lp:libdepixelize

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

187. By Vinipsmaker on 2014-05-13

Fix build under restrict glib (glib with deprecated features disabled).

Patch provided by Johan Engelen.

186. By Vinipsmaker on 2014-03-31

Bugfix: correctly following the proposed algorithmic extension in r184.

185. By Vinipsmaker on 2014-03-31

I forgot to argue why the change was an improvement in the latest commit.

184. By Vinipsmaker on 2014-03-30

Improved response to evil pattern.

Now the quality of the output images is better.

183. By Vinipsmaker on 2014-03-27

Fix bug in loop within the sanitize.sh test script

182. By Vinipsmaker on 2014-03-27

Add template script to instrumentize libdepixelize code

181. By Vinipsmaker on 2014-03-27

Fix some (not all) popt memory leaks.

These leaks don't affect the library, just the depixelize-kopf2011
binary.

Also, try to search for popt and memory leak and popt. You might be
surprised. I plan to move away from popt as long as Inkscape developers
move too.

Given popt errors, I wasn't able to run memory sanitizer on libdepixelize
itself, but address usage is fine (and leakage-free, thanks to RAII).

180. By Vinipsmaker on 2014-03-26

Fix of the previous commit confirmed.

Result is mostly the same as before.

On some cases is slightly slower and on other a little more performant. I
was wondering why the "*_safe" routine got faster on some large images
and I think it's because the use of more clever tests that makes a better
use of the cache. A second thought is that the lighter use of the vector
of crosses steals less cache from important places. The performance
regression was fixed and I won't bother to investigate any further.

179. By Vinipsmaker on 2014-03-26

New code that will hopefully help on performance regression done on
recent commits to fix behaviourial errors.

178. By Vinipsmaker on 2014-03-24

Stable method to apply heuristics on the pixel graph.

From now on, result will be correct and won't depend on iteration order.

Previous code could have problems on some cases.