from __future__ import * 2004/9

py2app development continues

I've done a bunch of work on py2app the past couple days. It's probably in a state where it could be used for a legitimate project, but I'm not going to provide any real support it yet. If you do play with it, you will need to also check out the latest macholib.

New stuff:

• dylib dependency hunting and rewriting

• standalone (for non-vendor python builds)

• data resources

• icon

• template plist file, plistlib.Plist, or dict

• alias build mode (similar to but better than bundlebuilder --symlink)

• can avoid modulefinder (when using alias build mode)

• Python 2.4 support

• dylib/framework includes

• semi-standalone option when using non-system Python

• argv emulation

• chdir to Resources folder (or not)

Still missing:

• Non-zip distribution mode

• Plugins (preference panes, etc.)

• Including full packages

• Still haven't tried anything non-trivial

• dylib/framework excludes

• baked-in "recipes" for common problematic packages (PIL, for example)

• a couple minor bundlebuilder options (creator, bundle-id, strip)

py2app builds its first .app

I started hacking on py2app again, and it's to the point where it can build applications now. What it does:

• Uses a bootstrap executable (like the PyObjC Xcode template, unlike bundlebuilder). It does not require a compiler to do this, it's precompiled with source included.

• Finds Python dependencies and stuffs them into a zip in the application bundle (some.app/Contents/Resources/Python/site-packages.zip). This is basically "semi-standalone" in bundlebuilder terms.

What it doesn't do, yet:

• Non-zip distribution mode

• dylib dependency hunting and rewriting (bundlebuilder2 / macholib)

• Standalone

• Plugins (preference panes, etc.)

• Haven't tried or tested including non-code resources or full packages

• Haven't yet tried doing anything non-trivial

• No way to avoid the modulefinder step (should have a way to manually specify dependencies)

• Take a template plist or additional plist arguments