SABnzbd Forums

Skypike, looks like the automated python handling in the package creates non-optimized compiled *.pyc files only. Will look into making it create pyo too as soon as I find the time for it.

jcfp wrote: Skypike, looks like the automated python handling in the package creates non-optimized compiled *.pyc files only. Will look into making it create pyo too as soon as I find the time for it.

For your REPOSITORY.... Would you be able to use the ICC version of PAR2 for x86 systems:
http://chuchusoft.com/par2_tbb/par2cmdl ... n32.tar.gz

...and if possible compile a AMD64 version for X86_64 computer?
(see my post @ http://forums.sabnzbd.org/index.php?topic=836.0)

Also, maybe use the newer WinRAR binaries from RARSOFT.com ?

32bit:
http://rarsoft.com/rar/rarlinux-3.8.b2.tar.gz

64it:
http://rarsoft.com/rar/rarlinux-x64-3.8.b2.tar.gz

Btw, your repo rocks !

Samushka wrote: For your REPOSITORY.... Would you be able to use the ICC version of PAR2 for x86 systems:
http://chuchusoft.com/par2_tbb/par2cmdl ... n32.tar.gz

Longer term, maybe, if it has real benefits, works on enough platforms, etc etc etc. But don't count on it to happen, let alone anytime soon.

Samushka wrote: Also, maybe use the newer WinRAR binaries from RARSOFT.com ?

Not unless sabnzbd+ would absolutely depend on any such recent unrar releases. Ubuntu and debian already have a well maintained unrar package; if you really insist on having the latest, just rebuild from the development release (intrepid) source package (currently at version 3.8.2).

jcfp wrote:

Samushka wrote: For your REPOSITORY.... Would you be able to use the ICC version of PAR2 for x86 systems:
http://chuchusoft.com/par2_tbb/par2cmdl ... n32.tar.gz

Longer term, maybe, if it has real benefits, works on enough platforms, etc etc etc. But don't count on it to happen, let alone anytime soon.

Samushka wrote: Also, maybe use the newer WinRAR binaries from RARSOFT.com ?

Not unless sabnzbd+ would absolutely depend on any such recent unrar releases. Ubuntu and debian already have a well maintained unrar package; if you really insist on having the latest, just rebuild from the development release (intrepid) source package (currently at version 3.8.2).

Pretty sure that the performance difference of the ICC on multicore systems is well documented. The windows bullds on SABnzbd.org and RAF's builds for MAC OS X all have it.

It's too bad the developer hasn't responded to my email, he has 64bit versions on every other platform, just seems odd that he doesn't have it on Linux. Maybe you can compile it and offer it as a standalone package?, i'd be willing to donate some money via PayPal for your time and effort. I would do it myself, but sadly my skill in this area is sorely lacking !

Samushka wrote: Pretty sure that the performance difference of the ICC on multicore systems is well documented. The windows bullds on SABnzbd.org and RAF's builds for MAC OS X all have it.

I think I remember looking at that code some time ago, finding the modifications made at that time were windoze-only. Good to known at least that has changed, although it lacking a 64bit linux build for the time being may very well not be a coincidence. But even if that all worked, it's still going to require time, which I just don't have right now; busy busy busy, with my vacation and yet more deadlines getting nearer means I'm probably not in a position to work on this par2 thing until somewhere near the end of august (at the earliest). Too bad .

jcfp wrote:

Samushka wrote: Pretty sure that the performance difference of the ICC on multicore systems is well documented. The windows bullds on SABnzbd.org and RAF's builds for MAC OS X all have it.

I think I remember looking at that code some time ago, finding the modifications made at that time were windoze-only. Good to known at least that has changed, although it lacking a 64bit linux build for the time being may very well not be a coincidence. But even if that all worked, it's still going to require time, which I just don't have right now; busy busy busy, with my vacation and yet more deadlines getting nearer means I'm probably not in a position to work on this par2 thing until somewhere near the end of august (at the earliest). Too bad .

It's all good dude, take your time...

0.4.0 final has been uploaded, should appear soon.

after doing an upgrade to 0.4.0 i get this in my cherrypy.log .. 

06/Jul/2008:20:52:27 HTTP INFO Page handler: 'No such file or directory'
Traceback (most recent call last):
  File "/var/lib/python-support/python2.5/cherrypy/_cphttptools.py", line 110, in _run
    applyFilters('before_finalize')
  File "/var/lib/python-support/python2.5/cherrypy/filters/__init__.py", line 151, in applyFilters
    method()
  File "/var/lib/python-support/python2.5/sabnzbd/utils/multiauth/filter.py", line 61, in beforeFinalize
    **rsrc.callable_kwargs)
  File "/var/lib/python-support/python2.5/sabnzbd/interface.py", line 287, in index
    return template.respond()
  File "_usr_share_sabnzbdplus_interfaces_Default_templates_main_tmpl.py", line 95, in respond
  File "/var/lib/python-support/python2.5/Cheetah/Template.py", line 1511, in _handleCheetahInclude
    nestedTemplateClass = compiler.compile(source=source,file=file)
  File "/var/lib/python-support/python2.5/Cheetah/Template.py", line 692, in compile
    fileHash = str(hash(file))+str(os.path.getmtime(file))
  File "/usr/lib/python2.5/posixpath.py", line 143, in getmtime
    return os.stat(filename).st_mtime
OSError: [Errno 2] No such file or directory: '/usr/bin/interfaces/Default/templates/inc_top.tmpl'

McM wrote: after doing an upgrade to 0.4.0 i get this in my cherrypy.log .. 

This looks very much like an error that I had taken care of. Anyway, please send the following to me by e-mail ([email protected]):

• output of: lsb_release -a
• output of: apt-cache policy sabnzbdplus
• output of: apt-cache madison sabnzbdplus
• the sabnzbd previous version you had installed from the repository (rc?)
• move any existing logs out of the way, run 'sabnzbdplus -w 1 -l 2', recreate the error, close sabnzbd, then:
• attach the resulting sabnzbd.log and cherrypy.log files
• ... and give a detailed description of exactly what you do to get this error

McM, just did some redesigning in the package and will upload the result in a few minutes. Please try that first, your problem might just be gone.

jcfp wrote: McM, just did some redesigning in the package and will upload the result in a few minutes. Please try that first, your problem might just be gone.

Tried the new package and now it works out-of-the-box.. 
Thank you!

Are the packages going to be updated to the new release? (0.4.2 at the time of this writing)

I'm back from my summer vacation, so 0.4.2 is finally being packaged. Changes in this version:

• Include statements in the default interface templates were fixed so the package design is back to normal.
• Optimized byte-compiled files (*.pyo) are now created on install for the default python version. Not certain how much of a performance gain this will bring (if any) but it never hurts to try.

The whole thing has been uploaded, but the build queue is 2+ hours at the moment so don't hold your breath.

Would love to see a debian repository also the same version of SABnzbd should work on both Ubuntu and debian distributions.

Any idea when SABnzbd might make it into the official repositories for debian or ubuntu?

You need to get an officially qualified package creator to enter a package into an official repository.
Qualification is quite an intense process, which nobody of the the SABnzbd team is looking forward to.

So, maybe JCFP can become a qualified packager or he can find a "sponsor" that accepts his packages.