Ronald Oussoren wrote:
>>> Unbelievable. Apple's "scripting bridge" push,
>>
>> You mean BridgeSupport
>
> There's also the Cocoa Scripting Bridge
Folks often use 'scripting bridge' to mean 'scripting language to ObjC
bridge' (PyObjC, RubyCocoa, etc). The misleadingly named Scripting
Bridge is Apple's attempt at ObjC to Apple event bindings, but isn't
relevant here beyond being one of the many APIs that PyObjC & co. can
talk to.
> The BridgeSupport project is a mess, the specification is too lax to be
> useful.
Not the first time I've heard this criticism, but there are enough
bridge and runtime developers around these days that they ought to be
able to pull together some improvements. I know with OSS it's common
for everyone to do their own thing, but a coherent group effort might
register better on Apple's radar when they're deciding what's worth
their support.
>> I would suggest dropping a note to Laurent Sansonetti (BridgeSupport +
>> MacRuby lead) and other bridge & runtime devs
>
>Ruby seems to be affected as well (although I'm not sure if I use the right
>syntax).
RubyCocoa and/or MacRuby? RubyCocoa uses the same method name swizzle
as PyObjC, minus the trailing underscore, e.g.
[obj foo: x bar: y] -> obj.foo_bar(x, y).
MacRuby uses heavier syntactic sugar to make positional args look like
keyword args, e.g.
[obj foo: x bar: y] -> obj.foo(x, bar: y)
>>> What other options are there? Specify your GUI by hand?
>>
>> Is it just the Read/Write Class Files functionality that's affected,
>> or has Xcode 4 also lost the ability to specify IB classes manually?
>> (IB 2 was good at this; IB 3 not so good but I think it was still
>> possible.)
>
>I haven't found a way to do "Read Class Files", probably because IB is now
>part of the Xcode IDE and no longer a standalone program (which means IB
>immediately knows when files are added or updated).
So can you still define class structures manually? Or is that gone too?
>There are probably good technical reasons why it is no longer easy to support
>scripting languages in IB, but the new behavior sucks nonetheless.
I doubt there's a good technical reason; my guess'd be that due to
some combination ignorance, arrogance, insufficient time and/or
managerial diktat [1] they've coupled it tightly to ObjC with little
thought for creating a more flexible, reusable solution. Apple have
always had a lukewarm attitude to any language not already called
"Objective-C", treating even their own AppleScript language as the
red-headed stepchild they'd club to death the moment everyone's backs
are turned.
Again, I'd recommend you have a pow-wow with all the other bridge and
runtime developers, as you're the folks who are best able to describe
the problem in both technical and practical terms and rally your user
bases in concerted protest.
Regards,
has (trundling web- and Linux-ward for a reason)
[1] This is what got appscript & co. rejected from 10.5 in favour of
the crappy in-house solution. I always thought the ObjC bridges would
be quite safe though, seeing as they hanged with the cool kids (ObjC &
Cocoa) rather than the old-skool losers (AppleScript + Carbon), but
sympathies if not.
_______________________________________________
Pythonmac-SIG maillist - Pythonmac-SIG@python.org
http://mail.python.org/mailman/listinfo/pythonmac-sig
unsubscribe: http://mail.python.org/mailman/options/Pythonmac-SIG
