pyobjc-dev — CVS commit notices go here

Re: [Pyobjc-dev] standalone-related questions

[<bb...@ma...>]

On Thursday, October 31, 2002, at 01:34 AM, Steven D. Arnold wrote:
> A couple questions related to the standalone nature of 
> double-clickable apps
> created using pyobjc.
>
> 1. Is the app truly standalone in the sense of not requiring a previous
> Python installation?  (My guess is no, at least not out-of-the-box)

The Cocoa-Python project template creates a Project Builder project 
that will build a completely standalone, double-click-to-launch, 
drag-and-drop-to-install application when the install target is used on 
a developer's machine that has PyObjC installed against the Apple 
supplied Python.

To the end user, there is absolutely no difference between an 
application built using PyObjC and an application built using 
Java-Cocoa, Objective-C, Carbon, or Swing other than a slightly long 
launch time for Cocoa-Python apps (which goes away once Apple ships an 
embeddable python).

> 2. If not, how could this be implemented?

It is is implemented by:

     - using a series of copy file phases to copy the PyObjC module into 
the 'pyobjc' subdirectory of the Resources directory of the application 
when the 'install' target is used ('pbxbuild install' from the command 
line).

     - the main.m uses execve() to transfer control to /usr/bin/python 
and exec Main.py in a fashion where OS X thinks that /usr/bin/python is 
the executable within the app wrapper (otherwise, mainBundle breaks).  
Unfortunately, this is what breaks gdb.  Fortunately, this will go away 
when Apple ships a new build of Python that includes a library for 
embedding.

     - the Main.py adds the pyobjc directory to sys.path and loads the 
PyObjC module before bootstrapping into the AppKit.

End result:  Standalone, double-clickable Python apps.   Grab the Web 
Services Tool disk image from http://www.friday.com/software/ and try 
it out.  Should just work without having a prior copy of pyobjc 
installed.  Requires 10.2

> We may want to consider this in light of people who use different 
> versions
> of Python for a program.  My app might need a feature available only in
> Python 2.3 for example.  Also, my app might use third-party modules 
> that are
> set up for a specific version of Python, which the end user may or may 
> not
> have.

Third party modules can be dropped into the pyobjc directory within 
Resources (or directly into the Resources) using a "copy files" phase.  
Those two directories are on sys.path and, therefore, modules within 
the directories will be found via the standard 'import' mechanism.

If you need to ship a custom build of Python (or a later build than the 
version supplied by Apple), you can use the framework build or ship the 
built python distribution [unix style] within the app wrapper.   Your 
app will jump from a minimum of 500k to a minimum of 5 to 9 megabytes 
(depending on build style).

b.bum

Re: [Pyobjc-dev] quick documentation format survey

[<sp...@al...>]

Le jeudi, 31 oct 2002, =E0 09:48 Europe/Paris, Steven D. Arnold a =E9crit =
:

> SimTex looks very impressive.  It seems to provide for all needed=20
> formats
> except Apple Help Format, for which we could add a Python backend.  It
> appears to be a fairly mature project from my brief look at it.

Thanks :)

SimTex has not been yet publicly released, but it is near a finalised=20
version. The only problem is that documentation is not completed yet,=20
and that some advanced features lack (custom markup, appendices).=20
Orchid=E9eNoire, the SimTex document processor, will also soon support=20=

DocBook as an output format, and it might be relatively easy to output=20=

Apple Help Format (conversion is made using XSLT) - if someone points=20
me out some doc on Apple Help Format, I would be able write the=20
transformation stylesheet soon!

> Overall, I like SimTex best because (1) It's written in Python; (2)=20
> Uses an
> OSI license (GPL); (3) Looks more-or-less like regular text before =
it's
> compiled to another format, hence any text editor is a good authoring=20=

> tool;
> and (4) The author is right here with us if questions arise.

If you use SimTex, you can be assured that I will do my best to make=20
SimTex meet your needs and fix remaining bugs. Same thing for=20
Orchid=E9eNoire.

BTW, the difference between SimTex and Orchid=E9eNoire might be unclear,=20=

so here is a small explaination:

  - SimTex is a structured text to XML processor, it allows to produce=20=

valid XML without writing XML
  - Orchid=E9eNoire is an XML application similar to DocBook, but aimed =
at=20
simplicity, and also a document processing chain, that uses XSLT as its=20=

transformation language. It currently supports HTML, PS, PDF, plain=20
text and Lout.

I personally also think that SimTex/Orchid=E9eNoire is one of the most=20=

versatile solution and that you will never get stuck with it ;). The=20
API is open and comprehensible, the format is XML and transformations=20
are easily made with XSLT. As I mentioned, apart from a few usability=20
bugs (command line options) and an incomplete reference manual,=20
SimTex/Orchid=E9eNoire would ease the process of editing and producing=20=

documentation.

If you want to subscribe to the SimTex mailing list, simply send a mail=20=

to <sim...@ty...>. If you want to have some help on=20
possible list manager commands, send to <sim...@ty...>. For=20
those having troubles installing 4Suite, I will try to make SimTex work=20=

with PyXML only.

Thanks for your interest in SimTex,

  -- S=E9bastien

--
=ABAnd never can a man be more disastrously in death than when death
itself shall be deathless.=BB
<http://www.type-z.org>                     -- St. Augustine

[Pyobjc-dev] quick documentation format survey

[Steven D. A. <st...@ne...>]

I looked into the formats suggested by list members as well as consulting
Google on the subject of documentation formats.  My search was brief and
cursory, so I may have missed many different projects.  I'm still open to
other suggestions should list members have any.  My criteria are:

(1) Must be in an easily-parseable format that can be compiled to many
different formats.
(2) Can't be proprietary.
(3) Must support graphics.
(4) Should be easy to read in native format.  XML is better than binary,
something analogous to plain text is better than XML.
(5) Should be fairly mature, not a lot of bugs.
(6) Written in Python is a plus.
(7) It would be beneficial if sophisticated editors existed on Mac OS X for
writing docs using the protocol/tool.


SimTex looks very impressive.  It seems to provide for all needed formats
except Apple Help Format, for which we could add a Python backend.  It
appears to be a fairly mature project from my brief look at it.

I also took a look at the TwistedMatrix project, as recommended by Donovan,
but I wasn't able to find a reference to their twisted.lore.  (Not even on
Google...I searched for "twisted.lore matrix" getting zero hits; "twisted
lore matrix" getting a bunch of unrelated hits, a couple others with similar
results...you get the picture.  If someone provides me with a specific
pointer to this project, I'll take another look.)

I looked at DocBook, a widely used XML DTD.  There are a number of tools for
writing DocBook documentation.  I only found one obvious tool for Mac OS X;
remember though that my search was cursory and there could be others I don't
know about, especially ones written in Java.

Overall, I like SimTex best because (1) It's written in Python; (2) Uses an
OSI license (GPL); (3) Looks more-or-less like regular text before it's
compiled to another format, hence any text editor is a good authoring tool;
and (4) The author is right here with us if questions arise.

The usage of this tool could influence or create more work for future
contributors, including the authors themselves, so if there are any concerns
about using SimTex, please let me know as soon as possible.  Bear in mind
that my search was very much a seat-of-the-pants affair, so one plausible
criticism is that we ought to do a more detailed survey of the options,
perhaps weighting the criteria above, before committing to one alternative.
I'm more in the mindset of "make a reasonable choice quickly and get it out
ASAP," which is what I have tried to do in the survey above, but if there's
a lot of concern about this I'll rethink that attitude.

Thanks again for the feedback,

steve
-- 

[Pyobjc-dev] standalone-related questions

[Steven D. A. <st...@ne...>]

Hi all,

A couple questions related to the standalone nature of double-clickable apps
created using pyobjc.

1. Is the app truly standalone in the sense of not requiring a previous
Python installation?  (My guess is no, at least not out-of-the-box)

2. If not, how could this be implemented?

We may want to consider this in light of people who use different versions
of Python for a program.  My app might need a feature available only in
Python 2.3 for example.  Also, my app might use third-party modules that are
set up for a specific version of Python, which the end user may or may not
have.

Thoughts?

steve
-- 

Re: [Pyobjc-dev] Trying to build WebServicesTool app OS X 10.2.2 (seed)

[<bb...@ma...>]

On Wednesday, October 30, 2002, at 11:38 PM, Stephen Steiner wrote:
>> Or set the PythonBinPath user default to point to the python 
>> interpreter binary that you want to use [that has pyobjc installed 
>> within its site-packages].
>
> Where would I find the 'PythonBinPath' user default (doc?)?

It is mentioned in the bin-python-main.m file;  it defaults to 
/usr/bin/python. As long as you can do the following, it should "just 
work" without setting that default.

[bumbox:~] bbum% /usr/bin/python
Python 2.2 (#1, 07/14/02, 23:25:09)
[GCC Apple cpp-precomp 6.14] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> from Foundation import *
 >>> x = NSMutableArray.array()
 >>> x
()
 >>>

>
> When I run from within PB I get:
>
> [Switching to process 2631 thread 0xb03]
> warning: ppc_frame_chain_valid: stack pointer from 0xbffffb1c to 
> 0x1000 grows upward; assuming invalid

Because of the way the control is transferred to the python 
interpreter, you can't directly run the application under GDB.  This 
limitation will go away once Apple ships a complete build of Python 
[the left out the ability to embed the interpreter into a binary and, 
as such, we have to execve() to pass control to the interpreter... this 
breaks gdb in the traditional sense].   See the mailing list archive if 
you want to run with gdb.

Try just running the app from within PB -- not running it in the 
debugger.

b.bum

[Pyobjc-dev] Major changes/additions to the bridging of collections

[Bill B. <bb...@co...>]

I spent a lot of time [that I didn't have to spend :-] 
changing/fixing/rewriting the convenience methods in the objc module 
and the associated tests.   Now, most of the various random slicing and 
dicing operations can be interchanged between Python and ObjC 
objects....

If this doesn't make sense, some examples.   NSDictionary can now do 
containment and iterations (using generators -- I'm happy that I found 
a place to use generators :-):

 >>> from Foundation import *
 >>> d = NSMutableDictionary.dictionary()
 >>> d['a'] = 1
 >>> d['b'] = 1
 >>> d[2] = 3
 >>> d[4] = 5
 >>> d
{b = 1; 2 = 3; 4 = 5; a = 1; }
 >>> for x in d: print '%s => %s' % (x, d[x])
...
b => 1
2 => 3
4 => 5
a => 1
 >>> 'a' in d
1
 >>> 'not' in d
0

And arrays -- note that the slicing operators work, including deletions 
and assignments of slices.

 >>> a = NSMutableArray.arrayWithArray_( range(10,30) )
 >>> x
'a'
 >>> a
(
     10,
     11,
     12,
     13,
     14,
     15,
     16,
     17,
     18,
     19,
     20,
     21,
     22,
     23,
     24,
     25,
     26,
     27,
     28,
     29
)
 >>> a[10:5]
[]
 >>> a[10:-5]
(20, 21, 22, 23, 24)
 >>> b = range(10,30)
 >>> b[10:-5]
[20, 21, 22, 23, 24]
 >>> a[-5]
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "/usr/lib/python2.2/site-packages/objc/_convenience.py", line 
119, in <lambda>
     CONVENIENCE_METHODS['objectAtIndex:'] = (('__getitem__', lambda 
self, index: self.objectAtIndex_( index )),
IndexError: NSRangeException - *** -[NSCFArray objectAtIndex:]: index 
(-5) beyond bounds (20)
 >>> a[-5:]
(25, 26, 27, 28, 29)
 >>> a[1:3]
(11, 12)
 >>> a[1:3] = [4,5,6,7,8,9]
 >>> a
(
     10,
     4,
     5,
     6,
     7,
     8,
     9,
     13,
     14,
     15,
     16,
     17,
     18,
     19,
     20,
     21,
     22,
     23,
     24,
     25,
     26,
     27,
     28,
     29
)
 >>> type(a)
<objective-c class NSCFArray at 0x1036a0>
 >>> del a[2:-2]
 >>> a
(10, 4, 28, 29)

And, because of the way Ronald implemented the convenience methods in 
the first place, NSSet also works:

 >>> s = NSMutableSet.setWithArray_(a)
 >>> s
<NSCFSet: 0xcb5970> (28, 10, 4, 29)
 >>> 10 in s
1
 >>> 900 in s
0

And, once I patched things to handle it:

 >>> from Foundation import *
 >>> x = NSSet.setWithArray_( [1,2,3,4,5] )
 >>> for i in x: print i
...
1
2
3
4
5

Fwd: [Pyobjc-dev] Trying to build WebServicesTool app OS X 10.2.2 (seed)

[Stephen S. <sst...@ma...>]

Begin forwarded message:

> From: Stephen Steiner <sst...@ma...>
> Date: Wed Oct 30, 2002  11:38:45  PM US/Eastern
> To: bb...@ma...
> Subject: Re: [Pyobjc-dev] Trying to build WebServicesTool app OS X 
> 10.2.2 (seed)
>
>
>> From the command line, can you do 'from Foundation import *'?
>
> Yes, that works fine.
>
>>     sudo /usr/bin/python setup.py install
>
> Yes, I did that thinking that I might not have correctly installed the 
> latest CVS build.
> No change.
>
>> Or set the PythonBinPath user default to point to the python 
>> interpreter binary that you want to use [that has pyobjc installed 
>> within its site-packages].
>
> Where would I find the 'PythonBinPath' user default (doc?)?
>
> When I run from within PB I get:
>
> [Switching to process 2631 thread 0xb03]
> warning: ppc_frame_chain_valid: stack pointer from 0xbffffb1c to 
> 0x1000 grows upward; assuming invalid
>
> etc...
>
> I'm not sure what to do next...
>
> Thanks for your help!
>
> Steve
>
>
Steve

Re: [Pyobjc-dev] Trying to build WebServicesTool app OS X 10.2.2 (seed)

[<bb...@ma...>]

 From the command line, can you do 'from Foundation import *'?

If not, I suspect that the pyobjc module is either not installed 
correctly or you have two different pythons on your system and you have 
pyobjc installed in a different python than the Apple supplied python.

Either try...

     sudo /usr/bin/python setup.py install

... in the pyobjc source directory.  Or set the PythonBinPath user 
default to point to the python interpreter binary that you want to use 
[that has pyobjc installed within its site-packages].

b.bum

On Wednesday, October 30, 2002, at 10:23 PM, Stephen Steiner wrote:

> I'm trying to build the Web Services Tool.
>
> I can build in PB with no errors , then when I try to run from within 
> PB I get:
>
> Traceback (most recent call last):
>   File "/Developer/_Projects/pyobjc/Examples/WebServicesTool/build/Web 
> Services Tool.app/Contents/Resources/Main.py", line 6, in ?
>     import objc
> ImportError: No module named objc
>
> Web Services Tool has exited with status 1.
>
> If I go to the command line and run Main.py with:
> 	python ./Main.py
>
> I get:
>
> Traceback (most recent call last):
>   File "./Main.py", line 11, in ?
>     pyFrameworkPathsIndex = sys.argv.index("-PyFrameworkPaths")
> ValueError: list.index(x): x not in list
>
> I'm sure this is some obvious setup mistake on my part...any help 
> would be appreciated!
>
> Thanks,
>
> Steve
>
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by: Influence the future of Java(TM) 
> technology. Join the Java Community Process(SM) (JCP(SM)) program now. 
> http://ads.sourceforge.net/cgi-bin/redirect.pl?sunm0004en
> _______________________________________________
> Pyobjc-dev mailing list
> Pyo...@li...
> https://lists.sourceforge.net/lists/listinfo/pyobjc-dev
>
b.bum
We gladly feast on those who would subdue us.

[Pyobjc-dev] Trying to build WebServicesTool app OS X 10.2.2 (seed)

[Stephen S. <sst...@ma...>]

I'm trying to build the Web Services Tool.

I can build in PB with no errors , then when I try to run from within 
PB I get:

Traceback (most recent call last):
   File "/Developer/_Projects/pyobjc/Examples/WebServicesTool/build/Web 
Services Tool.app/Contents/Resources/Main.py", line 6, in ?
     import objc
ImportError: No module named objc

Web Services Tool has exited with status 1.

If I go to the command line and run Main.py with:
	python ./Main.py

I get:

Traceback (most recent call last):
   File "./Main.py", line 11, in ?
     pyFrameworkPathsIndex = sys.argv.index("-PyFrameworkPaths")
ValueError: list.index(x): x not in list

I'm sure this is some obvious setup mistake on my part...any help would 
be appreciated!

Thanks,

Steve

Re: [Pyobjc-dev] Testing

[<bb...@ma...>]

On Wednesday, October 30, 2002, at 02:10 PM, Ronald Oussoren wrote:
....
> The reason for responding in this thread: Is it correct to place the 
> code above (including the class definitions) in a testcase? E.g.
>
> 	class TestSubclassing (unittest.TestCase):
> 		def testSubSubClass(self):
> 			# Code from above

Yes -- the code listed above is, in and of itself, the actual test case 
and, therefore, should be encapsulated in a test case such that it 
fails [outside of the regular parse errors] in the context of a test 
case.

I had a look at the code and added a test to ensure that a call to 
super works.   I also made a handful of other minor adjustments.

All of this should make support for GnuStep *a lot* easier to pull 
together as GnuStep includes an implementation of the Foundation 
classes.

The tests also make for a very simple and concise place for folks new 
to the bridge to see the various features that it implements.  Of 
course, for truly frightening testing potential, we could also use 
doctest....

b.bum

Re: [Pyobjc-dev] Testing

[Ronald O. <ous...@ci...>]

If just added Lib/objc/test/test_subclass.py to test some (actually 1) 
feature of subclassing Objective-C classes. It should be possible to do 
something like the code below, but that crashes the interpreter. I've 
also checked in a fix for this bug.

	NSObject = objc.lookup_class('NSObject')

	class LevelOneClass (NSObject): pass
	class LevelTwoClass (Level1Class): pass

	v = LevelTwoClass.alloc().init()

The reason for responding in this thread: Is it correct to place the 
code above (including the class definitions) in a testcase? E.g.

	class TestSubclassing (unittest.TestCase):
		def testSubSubClass(self):
			# Code from above


Ronald

[Pyobjc-dev] Testing

[<bb...@ma...>]

On Wednesday, October 30, 2002, at 11:21 AM, Ronald Oussoren wrote:
> On Wednesday, Oct 30, 2002, at 16:29 Europe/Amsterdam, bb...@ma... 
> wrote:
>> (Ronald:  I feel that the test/ directories should ship with the 
>> 'binary distribution' in the location that they are in.   Normally, 
>> the python distribution ships with the unittests and it has been 
>> incredibly helpful to be able to wander into an alien python 
>> installation, run the unit tests, and see exactly what was built into 
>> the distribution in a working fashion.... for us, it means that the 
>> user community can easily validate whether PyObjC still works after a 
>> system update from Apple.)
>
> Agreed. I've noticed how usefull this is w.r.t. testing of prereleases 
> of MacPython.
>
> Adding a script in that runs all tests would also be usefull. This 
> would have to  run all test_*.py scripts in a new interpreter: It is 
> highly likely that some tests will depend on having some modules 
> loaded or not loaded (e.g. testing if importing AppKit makes available 
> extra categories on a number of Foundation classes).

I have a test_all.py in my local workarea that I'm working on as a "run 
all tests" solution.  I hadn't thought of running each test in a new 
interpreter, but that certainly makes a lot of sense.

It also solves another problem.  Apple failed to ship the unit tests 
with their distribution of Python [bug filed] and, as such, the drivers 
for running all tests in a directory are missing.   I was going to 
figure out how to do another ssl/readline style 'add this missing piece 
to the Apple supplied python' package, but-- given that it doesn't 
support new interpreters anyway-- it wouldn't solve the problem at hand.

I'll work up a solution for running all the tests and commit it later 
today [tonight].

b.bum

Re: [Pyobjc-dev] PyObjc and GNUstep

[Ronald O. <ous...@ci...>]

On Wednesday, Oct 30, 2002, at 16:29 Europe/Amsterdam, bb...@ma... 
wrote:

> (Ronald:  I feel that the test/ directories should ship with the 
> 'binary distribution' in the location that they are in.   Normally, 
> the python distribution ships with the unittests and it has been 
> incredibly helpful to be able to wander into an alien python 
> installation, run the unit tests, and see exactly what was built into 
> the distribution in a working fashion.... for us, it means that the 
> user community can easily validate whether PyObjC still works after a 
> system update from Apple.)

Agreed. I've noticed how usefull this is w.r.t. testing of prereleases 
of MacPython.

Adding a script in that runs all tests would also be usefull. This 
would have to  run all test_*.py scripts in a new interpreter: It is 
highly likely that some tests will depend on having some modules loaded 
or not loaded (e.g. testing if importing AppKit makes available extra 
categories on a number of Foundation classes).

Ronald

[Pyobjc-dev] GnuStep on OS X

[<bb...@fr...>]

Found the instructions....  slightly out of date [the Apple runtime is, 
apparently, much faster in 10.2-- I haven't tested it and I don't know 
if the speed is in parity with the GNU runtime], but should still work.

http://www.gnustep.org/information/machines_1.html#SEC2

b.bum

Re: [Pyobjc-dev] PyObjc and GNUstep

[<bb...@ma...>]

As Ronald indicated, the current PyObjC module doesn't support=20
GNUStep's Obj-C runtime mostly because none of the primary developers=20
have that particular Obj-C runtime right now.

It appears that to use the GNU objc runtime on OS X would require=20
installing a different version of gcc than the one that Apple ships? =20
If that is true, it should be possible to have a both versions of=20
PyObjC on the same OS X box in parallel-- one compiled for the Apple=20
runtime and one for the GNU runtime.

Now that we have the beginnings of unit tests, it will be easier to=20
qualify against the GNU runtime.

For anyone that is going to contribute to this-- either to the GNU=20
runtime support or to PyObjC, in general-- I highly encourage you to=20
write a unit test in one of the appropriate test/ subdirectories first.=20=

  Make sure it fails, then write your feature/extension/bugfix such that=20=

it passes.

(Ronald:  I feel that the test/ directories should ship with the=20
'binary distribution' in the location that they are in.   Normally, the=20=

python distribution ships with the unittests and it has been incredibly=20=

helpful to be able to wander into an alien python installation, run the=20=

unit tests, and see exactly what was built into the distribution in a=20
working fashion.... for us, it means that the user community can easily=20=

validate whether PyObjC still works after a system update from Apple.)

b.bum

On Wednesday, October 30, 2002, at 04:11 AM, S=E9bastien Pierre wrote:

> Hi all,
>
> I was wondering if if was planned to add GNUstep and/or libFoundation=20=

> support to PyObjc ? As I mentioned in an earlier post, I am developing=20=

> an Objective-C library that I would like to use also with Python (that=20=

> is, use my library as a python module thanks to pyobjc). This library=20=

> should not be tied to OSX, but also work on any Unix with GNUstep.
>
> I did not manage to compile pyobjc on a Debian with GNUstep, mainly=20
> because the headers cannot be found:
>
> Modules/objc/pyobjc.h:7:31: objc/objc-runtime.h: No such file or=20
> directory
> Modules/objc/OC_PythonObject.h:24:31: Foundation/NSProxy.h: No such=20
> file or directory
> ...
>
> Is there any chance that pyobjc work on GNUstep too ?
>
>  -- S=E9bastien
>
> --
> =ABLife begins at the centerfold and expands outward.=BB
> <http://www.type-z.org>                     -- Miss November, 1966
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Pyobjc-dev mailing list
> Pyo...@li...
> https://lists.sourceforge.net/lists/listinfo/pyobjc-dev
>
b.bum
We gladly feast on those who would subdue us.

Re: [Pyobjc-dev] PyObjc and GNUstep

[<sp...@al...>]

Le mercredi, 30 oct 2002, =E0 16:03 Europe/Paris, Ronald Oussoren a =
=E9crit=20
:

> Aha, a volunteer ;-)
>
> I'd like to add support for GNUstep in the future, but at the moment=20=

> we're more concerned about getting first-rate support for MacOS X.=20
> What doesn't help of course is that I don't have access to a machine=20=

> with GNUstep at the moment (I had a disk-crash on my Linux PC at home=20=

> and I haven't gotten around to installing the new disk).

I seems like some people are willing to make a Fink port of GNUstep, so=20=

this may ease the port :)

  -- S=E9bastien

--
=ABLe soleil est Dieu=BB
<http://www.type-z.org>             -- Turner, avant de mourir

Re: [Pyobjc-dev] PyObjc and GNUstep

[Ronald O. <ous...@ci...>]

Aha, a volunteer ;-)

I'd like to add support for GNUstep in the future, but at the moment 
we're more concerned about getting first-rate support for MacOS X. What 
doesn't help of course is that I don't have access to a machine with 
GNUstep at the moment (I had a disk-crash on my Linux PC at home and I 
haven't gotten around to installing the new disk).

Ronald

Re: help file formats (was: Re: [Pyobjc-dev] problem accessing outlet)

[<sp...@al...>]

Le mercredi, 30 oct 2002, =E0 13:36 Europe/Paris, Ronald Oussoren a =
=E9crit=20
:

> I've used StructuredText for the small bits of documentation that I've=20=

> written upto now. This is more or less plain text and there are tool=20=

> available (on the 'net) to convert this to other formats.
>
> Using ST was an arbitrary choice, although I'm charmed by the fact=20
> that the documents look so much like plain text and can easily be read=20=

> from a text editor.
>

SimTex is actually a robust evolution of StructuredText - it is more=20
flexible, has a real object model, and can be easily extended. The API=20=

is almost completely documented, and of course it is written in Python=20=

:)

The only drawback yet is that I did not find the time to update the=20
reference manual...

  -- S=E9bastien

--
=ABIl vaut mieux suivre une voix stupide que l'on connait,
plut=F4t qu'une intelligente que l'on ne connait pas.=BB
<http://www.type-z.org>                -- No comment on Jacques C.

[Pyobjc-dev] PyObjc and GNUstep

[<sp...@al...>]

Hi all,

I was wondering if if was planned to add GNUstep and/or libFoundation=20
support to PyObjc ? As I mentioned in an earlier post, I am developing=20=

an Objective-C library that I would like to use also with Python (that=20=

is, use my library as a python module thanks to pyobjc). This library=20
should not be tied to OSX, but also work on any Unix with GNUstep.

I did not manage to compile pyobjc on a Debian with GNUstep, mainly=20
because the headers cannot be found:

Modules/objc/pyobjc.h:7:31: objc/objc-runtime.h: No such file or=20
directory
Modules/objc/OC_PythonObject.h:24:31: Foundation/NSProxy.h: No such=20
file or directory
...

Is there any chance that pyobjc work on GNUstep too ?

  -- S=E9bastien

--
=ABLife begins at the centerfold and expands outward.=BB
<http://www.type-z.org>                     -- Miss November, 1966=

Re: help file formats (was: Re: [Pyobjc-dev] problem accessing outlet)

[Ronald O. <ous...@ci...>]

I've used StructuredText for the small bits of documentation that I've=20=

written upto now. This is more or less plain text and there are tool=20
available (on the 'net) to convert this to other formats.

Using ST was an arbitrary choice, although I'm charmed by the fact that=20=

the documents look so much like plain text and can easily be read from=20=

a text editor.

Ronald

On Wednesday, Oct 30, 2002, at 09:53 Europe/Amsterdam, S=E9bastien =
Pierre=20
wrote:

>
> Le mardi, 29 oct 2002, =E0 23:42 Europe/Paris, Steven D. Arnold a =
=E9crit :
>
>> That makes a lot of sense.  It almost sounds like I should use a=20
>> version of
>> XML with some tool that can convert to HTML, text, or other formats. =20=

>> I
>> could then hack the script you mentioned so that the same XML source=20=

>> could
>> be converted to Apple Help format.
>>
>> I'll do a search for such tools; if anyone knows of such a tool,=20
>> please send
>> me a link.  In the absence of such a tool, I'll go ahead with HTML=20
>> for now,
>> and keep my eyes open.
>
> Hi!
>
> I'm actually writing such a tool: it's called "SimTex", it allows to=20=

> write a document in simply formatted text, which is then converted to=20=

> an XML document. This XML document conforms to the "Orchid=E9eNoire" =
XML=20
> application (also by me), which comes with a processing chain that=20
> allows to convert the input XML file to various formats including=20
> HTML, plain text, PS and PDF. It is rather easy to implement a new=20
> backend.
>
> You can have a look at <http://projects.type-z.org>. Documentation is=20=

> included into the source distros, under the Documentation folder.
>
> Hope this helps,
>
>  -- S=E9bastien
>
> --
> =ABToo old to be alternative, too alternative to be old.=BB
> <http://www.type-z.org> | Robert Smith, talking about his epitaph.
>
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Pyobjc-dev mailing list
> Pyo...@li...
> https://lists.sourceforge.net/lists/listinfo/pyobjc-dev
>

Re: help file formats (was: Re: [Pyobjc-dev] problem accessing outlet)

[<sp...@al...>]

Le mardi, 29 oct 2002, =E0 23:42 Europe/Paris, Steven D. Arnold a =E9crit =
:

> That makes a lot of sense.  It almost sounds like I should use a=20
> version of
> XML with some tool that can convert to HTML, text, or other formats.  =
I
> could then hack the script you mentioned so that the same XML source=20=

> could
> be converted to Apple Help format.
>
> I'll do a search for such tools; if anyone knows of such a tool,=20
> please send
> me a link.  In the absence of such a tool, I'll go ahead with HTML for=20=

> now,
> and keep my eyes open.

Hi!

I'm actually writing such a tool: it's called "SimTex", it allows to=20
write a document in simply formatted text, which is then converted to=20
an XML document. This XML document conforms to the "Orchid=E9eNoire" XML=20=

application (also by me), which comes with a processing chain that=20
allows to convert the input XML file to various formats including HTML,=20=

plain text, PS and PDF. It is rather easy to implement a new backend.

You can have a look at <http://projects.type-z.org>. Documentation is=20
included into the source distros, under the Documentation folder.

Hope this helps,

  -- S=E9bastien

--
=ABToo old to be alternative, too alternative to be old.=BB
<http://www.type-z.org> | Robert Smith, talking about his epitaph.

[Pyobjc-dev] Threading [Was: Re: Achieving 1.0....]

[Bill B. <bb...@co...>]

On Tuesday, October 29, 2002, at 04:17 PM, Jack Jansen wrote:
> Something else you may want to try though, if this is needed for 
> PyObjC, is to let the thread terminate. The comments in NSThread.h 
> lead me to believe that once you've made the transition to a 
> multithreaded program there's no going back. If that's indeed the case 
> then letting the thread terminate isn't a problem.

I wrote a little hunk of code to verify the behavior and you are 
correct-- a program can only become multithreaded, but will never 
revert to single threaded.   However, given that there is both a 
NSWillBecomeMultiThreadedNotification and a 
NSDidBecomeSingleThreadedNotification notification, it would seem that 
Apple intends to someday support the transition from multi->single 
threaded.

As such -- I would recommend leaving in the deadlocked thread.

NSDidBecomeSingleThreadedNotification is currently documented as "Not 
Implemented.".

(To test this case, create a new "foundation tool" project and replace 
the contents of main.m with the code shown below.

// --- start of code
#import <Foundation/Foundation.h>

@interface NSObject (TheVoid)
+ (void) intoTheVoid: aNothing;
@end
@implementation NSObject (TheVoid)
+ (void) intoTheVoid: aNothing;
{
/*  NSLock *l = [[NSLock alloc] init];
   [l lock];
   [l lock];*/
}

- (void) receiveNotification: aNotification
{
   NSLog(@"Recieved %@", aNotification);
}

- (void) timerFired: anObject
{
   NSLog(@"Timer fired.");
}
@end

int main (int argc, const char * argv[]) {
   NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init];

   [[NSNotificationCenter defaultCenter]
     addObserver: [NSObject class]
        selector: @selector(receiveNotification:)
            name: nil
          object: nil];

   [NSThread detachNewThreadSelector:@selector(intoTheVoid:) 
toTarget:[NSObject class] withObject: nil];
   [NSTimer scheduledTimerWithTimeInterval: 5.0
                                    target: [NSObject class]
                                  selector: @selector(timerFired:)
                                  userInfo: nil
                                   repeats: NO];
   [[NSRunLoop currentRunLoop] runUntilDate: [NSDate distantFuture]];

   [pool release];
   return 0;
}
// ------- end of code

2002-10-29 18:19:29.512 threadtest[14834] Recieved 
NSConcreteNotification 5a740 {name = 
NSWillBecomeMultiThreadedNotification}
2002-10-29 18:19:29.558 threadtest[14834] Recieved 
NSConcreteNotification 5a740 {name = _NSThreadDidStartNotification}
2002-10-29 18:19:29.629 threadtest[14834] Recieved 
NSConcreteNotification 5cec0 {name = NSThreadWillExitNotification; 
object = <NSThread: 0x5ce80>{num = 2, threadDictionary =
(null)
}}
2002-10-29 18:19:34.559 threadtest[14834] Timer fired.

threadtest has exited with status 0.


b.bum
We gladly feast on those who would subdue us.

Re: help file formats (was: Re: [Pyobjc-dev] problem accessing outlet)

[Donovan P. <ds...@ma...>]

On Tuesday, October 29, 2002, at 02:42 PM, Steven D. Arnold wrote:

> On 10/29/02 4:31 PM, "Jack Jansen" <Jac...@or...> wrote:
>
>>
>> Bill Fancher contributed a script to convert the Python manual
>> from it's HTML form to the form required by Apple Help Viewer
>> (see Mac/OSX/setupDocs.py in the Python CVS tree).
>>
>> Even though help viewer is a [censored] piece of [censored]
>> (although it's slightly better in Jaguar, admitted) it has the
>> advantage of the search engine and of everyone understanding it.
>> And I think it would be good to have all the Python
>> documentation together in one place, which would then have
>> sections for Python Language, IDE, PyObjC and anything else.
>
> That makes a lot of sense.  It almost sounds like I should use a 
> version of
> XML with some tool that can convert to HTML, text, or other formats.  I
> could then hack the script you mentioned so that the same XML source 
> could
> be converted to Apple Help format.
>
> I'll do a search for such tools; if anyone knows of such a tool, 
> please send
> me a link.  In the absence of such a tool, I'll go ahead with HTML for 
> now,
> and keep my eyes open.

There's been some work done recently on the documentation tool that 
Twisted uses (http://www.twistedmatrix.com). It used to be called 
generate-domdocs and it basically took XHTML and used some special 
attributes to indicate where function reference, python source files, 
etc should be placed. Moshe Zadka refactored it recently into a 
package, twisted.lore, and added LaTeX support (and thus, ps and pdf 
support)

You might want to drop by #twisted on irc.openprojects.net or email the 
twisted list to get more information. I personally would be interested 
in adding Apple Help Viewer support to lore. The only question is would 
I have time to do it :)

Donovan

help file formats (was: Re: [Pyobjc-dev] problem accessing outlet)

[Steven D. A. <st...@ne...>]

On 10/29/02 4:31 PM, "Jack Jansen" <Jac...@or...> wrote:

> 
> Bill Fancher contributed a script to convert the Python manual
> from it's HTML form to the form required by Apple Help Viewer
> (see Mac/OSX/setupDocs.py in the Python CVS tree).
> 
> Even though help viewer is a [censored] piece of [censored]
> (although it's slightly better in Jaguar, admitted) it has the
> advantage of the search engine and of everyone understanding it.
> And I think it would be good to have all the Python
> documentation together in one place, which would then have
> sections for Python Language, IDE, PyObjC and anything else.

That makes a lot of sense.  It almost sounds like I should use a version of
XML with some tool that can convert to HTML, text, or other formats.  I
could then hack the script you mentioned so that the same XML source could
be converted to Apple Help format.

I'll do a search for such tools; if anyone knows of such a tool, please send
me a link.  In the absence of such a tool, I'll go ahead with HTML for now,
and keep my eyes open.

steve
-- 

Re: [Pyobjc-dev] problem accessing outlet

[Jack J. <Jac...@or...>]

On dinsdag, oktober 29, 2002, at 05:16 , bb...@ma... wrote:

> Thank you! That would be brilliant -- we can drop it in the 
> Docs directory.   At some point, I really want to pull out the 
> web site and stick it in CVS.  It would be doubly-brilliant if 
> we could compile the web site and the shipped-with-the-module 
> docs from the same tree of source [Fink does this -- and Fink 
> compiles the content down to HTML, text, and-- I believe-- some 
> other formats all from the same source].
>
> But-- a "getting started writing Cocoa Python" apps would be a 
> HUGE step forward from where we are today!

Bill Fancher contributed a script to convert the Python manual 
from it's HTML form to the form required by Apple Help Viewer 
(see Mac/OSX/setupDocs.py in the Python CVS tree).

Even though help viewer is a [censored] piece of [censored] 
(although it's slightly better in Jaguar, admitted) it has the 
advantage of the search engine and of everyone understanding it. 
And I think it would be good to have all the Python 
documentation together in one place, which would then have 
sections for Python Language, IDE, PyObjC and anything else.
--
- Jack Jansen        <Jac...@or...>        
http://www.cwi.nl/~jack -
- If I can't dance I don't want to be part of your revolution -- 
Emma Goldman -