Planning for 0.77 Release

[Crayfish]

What was it?

[imbaczek]

several issues with pathing. it's not fixed yet, but a workaround that stops desyncs is in place. full-scale testing can commence.

[el_matarife]

Excellent, when is everyone planning the first large scale multiplayer tests?

[imbaczek]

ASAP, maybe tonight? (that is, like, in 20 minutes?)

What's the ubuntu timelimit btw? (with timezone?)

[panzeriv2]

A teeny tiny question:
Is bumpwater designed to work using ATi Radeon? It can't seem to work on my comp though...

[jK]

Designed for ATi is nothing, there is just OpenGL and stuff that works on one card should work on any other, too. (as long it supports the same extensions)
So you should call it ati driver bugfixing :p

det helped a bit, so most parts of it should work on HD-series cards (except shorewaves). Because ATi only fixes their drivers for newer cards it doesn't have to work on older series.
So you can just try to disable parts of it like shorewaves etc. and check if it works then.

[panzeriv2]

OK did a bit more testing on your Bumpwater offline (mind you) and....
here are the results...

First test:
Everything:- No no. WTH

Second test:
Everything without shorewaves:-Got a big blank. WTH

Third test:
without softening of the beach/whatever:- Got a big blank. WTH

Forth test:
Without blur reflections:- Got a big blank. WTH

Fifth test:
Just ol' bumpwater: Loaded. Hooray!

Sixth test:
reflections:- Got a big blank.WTH

Seventh test:
Everything without shorewaves and reflections:- Loaded. Hooray!

So it seems that reflections cannot be loaded on my Radeon...which is a X1050, typically a 2007 model, but it works. Hooray!

Strangely, when I entered "/water 4" into Spring, it shows nuisance (for me) except this "no vertex shader defined." X1050 HAS support for this vertex shader thing-thing-thing. Oh well, that's just the shorewaves...

And one more thing: what is Atihacks?

[jK]

"no vertex shader defined."

Seems as ATi's have issues with shader programs, which only use one of shader stages and rely in the other on the Fixed Function Pipeline.
Good to know, perhaps I can get the other stuff now working, too :)
(btw ATi's error messages are always such vacant, that makes the bugfixing so hard on ATi's)

AtiHacks enables a number of ATi driver bugfixes, it should be enabled by default on ATi's as long as you didn't run Settings.exe before Spring.exe or already started an older svn build. If it was turned off, it is also possible that reflections work with bumpwater after you enabled it.

[tberthel]

6485 seems to work. What is needed?

[panzeriv2]

AtiHacks enables a number of ATi driver bugfixes, it should be enabled by default on ATi's as long as you didn't run Settings.exe before Spring.exe or already started an older svn build. If it was turned off, it is also possible that reflections work with bumpwater after you enabled it.

So how do I enable it?

[hoijui]

with settings.exe

[panzeriv2]

Where....

[Patch]

Installed a fresh windows XP, 8.9 catalyst, svn6486, 6492

Now, our problem with crashing mothership (or something) is fixed, but i get some visual problems, ulpoaded 3 pictures from the game, and 1 from settings.

Maybe it's helps in ati problems.

[Argh]

1. Bug found, "Unit Reply Volume" registry setting does NOT result in changes to actual volume of those sounds.

2. Bug found, Lua reports a "recursive" error with Spring.DestroyUnit() callouts in Lua, if included within a UnitKilled() callin. This error is displayed whether or not the DestroyUnit callout refers back to the UnitID of the unit that has been killed this gameframe.

It's just an error message, and does not cause a halting state or serious problems, since I'm not trying to destroy units that are already killed, for obvious reasons, but I'd appreciate if somebody could find it and kill it.

[Argh]

Annoyance found:

When a Unit is selected, Spring draws the text, "Selected units #". I can't seem to find a reference to this anywhere in Lua, so I'm guessing that this hardcoded as to position and value. Can this get removed or at least turn-offable? Sorry, it really appears that this is hardcoded, I've searched through pretty much every file in Spring, trying to find a reference to this behavior, and came up with nothing.

[Kloot]

Yes, it is hardcoded (CGuiHandler::DrawSelectionInfo()).

[jK]

no, it's not ..

PS for the lazy ones: ctrlpanel.txt -> [x|y]SelectionPos

[Argh]

I don't want the position... I just want the text removed entirely. Duplicating it in Lua is trivial.

[imbaczek]

lua callout provided for control, see one of the recent revisions.

[jK]

I don't want the position... I just want the text removed entirely. Duplicating it in Lua is trivial.

negative values, values >1, ... be a bit creative