Just to make sure I am not crazy, I checked 4.2.3. getimagesize() on that azj.jpg image returns:
array(6) { [0]=> int(1536) [1]=> int(1024) [2]=> int(2) [3]=> string(26) "width="1536" height="1024"" ["bits"]=> int(8) ["channels"]=> int(3) } which is correct. So we definitely do have breakage The streams related change was: - FP_FREAD(temp, sizeof(temp), socketd, fp, issock); - FP_FREAD((char*) &dim, sizeof(dim), socketd, fp, issock); + php_stream_read(stream, temp, sizeof(temp)); + php_stream_read(stream, (char*) &dim, sizeof(dim)); - if((FP_FREAD(a, sizeof(a), socketd, fp, issock)) <= 0) return 0; + if((php_stream_read(stream, a, sizeof(a))) <= 0) return 0; Along with a few more, but they look correct to me, so I don't think the problem is there. I am more suspicious of the php_next_marker() changes.. But who knows, the unified diff of image.c is 958 lines between 4.2.3 and current HEAD. -Rasmus On Wed, 2 Oct 2002, Andrei Zmievski wrote: > On Wed, 02 Oct 2002, Rasmus Lerdorf wrote: > > Hrm... Actually, after a bit more checking it is only some jpegs that > > don't work. Specifically the ones coming from my digital camera are no > > longer working. Images that used to work with getimagesize() are now not, > > so I think something changed. Try running getimagesize() on this image: > > > > http://lerdorf.com/azj.jpg > > Yes, it doesn't work for me either. I don't see anything in image.c that > would have changed, so it must be something in the streams code, > perhaps? > > -Andrei http://www.gravitonic.com/ > > Commitment, n.: > Commitment can be illustrated by a breakfast of ham and eggs. The chicken > was involved, the pig was committed. > -- PHP Development Mailing List <http://www.php.net/> To unsubscribe, visit: http://www.php.net/unsub.php