Mr. Picklesworth
October 18th, 2007, 03:31 PM
Okay, this forum is going to bug reports early ;)
There are still some Gutsy issues to clear up, and if they aren't fixed there with updates let's hope they get into Hardy.
Somebody mentioned in a thread that hardly anyone plays gnubots. This led me to a thought: Since people who ran the Gutsy devel versions had probably played the game before (it's fun the first few times), it was unlikely to have been tested much at all. Indeed, playing it just now, I have discovered a horrifying mess of bugs that I can only hope stem from a single source...
The trouble is that this is such a horrifying mess of bugs I cannot figure out where one ends and the other begins. I need help! We must return gnubots2 to its past glory by solving these issues.
-Run gnobots2 in the terminal
-Press Teleport Randomly continually, until the player loses
-(Repeat if necessary?)
You will notice that the game crashes when you die. Oddly, this seems to happen only after a certain number of teleports. I can't pin it to any exact pattern, but it happens very frequently! (While we're here: Gameplay failure! The only way to die now is to teleport randomly into an enemy. Since when were arcade games meant to be idiot-proof?).
When the game crashes, the following appears for me in the output:
(gnobots2:11221): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `start <= stop' failed
sh: jackd: command not found
Segmentation fault
The jackd not found error has been reported here (https://bugs.launchpad.net/ubuntu/+source/ubuntu-meta/+bug/154028), and happens absolutely everywhere that a sound is played with gstreamer, with no visible consequences. After installing the jackd package, the crash still occurs. Of interest may be that the segfault occurs before the program tries to run jackd.
There's more!
Whenever I teleport (the - quite unfortunate sounding - teleport sound plays), that critical GStreamer error appears in the output. (gnobots2:11711): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `start <= stop' failed
Actually, that critical error seems to happen with every game in the gnome-games package, almost whenever a sound is played. (Iagno, gnobots2, gnibbles, gnometris with sounds enabled). Teleporting in gnobots2 is just the easiest and quickest way to reproduce that...
Interestingly, the startup noise in gnometris does not give the critical error, though every sound in-game does.
Finally, the AIEEEE sound when the player dies in gnobots2. This sound plays whenever the player dies by Waiting via the toolbar or menu options. (Bug with Wait command: Clicking on the player to wait behaves differently from the toolbar / menu actions. The clicking on the player wait is designed such that somebody cannot lose by doing it).
However, the sound does not play if the play if the player dies by teleporting randomly into an enemy, even though the visual feedback is the same. Does not seem related to the crash, as in the rare instance that one dies and the game does not crash, that sound still does not play.
However, when the player dies by using the Wait toolbar / menu button, the critical gStreamer error does not happen! (Though the game still seg faults, on occasion).
sh: jackd: command not found is output when the AIEEE sound starts playing. The seg fault occurs when the sound finishes playing.
So, lots of corners. I can't find my way around this :(
Anyone here good at bug reporting? I can't decide how to report all of it...
There are still some Gutsy issues to clear up, and if they aren't fixed there with updates let's hope they get into Hardy.
Somebody mentioned in a thread that hardly anyone plays gnubots. This led me to a thought: Since people who ran the Gutsy devel versions had probably played the game before (it's fun the first few times), it was unlikely to have been tested much at all. Indeed, playing it just now, I have discovered a horrifying mess of bugs that I can only hope stem from a single source...
The trouble is that this is such a horrifying mess of bugs I cannot figure out where one ends and the other begins. I need help! We must return gnubots2 to its past glory by solving these issues.
-Run gnobots2 in the terminal
-Press Teleport Randomly continually, until the player loses
-(Repeat if necessary?)
You will notice that the game crashes when you die. Oddly, this seems to happen only after a certain number of teleports. I can't pin it to any exact pattern, but it happens very frequently! (While we're here: Gameplay failure! The only way to die now is to teleport randomly into an enemy. Since when were arcade games meant to be idiot-proof?).
When the game crashes, the following appears for me in the output:
(gnobots2:11221): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `start <= stop' failed
sh: jackd: command not found
Segmentation fault
The jackd not found error has been reported here (https://bugs.launchpad.net/ubuntu/+source/ubuntu-meta/+bug/154028), and happens absolutely everywhere that a sound is played with gstreamer, with no visible consequences. After installing the jackd package, the crash still occurs. Of interest may be that the segfault occurs before the program tries to run jackd.
There's more!
Whenever I teleport (the - quite unfortunate sounding - teleport sound plays), that critical GStreamer error appears in the output. (gnobots2:11711): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `start <= stop' failed
Actually, that critical error seems to happen with every game in the gnome-games package, almost whenever a sound is played. (Iagno, gnobots2, gnibbles, gnometris with sounds enabled). Teleporting in gnobots2 is just the easiest and quickest way to reproduce that...
Interestingly, the startup noise in gnometris does not give the critical error, though every sound in-game does.
Finally, the AIEEEE sound when the player dies in gnobots2. This sound plays whenever the player dies by Waiting via the toolbar or menu options. (Bug with Wait command: Clicking on the player to wait behaves differently from the toolbar / menu actions. The clicking on the player wait is designed such that somebody cannot lose by doing it).
However, the sound does not play if the play if the player dies by teleporting randomly into an enemy, even though the visual feedback is the same. Does not seem related to the crash, as in the rare instance that one dies and the game does not crash, that sound still does not play.
However, when the player dies by using the Wait toolbar / menu button, the critical gStreamer error does not happen! (Though the game still seg faults, on occasion).
sh: jackd: command not found is output when the AIEEE sound starts playing. The seg fault occurs when the sound finishes playing.
So, lots of corners. I can't find my way around this :(
Anyone here good at bug reporting? I can't decide how to report all of it...