Positron 1.1
Printed from: Neuros Forums
Topic URL: http://www.neurosaudio.com/community/forum/topic.asp?TOPIC_ID=3185
Printed on: 03/24/2006
Topic:
Topic author: nxg125
Subject: Positron 1.1 Just got this from the neurosetta mailing list...
Posted on: 04/01/2004 9:52:02 PM
Message:
--Nick
I'm happy to announce that after *mumble,mumble* months, we have
released a new version of positron at:
http://www.xiph.org/positron/
I know it's April Fool's Day, but the only joke (and not the funny kind)
is how long it took to release this version.
Here's the changes:
* mp3_support option added to permit MP3 support to be disabled.
Handy for testing firmwares that lack MP3 support.
* Fixed serious bug that caused deletion of multiple files to be done
incorrectly.
* Positron now preserves the access and modification times of files
when copying them to the Neuros, allowing the use of standard
synchronization tools like rsync.
* Major improvement of MP3 detection and parsing code.
* Vastly improved database sorting, including sorting by track number.
Note that you will need to use "positron rebuild" to resort your
existing databases.
* Database handling fix that forces firmwares after 1.41 to cache
the database when it is modified.
* Informational messages when scanning directories for files so
you know something is going on.
* Better error reporting when database directory is not present.
* Ogg Vorbis support turned on by default.
* Eliminated warnings in Python 2.3 about character encodings
in *_zip.py files.
Enjoy!
---
Stan Seibert
Replies:
Reply author: nxg125
Replied on: 04/01/2004 9:55:31 PM
Message:
Once I read the message more closely, this jumped out at me:
quote:
* mp3_support option added to permit MP3 support to be disabled. Handy for testing firmwares that lack MP3 support.
Does this mean we're on the road to a free Neuros?
--Nick
Reply author: alecm
Replied on: 04/01/2004 10:20:34 PM
Message:
quote:
Originally posted by nxg125
Once I read the message more closely, this jumped out at me:quote:
* mp3_support option added to permit MP3 support to be disabled. Handy for testing firmwares that lack MP3 support.
Does this mean we're on the road to a free Neuros?
--Nick
Unfortunately it was added back in July 2003, when the preliminary Ogg firmware didn't support MP3 playback.
<egotism>He forgot to mention the new WAV support.</egotism>
It's great to see a new release.
Reply author: nxg125
Replied on: 04/02/2004 10:04:46 AM
Message:
quote:
Unfortunately it was added back in July 2003, when the preliminary Ogg firmware didn't support MP3 playback.
Oh well, maybe it will be needed again some day!
--Nick
Reply author: hal
Replied on: 04/02/2004 2:51:26 PM
Message:
quote:
Originally posted by nxg125
Just got this from the neurosetta mailing list...
--Nick
[code]
I'm happy to announce that after *mumble,mumble* months, we have
released a new version of positron at:
http://www.xiph.org/positron/
Does this version incorporate the playlist patch? I tried to wade thru the source, but I'm not a Python person, and didn't see anything resembling the patches to 1.0.
Also tried to build 1.1 on a Turbolinux Workstation 8.0 system with Python 2.2, and received the following when doing 'positron config':
Traceback (most recent call last):
File "/usr/bin/positron", line 24, in ?
from positron.neuros import Neuros
File "/usr/lib/python2.2/site-packages/positron/neuros.py", line 17, in ?
import db
File "/usr/lib/python2.2/site-packages/positron/db/__init__.py", line 17, in ? from MDB import *
File "/usr/lib/python2.2/site-packages/positron/db/MDB.py", line 101, in ?
class MDB:
File "/usr/lib/python2.2/site-packages/positron/db/MDB.py", line 425, in MDB
def set_modified_flag(self, new_state=True):
NameError: name 'True' is not defined
Research on the xiph mail archives shows this error last year, but no fix. The docs say that 'True' was a new addition in Python 2.2 (which seems to be running), and the same error is reported to; '/usr/bin/python2.2 config'. The positron 1.0 (with playlist patch) I have running is under Turbolinux 10 Desktop and I hesitate to risk that working system by installing 1.1.
Hal
Reply author: alecm
Replied on: 04/02/2004 4:54:40 PM
Message:
quote:
Originally posted by halquote:
Originally posted by nxg125
Just got this from the neurosetta mailing list...
--Nick
[code]
I'm happy to announce that after *mumble,mumble* months, we have
released a new version of positron at:
http://www.xiph.org/positron/
Does this version incorporate the playlist patch? I tried to wade thru the source, but I'm not a Python person, and didn't see anything resembling the patches to 1.0.
Also tried to build 1.1 on a Turbolinux Workstation 8.0 system with Python 2.2, and received the following when doing 'positron config':
Traceback (most recent call last):
File "/usr/bin/positron", line 24, in ?
from positron.neuros import Neuros
File "/usr/lib/python2.2/site-packages/positron/neuros.py", line 17, in ?
import db
File "/usr/lib/python2.2/site-packages/positron/db/__init__.py", line 17, in ? from MDB import *
File "/usr/lib/python2.2/site-packages/positron/db/MDB.py", line 101, in ?
class MDB:
File "/usr/lib/python2.2/site-packages/positron/db/MDB.py", line 425, in MDB
def set_modified_flag(self, new_state=True):
NameError: name 'True' is not defined
Research on the xiph mail archives shows this error last year, but no fix. The docs say that 'True' was a new addition in Python 2.2 (which seems to be running), and the same error is reported to; '/usr/bin/python2.2 config'. The positron 1.0 (with playlist patch) I have running is under Turbolinux 10 Desktop and I hesitate to risk that working system by installing 1.1.
Hal
No playlist patch. And, that's probably for the best. As I was fooling around with positron's DB support today (which was somewhat playlist related), I learned that the playlist support is not just incredibly slow, but also incorrect in a number of ways. The patch still applies though, and works, so feel free to use it. You might want to consider also adding the DB patch I just submitted to positron Bugzilla, so that you can use NDBM to manage your playlists, and still use positron.
ETA: Those values (True, False) should supported in 2.2, and I never had any problems with CVS under 2.2 (though I run 2.3 now). There must be something wrong with your python setup, try looking at the positron script in /usr/bin to make sure the first line is:
#!/usr/bin/python2.2
Reply author: hal
Replied on: 04/02/2004 7:14:46 PM
Message:
quote:
Originally posted by alecmquote:
Originally posted by hal
<snip>
def set_modified_flag(self, new_state=True):
NameError: name 'True' is not defined
No playlist patch. And, that's probably for the best. As I was fooling around with positron's DB support today (which was somewhat playlist related), I learned that the playlist support is not just incredibly slow, but also incorrect in a number of ways. The patch still applies though, and works, so feel free to use it. You might want to consider also adding the DB patch I just submitted to positron Bugzilla, so that you can use NDBM to manage your playlists, and still use positron.
This will probably be the best way. I prefer Positron for the basic data upload/maintenance, but NDBM works better for playlist support to date anyway. I'll try to get the DB patch loaded.
quote:
ETA: Those values (True, False) should supported in 2.2, and I never had any problems with CVS under 2.2 (though I run 2.3 now). There must be something wrong with your python setup, try looking at the positron script in /usr/bin to make sure the first line is:
#!/usr/bin/python2.2
Mine did not have the 2.2, so I edited it in, and received the same error, so I guess it's time to upgrade the distro (or at least Python). Thanks.
Hal
Reply author: alecm
Replied on: 04/02/2004 7:24:39 PM
Message:
quote:
Originally posted by hal
Mine did not have the 2.2, so I edited it in, and received the same error, so I guess it's time to upgrade the distro (or at least Python). Thanks.
That's really strange, because that line was added way back in October, and there have been True and False references in the code since version 1.0. Good luck with the update.
Reply author: hal
Replied on: 04/04/2004 4:14:09 PM
Message:
quote:
Originally posted by alecmquote:
Originally posted by hal
Mine did not have the 2.2, so I edited it in, and received the same error, so I guess it's time to upgrade the distro (or at least Python). Thanks.
That's really strange, because that line was added way back in October, and there have been True and False references in the code since version 1.0. Good luck with the update.
The fix turned out to be downloading source for Python 2.3.3 and compiling it in. Fixed the above problem, but after adding playlists with NDBM, positron barfs. Since there were references to this being fixed in 1.1, I tried a test.
I rebuilt Neuros database with NDBM 1.31 (playlist deleted by this
action). 'positron list' (both 1.0 and 1.1) showed entries of form:
Title: Oh chansayo <Oh my Angel>
Artist: Pearl Sisters Album: Pearl Sisters Hits #3, KLS-29
Genre: None Time: 213s Size: 7754kB
Filename: /mnt/neuros/music/kls29/Track11.ogg
I then built a portion of the deleted playlist with NDBM and executed a sync (with music directories set to not touch the original Positron built music directory on Neuros), after which neither version of positron (1.0 or 1.1) would list the database. The errors differed only in line numbers with 1.1 giving:
===== Database "audio" =====
Traceback (most recent call last):
File "/usr/bin/positron", line 199, in ?
main(sys.argv)
File "/usr/bin/positron", line 181, in main
cmd.run(config, myNeuros, remaining[1:])
File "/usr/local/lib/python2.3/site-packages/positron/cmd_list.py", line
74, in run
records = database.get_records()
File "/usr/local/lib/python2.3/site-packages/positron/db/WOID.py",
line 154, in get_records
return [self._get_record_at(pointer[0]) for pointer in self.sai
File "/usr/local/lib/python2.3/site-packages/positron/db/WOID.py",
line 112, in _get_record_at
mdb_record = self.mdb.read_record_at(pointer)[0]
File "/usr/local/lib/python2.3/site-packages/positron/db/MDB.py", line
212, in read_record_at
values = [struct.unpack(">I", item)[0]
struct.error: unpack str size does not match format
Hal
Reply author: alecm
Replied on: 04/04/2004 11:52:25 PM
Message:
quote:
Originally posted by hal
The fix turned out to be downloading source for Python 2.3.3 and compiling it in. Fixed the above problem, but after adding playlists with NDBM, positron barfs. Since there were references to this being fixed in 1.1, I tried a test.
I submitted a patch for this post 1.1 unfortunately. It is in positron bugzilla at http://bugs.xiph.org/show_bug.cgi?id=518. Should fix your problem. Good luck.
Reply author: hal
Replied on: 04/05/2004 04:17:59 AM
Message:
quote:
Originally posted by alecm
I submitted a patch for this post 1.1 unfortunately. It is in positron bugzilla at http://bugs.xiph.org/show_bug.cgi?id=518. Should fix your problem. Good luck.
Your patch does the trick, Alec. Hope they pick it up soon.
Hal
Neuros Forums : http://www.neurosaudio.com/community/forum/
© Copyright ©2002-05 Neuros Technology International, LLC all rights reserved.