John Bowler

The fix seems reasonable and correct.

``` If I change line 421 in pngrutil.c from #if ZLIB_VERNUM >= 0x1290 && to #if ZLIB_VERNUM < 0x1290 && \ ``` and from the original report: ``` Undefined symbols...

This seems to be an assert but because the condition (class with itself as an ancestor) may perhaps be produced by code outside PythonQt a qFatal is probably a better...

I didn't even know about that new repo...

Incidentally I don't think it can be _just_ archived. I still have the canonical repo as git:://git.code.sf.net/p/libpng/code; I assume that is what should be used by distros, however glennrp is...

[email protected] Or, if you want to make it so I can classify automagically [email protected], but there's more chance of borkage with @jbowler.com because I maintain the MX records myself to...

Not me :-) I'm not actually on any of the libpng mailing lists (I think) or at least I just file messages sent to them.

#293 is a duplicate of this. It contains this proposed fix: https://github.com/openembedded/openembedded-core/commit/38c6b26d5543d75304aec58adb3e2a54253afc25 However there is no bug to fix here; the memory is freed immediately because the program exits immediately....

EDIT: revised patch. There is a much more pervasive version of the same behavior: pngcp --nowrite --search *.png or to make it really bad: find / -name '*.png' -print0 |...

There have been many changes to the cmake build system which was unmaintained at the time you wrote the original bug report. It should be fixed now and it would...