[Hendricks266]

Mblackwell, on 15 September 2016 - 09:07 AM, said:

This was broken before.

"Bouncing in one spot" sounds pretty broken in its own right.

[Mblackwell]

Trooper Dan, on 15 September 2016 - 10:18 AM, said:

Just to be clear: what exactly was broken? Projectiles that explode after a certain number of bounces have worked for as long as I can remember.

Timed projectiles.

[Mblackwell]

Hendricks266, on 15 September 2016 - 03:39 PM, said:

"Bouncing in one spot" sounds pretty broken in its own right.

If you have 16384 (EXPLODEONTIMER) set but not 64 (TIMED) set then the game doesn't read RANGE as the TIMER so it can't figure out when it should explode.

[Danukem]

Mblackwell, on 15 September 2016 - 09:19 PM, said:

If you have 16384 (EXPLODEONTIMER) set but not 64 (TIMED) set then the game doesn't read RANGE as the TIMER so it can't figure out when it should explode.

Got it. I'm pretty sure that I have always had both of those set on my timed projectiles because the documentation said to do so.

By the way, what would be the use case for having one flag set but not the other?

[Jblade]

I don't know what the conclusion here is but the projectile in question isn't set to explode on any kind of timer, they're supposed to just bounce 6 times or whatever and then explode at the end. This isn't working anymore.

[Danukem]

Jblade, on 15 September 2016 - 11:47 PM, said:

I don't know what the conclusion here is but the projectile in question isn't set to explode on any kind of timer, they're supposed to just bounce 6 times or whatever and then explode at the end. This isn't working anymore.

I have been using the latest synthesis build (r5850, 64 bit) while playing Attrition, which has several different custom bouncing projectiles, and I have not seen the problem you described. It must be something more specific than just any bouncing custom projectile. Can you post the definitions on one of your problem cases so we can compare flags?

[RichardStorm]

I saw that timed AND bouncing projectiles (like grenades) will not explode on timer, but on bounces left, so i put a counter variable and coded projectile death-on-timer by myself.

[Jblade]

defineprojectile 5375 PROJ_TRAIL GUNSMOKE
defineprojectile 5375 PROJ_WORKSLIKE 18510
defineprojectile 5375 PROJ_HITRADIUS 2048
defineprojectile 5375 PROJ_VEL 700
defineprojectile 5375 PROJ_DROP -140
defineprojectile 5375 PROJ_ISOUND 630
defineprojectile 5375 PROJ_BSOUND 629
defineprojectile 5375 PROJ_BOUNCES 3
defineprojectile 5375 PROJ_XREPEAT 32
defineprojectile 5375 PROJ_YREPEAT 32
defineprojectile 5375 PROJ_EXTRA 280
defineprojectile 5375 PROJ_EXTRA_RAND 125
defineprojectile 5375 PROJ_SPAWNS GROUNDEXPLOSION

it would bounce 3 times and then explode, now it just stops and bounces indefinitely on the spot.


I'm gonna work through snapshots and let you know which one it breaks in.

EDIT:
http://dukeworld.duk.../20160704-5811/

This is the last snapshot it worked in. I couldn't test 5839 since it crashed on startup but in 5844 it stops working properly and just bounces as I've reported.

[Mblackwell]

Yes you have 16384 in WORKSLIKE. It's the flag EXPLODEONTIMER.

The reason it exploded after a few bounces before was because that bit (EXPLODEONTIMER) wasn't working.

[Mblackwell]

Trooper Dan, on 15 September 2016 - 10:47 PM, said:

Got it. I'm pretty sure that I have always had both of those set on my timed projectiles because the documentation said to do so.

By the way, what would be the use case for having one flag set but not the other?

I honestly have no idea! Let's find out!

Alright here's the skinny:

If WORKSLIKE flag 64 (TIMED) and not 16384 (EXPLODEONTIMER) is set then the sprite will delete itself when the timer reaches RANGE (without spawning anything). If you have BOUNCES (and appropriate flags) set then it will either try to reach max bounces and explode, or it will still get deleted after RANGE if it can't max out its bounces before then.

[Jblade]

Mblackwell, on 16 September 2016 - 06:33 AM, said:

Yes you have 16384 in WORKSLIKE. It's the flag EXPLODEONTIMER.

The reason it exploded after a few bounces before was because that bit (EXPLODEONTIMER) wasn't working.

alright, in my defence the circumstances did really make it come across like some kind of bug.

[Mblackwell]

[Jblade]

A quick question about the cachesize command - it doesn't seem to work for mapster32 if used in a def file, is this intended? If I set it via a shortcut it works fine but putting cachesize into the def file doesn't seem to do anything. The rest of the def file is working and the console isn't giving me errors, it's just the log file still reports the default cache size being used. I don't know if this is a bug or whatever but I swore I've seen somebody say it can be set in the def file.

[Tea Monster]

Are those render issues likely to get fixed?

[Hendricks266]

I just updated the SDL with which synthesis builds to 2.0.5. Let me know if you experience any problems. At the very least, the taskbar icon should look better while running either exe.

[Danukem]

Hendricks266, on 19 October 2016 - 10:28 PM, said:

I just updated the SDL with which synthesis builds to 2.0.5. Let me know if you experience any problems. At the very least, the taskbar icon should look better while running either exe.

So is there actually a build in synthesis right now that uses the updated SDL? I'm not seeing anything new there.

[Hendricks266]

As of this post I committed nine minutes ago. IIRC the synthesis check interval is 15 minutes. I'm not sure if automatic building is fixed yet or if we still have to kick the box. Time will tell.

[Danukem]

Looks like it needs a kick.

[Danukem]

Trooper Dan, on 19 October 2016 - 11:02 PM, said:

Looks like it needs a kick.

Bump. Please "kick the box" so that we can check out a build with the updated SDL.

[TerminX]

Kick delivered.

[Apiai]

Hi,

I've got a render bug :
EVERY transparent texture is divided into two triangles which flash in a desynchronized way.
Tested with r5888 and r5905 (in polymost not polymer)

[AdrenalinDragon]

Apiai, on 20 October 2016 - 11:20 PM, said:

Hi,

I've got a render bug :
EVERY transparent texture is divided into two triangles which flash in a desynchronized way.
Tested with r5888 and r5905 (in polymost not polymer)

Yeah, I mentioned it in the World Tour EDuke32 thread. Really needs fixing. Using an old version of EDuke32 for now.

[Tea Monster]

AdrenalinDragon, on 22 October 2016 - 06:22 AM, said:

Yeah, I mentioned it in the World Tour EDuke32 thread. Really needs fixing. Using an old version of EDuke32 for now.

I take that as a no then


EDIT: Why are you encouraging people to report bugs and then downvoting them for doing so? I'm not a programmer, otherwise I'd fix it myself.

[Jblade]

Completing a level in episode 6 or 7 with a standard nuke button now just takes you back to the first level of the episode rather than the next one. I'll see if I can track down when this change happened.


Just tested with some vanilla cons, problem is probably on my end. gonna do some more investigating.

[Hendricks266]

Tea Monster, on 23 October 2016 - 12:00 AM, said:

EDIT: Why are you encouraging people to report bugs and then downvoting them for doing so? I'm not a programmer, otherwise I'd fix it myself.

This post isn't reporting a bug, it's pouring salt on me.

The security camera bug started in r5825. It's a huge diff that I haven't gone through yet.
I've narrowed your Polymer regressions to r4600.
The alpha/blinking bug is limited to 32-bit builds with compiler optimizations and started between 5609 and 5617 inclusive for no reason that I can tell.
I've already fixed both bugs in http://bugs.eduke32....ails&task_id=35 locally.
Synthesis needs fixing but that's up to Plagman and TX; I don't have any shell access to the server.

I could go on. I'm overflowing with requests and responsibilities for more projects than I can count on one hand.

[Jblade]

Quote

I've already fixed both bugs in http://bugs.eduke32....ails&task_id=35 locally.

you're 100% spot with the EX name since I was getting those same warnings with a bunch of sounds I had named EXPLOSION_somethingsomething - I wasn't sure what was up with it so I never mentioned it but glad to see I wasn't just crazy.

[Hendricks266]

Pushed my fix for that. Normally it wouldn't be enough for me to want to trigger a build, but hey, it's broken.

[AdrenalinDragon]

After going through hundreds of EDuke32 versions, here's when the flickering bug on Polymost starts for me:

r5084 and after - Flickering bug starts
r4908 to r5064 - Button disappears
r4886 to r4904 - Can't load game
r4882 and before - Flickering bug not present

[Hendricks266]

AdrenalinDragon, on 23 October 2016 - 12:44 PM, said:

After going through hundreds of EDuke32 versions, here's when the flickering bug on Polymost starts for me:

Hey thanks! Always good to have more information. A 200 revision window is still large, but better than nothing.

AdrenalinDragon, on 23 October 2016 - 12:44 PM, said:

r4908 to r5064 - Button disappears

What do you mean?

AdrenalinDragon, on 23 October 2016 - 12:44 PM, said:

r4886 to r4904 - Can't load game

Yeah, we bumped BYTEVERSION at least once since then. I recommend avoiding savegames for any kind of bisecting. Instead, use the editor to change a map's start point to the exact place a bug occurs by pressing Scroll Lock, and play the map to test.

AdrenalinDragon, on 23 October 2016 - 12:44 PM, said:

r5084 and after - Flickering bug starts

r4882 and before - Flickering bug not present

Candidates include:
5079
5075
5001
4904

Regrettably, the first two are another instance of massive diffs with several unrelated functional changes that should have been multiple commits, which impedes systematic testing of each individual change for bugs, and overall wastes my time, causing less development to get done.

If possible, could you try my map suggestion with the nearest builds bookending these commits? It doesn't look like a build was produced between 5075 and 5079, so I will follow up with you if those end up being the problem.

[AdrenalinDragon]

Hendricks266, on 23 October 2016 - 02:16 PM, said:

What do you mean?

It's when some of the buttons normally visible in the game are now hidden and you can't see them, but they still function. I thought it might be related to the flickering bug, so I gave it a mention.

As for "Can't load game", I meant as in not being able to start the game at all, so that's my bad. Again, just thought it was worth mentioning in case it was related, maybe.

I haven't bumped into the flickering bug with eduke32_win32_debug_20161020-5905 yet, but I'll test out that flickering button in Duke Carribean's "The Wavemistress" level.