Search found 73 matches
- 2014-05-26T06:46:42-07:00
- Forum: Bugs
- Topic: Repeatedly using DrawSetClipPath shows corruption
- Replies: 13
- Views: 14672
Re: Repeatedly using DrawSetClipPath shows corruption
Hi Dlemstra, The problem does still occur when the 'extraClipping' is totally disabled, e.g. http://www.basereality.com/image/2409/clipSecond_2.png Perhaps I didn't make one thing clear though - in the code half the images are produced with the 'extraClipping' enabled and half are not, depending on ...
- 2014-05-24T09:14:03-07:00
- Forum: Bugs
- Topic: Repeatedly using DrawSetClipPath shows corruption
- Replies: 13
- Views: 14672
Re: Repeatedly using DrawSetClipPath shows corruption
Did you use the php code to create the images? No. I used the code I said I was using. For the record it was tested on : CentOS release 6.4 gcc version 4.4.7 20120313 (Red Hat 4.4.7-4) (GCC) I am getting the following images: Just to note, those images also appear incorrect. Although they don't hav...
- 2014-05-24T08:33:42-07:00
- Forum: Bugs
- Topic: Repeatedly using DrawSetClipPath shows corruption
- Replies: 13
- Views: 14672
Re: Repeatedly using DrawSetClipPath shows corruption
For completeness, the issue was found originally through the PHP Imagick extension under PHP-FPM. That process manager processes multiple requests before stopping and restarting the PHP worker instance. The test for setClipPath works a couple of times, and then goes wonky. Original code: function se...
- 2014-05-24T08:27:27-07:00
- Forum: Bugs
- Topic: Repeatedly using DrawSetClipPath shows corruption
- Replies: 13
- Views: 14672
Re: Repeatedly using DrawSetClipPath shows corruption
I'm currently on the beta of 6.8.9-2 from a few days ago, but this also happens on 6.8.8-9 /usr/local/bin/convert --version Version: ImageMagick 6.8.9-2 Q16 x86_64 2014-05-21 http://www.imagemagick.org Copyright: Copyright (C) 1999-2014 ImageMagick Studio LLC Features: DPC OpenMP Delegates: bzlib fr...
- 2014-05-24T07:07:45-07:00
- Forum: Bugs
- Topic: Repeatedly using DrawSetClipPath shows corruption
- Replies: 13
- Views: 14672
Repeatedly using DrawSetClipPath shows corruption
Hi, I think I've found a pretty serious bug where DrawSetClipPath shows data corruption. The code example below is doing the following: * Create a DrawingWand. * Defining a clip path which is either just a rectangle in the top left quadrant of the image, or the same but with extra clip regions acros...
- 2014-05-21T15:54:01-07:00
- Forum: Bugs
- Topic: MagickIdentifyImage + pseudo image crash
- Replies: 1
- Views: 5171
MagickIdentifyImage + pseudo image crash
Hi, Calling MagickIdentifyImage on an image created through MagickNewImage appears to crash with the message: identify: magick/magick.c:662: GetMagickMimeType: Assertion `magick_info != (MagickInfo *) ((void *)0)' failed. about to identifyAborted Code example is below, or can be downloaded from http...
- 2014-05-21T09:16:41-07:00
- Forum: Bugs
- Topic: gradient with rgba color with spaces
- Replies: 1
- Views: 5193
gradient with rgba color with spaces
Hi, Are spaces meant to be allowed in rgba color strings when making gradients? I'm seeing different results when I run: convert -size 400x400 'gradient:red-rgba(0, 0, 255, 0.1)' test_1.png convert -size 400x400 'gradient:red-rgba(0,0,255,0.1)' test_2.png Produce the two images: http://cdn5.www.base...
- 2014-05-18T05:16:00-07:00
- Forum: Bugs
- Topic: Unitialized memory access corrupting images
- Replies: 6
- Views: 11616
Re: Valgrind reports issues.
Hi, I'm reasonably sure that this issue is causing some image corruption, or at least it's highly coincidental to some image corruption. I don't have a 100% reproducible test case, but what I'm seeing is a that sometimes there is a small amount of corruption when calling the ImageMagick wand methods...
- 2014-05-15T07:50:40-07:00
- Forum: Bugs
- Topic: Equality operator in fx image
- Replies: 1
- Views: 5037
Equality operator in fx image
Hi, I think I've discovered a bug in the equality operator in the fx operator. convert -size 256x256 "xc:pink" input1.jpg convert -size 256x256 "gradient:black-white" input2.jpg convert input1.jpg input2.jpg -fx "colorInt=int(256 * v.p{0,j}.lightness); pos = int(256 * i/w); ...
- 2014-05-08T11:24:55-07:00
- Forum: Bugs
- Topic: Unitialized memory access corrupting images
- Replies: 6
- Views: 11616
Unitialized memory access corrupting images
*Edit - was "Valgrind reports issues." but there's a repo case now.* While investigating some issue with PHP Imagick library with Valgrind, it reported that there were some uninitialised variables being used by ImageMagick. These are also shown when running ImageMagick directly from the co...
- 2013-10-18T16:17:53-07:00
- Forum: Bugs
- Topic: Possible bug in IsPixelWandSimilar
- Replies: 3
- Views: 4206
Re: Possible bug in IsPixelWandSimilar
Thanks for the reply. Just to be clear then, if I want to know if two RGB colours are within a distance of 5 units (with the axes being 255 units long), then I should do: #define ROOT_3 1.732050807568877 fuzz = (5 * QuantumRange / 255.0) / ROOT_3; That's correct isn't it? cheers Dan *edit* clarified...
- 2013-10-18T13:19:12-07:00
- Forum: Bugs
- Topic: Possible bug in IsPixelWandSimilar
- Replies: 3
- Views: 4206
Possible bug in IsPixelWandSimilar
Hi, I think I may have found a bug in the IsPixelWandSimilar() function. When comparing the two colours 'rgb(255, 0, 0)' and 'rgb(250, 0, 0)' with a fuzz factor of 4 (before scaling the fuzz factor into the appropriate scale) IsPixelWandSimilar() returns true, when it should return false, as those c...
- 2013-10-14T10:12:31-07:00
- Forum: Developers
- Topic: Force 64bit compilation?
- Replies: 0
- Views: 4608
Force 64bit compilation?
Hi, I'm compiling from the ImageMagick-6.8.7-0 source on a Centos 64bit vagrant box, http://developer.nrel.gov/downloads/vagrant-boxes/CentOS-6.4-x86_64-v20130427.box It appears that only the 32bit libraries are being built. Is there any way to force the 64bit libraries to be built? This is the conf...