SRB2 v2.0.5 is now available!

Status
Not open for further replies.
So, given the number of stupidly misplaced things in every other map, how far off are we from a 2.0.6? Five months?
 
Alam's build runs as smooth as silk. Castle Eggman Act 2 doesn't lag in the slightest.

So, given the number of stupidly misplaced things in every other map, how far off are we from a 2.0.6? Five months?


Wasn't 2.0 super buggy when it came out? Most of these "bigger" patches end up buggy at first, 'till another (usually quicker to produce) patch goes and fixes them.
 
犬夜叉;672467 said:
Do me a favor and try Alam's Mingw build of 2.0.5 and tell me if the lag is still there.

If it isn't... well, that gives us a perfectly valid reason to FINALLY start using GCC.

(I wouldn't vouch for this being compatible with the released 2.0.5 EXE due to floating point math differences, like before, so you might want to rename this EXE instead of just replacing the normal one.)

Are you not keeping up with developments? Check tables.c for FIXEDPOINTCONV. When #define'd, it prevents these math issues. Also, it's only VC6 that handled floating point math differently - VC9 wouldn't have this issue even if FIXEDPOINTCONV wasn't #define'd.
 
Wasn't 2.0 super buggy when it came out? Most of these "bigger" patches end up buggy at first, 'till another (usually quicker to produce) patch goes and fixes them.
The problem is, most of the faults in this version seem to stem from the IWAD, which is harder to update than the executable. Unless we have some sort of diff-based patching system working which I don't know about.
 
Are you not keeping up with developments? Check tables.c for FIXEDPOINTCONV. When #define'd, it prevents these math issues. Also, it's only VC6 that handled floating point math differently - VC9 wouldn't have this issue even if FIXEDPOINTCONV wasn't #define'd.

I'm covering my ass just in case, thank you very much.

The last thing I want to do is make hosting netgames even more hellish.

EDIT: And, in fact, the consistency failures do STILL happen, even with it defined. Just look at the Official Dedicated Server.
 
Last edited:
I felt that there were netgaming problems with my own experience using SRB2ALAM. So, I stopped using it.

Now just make it the official SRB2 Build....
 
犬夜叉;672467 said:
Do me a favor and try Alam's Mingw build of 2.0.5 and tell me if the lag is still there.

If it isn't... well, that gives us a perfectly valid reason to FINALLY start using GCC.

(I wouldn't vouch for this being compatible with the released 2.0.5 EXE due to floating point math differences, like before, so you might want to rename this EXE instead of just replacing the normal one.)

Tried it, seems the same, but the menu has 0 lag just as 2.0.4 had.
 
Okay.

This is a SOC. I didn't make it, but long story short, it's supposed to block players from firing red rings by suppressing the projectile itself and incrementing the player's ring count by one (so players don't lose rings for nothing).

This works perfectly fine in Alam's MinGW build, but it fails to increment the player's health in the official EXE. Assuming that the two versions are identical save for the compiler used, this means we don't just have slight performance issues, it means we have a crummy compiler fucking things up.

That is all.
 
The official 2.0.5 EXE was compiled with MSVC9, and without the full optimizations that we normally use (note that the latter wasn't done intentionally).

We'll be sending out a more stable release as soon as we can.
 
Ulrich, that's from before FIXEDPOINTCONV and even then, Alam recompiled it with MSVC6 to prevent cfail.

Also Inuyasha, that is a misdiagnosis. The new consfail protection system is based on taking action after a set number of consistency failures in a set amount of time. I'm not sure quite what's triggering it, but just about every server is suffering from CKick issues.
 
I highly doubt that, Cinefast, considering A: These consistency failures did not happen on maps lacking water (hmm...) and B: I was sitting in there in the server with a GCC build and watching literally everyone else get C-kicked while I stayed on the server until I finally got tired of waiting and quit.
 
Status
Not open for further replies.

Who is viewing this thread (Total: 2, Members: 0, Guests: 2)

Back
Top