Messages

1

Linux / Re: gtksourceview2 is unmaintained, port to gtksourceview3?

« on: October 31, 2018, 09:37:41 pm »

Hi,

ok, it seems the removal of gtksourceview2 has been postponed for another release cycle. So there is no need to change anything at the moment. However expect that in one or two years, the library is history and has to be replaced by something. 

2

Linux / Re: gtksourceview2 is unmaintained, port to gtksourceview3?

« on: October 28, 2018, 06:45:23 pm »

Hi,

I would be fine with an embedded version of GtkSourceView2. Normally this is not good packaging practice but in this case I think it is OK. I doubt there is a great risk of security vulnerabilities.

3

Linux / Re: gtksourceview2 is unmaintained, port to gtksourceview3?

« on: October 19, 2018, 01:32:45 pm »

Hi,

was there any progress with porting Radiant to GTKSourceView3 or even GTKSourceView4? The moment of truth has arrived and our Gnome team wants to remove GTKSourceView2 which has seen its last release in 2010.

Alternatively I could probably delay the removal by taking over maintainership of GtkSourceView2 but then in two years it really has to go.

Btw would it be possible to upgrade libpng in Buster? As I see testing still has 6.2 which is quite old. (8.0 we are compatible with for sure, 9.0 maybe not..)

hth,
-geever

libpng in Debian is pretty much up-to-date (current upstream version is 1.6.35 and Debian has 1.6.34)

4

Bugs in stable version (2.5) / Re: Game crashing on base attack that has antimatte storage building

« on: March 04, 2018, 05:10:44 am »

Just FYI: Ubuntu uses the same packages as Debian in case of UFO:AI (and many, many other software) That's the cool thing, if you get something into Debian it automatically lands in all derivatives at some time. 

The game itself is built without debugging symbols but...there are dedicated -dbgsym packages which contain all those information. Just install ufoai-dbgsym and you should be fine. All architecture-dependent packages have a -dbgsym package nowadays.

5

Linux / Re: gtksourceview2 is unmaintained, port to gtksourceview3?

« on: March 04, 2018, 04:44:49 am »

Thanks for the info. Do you happen to know if all GTK2 libraries will be gone or only the sourceview one is the problem? Also how much time we have to resolve the dependency issue?

I believe to use gtksourceview3 we need to port the whole mapeditor to gtk3 which is quite a big task - especially for me, I never worked with GTK before.

We have a few options: (from best to worst)

• Port UFORadiant to GTK3
• Switch to GTK-SourceView3 but stay on GTK2 (90% sure it is not possible, helps only if GTK2 will be still supported)
• Remove the TextEditor feature from UFORadiant (helps only if GTK2 will be still supported)
• Remove UFORadiant debian package (the game runs fine without it), upgrade to GTK3 later

So, worst case Debian repo will lose the map editor (temporarily), but obviously not ideal.

- geever

Here is the complete list with all obsolete Gnome libraries:
https://lists.debian.org/debian-devel/2017/10/msg00300.html

At the moment only gtksourceview2 is a problem. I expect the next Debian development freeze at the end of the year so there is still some time left. We could also wait a little bit longer. At the moment the bug is not release-critical but it is very likely that this will change in the future.

I agree that porting to GTK3 would be ideal but I guess we could live with removing the Texteditor feature which seems to be the most reasonable and efficient solution for now. I would also take solution 2 if that's doable.  Removing UFORadiant is really a worst case scenario and I don't think it will be necessary.

P.S.: I know that you're short on contributors but if you want to get 2.6 into Debian I suggest to set autumn 2018 at the latest as a target otherwise it would be probably too late for Debian 10 aka Buster. It depends mostly on how quickly I can wrap my mind around the new build system and if there are any new (build-)dependencies although I believe we have everything related to C++ already 

6

Linux / gtksourceview2 is unmaintained, port to gtksourceview3?

« on: March 03, 2018, 04:48:26 pm »

Hi,

I just wanted to forward a bug report to you. Apparently gtksourceview2 is obsolete and unmaintained upstream and gtksourceview3 is now the next best thing.

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=885685

AFAICT SourceView.cpp is the only file that makes use of this library. Not sure how much work is required to port radiant to gtksourceview3 but I wanted to let you know anyway.

7

Linux / Re: Official Debian release for UFO:AI

« on: July 21, 2017, 01:01:56 am »

It seems a lot has changed in regard to the build system since 2.5. I presume it will take some time to update the current Debian packages and to understand how everything is supposed to work. I don't intend to provide Git snapshots as *.deb packages but if the release of 2.6 is imminent, please ping me and I will try my best to package the new version.

8

Bugs in stable version (2.5) / Re: Segmentation fault Item::getReactionFireWeaponType (this=0x0)

« on: May 07, 2017, 07:36:22 pm »

Thank you for the quick fix!

9

Bugs in stable version (2.5) / Segmentation fault Item::getReactionFireWeaponType (this=0x0)

« on: May 07, 2017, 02:36:58 pm »

Hi,

we had to recompile ufoai a few days ago because of the bug in ufo2map which we had discussed in a separate thread. Now the following happens: If I start a skirmish, I can already see the map but before all team members appear ufoai crashes with a segmentation fault. This is Debian bug https://bugs.debian.org/861979.

gdb shows a null pointer and some memory that can't be accessed. Any ideas why the game is suddenly crashing at this position? My debug.log is attached.

10

Linux / Re: ufo2map segmentation fault

« on: April 25, 2017, 05:49:50 pm »

It turned out that there was an uninitialized variable but it got already fixed

https://sourceforge.net/p/ufoai/code/ci/bae827ba025ff1d3a56b402c27aa84401914acc4/

All good now

11

Linux / ufo2map segmentation fault

« on: April 19, 2017, 07:40:14 pm »

Hi,

In Debian I build the maps for UFO:AI 2.5 from source. This works most of the time. However on i386 (32bit) and rarely even on amd64 (64bit) I get a segmentation fault, something like

/usr/bin/ufo2map -v 4 -nice 19 -quant 4 -soft maps/oriental.map
> build/maps.mk:46: recipe for target 'base/maps/oriental.bsp' failed
> make[2]: *** [base/maps/oriental.bsp] Segmentation fault

Is this a known issue and possibly fixed in the 2.6 branch? Is there a compile time option to increase the verbosity of ufo2map?
You should be able to reproduce this with the official Debian package (ufoai-maps) https://tracker.debian.org/pkg/ufoai-maps.
I'm reporting this issue because there is currently an open Debian bug #860680 https://bugs.debian.org/860680 and I would like to fix it.

Any ideas?

12

Linux / Re: Official Debian release for UFO:AI

« on: January 14, 2016, 11:59:53 pm »

Once 2.6 is released and has been declared stable, I will package it for Debian.

13

Linux / Re: Official Debian release for UFO:AI

« on: January 13, 2016, 01:32:38 pm »

Finally UFO:AI is now officially in Debian and can be installed the usual way with

Code: [Select]

apt install ufoai
It will automatically migrate to Ubuntu and will be part of the next release 16.04.

https://tracker.debian.org/pkg/ufoai

14

Linux / Re: PlayDeb.net has 2.4.1 on it instead of 2.5

« on: October 17, 2014, 07:23:45 pm »

You might also find my Personal Package Archive (PPA) useful. It is based on the not-yet-official Debian package.

https://launchpad.net/~apoleon/+archive/ubuntu/ufoai-ppa

It is targetted for Ubuntu 14.10 though.

15

Linux / Re: Debian: feedback and tests welcome

« on: October 09, 2014, 09:17:58 pm »

So, any other issues remaining?

Thanks. I'm currently not aware of other issues except the random segmentation fault when I compile the maps from scratch. Since I can't reproduce it I have a hard time with debugging the problem. Any idea what needs to be done to make the map compilation as verbose as possible? Perhaps this would provide the necessary hint in the future.