[Pkg-gmagick-im-team] Bug#656942: closed by Bastien Roucariès <roucaries.bastien+debian at gmail.com> (Bug#656942: fixed in imagemagick 8:6.7.7.2-1)

[Jason Woofenden]

On 2012-05-31 09:51PM, Bastien ROUCARIES wrote:
> Please could you explain me, better. Particularly what are you
> expecting this time.

Here's the bit from my original bug report:

> >> $ identify -verbose rock48.png | grep Class
> >> Â  Class: DirectClass
> >>
> >> Oops! should be:
> >>
> >> Â  Class: PseudoClass

This is still broken.

Test case #1
============

This test produces the wrong result in both these versions of
imagemagic package:

+	8:6.6.9.7-5+b2 (the version I reportbuged)

+	8:6.7.7.2-1 (the version you posted as fixing this bug)

Command:

	identify -verbose rock48.png | grep Class

Current (incorrect) output:

	Class: DirectClass

Correct/Expected output:

	Class: PseudoClass


Test case #2
============

This test produced the correct results before (8:6.6.9.7-5+b2) but
now (8:6.7.7.2-1) produces incorrect output.

Command:

	identify rock48.png

Output (correct) from version 8:6.6.9.7-5+b2:

	rock48.png PNG 20x20 20x20+0+0 8-bit PseudoClass 981B 0.000u 0:00.000

Output (incorrect) from version 8:6.7.7.2-1:

	rock48.png PNG 20x20 20x20+0+0 8-bit DirectClass 981B 0.000u 0:00.000

Correct/Expected result:

	rock48.png PNG 20x20 20x20+0+0 8-bit PseudoClass 981B 0.000u 0:00.000


Notes
=====

This is assuming that my understanding (that PseudoClass means the
image has a palette) is correct. Perhaps I am wrong about this. If
so: please tell me. I said in my original report that I wasn't
sure, and asked in my follow-up.

Please run the above commands on the rock48.png that I attached to
my original bug report.

     - Jason