Patch 0.80.4.2 Released

[Auswaschbar]

After just 2 weeks, another release with a few more bugfixes:

• fix crash when doing lots of mapdrawing (#1632)
• fix crash when loading certain cursornormal files
• fix drawing of ghosted buildings
• properly initialise SDL timer for dedicated server
• fixed GatherWait cursor
• don't send resign-command when already spectating
• fix units with vlaunch missiles (#1647)
• Spring compiles with boost 1.40 (no, doesn't...)

Executable only, replace spring.exe with this
Complete installers can be found on the download page like usual.

Also, if you have a dedicated server running, you should upgrade to this version.

[cleanrock]

i think you forgot to increase version number to .2 .. splash screen still says 0.80.4.1

[AF]

If you issue a .3 to fix that, can you also include the necessary 1 liner needed to make the AI ladder implementation easy?

[aegis]

If you issue a .3 to fix that, can you also include the necessary 1 liner needed to make the AI ladder implementation easy?

wait what

edit: also, AI ladder can easily run a custom version of spring

[AF]

imbaczek iirc said it was a simple console output 1 liner on AI death.

As for custom spring, erm, nope, I'd rather that anybody could do it with the task at hand. Maintaining a fork of the spring project that has a single line changed seems like an absurd idea

[aegis]

you could also do it with a widget?
the only thing I'm waiting for is headless spring - tis what will really make the ai ladder implementation easy

[AF]

Then write said widget or tell how to make it, lua widgets and gadgets aren't my forte

[hoijui]

why not play a game with AIs, and wait til one dies.. you may see that there are some messages already, and maybe they are what you want.

[Umrug]

Weak machine owners committee is asking: why isn't this http://springrts.com/phpbb/viewtopic.php?f=1&t=20338 in? (The planes on patrol fix)?

[Tobi]

Would require bumping required version (ie. 0.80.5.0 and forcing upgrade to everyone), as it is a change to 'synced' code.

(Although the change should sync, but it' s very risky.)

[hoijui]

i guess cause the change is not trivial, and Auswaschbar did not want to do sync testing just for that change. Or maybe he knows that it is sync relevant. Though, if it is not the later, it would be worth to test syncness, as it would really help a lot of users. So if it is not the later, then maybe Aus can say so, and we could test the commit for syncedness.

[Tobi]

Maybe hard to test all border cases though.. if there's just one place in the code where CQuadField::Quad::teamUnits doesn't match exactly with CQuadField::Quad::units then it may break, and it may be very hard to find this.

Personally I rather go for 0.80.5 / 0.81.0 anyway in not too much time, as I did many changes to Lua unit scripts too