[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gnump3d-users] Some observations to the current CVS version

From: C. Menge
Subject: [Gnump3d-users] Some observations to the current CVS version
Date: Thu, 15 Dec 2005 20:32:48 +0100
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b) Gecko/20050217

Migration to new checkout server successful, new observations follow:

- The biggest new change is that the song titles aren't displayed
  Directory names appear as usually, but all the audio files are
  displayed like that, completely missing their names:

        5.3Mb ]         [Info] [Download]
        7.9Mb ]         [Info] [Download]
        5.7Mb ]         [Info] [Download]
        6.8Mb ]         [Info] [Download]

- File titles containing an '&' are displayed with '&' in the
  streamed file titles, still some HTML residues here :-)

  Oh, and BTW, I believe according to the usual nomenclature the sizes'
  unit should be 'MB' with capital 'B'.

The other observation is about random crashes on windows machines:

I've collected some --debug output over the last days, here it is:

- Rare and unreproducable occurrence:
| Use of implicit split to @_ is deprecated at
| D:/Tools/gnump3d/lib/gnump3d/plugins/ line 313.

- Occurring now and then, I don't know why.
  *Always* in exactly this order and together,
  the server continues to run:
| print() on closed filehandle GEN64 at
| D:\Tools\GNUmp3d\bin\gnump3d2 line 2037.
| Use of uninitialized value in vec at
| D:\Tools\GNUmp3d\bin\gnump3d2 line 1965.
| Use of uninitialized value in vec at
| D:\Tools\GNUmp3d\bin\gnump3d2 line 1967.

- Crash occurring reproducably under "high load", which means streaming
  at least one file and at the same time using the web interface to
  browse the collection:
| Use of uninitialized value in numeric eq (==) at
| D:\Tools\GNUmp3d\bin\gnump3d2 line 549.
|   Fatal error - cannot fork().
|   The system error was Resource temporarily unavailable

If I comment out the 'exit;' in line 562, this stops happening without
visibly affecting the running server.
My idea (without any deeper knowledge, just speculating) about this is
that it may be caused by a difference in the perl implementations for
*ix and windows.
I think the server shouldn't die because of a *temporary* problem.

perl --version on the machine gives:
| This is perl, v5.8.7 built for MSWin32-x86-multi-thread
| Binary build 815 [211909] provided by ActiveState
| Built Nov  2 2005 08:44:52

[end of observations]

Hope this helps. Especially me ;-)


reply via email to

[Prev in Thread] Current Thread [Next in Thread]