issue 3
(Hang with exception) reported by wac
- Looks like the worker thread is crashing.
gawker_img_fetch.py [19859]: No host and port for service.
Traceback (most recent call last):
File "gawker-image-fetch/Zeroconf.py", line 863, in run
self.readers[socket].handle_read()
File "gawker-image-fetch/Zeroconf.py", line 908, in handle_read
msg = DNSIncoming(data)
File "gawker-image-fetch/Zeroconf.py", line 469, in __init__
self.readQuestions()
File "gawker-image-fetch/Zeroconf.py", line 495, in readQuestions
question = DNSQuestion(name, info[0], info[1])
File "gawker-image-fetch/Zeroconf.py", line 262, in __init__
raise NonLocalNameException
NonLocalNameException
Looks like the worker thread is crashing.
gawker_img_fetch.py [19859]: No host and port for service.
Traceback (most recent call last):
File "gawker-image-fetch/Zeroconf.py", line 863, in run
self.readers[socket].handle_read()
File "gawker-image-fetch/Zeroconf.py", line 908, in handle_read
msg = DNSIncoming(data)
File "gawker-image-fetch/Zeroconf.py", line 469, in __init__
self.readQuestions()
File "gawker-image-fetch/Zeroconf.py", line 495, in readQuestions
question = DNSQuestion(name, info[0], info[1])
File "gawker-image-fetch/Zeroconf.py", line 262, in __init__
raise NonLocalNameException
NonLocalNameException
Jun 07, 2009
issue 2
(Script crashes while unattended) reported by wac
- What steps will reproduce the problem?
1. Leave script running.
2. Have socket failure for some reason.
What is the expected output? What do you see instead?
Expect continued operation, instead termination with the following traceback:
gawker_img_fetch.py [90316]: Error in read loop
Traceback (most recent call last):
File "gawker_img_fetch.py", line 339, in main
obj.read_loop()
File "gawker_img_fetch.py", line 99, in read_loop
elif ready_dict[self.sock.fileno()] & select.POLLIN:
File "<string>", line 1, in fileno
File "/usr/local/lib/python2.5/socket.py", line 146, in _dummy
raise error(EBADF, 'Bad file descriptor')
error: (9, 'Bad file descriptor')
Please use labels and text to provide additional information.
What steps will reproduce the problem?
1. Leave script running.
2. Have socket failure for some reason.
What is the expected output? What do you see instead?
Expect continued operation, instead termination with the following traceback:
gawker_img_fetch.py [90316]: Error in read loop
Traceback (most recent call last):
File "gawker_img_fetch.py", line 339, in main
obj.read_loop()
File "gawker_img_fetch.py", line 99, in read_loop
elif ready_dict[self.sock.fileno()] & select.POLLIN:
File "<string>", line 1, in fileno
File "/usr/local/lib/python2.5/socket.py", line 146, in _dummy
raise error(EBADF, 'Bad file descriptor')
error: (9, 'Bad file descriptor')
Please use labels and text to provide additional information.
May 31, 2009
Revision e6b8709127
(Fix patch to not have absolute paths.)
pushed by wac
- Fix patch to not have absolute paths.
Fix patch to not have absolute paths.
May 30, 2009
issue 1
(OS X 10.5 Python build lacks select.poll()) Status changed by wac
- This should now be fixed in revision 2b4b6cd605
Revision 2b4b6cd605
(Use select.select() on platforms where select.poll() is unav...)
pushed by wac
- Use select.select() on platforms where select.poll() is unavailable.
Use select.select() on platforms where select.poll() is unavailable.
14 new revisions
pushed by wac
- Revision 2dbfee8a43:Initial directory structure.Revision 0a1b0b9789:Now with zeroconf features and supporting library.Revision d424955f16:More bug fixes in the Zeroconf module.Revision 01ccc87a75:Ignore pyc generated by normal running.Revision 4104027674:Fix a file descriptor leakRevision eaa7f32e1b:* New feature to save a time-stamped copy of every image to a directory. It's
worth noting that this can consume dozens of megabytes per day, be careful
that there is adequate capacity before turning it on.
* Fixed a DeprecationWarning in the Zeroconf code. (A float was being passed
to struct.pack for an integer field, there are probably more like this.)Revision 5a65b79568:* Add scripts to take the collected movie files and turn them into a
H.264-encoded mpeg4 file. These scripts presume a 640x480 source among other
things.Revision e8702b9885:Sticking a fifo in to work around the not-so-smartness of x264.Revision 97f5ec6a3b:x264 is more ridiculous than I had first anticipated.Revision 586ebe1786:More working around x264 craziness.Revision e903a4e657:Go back to using pipelining with a patch to x264 so that it treats stdin as a
YUV4MPEG stream even if "-" doesn't end in .y4m. This results in a faster run
and avoids occupying several hundred megs of disk with the intermediate
YUV4MPEG file.Revision 5aefaf6353:jpeg2yuv is apparently not recommended for non-debugging use, so now there is
a pipeline with jpegtopnm, ppmtoy4m, and x264 to generate the movie. This also
resolves the need for all the hardlinks to make jpeg2yuv do the right thing.Revision a6e0d4574d:Some refactoring and add use of qt-faststartRevision 543003cfb3:x264 changes and other cleanup
Revision 2dbfee8a43:Initial directory structure.Revision 0a1b0b9789:Now with zeroconf features and supporting library.Revision d424955f16:More bug fixes in the Zeroconf module.Revision 01ccc87a75:Ignore pyc generated by normal running.Revision 4104027674:Fix a file descriptor leakRevision eaa7f32e1b:* New feature to save a time-stamped copy of every image to a directory. It's
worth noting that this can consume dozens of megabytes per day, be careful
that there is adequate capacity before turning it on.
* Fixed a DeprecationWarning in the Zeroconf code. (A float was being passed
to struct.pack for an integer field, there are probably more like this.)Revision 5a65b79568:* Add scripts to take the collected movie files and turn them into a
H.264-encoded mpeg4 file. These scripts presume a 640x480 source among other
things.Revision e8702b9885:Sticking a fifo in to work around the not-so-smartness of x264.Revision 97f5ec6a3b:x264 is more ridiculous than I had first anticipated.Revision 586ebe1786:More working around x264 craziness.Revision e903a4e657:Go back to using pipelining with a patch to x264 so that it treats stdin as a
YUV4MPEG stream even if "-" doesn't end in .y4m. This results in a faster run
and avoids occupying several hundred megs of disk with the intermediate
YUV4MPEG file.Revision 5aefaf6353:jpeg2yuv is apparently not recommended for non-debugging use, so now there is
a pipeline with jpegtopnm, ppmtoy4m, and x264 to generate the movie. This also
resolves the need for all the hardlinks to make jpeg2yuv do the right thing.Revision a6e0d4574d:Some refactoring and add use of qt-faststartRevision 543003cfb3:x264 changes and other cleanup
May 27, 2009
issue 1
(OS X 10.5 Python build lacks select.poll()) reported by wac
- What steps will reproduce the problem?
1. Attempt to run gawker-image-fetch script on OS X 10.5 default python install.
What is the expected output? What do you see instead?
Normal operation.
gawker_img_fetch.py [77756]: Starting zeroconf service browser
gawker_img_fetch.py [77756]: Creating reader object
gawker_img_fetch.py [77756]: <__main__.GawkerListener object at 0x2ed890>
Traceback (most recent call last):
File "gawker_img_fetch.py", line 348, in < module >
if __name__ == "__main__": sys.exit(main())
File "gawker_img_fetch.py", line 335, in main
options.directory)
File "gawker_img_fetch.py", line 60, in __init__
self.poller = select.poll()
AttributeError: 'module' object has no attribute 'poll'
Please use labels and text to provide additional information.
Python 2.5.1 (r251:54863, Feb 6 2009, 19:02:12)
[GCC 4.0.1 (Apple Inc. build 5465)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import select
>>> select.poll
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'poll'
>>> dir(select)
['__doc__', '__file__', '__name__', 'error', 'select']
Should probably fall back to select.select() in this case, or display a meaningful error message.
(Especially for OS X users who can work around this by getting a noncrippled python build from
python.org.)
What steps will reproduce the problem?
1. Attempt to run gawker-image-fetch script on OS X 10.5 default python install.
What is the expected output? What do you see instead?
Normal operation.
gawker_img_fetch.py [77756]: Starting zeroconf service browser
gawker_img_fetch.py [77756]: Creating reader object
gawker_img_fetch.py [77756]: <__main__.GawkerListener object at 0x2ed890>
Traceback (most recent call last):
File "gawker_img_fetch.py", line 348, in < module >
if __name__ == "__main__": sys.exit(main())
File "gawker_img_fetch.py", line 335, in main
options.directory)
File "gawker_img_fetch.py", line 60, in __init__
self.poller = select.poll()
AttributeError: 'module' object has no attribute 'poll'
Please use labels and text to provide additional information.
Python 2.5.1 (r251:54863, Feb 6 2009, 19:02:12)
[GCC 4.0.1 (Apple Inc. build 5465)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import select
>>> select.poll
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'poll'
>>> dir(select)
['__doc__', '__file__', '__name__', 'error', 'select']
Should probably fall back to select.select() in this case, or display a meaningful error message.
(Especially for OS X users who can work around this by getting a noncrippled python build from
python.org.)
