From fredrik.lundh@image.combitech.se Mon Jun 2 18:28:59 1997 From: fredrik.lundh@image.combitech.se (Fredrik Lundh) Date: Mon, 2 Jun 1997 19:28:59 +0200 Subject: [PYTHON IMAGE-SIG] PIL 0.3a1 now available Message-ID: <9706021728.AA23747@arnold.image.ivab.se> Folks, In the great PIL tradition of only releasing alphas and betas, I've decided to skip 0.2 final and go directly on to 0.3. The reason is simple: I've added way too many features during the 0.2 beta phase (cannot really complain about that, can you?). Anyway, to make up for this, I plan to make a quick test cycle for 0.3. There will probably be only one beta (some cool user-provided patches, a few other fixes, a handbook), in mid-June, followed by a 0.3final towards the end of the month. The alpha release can be found here: http://starship.skyport.net/crew/fredrik/Imaging-0.3a1.tar.gz (240k) A prebuilt Windows binary will be available later this week. Enjoy /F ==================================================================== The Python Imaging Library 0.3a1 ==================================================================== This is an alpha release of the new Python Imaging Library (PIL). News highlights in 0.3a1 (as compared to 0.2b4): + Support floating point images, including new decoders to deal with virtually any kind of input data. + The Tk interface now supports transparent images. *** NOTE: the Tk interface has been changed; you need to rebuild *** your Tkinter properly handle the new interface. + Added McIdas read support. + Added PIXAR raster read support. + Added FlashPix (FPX) read support. + Added Microsoft Image Composer (MIC) read support. + Fixed the FLI/FLC reader to properly read FLC files. + Fixed the PDF writer. It now properly writes "L" and "RGB" images. + Fixed various minor problems with 0.2b4 For more information on these and other changes, see the CHANGES document included in the distribution. _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Zachary_Roadhouse@brown.edu Mon Jun 2 19:37:55 1997 From: Zachary_Roadhouse@brown.edu (Zachary_Roadhouse@brown.edu) Date: Mon, 02 Jun 1997 14:37:55 -0400 (EDT) Subject: [PYTHON IMAGE-SIG] [Announce] PythonMagick 0.1a6 Message-ID: In the tradition of releasing things before going away, I am going to release PythonMagick 0.1a6. Although it is far from complete, almost all of the ImageMagick image manipulation routines have been finished. You can't set any of the image attributes but getting them should work. I have created routines to convert from MagickImage objects to the PIL image object and have some array conversion utilities as well. The MagickImage object is similar to the PIL object in that it offers mode, size, fromstring, and tostring members. This should make it flexible enough to get by for now. There is a web page detailing the release which should correspond to what the release actually contains. It is located at: http://starship.skyport.net/crew/zack/pymagick/ The release can be ftp'd from ftp://ftp.lems.brown.edu/pub/turtles/magick/ I'm leaving for the summer, hopefully I'll have e-mail and will be able to continue to improve PythonMagick. --Zack /--------------------------------------------------------------------- Zachary F. Roadhouse Home: (401)863-xxxx Box 220, Brown University, Providence, RI 02912 Work: (401)863-2177 E-MAIL: Zachary_Roadhouse@brown.edu WWW: http://www.techhouse.brown.edu/~zack --------------End of forwarded message------------------------- --Zack /--------------------------------------------------------------------- Zachary F. Roadhouse Home: (401)863-xxxx Box 220, Brown University, Providence, RI 02912 Work: (401)863-2177 E-MAIL: Zachary_Roadhouse@brown.edu WWW: http://www.techhouse.brown.edu/~zack _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Jack.Jansen@cwi.nl Tue Jun 3 09:37:43 1997 From: Jack.Jansen@cwi.nl (Jack Jansen) Date: Tue, 03 Jun 1997 10:37:43 +0200 Subject: [PYTHON IMAGE-SIG] PIL 0.3a1 now available In-Reply-To: Message by Fredrik Lundh , Mon, 2 Jun 1997 19:28:59 +0200 , <9706021728.AA23747@arnold.image.ivab.se> Message-ID: <9706030837.AA24117=jack@snelboot.cwi.nl> Fredrik, are your alpha and beta releases really alpha and beta? If they're all tested equally well and it's just a question of more features and such it might be better to adopt an 0.3.1 version numbering scheme: I didn't build any of the 0.2b?? for the mac because I was waiting for the final 0.2, but now it appears to have been skipped... -- Jack Jansen | ++++ stop the execution of Mumia Abu-Jamal ++++ Jack.Jansen@cwi.nl | ++++ if you agree copy these lines to your sig ++++ http://www.cwi.nl/~jack | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Jack.Jansen@cwi.nl Tue Jun 3 09:37:43 1997 From: Jack.Jansen@cwi.nl (Jack Jansen) Date: Tue, 03 Jun 1997 10:37:43 +0200 Subject: [PYTHON IMAGE-SIG] PIL 0.3a1 now available In-Reply-To: Message by Fredrik Lundh , Mon, 2 Jun 1997 19:28:59 +0200 , <9706021728.AA23747@arnold.image.ivab.se> Message-ID: <9706030837.AA24117=jack@snelboot.cwi.nl> Fredrik, are your alpha and beta releases really alpha and beta? If they're all tested equally well and it's just a question of more features and such it might be better to adopt an 0.3.1 version numbering scheme: I didn't build any of the 0.2b?? for the mac because I was waiting for the final 0.2, but now it appears to have been skipped... -- Jack Jansen | ++++ stop the execution of Mumia Abu-Jamal ++++ Jack.Jansen@cwi.nl | ++++ if you agree copy these lines to your sig ++++ http://www.cwi.nl/~jack | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From fredrik.lundh@image.combitech.se Tue Jun 3 09:56:30 1997 From: fredrik.lundh@image.combitech.se (Fredrik Lundh) Date: Tue, 3 Jun 1997 10:56:30 +0200 Subject: [PYTHON IMAGE-SIG] PIL 0.3a1 now available In-Reply-To: <9706030837.AA24117=jack@snelboot.cwi.nl> (message from Jack Jansen on Tue, 03 Jun 1997 10:37:43 +0200) Message-ID: <9706030856.AA19068@arnold.image.ivab.se> Jack asked: > are your alpha and beta releases really alpha and beta? If they're > all tested equally well and it's just a question of more features > and such it might be better to adopt an 0.3.1 version numbering > scheme The jump from 0.2b5 to 0.3a1 was an attempt to get back on track; the 0.2 beta phase became much longer than I expected, and loads of features made their way into the library (we do apologize for the inconvenience) without the few remaining 0.2 issues being resolved. And when I redesigned the Tk interface to support transparency, and added support for floating point data, I found it somewhat hard to continue calling it 0.2... As for 0.3, I now have a well defined list of things to look at before 0.3b1 (mid June), and there will probably be no more releases before 0.3 final (end of June). So for the Mac, you can just sit back and wait for 0.3b1. The next release may be named 0.3.1 or 0.4, I don't really know... Cheers /F _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From mikec@math.ubc.ca Sat Jun 7 00:38:44 1997 From: mikec@math.ubc.ca (Michael Cherkassoff) Date: Fri, 06 Jun 1997 16:38:44 -0700 Subject: [PYTHON IMAGE-SIG] Bug ? Message-ID: <33989F84.3D@math.ubc.ca> Hi, there. It looks like a bug: When I create new image with Image.new("RGB",(sizex,sizey)) then it usually does not get initialized to all black as the manual claims. Usually it has some garbage pixels or even some previous image (in color) if you work interactively and have created some before. With explicit color initializer it works fine though. Michael. -------------------------------------------------------------- Michael Cherkassoff (mikec@math.ubc.ca) Department of Mathematics, University of British Columbia, B.C. _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From mikec@math.ubc.ca Sat Jun 7 00:38:44 1997 From: mikec@math.ubc.ca (Michael Cherkassoff) Date: Fri, 06 Jun 1997 16:38:44 -0700 Subject: [PYTHON IMAGE-SIG] Bug ? Message-ID: <33989F84.3D@math.ubc.ca> Hi, there. It looks like a bug: When I create new image with Image.new("RGB",(sizex,sizey)) then it usually does not get initialized to all black as the manual claims. Usually it has some garbage pixels or even some previous image (in color) if you work interactively and have created some before. With explicit color initializer it works fine though. Michael. -------------------------------------------------------------- Michael Cherkassoff (mikec@math.ubc.ca) Department of Mathematics, University of British Columbia, B.C. _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Dirk.Engelmann@iwr.uni-heidelberg.de Sat Jun 7 23:28:53 1997 From: Dirk.Engelmann@iwr.uni-heidelberg.de (Dirk Engelmann) Date: Sun, 8 Jun 1997 00:28:53 +0200 (MET DST) Subject: [PYTHON IMAGE-SIG] tkinter without tkappinit.c ? Message-ID: Is there a way to avoid linking tkappinit.c in _tkinter module ? I've got problems with _tkinter.so using tix (some tix elements dont work). Once I read something about avoiding tkappinit.c - but don't remember how it was. Thanks for help! Dirk Engelmann _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From fredrik.lundh@image.combitech.se Sun Jun 8 18:18:44 1997 From: fredrik.lundh@image.combitech.se (Fredrik Lundh) Date: Sun, 8 Jun 1997 19:18:44 +0200 Subject: [PYTHON IMAGE-SIG] Bug ? In-Reply-To: <33989F84.3D@math.ubc.ca> (message from Michael Cherkassoff on Fri, 06 Jun 1997 16:38:44 -0700) Message-ID: <9706081718.AA24875@arnold.image.ivab.se> > When I create new image with > > Image.new("RGB",(sizex,sizey)) > > then it usually does not get initialized to all black > as the manual claims. Usually it has some garbage pixels or > even some previous image (in color) if you work interactively > and have created some before. Yep, it's definitely a buglet (the documentation and the code doesn't match). In 0.2b4, the syntax is: Image.new(mode, size) # don't initialize Image.new(mode, size, ink) # initialize to ink In 0.3, this has been changed to: Image.new(mode, size) # initialize to black Image.new(mode, size, ink) # initialize to ink Image.new(mode, size, None) # don't initialize to make life a bit easier... Cheers /F http://hem1.passagen.se/eff (python stuff) http://starship.skyport.net/crew/fredrik (more python stuff) _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Fred L. Drake, Jr." References: Message-ID: <199706091811.OAA06321@weyr.cnri.reston.va.us> Dirk Engelmann writes: > > Once I read something about avoiding tkappinit.c - but don't > remember how it was. I've forwarded my recent post on this topic to Dirk, but he brings up a good point; the changes didn't make it in for the 0.3a1 release. Fredrik, were there problems with my solution or were they incompatible with other changes to PIL? I've not had time to play with the new release yet. -Fred -- Fred L. Drake, Jr. fdrake@cnri.reston.va.us Corporation for National Research Initiatives 1895 Preston White Drive Reston, VA 20191-5434 _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Fredrik Lundh" > Fredrik, were there problems with my solution or were they >incompatible with other changes to PIL? I've not had time to play >with the new release yet. And I haven't had time to look them over yet; the 0.3 interface is changed, but I don't think it will be much of problem to adapt your fixes. Will make it into 0.3b1 (next weekend, if everything turns out well). Cheers /F _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From LEHMANNJ@saatchi.com.au Tue Jun 10 21:29:00 1997 From: LEHMANNJ@saatchi.com.au (John Lehmann (SSASyd)) Date: Tue, 10 Jun 97 20:29:00 S Subject: [PYTHON IMAGE-SIG] Save a 5 bit gif as a 5 bit gif? Message-ID: <339D1ED5@smtp.saatchi.com.au> Hi, I have a large 32 colour (5bit) gif (a map), from which I crop a smaller image in response to a click (zooming in on a square of an image-map). When I save this smaller image as a GIF, PIL seems to default to using an 8-bit palette (filling the remaining space with shades of grey) which has the side effect of making the smaller image about 55k - almost as big as the original. As a 5-bit image, the gif is more like 15k. However, I cannot find any obvious way of setting the bit depth that I want for the gif. I tried looking through the source but got confused. There does seem to be some support for setting the bit depth of a gif... is this likely to turn out to be a fruitful line of inquiry? -- John J Lehmann--lehmannj@saatchi.com.au _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Fredrik Lundh" >When I save this smaller image as a GIF, PIL seems to default to using = an =20 >8-bit palette (filling the remaining space with shades of grey) which = has =20 >the side effect of making the smaller image about 55k - almost as big = as =20 >the original. As a 5-bit image, the gif is more like 15k. However, I = >cannot find any obvious way of setting the bit depth that I want for = the =20 >gif. The problem here is not really the bitdepth, it's the LZW patentl... Since reaching an agreement with Unisys turned out to be impossible (it would have meant me *and* the PIL users paying them for each *instance* of PIL being in use, and also me keeping a registry of all PIL users; they also consider all correspondence on the matter as being their property ;-), I instead added a mechanism to write "uncompressed" GIF files. Such files can still be read by any LZW decoder, but are, as you noted, very large. And using this technique only works well with 8-bit images... If you're on a unix platform, you can change the GifImagePlugin to use NetPBM instead (see comments in the file for details). Otherwise, I have no good ideas (if Netscape had cared, we'd all be using PNG instead by now...) Cheers /F _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From tismer@appliedbiometrics.com Tue Jun 10 12:20:10 1997 From: tismer@appliedbiometrics.com (Christian Tismer) Date: Tue, 10 Jun 1997 13:20:10 +0200 Subject: [PYTHON IMAGE-SIG] Save a 5 bit gif as a 5 bit gif? References: <9706101016.AA19689@arnold.image.ivab.se> Message-ID: <339D386A.416E3385@tismer.com> ... > The problem here is not really the bitdepth, it's the LZW patentl... Oops! Does this apply to the zlib also? I thought it was used in PIL, and there were no license prob? Or do I talk about different things. cheers - pirx -- Christian Tismer :^) Applied Biometrics GmbH : Have a break! Take a ride on Python's Carmerstr. 2, 10623 Berlin : *Starship* http://starship.skyport.net we're tired of banana software - shipped green, ripens at home _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Fredrik Lundh" >> The problem here is not really the bitdepth, it's the LZW patentl... > >Oops! Does this apply to the zlib also? I thought it was used in PIL, >and there were no license prob? Or do I talk about different things. LZW (as used in GIF, TIFF, PDF, Postscript, and compress) is patented by Unisys (and also by IBM ;-). Unisys simply cannot seem to understand that there's such things as free software (they basically told one freeware author that he was an idiot since he gave away software; had a quite absurd mail dialogue with them myself which they explicitly told me I wasn't allowed to post ;-). The ZIP (aka deflate) compression (used in ZLIB, PNG, PDF, recent Microsoft apps, various ZIP applications, squeeze, and everyone else) is (most likely) patent free. So, wherever you possible can, use PNG instead of GIF. Unfortunately, too many browsers still don't support it (MSIE 4 does, Netscape doesn't). Cheers /F (yet another idiot ;-) _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From tismer@appliedbiometrics.com Tue Jun 10 15:12:48 1997 From: tismer@appliedbiometrics.com (Christian Tismer) Date: Tue, 10 Jun 1997 16:12:48 +0200 Subject: [PYTHON IMAGE-SIG] PNG support (was: Save a 5 bit gif as a 5 bit gif?) References: <9706101212.AA08847@arnold.image.ivab.se> Message-ID: <339D60E0.DDF2E451@tismer.com> oh-oh. > The ZIP (aka deflate) compression (used in ZLIB, PNG, PDF, > recent Microsoft apps, various ZIP applications, squeeze, and > everyone else) is (most likely) patent free. > > So, wherever you possible can, use PNG instead of GIF. > Unfortunately, too many browsers still don't support it > (MSIE 4 does, Netscape doesn't). Just on this one: I use the NS Communicator 4.0 now, and it performs very well yet, as far I can tell from 4 days. Also the mailer is much improved, can use HTML, and has mutlltiple users and rules to file your mail from the inboy automatically. There is a PNG plugin for it from SIEGEL & GALE: http://codelab.siegelgale.com/solutions/index.html They have it for W95, NT and Power Mäc. I didn't try it yet, but people might like to know it. Uh, if I had time to write my own Plugin for NS.. There's another one: Did you ever look at FIF? Fractal imaging? I have no idea what the policy is, but the images are great and scalable. http://www.iterated.com I follow this development since years now, but they didn't break through. > Cheers /F (yet another idiot ;-) I'm in the club :-] -- Christian Tismer :^) Applied Biometrics GmbH : Have a break! Take a ride on Python's Carmerstr. 2, 10623 Berlin : *Starship* http://starship.skyport.net we're tired of banana software - shipped green, ripens at home _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From kajiyama@grad.sccs.chukyo-u.ac.jp Fri Jun 13 03:32:00 1997 From: kajiyama@grad.sccs.chukyo-u.ac.jp (Tamito Kajiyama) Date: Fri, 13 Jun 97 02:32:00 JST Subject: [IMAGE-SIG] Compiling PIL 0.3a1 on SunOS 4.1.4_JL Message-ID: <9706121732.AA15521@pseudo.grad.sccs.chukyo-u.ac.jp> Hi, PIL maintainers. I tried out the PIL extension version 0.3a1 last night. I've used a Sun4/5 running SunOS 4.1.4_JL. I compiled the extension with gcc 2.7.2.1, and installed it in combination with Python 1.4 final. To compile the source files and obtain `_imagingmodule.so', I needed to edit `Imaging/Makefile' and `Imaging/libImaging/Makefile' manually, since the configure script did not recognize the needs of the compiler option `-fPIC' which make the compiler generate position-independent code. Moreover, I edited the former Makefile so that the linker `ld' linked a shared form of the libjpeg-6a library (I also re-installed the libjpeg-6a library, since I'd installed only `libjpeg.a'). I hope this message help you. If these are already known problems, please excuse me ;-) Regargs, -- Kajiyama Tamito (e-mail: kajiyama@grad.sccs.chukyo-u.ac.jp) _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From klm@CNRI.Reston.Va.US Thu Jun 12 20:01:24 1997 From: klm@CNRI.Reston.Va.US (Ken Manheimer) Date: Thu, 12 Jun 1997 15:01:24 -0400 (EDT) Subject: [IMAGE-SIG] test Message-ID: <199706121901.PAA12429@glyph.cnri.reston.va.us> ing. _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From amk@magnet.com Wed Jun 18 19:02:37 1997 From: amk@magnet.com (Andrew Kuchling) Date: Wed, 18 Jun 1997 14:02:37 -0400 (EDT) Subject: [IMAGE-SIG] Creating a new Image object from C code Message-ID: <199706181802.OAA28299@lemur.magnet.com> I'm working on a PIL/SANE interface, and am wondering how I should create a new Image object in C code, to be returned as the scanner's output. There's the .putdata() image method, which wants a sequence object containing the pixel data. However, the SANE interface works in chunks (the processing loop looks like this: st=SANE_STATUS_GOOD; while (st!=SANE_STATUS_EOF) { st=sane_read(self->h, buffer, 8192, &len); if (st && st!=SANE_STATUS_EOF) return PySane_Error(st); /* XXX Do something with the len bytes of data in buffer[] */ } ) and I'm wondering if I have to accumulate all the data until the whole frame is done. That's probably not optimal, since an 8.5x11 page, scanned at 300dpi, takes a lot of room. Andrew Kuchling amk@magnet.com http://people.magnet.com/%7Eamk/ _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From amk@magnet.com Thu Jun 19 22:22:33 1997 From: amk@magnet.com (Andrew Kuchling) Date: Thu, 19 Jun 1997 17:22:33 -0400 (EDT) Subject: [IMAGE-SIG] PIL/SANE interface Message-ID: <199706192122.RAA02962@lemur.magnet.com> I've completed a first version of a SANE interface for the Python Imaging Library. SANE is a freeware API for various raster devices like scanners and digital cameras; it runs on various Unixen. The Python Imaging Library, of course, allows you to read and write images in various formats, and manipulate them in various ways. It seems like a good fit to me... You can get the code from http://starship.skyport.net/crew/amk/projects/ . If you try it out, please let me know how it works for you. SANE home page: http://www.azstarnet.com/~axplinux/sane/ PIL home page: http://www.python.org/sigs/image-sig/Imaging.html Here's a sample script that gets a single image from a QuickCam, and writes the image out to a GIF file. The script's output is: SANE version: (327680, 0, 5, 0) Available devices= [('qcam:0x378', 'Connectix', 'B&W QuickCam', 'video camera'), ('pnm:0', 'Noname', 'PNM file reader', 'virtual device'), ('pnm:1', 'Noname', 'PNM file reader', 'virtual device')] SaneDev object= Device parameters: ('L', 1, (330, 246), 8, 330) And the script itself is: #!/bin/env python # Get the path set up properly import sys ; sys.path.append('.') import sane print 'SANE version:', sane.init() print 'Available devices=', sane.get_devices() scanner=sane.open('qcam:0x378') print 'SaneDev object=', scanner print 'Device parameters:', scanner.get_parameters() # Set scan parameters scanner.contrast=170 ; scanner.brightness=150 ; scanner.white_level=190 scanner.depth=6 scanner.br_x=320 ; scanner.br_y=240 # Initiate the scan scanner.start() # Get an Image object containing the scanned image # (For my B&W QuickCam, this is a grey-scale image. Other scanning devices # may return an RGB image, or a single channel of an RGB image.) im=scanner.snap() # Write the image out as a GIF file im.save('foo.gif') # The show method() simply saves the image to a temporary file and calls "xv". #im.show() Andrew Kuchling amk@magnet.com http://people.magnet.com/%7Eamk/ _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From hewitt@pootle.cgu.mcc.ac.uk Wed Jun 25 08:05:08 1997 From: hewitt@pootle.cgu.mcc.ac.uk (W T Hewitt) Date: Wed, 25 Jun 1997 08:05:08 +0100 (BST) Subject: [IMAGE-SIG] Bugs in PIL Message-ID: I'm trying to compile the 0.2b4 release on IRIX 6.3: --------------------------------------------------- The Python Imaging Library $Id: README,v 1.9 1997/01/14 18:01:06 fl Exp fl $ Release 0.2b4 (January 10, 1997) --------------------------------------------------- 1) SG now seem to supply a library called /usr/lib/libjpeg.so, but there is no signs of the header files (jpeglib.h) 2) I got jpeg-6a and installed it 3) JpegEncode.c and JpegDecode.c had compilation errors... 1) calls to macros GLOBAL and METHODDEF need their parameters in (), e.g., GLOBAL void should be GLOBAL(void) 2)METHODDEF is misspelt in JpegDecode. Sorry I can't remember which one. I forgot to keep an original Hope this can be included in the next release. Please note I'm not a member of the mailing list Best wishes Terry W T Hewitt Telephone: +44 161 275 6095 Manchester Visualization Centre Fax: +44 161 275 6800 Manchester Computing Electronic mail: w.t.hewitt@mcc.ac.uk University of Manchester Manchester M13 9PL United Kingdom _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________ From Fredrik Lundh" >2) I got jpeg-6a and installed it > >3) JpegEncode.c and JpegDecode.c had compilation errors... > > 1) calls to macros GLOBAL and METHODDEF need their parameters in (), e.g., > GLOBAL void should be GLOBAL(void) Yep, this is a change between 6 and 6a (no real changes affects PIL's use of the JPEG library, so I've just been lazy...). In the 0.3a1 release, the README points this out (not sure about the 0.2 release), but okay, I better update the code. It's just that libjpeg doesn't build out of the box under Windows that has held me back. > 2)METHODDEF is misspelt in JpegDecode. > Sorry I can't remember which one. I forgot to keep an original Not in my copy ??? Cheers /F _______________ IMAGE-SIG - SIG on Image Processing with Python send messages to: image-sig@python.org administrivia to: image-sig-request@python.org _______________