On Fri, 2007-08-10 at 14:43 -0400, Gail Banaszkiewicz wrote:

Christopher Brown wrote:

...

You'll have to install ImageMagick with Magick++ support (which is default) in order for these new extensions to work.

Guess that means I'd have to install ImageMagick just to compile too, eh? Just updated and couldn't compile because of the ImageMagick stuff. Is there some option I can use to turn this off? I don't really want to bother installing it right now. Thanks!

This is on Windows? I tested to ensure you don't need it on Linux...

--Ted

I don't see how the windows build tool (btool) references the configure.acfile at all. I'll keep looking into it.

On 8/10/07, Gail Banaszkiewicz <gbanaszk@...1686...> wrote:

Ted Gould wrote:

On Fri, 2007-08-10 at 14:43 -0400, Gail Banaszkiewicz wrote:

Guess that means I'd have to install ImageMagick just to compile too, eh? Just updated and couldn't compile because of the ImageMagick stuff. Is there some option I can use to turn this off? I don't really want to bother installing it right now. Thanks!

This is on Windows? I tested to ensure you don't need it on Linux...

Yes, this is on Windows. I rolled back for now.

On 8/11/07, Bob Jamison <rwjj@...127...> wrote:

Sounds great. Sorry I missed you when you were asking about Win32 and the Buildtool. So I tried building it and have it mostly done now.

From a portability side, I have found a couple of thing that need to be tweaked.

In imagemagick.cpp, the first include should be #include <libintl.h> This will work on both Linux and windows and will provide the definition for libintl_printf() that Windows needs.

strndup, stpcpy and strnlen are Linux-only

strndup can be replaced with g_strndup()

For stpcpy() , which returns the tail of the destination string, I would either replace it with some pointer-walking code, like below, or preferably use a Glib::ustring as an output buffer and append the data to it. Slightly slower, but a LOT safer.

I hope Thunderbird doesn't screw up the formatting too much :)

int formatted_len = raw_len * 78 / 76 + 100; char *formatted = new char[formatted_len]; char *formatted_i = formatted; const char *src;

for (src = "data:image/" ; *src ; ) *formatted_i++ = *src++; for (src = effectedImage.magick().c_str() ; *src ; ) *formatted_i++ = *src++; for (src = ";base64, \n" ; *src ; ) *formatted_i++ = *src++;

int col = 0; while (*raw_i) { *formatted_i++ = *raw_i++; if (col++ > 76) { *formatted_i++ = '\n'; col = 0; } }

if (col) *formatted_i++ = '\n';

*formatted_i = '\0';

======================= /* Being lazy, I prefer this */ Glib::ustring buf = "data:image/"; buf.append(effectedImage.magick()); buf.append(";base64, \n"); int col = 0; while (*raw_i) { buf.push_back(*raw_i++); if (col++ > 76) { buf.push_back('n'); col = 0; } } if (col) buf.push_back('\n');

const char *formatted = (const char *)buf.c_str();

======================

k, did the above, replacing the strndup with the g_ version, and using the /* being lazy */ code for the stpcpy and it compiles ok, but I'm not getting any raster extensions showing on the effects menu here. Is it checking for IM somewhere? I'm gonna guess that the test isnt working right on windows... Let me know and I'll help you try get this working.

Cheers

Sim

On 8/13/07, Ted Gould <ted@...11...> wrote:

On Sun, 2007-08-12 at 20:55 +0100, john cliff wrote:

...

k, did the above, replacing the strndup with the g_ version, and using the /* being lazy */ code for the stpcpy and it compiles ok, but I'm not getting any raster extensions showing on the effects menu here. Is it checking for IM somewhere? I'm gonna guess that the test isnt working right on windows... Let me know and I'll help you try get this working.

Yeah, there will need to be a "#define WITH_IMAGE_MAGICK 1" somewhere. In the autotools environment it's in config.h if ImageMagick is detected. I'm not sure where this check exists in the custom Windows build tool, sorry, that's all I know.

            --Ted

I'd imagine that wants to go in build.xml or somwhere, but for now I just deleted the checks in init.cpp, thing is I now get a bunch of link errors similar to these:

Make error line 348: LINK problem: build\libinkscape.a(adaptiveThreshold.o):adap tiveThreshold.cpp:(.text+0x124): undefined reference to `Magick::Image::adaptive Threshold(unsigned int, unsigned int, unsigned int)' build\libinkscape.a(addNoise.o):addNoise.cpp:(.text+0x117): undefined reference to `Magick::Image::addNoise(MagickLib::NoiseType)' build\libinkscape.a(blur.o):blur.cpp:(.text+0x11e): undefined reference to `Magi ck::Image::blur(double, double)' build\libinkscape.a(channel.o):channel.cpp:(.text+0x11f): undefined reference to `Magick::Image::channel(MagickLib::ChannelType)' build\libinkscape.a(charcoal.o):charcoal.cpp:(.text+0x11e): undefined reference to `Magick::Image::charcoal(double, double)' build\libinkscape.a(colorize.o):colorize.cpp:(.text+0x21d): undefined reference to `Magick::ColorRGB::ColorRGB(double, double, double)' build\libinkscape.a(colorize.o):colorize.cpp:(.text+0x240): undefined reference to `Magick::Image::colorize(unsigned int, Magick::Color const&)' build\libinkscape.a(colorize.o):colorize.cpp:(.text+0x252): undefined reference to `Magick::ColorRGB::~ColorRGB()' build\libinkscape.a(colorize.o):colorize.cpp:(.text+0x283): undefined reference to `Magick::ColorRGB::~ColorRGB()' build\libinkscape.a(contrast.o):contrast.cpp:(.text+0x10a): undefined reference to `Magick::Image::contrast(unsigned int)' build\libinkscape.a(convolve.o):convolve.cpp:(.text+0x10e): undefined reference to `Magick::Image::convolve(unsigned int, double const*)'

Any ideas anyone?

Cheers

Sim

On Mon, 2007-08-13 at 08:07 -0500, Bob Jamison wrote:

2. Add the /bitmap directory to the <excludes> list for the cc task.

But, yes, if the rest of Inkscape has a code hook to the /bitmap directory, then somewhere you will need to have a switch like that.

I think this makes sense for now. The init.cpp file won't call the stuff in the bitmap directory unless WITH_IMAGE_MAGICK is set.

--Ted

So does all the stuff below mean that the Win32 build should be working again? It's still not working for me.

Make error line 179: problem compiling: In file included from src/extension/internal/bitmap/adaptiveThreshold.h:9, from src/extension/internal/bitmap/adaptiveThreshold.cpp:12: src/extension/internal/bitmap/imagemagick.h:14:22: error: Magick++.h: No such file or directory In file included from src/extension/internal/bitmap/adaptiveThreshold.h:9, from src/extension/internal/bitmap/adaptiveThreshold.cpp:12: src/extension/internal/bitmap/imagemagick.h:28: error: 'Magick' has not been declared src/extension/internal/bitmap/imagemagick.h:28: error: ISO C++ forbids declaration of 'Image' with no type src/extension/internal/bitmap/imagemagick.h:28: error: expected ';' before '*' token src/extension/internal/bitmap/imagemagick.h:33: error: 'Magick' has not been declared src/extension/internal/bitmap/imagemagick.h:33: error: expected ',' or '...' before '*' token src/extension/internal/bitmap/imagemagick.h:40: error: 'Magick' has not been declared src/extension/internal/bitmap/imagemagick.h:40: error: expected ',' or '...' before '*' token In file included from src/extension/internal/bitmap/adaptiveThreshold.cpp:12: src/extension/internal/bitmap/adaptiveThreshold.h:22: error: 'Magick' has not been declared src/extension/internal/bitmap/adaptiveThreshold.h:22: error: expected ',' or '...' before '*' token src/extension/internal/bitmap/adaptiveThreshold.cpp:20: error: variable or field 'applyEffect' declared void src/extension/internal/bitmap/adaptiveThreshold.cpp:20: error: 'Magick' has not been declared src/extension/internal/bitmap/adaptiveThreshold.cpp:20: error: 'image' was not declared in this scope

--bb

On 8/12/07, Bob Jamison <rwjj@...127...> wrote:

Sounds great. Sorry I missed you when you were asking about Win32 and the Buildtool. So I tried building it and have it mostly done now.

From a portability side, I have found a couple of thing that need to be tweaked.

In imagemagick.cpp, the first include should be #include <libintl.h> This will work on both Linux and windows and will provide the definition for libintl_printf() that Windows needs.

strndup, stpcpy and strnlen are Linux-only

strndup can be replaced with g_strndup()

For stpcpy() , which returns the tail of the destination string, I would either replace it with some pointer-walking code, like below, or preferably use a Glib::ustring as an output buffer and append the data to it. Slightly slower, but a LOT safer.

I hope Thunderbird doesn't screw up the formatting too much :)

int formatted_len = raw_len * 78 / 76 + 100; char *formatted = new char[formatted_len]; char *formatted_i = formatted; const char *src;

for (src = "data:image/" ; *src ; ) *formatted_i++ = *src++; for (src = effectedImage.magick().c_str() ; *src ; ) *formatted_i++ = *src++; for (src = ";base64, \n" ; *src ; ) *formatted_i++ = *src++;

int col = 0; while (*raw_i) { *formatted_i++ = *raw_i++; if (col++ > 76) { *formatted_i++ = '\n'; col = 0; } }

if (col) *formatted_i++ = '\n';

*formatted_i = '\0';

======================= /* Being lazy, I prefer this */ Glib::ustring buf = "data:image/"; buf.append(effectedImage.magick()); buf.append(";base64, \n"); int col = 0; while (*raw_i) { buf.push_back(*raw_i++); if (col++ > 76) { buf.push_back('n'); col = 0; } } if (col) buf.push_back('\n');

const char *formatted = (const char *)buf.c_str();

======================

Christopher Brown wrote:

...

Dear fellow Inkscape users,

My Google Summer of Code project is now in SVN. My project is "Raster Functionality in Inkscape," which involved integrating ImageMagick libraries into Inkscape, in order to manipulate bitmaps from inside Inkscape.

You'll have to install ImageMagick with Magick++ support (which is default) in order for these new extensions to work.

They can be accessed under the sub-menu "Raster" in the extensions drop-down menu.

The code is not quite in its final state--I am still perfecting and smoothing it out--but it's functional and I'd love to get feedback from all of you, especially from Windows users (because I have not tested it on that OS yet).

Thank you all,

-Christopher Brown

---

This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel

On Sun, 2007-08-12 at 02:52 -0300, bulia byak wrote:

• Is it possible to disable the effects unless a bitmap is selected?

Ted, I think you planned to work on something like that for verbs - what is the status of that?

The biggest is to change SPAction to handle the expose event, and then use it for something useful. But, I started working on other things because I thought everything was going to be changed to GTK Actions... which doesn't seem to have panned out. So, I'll re-add it to my TODO list. It'd be nice for all the effects -- especially those that require paths too.

--Ted

On 8/25/07, Ted Gould wrote:

While some of these things can be done with SVG filters, that isn't the goal here. It is to adjust the bitmaps. This is useful for lots of things, but one of them off the top of my head is use an SVG file on a renderer that doesn't support filters. The goal is to provide basic bitmap editing functionality so that you don't have to take your bitmaps into GIMP.

Probably I didn't express clearly what I was talking about :)

What I was going to say is that with current approach it's not possible to adjust bitmaps in a non-destructive way like e.g. Scribus does it:

http://img29.picoodle.com/img/img29/9/8/25/f_ScreenshotIm_1755ef2.png

This is not about competing with GIMP or Krita, this is about not breaking creative process in the middle :)

Alexandre