8:11 a.m.
Hi all,
I am Tecnoballz maintainer.
Currently Tecnoballz is failing to build in rawhide:
https://bugzilla.redhat.com/show_bug.cgi?id=1308179
There is a new upstream version (0.93.1) and it builds without problems. I
also added Debian/upstream patches.
In order to allow a shared scoreboard file, the Games Packaging guidelines
strongly suggest to make the executable setgid 'games' and to drop setgid
privileges as soon as possible:
https://fedoraproject.org/wiki/SIGs/Games/Packaging
I tried to port the patch Hans De Goede did for version 0.92:
https://bugzilla.redhat.com/show_bug.cgi?id=234563#c4
http://pkgs.fedoraproject.org/cgit/tecnoballz.git/tree/tecnoballz
-0.92-dropsgid.patch
But since I don't know C++ I didn't succeed.
I hope someone can help me with this task.
Work in progress spec and src.rpm can be found here:
https://dl.dropboxusercontent.com/u/12575912/reviews/tecnoballz.spec
https://dl.dropboxusercontent.com/u/12575912/reviews/tecnoballz-0.93.1-1....
Thanks,
Andrea
10:51 a.m.
Hi Andrea,
known issue with gcc6 enforcing narrowing.
I've uploaded a patch into bugzilla. Please poke upstream and send the patch asking
for removal of the dirty variable typing.
~R.
11:21 a.m.
Hi Raphael,
thanks for the patch but I was aiming at having v0.93.1 into Fedora.
This already compiles fine (not as v0.92).
What I miss is to port the dropsgid patch to 0.93.1 (I tried and I failed).
Please have a look at:
https://dl.dropboxusercontent.com/u/12575912/reviews/tecnoballz-0.93.1-1....
If you don't think it is feasible I'll release another v0.92 with your
patch to fix FTBFS.
Thanks.
BR,
Andrea
On Sat, Mar 26, 2016 at 4:51 PM, Raphael Groner <raphgro(a)fedoraproject.org>
wrote:
Hi Andrea,
known issue with gcc6 enforcing narrowing.
I've uploaded a patch into bugzilla. Please poke upstream and send the
patch asking for removal of the dirty variable typing.
~R.
_______________________________________________
games mailing list
games(a)lists.fedoraproject.org
http://lists.fedoraproject.org/admin/lists/games@lists.fedoraproject.org
11:29 a.m.
Hi Andrea,
Am 26.03.2016 um 17:21 schrieb Andrea Musuruane:
Hi Raphael,
thanks for the patch but I was aiming at having v0.93.1 into
Fedora. This already compiles fine (not as v0.92).
What I miss is to port the dropsgid patch to 0.93.1 (I tried and I
failed).
Please have a look at:
https://dl.dropboxusercontent.com/u/12575912/reviews/tecnoballz-0.93.1-1....
If the source changes a lot, the patch command is not able any more to
(re)locate the hunks. In such a case a patch needs to get rebased, as
you already suggested. You can do so by manually comparing the hunks
with the new source.
If you don't think it is feasible I'll release another v0.92
with your
patch to fix FTBFS.
Sorry I can not help any further due to lack of free time ATM.
~R.
11:39 a.m.
On Sat, Mar 26, 2016 at 5:29 PM, Raphael Groner <projects.rg(a)smart.ms>
wrote:
Hi Andrea,
Am 26.03.2016 um 17:21 schrieb Andrea Musuruane:
Hi Raphael,
thanks for the patch but I was aiming at having v0.93.1 into Fedora.
This already compiles fine (not as v0.92).
What I miss is to port the dropsgid patch to 0.93.1 (I tried and I failed).
Please have a look at:
https://dl.dropboxusercontent.com/u/12575912/reviews/tecnoballz-0.93.1-1....
If the source changes a lot, the patch command is not able any more to
(re)locate the hunks. In such a case a patch needs to get rebased, as you
already suggested. You can do so by manually comparing the hunks with the
new source.
I rebased the patch but it seems that either gcc behaviour changed (I have
error when compiling) or I rebased it incorrectly. My attempt is in the
src.rpm
Thanks.
Andrea
11:50 a.m.
Personally I don't see the point of going through the effort for a
shared scoreboard. When I fixed up Ostrich Riders, I specifically put
the scoreboard in the user directory rather than go through that
effort.
The failed build gives this error:
tiles_background.cc:748:35: error: narrowing conversion of '128' from
'int' to 'char' inside { } [-Wnarrowing]
From looking at the code you have a big array defined in
tiles_background.cc. If you add in casts it get rid of the error. So
for this:
char x[2] = {0x30, 0xF2};
change it to:
char x[2] = {0x30, (char)0xF2};
Granted it might be easier to just upgrade to the new version.
On Sat, 2016-03-26 at 14:11 +0100, Andrea Musuruane wrote:
Hi all,
I am Tecnoballz maintainer.
Currently Tecnoballz is failing to build in rawhide:
https://bugzilla.redhat.com/show_bug.cgi?id=1308179
There is a new upstream version (0.93.1) and it builds without
problems. I also added Debian/upstream patches.
In order to allow a shared scoreboard file, the Games Packaging
guidelines strongly suggest to make the executable setgid 'games' and
to drop setgid privileges as soon as possible:
https://fedoraproject.org/wiki/SIGs/Games/Packaging
I tried to port the patch Hans De Goede did for version 0.92:
https://bugzilla.redhat.com/show_bug.cgi?id=234563#c4
http://pkgs.fedoraproject.org/cgit/tecnoballz.git/tree/tecnoballz-0.9
2-dropsgid.patch
But since I don't know C++ I didn't succeed.
I hope someone can help me with this task.
Work in progress spec and src.rpm can be found here:
https://dl.dropboxusercontent.com/u/12575912/reviews/tecnoballz.spec
https://dl.dropboxusercontent.com/u/12575912/reviews/tecnoballz-0.93.
1-1.fc23.src.rpm
Thanks,
Andrea
_______________________________________________
games mailing list
games(a)lists.fedoraproject.org
http://lists.fedoraproject.org/admin/lists/games@lists.fedoraproject.
org
1:54 p.m.
On Sat, Mar 26, 2016 at 5:50 PM, Dennis Payne <dulsi(a)identicalsoftware.com>
wrote:
Personally I don't see the point of going through the effort for
a shared
scoreboard. When I fixed up Ostrich Riders, I specifically put the
scoreboard in the user directory rather than go through that effort.
The failed build gives this error:
[..]
Granted it might be easier to just upgrade to the new version.
I just submitted and built tecnoballz-0.92 with the GCC6 patch Raphael
hinted for F24 and rawhide.
My aim is to have 0.93.1 in Fedora. I think having a user scoreboard
instead of a shared one is also fine. But I need help with this too.
Bye,
Andrea
2947
days inactive
2947
days old
6 comments
4 participants
participants (4)
-
Andrea Musuruane -
Dennis Payne -
Raphael Groner -
Raphael Groner