xit's demos. Neither the NM or UV Speed demos mention it(that's why I didn't post them at the DSDA).

If anyone else has been able to get any of these to sync, I'd love to know how...

[Hitherto]

As I suppose, Xit was using another version of rq21fix.wad, perhaps own made, with right reject table like Andy's fix, but without fixed monsters.

---
Heh, I'm wrong about Andy's action, his fix is just right zennode recompilation with zero reject, and hadn't any special "fixed monsters". I'm sorry.

Indeed Rudy's "fix" (recompiled with lame bsp) has bad reject table in another way, and it's cause for that 8-monsters sleepiness, and perhaps cause for additional desynches.

Anyway, only Xit himself may say how it was recorded and how it should be playbacked.

[xit-vono]

I think that there may be another version of the fix out there. I know that Kai used a fix that allowed for maximum kills, and that Rudy used a different one but didn't get 100% kills. This fix was used at compet-n for a while. Then I think Adam decided to use a new compet-n fix, and that I didn't have to beat Rudy's time. I don't have the file on my computer right now, have you checked at the compet-n website? I may have it on my other computer at home, or maybe someone else has it. In any case I don't think it will be a huge loss for humanity if the ability to watch those demos is lost

[Andy Olivera]

Technically, the version Rudy used for the newer demopack(and the one that was at the Compet-N) had a Nodes problem. Reject only affects what monsters can see, but the dormant monsters in Rudy's demo are supposed to be awoken by sound. Rudy used BSP to rebuild the level, which, for some reason related to the self-referencing sectors used in the sky segment, prevents sound from being relayed to the teleporter room.

Also, the DSDA version(for Kai's demo) isn't a simple zeroed Reject Map; all the resources are rebuilt. It's just that for some reason it worked with Zennode, but not BSP.

I don't remember reading about a third fix for the level, but the fact that Xit's UV Max makes it nearly halfway through leads me to believe it used a zeroed Reject, as well. I tried zeroing the Reject on the original WAD(not rebuilding anything else), but it desyncs at the 'face' room(right after the teleport), which means the Nodes were probably rebuilt. If there was another version, we can either find it, or try rebuilding the WAD with a bunch of different nodebuilders until we find the one that was used(maybe another version of Zennode).

Maybe I'll test the old versions of Zennode this weekend...

[Ryback]

Xit is right, there were two different compet-n fixes. On my Compet-n /pwads/requiem directory there are two files: rq21oldf.zip and req21fix.zip.

rq21oldf.zip contains req21fix.wad, dated 7/20/1998, 591488 bytes long, and is presumably Rudy's fix.
req21fix.zip also contains req21fix.wad, this one dated 10/01/2001, 571446 bytes long. That was created when Requiem and the other pwads were added to the records table.

By comparison req21fix_dsda.wad is dated 03/22/2009, and 571146 bytes long. Almost the same as the second compet-n version. The similar size does suggest they were created the same way just with different versions of Zennode.

Unfortunately the only version available at the Compet-n page is the first requiem fix, which has inexplicably been renamed from rq21oldf.zip to req21fix.zip.

Edit: sent you an email Andy.

[Never_Again]

[Andy Olivera]

Three cheers for Ryback! I'll add Xit's other two Compet-N records and update the Requiem download with this new version.

Another DooM demo artifact recovered! SWEET!!!