[ python-Bugs-765903 ] bundlebuilder Info.plist files.
SourceForge.net
noreply at sourceforge.net
Fri Jul 16 14:41:01 CEST 2004
Bugs item #765903, was opened at 2003-07-04 14:11
Message generated for change (Comment added) made by jackjansen
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=105470&aid=765903&group_id=5470
Category: Macintosh
Group: None
>Status: Closed
Resolution: Fixed
Priority: 5
Submitted By: Jack Jansen (jackjansen)
Assigned to: Just van Rossum (jvr)
Summary: bundlebuilder Info.plist files.
Initial Comment:
Spotted by Edward Moy:
-----------
8) In BuildApplet.app/Contents/Info.plist, the
CFBundleIdentifier needs to be a reverse-dns format
(org.python.BuildApplet; in the WWDC preview, I have
accidentally made it com.apple.BuildApplet). It also needs a
CFBundleShortVersionString entry (I just used 1.0 for the
WWDC preview).
------------
This needs a couple of changes, really:
- bundlebuilder needs a way to specify these, plus sensible
defaults.
- the same for BuildApplet, when it calls bundlebuilder
- Mac/OSX/Makefile needs to be adapted to use the new
options.
If you don't have the time to fix all of these please go as far
as you can and reassign back to me.
----------------------------------------------------------------------
>Comment By: Jack Jansen (jackjansen)
Date: 2004-07-16 14:41
Message:
Logged In: YES
user_id=45365
It appears this is all fixed.
----------------------------------------------------------------------
Comment By: Jack Jansen (jackjansen)
Date: 2003-07-07 12:08
Message:
Logged In: YES
user_id=45365
Yes, I guess you're right. bundlebuilder should put valid default
values in the plist files and that's it. Our applets will just have to
come with their own plist files. So I'd say rip out the bundle_id
option but put in a valid bundle ID and version number, and please
either create plist files for our applets (and use these in Mac/OSX/
Makefile) or put in a bug report to that effect on my name.
----------------------------------------------------------------------
Comment By: Just van Rossum (jvr)
Date: 2003-07-04 18:24
Message:
Logged In: YES
user_id=92689
Oops, you're right, I was too fast: I indeed didn't do BuildApplet or
CFBundleShortVersionString.
I'm worried about bundlebuilder feature-bloat, though: the net --
bundle-id flag is only a convenience to avoid creating an explicit
plist. BuildApplet could easily create the plist "manually", and that
would in fact be the more general solution, so I'm tempted to rip
--bundle-id out again.
Do all bundles *have* to specify CFBundleShortVersionString? If
so, should bundlebuilder simply use "1.0" if it's not specified in the
plist?
----------------------------------------------------------------------
Comment By: Jack Jansen (jackjansen)
Date: 2003-07-04 17:29
Message:
Logged In: YES
user_id=45365
I got the impression that you only did step 1 (bundlebuilder), not
BuildApplet and Mac/OSX/Makefile, right? Also, did you do the
CFBundleShortVersionString bit too?
Again, reassign to me if you don't feel like hacking this.
----------------------------------------------------------------------
Comment By: Just van Rossum (jvr)
Date: 2003-07-04 16:22
Message:
Logged In: YES
user_id=92689
Done, except they're named bundle_id and --bundle-id.
----------------------------------------------------------------------
Comment By: Just van Rossum (jvr)
Date: 2003-07-04 15:28
Message:
Logged In: YES
user_id=92689
IMO the sensible default is what we have now (and what a _lot_ of
commercial apps do as well, just have a look at ~/Library/
Preferences/*.plist). For the applets shipping with Python it would
indeed be much better if the bundle id started with "org.python.". I
suggest to add a --bundle-identifier switch (and a bundle_identifier
kwarg). I will do this unless you disagree.
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=105470&aid=765903&group_id=5470
More information about the Python-bugs-list
mailing list