Skip to content

Issue 161#127

Open
btakita wants to merge 159 commits into
rmagick-temp:masterfrom
btakita:issue-161
Open

Issue 161#127
btakita wants to merge 159 commits into
rmagick-temp:masterfrom
btakita:issue-161

Conversation

@btakita

@btakita btakita commented Jan 7, 2015

Copy link
Copy Markdown

This fixes the install on Debian Sid (Crunchbang Linux).

rmagick/rmagick#161

bf4 and others added 30 commits July 31, 2014 20:43
Compile with rake-compiler.
Make rake-compiler easily installable by `bundle install`.
Ignore all build artifacts (.gitignore from Bundler template).
Fixes test failures of the type:

Magick::ImageMagickError: unable to open image
`../doc/ex/images/Button_0.gif':  @ error/blob.c/OpenBlob/2587
It was expected to be there as of 962d56c in Feb 2013
Yet was never commit as far as I can tell.

I haven't found what that image might have been, so I grabbed
an image with ICC version 2 and ICC version 4 profiles from
http://www.color.org/version4html.xalter and confirmed
the code works as expected

Magick::Image.read("./doc/ex/images/image_with_profile.jpg").first.color_profile
returns a string of characters. (Otherwise it would return nil)
Improve the installation instructions somewhat and organize the text
a bit better.
I took the "correct" value right from the RSpec output on Travis CI:

    test_limit_resources(Magick_UT)
    [/home/travis/build/gemhome/rmagick/test/Magick.rb:268]:
    <782635008> expected but was <3221225472>.

In fact, this value is machine and OS-dependent.
ImageMagick 6.8.9 (r15243) has the following definition:

magick/resource.c:111:

    MagickULLConstant(1536)*1024*1024, /* memory limit */

magick/magick-type.h:36:

    #  define MagickULLConstant(c)  (MagickSizeType) (c ## ULL)

magick/magick-type.h:151:

    typedef unsigned long long MagickSizeType;

When building on Windows, other definitions take place.
So, I think the initial value should not be in the test at all.
But let it be there for now, to test that the Ruby library correctly
gets the value from the C extension that correctly gets it from
ImageMagick that is running on Travis machines.
Do not test machine and OS-specific integers
Fixed: build error on Windows Ruby x64 (with ImageMagick 6.8.0-10 or Ima...
vassilevsky and others added 30 commits November 28, 2014 16:19
People still use it in production (see #150 for example)
Only require RuboCop on Ruby versions >= 1.9.2
Deprecate requiring 'RMagick' and avoid filename case conflicts
I had to change a bunch of `require`s to `require_relative` to resolve
pathing issues when running this. You can run the coverage with
`ruby test/test_all_basic.rb`, then open up `coverage/index.html`
Make extconf.rb (a little) easier to understand
PR rmagick-temp#125 added the string Magick::Magick_features

Issue #148 reported compilation failures on
  Mac OSX Yosemite
  Rails 4.2.0.beta4
  rake, version 10.4.0
  ImageMagick 6.8.9-8 Q16 x86_64 (via macports)

in rmmain.c:1724:28:
error: use of undeclared identifier 'MagickSupport'

This PR now makes the flow to use if defined:
GetMagickFeatures() or
  MagickFeatures or
  MagickSupport or
  "unknown"
Change all require RMagick to rmagick
Fix #148; More robust feature detection
This reverts commit 2b91e03, reversing
changes made to e587bdc.
Post #157, some were introduced by reverting #147
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.