Mailman 3 August 2003 - Pythonmac-SIG

Re: [Pythonmac-SIG] Python for OSX programming beginner?
by Russell E Owen March 20, 2005

March 20, 2005

>Been lurking here for a while and quietly observing. > >I am looking for a programming language and preferably and IDE for a >young nephew wanting to dip his toes in the programming pool... >Is Python worth further investigating and how good are the IDEs? >Is there something along the lines of REALbasic VisualBasic >integrated interface/code style IDE available? >What are the specialities limits of the language? >Why learn Python rather than battle on and master C or more >particularly Objective C ? >Rather difficult to answer, but how different is the learning curve? > >Many thanks for any help. Hmmm...I suggest the following: Smalltalk: Squeak and VisualWorks NC are both free (the latter only for noncommercial use). Smalltalk is a lovely, simple really powerful language that includes: - powerful IDE and class browser with integrated debugger; smalltalk has the best development environment I've ever seen for a language because it includes live debugging, live changing of classes (want to fix a method? go ahead; the change propagates; fix it and continue on the fly) - really simple yet powerful language; one of the most elegant languges I've seen, though the syntax is a bit different than C and Python programmers are used to - very rich class library, including one of the nicest collection classes (lists, sets, etc.) available Between these two, squeak is smaller and more fun and includes a nifty GUI called Morphic; however it all runs in one big window (very non-Mac like). VisualWorks is more Mac-like (at least on a Mac) and more mature, but is much larger. Both are fully cross-platform (write once, run everywhere). One limitation to smalltalk is that it's not much good for simple scripting, hence simple text processing scripts and such. Python really shines there, so... Python. This is a really nice language and he'll be able to go far with it. However, based on what you said above, some of its limitations make me nervous: - there is no really well integrated GUI, unlike the suggestions above. You can do GUI programming, but you have to choose a 3rd party package and they are fully powerful, but the two main ones (Tkinter and wxPython) are somewhat clumsy and poorly documented. - hence there is no really first rate IDE (not to put down the IDEs that are there, but in comparison to smalltalk or RealBasic...) Note: python is built into MacOS X 10.2, but it's an outdated version missing some of the things that make the language sane (such as readlines support and a GUI package). Install the missing items or a better package (if you've been reading you've seen various options for that including at least one recent binary installer). I suggest learning those two (since their few weaknesses are complimentary), but you could also consider: Buy RealBasic. The knowledge is well worth the money! It's not a wonderful language, but not so bad either and if he already knows it, well...that's a huge leg up. Java. Not sure what the IDE situation is for MacOS X, but the language is reasonable (much saner than C++). The problem is that the need to declare the type of all objects makes it a clumsy and messy language compared to Python and Smalltalk. I suggest NOT using C++. The language is too arcane and messy. The object model is insanely complicated and the pitfalls fill volumes. Learn to program first, then learn C++ later if you really have to. -- Russell

3 3

Quicktime, Carbon and MacPython
by Mark Kim Dec. 3, 2003

Dec. 3, 2003

Hello all, I'm a developer with Transana (www.transana.org) working on integrating Quicktime into our new Python version of Transana. I'm currently using MacPython 2.2.2 (or 3) on a G4 1.0Ghz with 1GB of RAM running 10.2.6. I'm having a problem with the MovieController. Currently I have disabled it in the MovieTask loop because I have created a second thread to play the loop and the combination causes the interpreter to crash. A rough outline of the code looks like (this is not syntactically correct): Qt.EnterMovies() self.theWindow = Win.CreateNewWindow(6,31,bounds) Qd.SetPort(...) self.theMovie ... = Qt.NewMovieFromFile(....) ... self.MoviePlayer = MoviePlayer(self.theMovie, self.theController) self.MoviePlayer.start() #new thread Within the class MoviePlayer(threading.Thread) while not done and not stop: gotone, evt = Evt.WaitNextEvent(0xffff, 0) etc...etc... looks almost identical to example given in MacPython I'm assuming because I've spawned a second thread and there's a Carbon event, this is causing the crash. Is my assumption correct? Is there a way to spawn a thread and have the MovieController at the same time? Or maybe I'm looking at this completely wrong? Thanks in advance, sorry if this is long, Mark Kim

2 1

Python 2.3 vs 2.2, VTK, and Carbon
by Yves Starreveld Sept. 18, 2003

Sept. 18, 2003

Hi, folks, I have been sorting through the problems with 2.3 and VTK. One odd thing is that the importation of both Carbon/Carbon.h and Python.h from 2.3 leads to the following errors: In file included from /System/Library/Frameworks/CoreServices.framework/Frameworks/ CarbonCore.framework/Headers/CarbonCore.h:165, from /System/Library/Frameworks/CoreServices.framework/Headers/ CoreServices.h:21, from /System/Library/Frameworks/Carbon.framework/Headers/Carbon.h:20, from /Users/ystarrev/Development/VTK/Rendering/ vtkCarbonRenderWindowInteractor.h:30, from /Users/ystarrev/Development/VTKRelease/Rendering/ vtkCarbonRenderWindowInteractorPython.cxx:7: /System/Library/Frameworks/CoreServices.framework/Frameworks/ CarbonCore.framework/Headers/fp.h: In function `long double scalbl(long double, long int)': /System/Library/Frameworks/CoreServices.framework/Frameworks/ CarbonCore.framework/Headers/fp.h:1896: error: ` scalb' undeclared (first use this function) /System/Library/Frameworks/CoreServices.framework/Frameworks/ CarbonCore.framework/Headers/fp.h:1896: error: (Each undeclared identifier is reported only once for each function it appears in.) Other files in vtk import Carbon.h without difficulty, so I am surmising a conflict. Has anyone run into anything similar? Thanks for any comments. Yves Starreveld

4 5

Additional binary packages for Python2.3rc1 on 10.2.6
by Bob Ippolito Sept. 4, 2003

Sept. 4, 2003

I've setup a repository for several packages that I use that aren't in jack's list yet, or weren't compiled correctly for the non-fink user. The Package Manager URL is: http://undefined.org/python/pimp/darwin-6.6-Power_Macintosh.plist I couldn't find what jack has been using to build binary packages, so I wrote one that makes it easy. It's at: http://undefined.org/python/makepimp.py Usage of makepimp is: makepimp.py modulename versionnumber It spits out a plist template of what you'd paste into the package manager file. If you use it, you'll want to change HTTPBASE and UPLOADCMD up at the top. modulename must already be installed in your site-packages, I have not tested userdir installation, but it works for me. The packages I have up so far are: readline 2.3c1 PIL 1.1.4 (with static libjpeg support, no tkinter support) egenix-mx-base 2.0.4 ctypes 0.6.2 I'll be adding more sometime (primarily: bsddb3, pyPgSQL, pygame, twisted), but the others I commonly use have non-python dependencies and I'm too lazy to write the detection stuff for all of them at the moment :) I may also add source installers and panther dp versions, but that's a lower priority since my primary machine is OS X 10.2.6. I don't like the current "you have to go fetch this dependency" way of doing things, so I'll probably end up writing some python-based installers for libraries like SDL, Tcl/Tk, BerkeleyDB, PostgreSQL, etc. that will download the binary from somewhere and install, or at least open up the .pkg for you. -bob

4 9

unixy python and applescript
by Chris Lee Sept. 2, 2003

Sept. 2, 2003

Hi All, I use an application called Pro Fit for graphing, data analysis and some numerical analysis. However I do most of my numerical work in python and now I want to link the two :) What I am doing at the moment is using an applescript to get the variables out of ProFit and then I am calling do script 'thescript v1 v2 v3....' I have a python function which extracts the variables and then returns the result to applescript which in turn passes it back to Pro Fit. It all works just fine provided you have set up all your paths in environment.plist file. The problem of course is that it is very slow. I know that Macpython can be linked directly with applescript but what I really want to do is to allow the user to choose which python they use (so they install the minimum of stuff to use it). What modules have to be installed for the UNIX pythons (either apple or third party installed) to communicate with an applescript ? Cheers Chris ################################# Chris Lee Physics Department Otago University PO Box 56 Dunedin New Zealand Phone ++64 3 479 7749 Fax ++64 3 479 0964 #################################

3 7

Gordon McMillan's installer question for the Mac
by Python Development Aug. 31, 2003

Aug. 31, 2003

Hi, Currently in our PC product, the Gordon McMillan wrapper is ran within a thread, while the other thread runs the UI. On the Mac, we had some problems with that, so we are thinking of having the UI in one process, while the wrapper app is running along side as a second process. So on the PC there is one process running two threads, and on the Mac their will be two processes running each one thread and communicating between each process other with unix sockets on the local machine. But before we went any further in the design work and time, I am checking here to see if it is going to be possible or if others have tried but failed a two process approach. I have asked Gordon McMillan, but he has not done anything with the Mac since OS 8/9. Our current product is for OS X, with an objective-C UI. And just so you know, I am having to go this way because this is an 80% PC market solution. Thank you for your time, Randy

2 2

Licensing issues
by Glenn Andreas Aug. 30, 2003

Aug. 30, 2003

So I'm getting closer to the next release of PyOXIDE, which will have full source available. It is broken into two parts: IDEKit.framework, and PyOXIDE itself. The framework is a reusable framework that handles running the editors (syntax coloring, language support, split panes, macro expansion/completion, etc...), provides preference settings, has infrastructure for "projects" (grouped listings of files, file orders, multiple targets), and various other little things. The PyOXIDE code is mostly a shim between IDEKit.framework and the underlying python code, as well as some refinements on IDEKit.framework (for example, there are no "headers" to provide a popup header function, and it only supports python). The plan was to have IDEKit.framework handled much like MySQL - released under (L)GPL, and any software that is released under GPL (or compatible OSI license) can use it for free, with a commercial license available for everybody else. The intent here is that there is a lot of work in this thing, and I don't mind sharing, but I'm not too keen on somebody else making a buck off of my work without me getting something in return. The PyOXIDE part will be either GPL or BSD. So the question is, if PyOXIDE is to aspire to become the standard MacPython IDE (which seems like a good thing), and possibly even allow for Apple to include in some post-panther cat, will this work? Jack mentioned that GPL might be a no-no for standard distribution, though fine for developer stuff. Anybody have any ideas/insights/opinions on this? -- Glenn Andreas gandreas(a)delver.com Author of Macintosh games: Theldrow 2.3, Blobbo 1.0.2, Cythera 1.0.2 Be good, and you will be lonesome

1 0

Re: [Pythonmac-SIG] Restrictions on what you can do in a Python module?
by Jack Jansen Aug. 29, 2003

Aug. 29, 2003

On 29-aug-03, at 18:38, Robb Brown wrote: >> On woensdag 27 augustus 2003, at 20:12PM, Robb Brown wrote: >>> Anyway, here's the code. The crash is on line 104, where all the >>> setup is finished and I actually ask dcmtk to connect. >> >> That line looks harmless enough. > > You'd think.... How would one go about using a debugger on a Python > module (remember, the code does not crash when not compiled as a > module). What I tend to do is as follows: % gdb python ..... (gdb) run ..... >>> import themoduletodebug >>> control-C (gdb) break thefunctioninthemodulethatIwanttosee (gdb) continue >>> do_something_that_calls_that_function Then gdb should break on the function, and you can start single-stepping, etc. Again, let me emphasize that -O0 is required if you want to keep your sanity. gcc moves code all over the place without it. -- Jack Jansen, <Jack.Jansen(a)cwi.nl>, http://www.cwi.nl/~jack If I can't dance I don't want to be part of your revolution -- Emma Goldman

1 0

Re: [Pythonmac-SIG] Restrictions on what you can do in a Python module?
by Robb Brown Aug. 29, 2003

Aug. 29, 2003

>> On woensdag 27 augustus 2003, at 20:12PM, Robb Brown wrote: >>> Anyway, here's the code. The crash is on line 104, where all the >>> setup is finished and I actually ask dcmtk to connect. >> >> That line looks harmless enough. > > You'd think.... How would one go about using a debugger on a Python > module (remember, the code does not crash when not compiled as a > module). I've been debugging with printf's and it was a trip down the > rabbit hole. That line winds through a couple of functions to a > finite state machine that uses a doubly linked list. The actual error > seems to be when one of the nodes in this list is deleted. I don't > know why this would be different in and out of a module. My next step > is to try running it as an executable to see how it SHOULD work. > > >> A few ideas: >> - Try building with -O0. The crash may actually be on a different line >> if you've compiled with optimizing on. >> - How are you compiling and linking? If you're using distutils you >> should be >> fine, otherwise please send the Makefile. Actually: don't send the >> Makefile, >> but use distutils and see if the problem goes away;-) >> - Try building the working C++ version with the exact same libraries >> and defines and other compiler flags that are used for compile/link >> of the Python module. See if it still works. >> >> Finally, you probably know what you're doing, but the treatment >> of some objects strikes me as funny. For instance, the "params" >> structure appears to be allocated by your library (because you >> initialize >> it with "ASC_createAssociationParameters(&params, ASC_DEFAULTMAXPDU)", >> but there isn't a corresponding ASC_freeAssociationParameters() or >> something >> similar. This looks suspicious. You could examine the values of the >> various pointers in gdb to make sure that they make sense. Or, even >> better, >> initialize them all to NULL at function entry, so you know you'll get >> a segmentation >> violation if there's a misunderstanding about the API. >> -- >> Jack Jansen, <Jack.Jansen(a)cwi.nl>, http://www.cwi.nl/~jack >> If I can't dance I don't want to be part of your revolution -- Emma >> Goldman >> >> _____________________________ Robb Brown Seaman Family MR Center Calgary, AB

1 0

Re: [Pythonmac-SIG] Restrictions on what you can do in a Python module?
by Robb Brown Aug. 29, 2003

Aug. 29, 2003

I thought you might be interested... dctk is a DICOM (medical image file format / network protocol for those not in the know) toolkit -- file read and write as well as network connectivity. My idea is to wrap the basic functions in Python and then writing DICOM enabled apps is no problem. Plus, I'm frequently annoyed by our expensive research PACS (a medical image archiving server)... a good Python DICOM package and existing Python database modules would let you write your own PACS in an afternoon. If it works I'm sure there would be a lot of interest... we've been considering the MERGE toolkit which is several thousand dollars a seat and is only C++. On Wednesday, August 27, 2003, at 09:15 PM, Yves Starreveld wrote: > Hi, Robb, > > That stuff sounds like it would be of fairly general interest.... > > What kind of thing are you putting together? We are stringing > something together using php at this time. > > Yves > On Wednesday, August 27, 2003, at 11:40 AM, Robb Brown wrote: > >> I'm trying to write wrappers for dcmtk (a toolkit for the medical >> imaging DICOM standard). At the moment I have three functions, to >> open an association (basically a network connection), close it, and >> send an echo request (like a ping). A test program written entirely >> in C works fine but when compiled as a module and called from Python >> I get an illegal instruction when the dcmtk function to actually open >> the association is called. I believe this function is simply opening >> a socket, connecting to a server and negotiating what flavour of the >> protocol to use. >> >> Are there things that simply can't be done in a Python module? What >> exactly is an illegal instruction? >> >> Thanks for any help, >> >> Robb >> >> _____________________________ >> Robb Brown >> Seaman Family MR Center >> Calgary, AB >> >> >> _______________________________________________ >> Pythonmac-SIG maillist - Pythonmac-SIG(a)python.org >> http://mail.python.org/mailman/listinfo/pythonmac-sig >> > > _____________________________ Robb Brown Seaman Family MR Center Calgary, AB

4 4