Exempi 1.99.1
I forgot, but Thursday night I released Exempi 1.99.1. The ChangeLog is probably not very obvious be the 2 notable changes are xmp_register_namespace() and NS_CC. The first is to register a new namespace to add to a metadata block. The second is actually the constant for the Creative Commons XMP namespace.
This release is brought to you especially for our friend at Creative Commons.
PresentationML/DrawingML text import
One of the good things with being employed to work on a Free Software project like OpenOffice.org is that you can blog about it, since the code is publicly available. My day to day work is current PowerPoint 2007 import for Impress. This is a joint work between Sun and Novell.
A few weeks ago I implemented the text import from PresentationML (actually from DrawingML). It is far from perfect, but it is getting there. See the example:
It imports the text, the characters and paragraph properties, etc.
... solved
I went ahead and solved the problem on my own with planet OpenOffice.org: I changed the RSS template files to escape their content. I don't know why the behaviour is different between two Python installation, and actually I don't wanna know ;-)
It looks like grep is still one of my best friends.
People aggregating the planet should no longer have issue. Sorry for the mess.
Planet issues...
Dear Lazyweb,
Since I moved planet OpenOffice, it generates invalid RSS feed because it does not escape the content of <dc:author>. The planet version is the same, but the Python version likely is different. I have no power on the version of Python that is installed.
This invalid feed cause problems for people aggregating planet in a feed reader (like I do).
Oh Lazyweb, do you have a magical solution? Your friend Google seems to not have been helpful.
Thanks, love,
Hub
Planet OpenOffice.org and go-oo.org service announcement
I just migrated Planet OpenOffice.org to a new ISP. Nothing should be changed for you readers, but if you notice any problem, feel free to contact me.
We will also be moving the rest of go-oo.org progressively, there might be disturbances. We will try to minimize them.
Thanks.
Lossless JPEG
Last night I reached a new milestone in libopenraw: being able to decompress the RAW data from CR2 files using Lossless JPEG. The lazyweb has been helpful in finding code for it (libjpeg does not support it) code that I ported from old K&R to C++. I'm still not completely statisfied with the port but it works much better than my attempt to port some MIT-licensed Java code.
Above is the PGM I generated, that contain the RAW data, viewed in eog. The look is due to the fact that we have a GRBG 2x2 color filter array diplayed as a graymap, and that eog does sone dithering. This is the data I should be passing to the demosaicing code. The picture is one of the random sample I took from my own photo library, taken around the corner from where I was living in Montréal.
DNG support should come The code is in the git repository.
Eight years ago
8 years ago, back when AbiWord was a commercial venture, the AbiWord folks from Sourcegear had a booth at Linux Expo 99.
Abi the ant was there. Marc dug up pictures that were still on the server. One did catch our eyes (along with the caption):
Miguel and his partners are starting a new company, International GNOME Support
Exempi 1.99.0
I think that a release of Exempi 1.99.0 is due.
This release is the first release towards Exempi 2.0. It is based on Adobe own XMP SDK and unlike the original Adobe code, it builds fine on Linux, and even more on x86_64 (I'm not proud of the fix, but it is as good as it gets for now). This version is much more feature complete than Exempi 0.5.1 as it read and writes (serialize) XMP metadata and load from and to files, including, for supported file types, reconciling the existing metadata.
Download: .tar.gz - GPG signature
It is licensed under the BSD license.
The current API is somewhat not definitive, but I hope it is OK. The documentation still has to be written / generated.
One of the question I got asked is how do you add XMP metadata like CC:Licensing to a PDF[1]. Here is a snippet:
#include <exempi/xmp.h>
...
XmpFilePtr f;
f = xmp_files_open_new("test.pdf", XMP_OPEN_FORUPDATE);
XmpPtr xmp = xmp_files_get_new_xmp(f);
xmp_set_property(xmp, NS_XAP_RIGHTS, "Copyright", "(c) ACME Inc., some rights reserved"
" - This work is licensed to the public under the Creative Commons Attribution-ShareAlike "
"license http://creativecommons.org/licenses/by-sa/2.0/");
xmp_files_put_xmp(f, xmp);
xmp_free(xmp);
xmp_files_close(f, XMP_CLOSE_SAFEUPDATE);
Now a little call: Exempi 1.99.0 supports several media format, but not vorbis... I wish I could add support for it. Any idea? I haven't actually dug into it yet.
[1] could be any of the other supported file types
exiv2 and Python
I discovered today pyexiv2, which is a Python interface to the infamous Exiv2, a GPL library to read and write Exif and IPTC/IIM metadata from JPEG and various camera RAW file. Too bad I can't reuse the exiv2 code in libopenraw as it is GPL and not LGPL.[1]
pyexiv2 is actually an exemple on how easy it is to write Python module in C++ using Boost.Python
[1] The decision to use LGPL for libopenraw is in complete relation to the policies from both Gnome and KDE when it come to libraries, and nothing else.
Last week-end (or was it the one before ?)
Last week-end (or maybe the one before, I keep delaying this post), for Victoria day, we took the train to Montréal. It was actually a meetup of Range Finder Forums members.
Without order, occurred during the week-end:
I really miss living in that city. I felt that last time when attending LGM, this time just confirmed it. I miss walking down Sainte-Catherine with people in the street, in the evening or on week-ends, I miss the Metro.
Let's talk about changes. When I left they were starting to tear down the building that was hiding Saint-James-United church. Now the front of the church is free, visible. Here is what that give at dusk, from the street:
taken with an EF 50mm f1.4 USM, 1/160 f/2.5 @1600 ISO
Before that, I could only photograph this from the inside, like this, with very little distance:
taken with an EF 24-85mm f/3.5-4.5 USM, 1/125 f/3.5 @400 ISO
I'll go further in depth for the pictures, but the workflow for digital photography on Linux is still pretty steep...
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!