Category: Open Source

FluTE makefile for wxDev-C++ (Windows)

FluTE is an influenza epidemic simulation model written by Dennis L. Chao at CSQUID. It works out-of-the box on GNU/Linux (just type make and run it).

I wanted to see how it works. But since I’m temporarily stuck with a Windows laptop, I downloaded a free C++ compiler for Windows (wxDev-C++), imported all the files in a project and compiled. For those who want to try, here is the project file and the specific makefile in a zip file (2 kb). Just decompress the FluTE archive (I used version 1.15), copy the two files from the zip file above and launch the IDE. In the project options (Alt+P), specify the custom makefile (in the "Makefile" tab) as the one from the zip file above. Compile (Ctrl+F9). Done.

On my Intel Core2 Duo T5450 (2Gb RAM), it took 6 minutes to simulate the "two-dose" example.

Please note that I didn’t try to compile with OpenMPI. Maybe for next time.

Waiting for PDF comments in Evince

Evince defines itself as “simply a document viewer” (for Linux/Gnome and now for Windows too). However it can already read a lot of formats: PDF, TIFF, PS, DVI, DJvu and plans to support a lot more in the future.

But for me there is one important feature missing: the ability to read comments in PDF files. I sent PDF versions of draft documents to my PhD thesis promoters and they send them back with their comments. Open them in Evince: you’ll only get the balloons but no possibility to click on them (see Figure 1 below). Open them in Acrobat Reader and not only you can see that there are comments but you can also see their content (see Figure 2 below).

Reading a PDF with comments in Evince
Figure 1: reading a PDF with comments in Evince

Reading a PDF with comments in Acrobat Reader
Figure 2: reading the same PDF with comments in Adobe Acrobat

It’s in the roadmap and Carlos García Campos already has an unstable release that includes annotations. So I’ll keep Acrobat Reader for the moment. As soon as Fedora packages Evince with annotations, I’ll not see any reasons to keep Acrobat Reader on my laptop 🙂

Btw, both Okular and KPDF also seem to miss this feature.

postr, simply puts your pictures on Flickr

I really like gthumb to have a look at my photos, quickly perform some basic modifications or effects and display all the photos to people around me. But there is one thing that is annoying me: it seems impossible for my gthumb version (2.10.11) to upload to Flickr, where I put some of my pictures. There is an “issue 73” in the GNOME’s GHOP Contest page from 2007 and the development seems to be done ; it’s just not yet in the main branch.

Now enters Ross Burton’s postr, a Flickr uploading tool for the GNOME desktop. It’s simple. It has all the functionalities you need when posting pictures: title, description, tag, sets, groups (and privacy) fields and settings. Voilà! And if you need more advanced features (like geotagging – which isn’t anyway in the Flash-based Flickr upload tool), the Organise tool from Flickr is still there.

Don’t be afraid by the fact it’s still in version 0.12.3 nor by the fact it wasn’t updated since December 2008: it’s fully working, already in your favourite Linux distribution (at least in Fedora), I adopt it 🙂

Postr 0.12.3 screenshot
postr screenshot. The picture that will be uploaded is here.

WordPress problem with permalinks after upgrade

If you get error 404 with your permalinks and RSS feed after an upgrade of your WordPress installation to version 2.8.3, it’s worth to check the “Permalinks” section (under the “Settings” tab in the admin panel). Try set it up to “Common”, save changes and then put it back to your previous structure (“Month and name” in my case). This should solve most of current 404 errors after upgrade. If not, check the WordPress support forum.

Installing Pwytter on Fedora 11

This morning, it was impossible to post tweets on Twitter so I finally gave in to install a Twitter client. Amongst many software available, Pwytter seemed interesting to try: free software, written in Python were my two criteria.

Unfortunately, the installation process is not straightforward (although its use of the general python setup procedure). Here is how to install it on Fedora 11 from the command line:

  1. Download Pwytter, unzip it, enter directory pwytter-0.8
  2. Install ImageTK: as root, type “yum install python-imaging-tk” (in Fedora, ImageTK was renamed python-imaging-tk)
  4. Install simplejson: as root, type “yum install python-simplejson“
  5. (optional) launch: python setup.py build
  6. launch: python setup.py install
  7. Complete the installation by copying some files with the 4 lines below (type them as root too) (a comment in pwytter blog helps to solve the pwCache installation bug):
cp pwCache.py /usr/lib/python2.6/site-packages/pwytter-0.8-py2.6.egg
cp pwCache.pyc /usr/lib/python2.6/site-packages/pwytter-0.8-py2.6.egg
cp -r media/ /usr/lib/python2.6/site-packages/pwytter-0.8-py2.6.egg
cp -r theme/ /usr/lib/python2.6/site-packages/pwytter-0.8-py2.6.egg

Now you can launch pwytter from any user! In addition, since the source code is available and Identi.ca supports a Twitter-compatible API, let’s see if it’s easy to modify pwytter for Identi.ca 🙂

pwytter screenshot

A new home for IPGphor2reader

IPGphor2reader is a software meant to parse log (text) files resulting from an experiment with the IPGPhor and to plot graphs. I previously hosted it on my personal website and just moved it to Sourceforge, here. Amongst the various reasons for this move, I wanted the possibility for anyone to participate in the project and no hassle to manage this.

Slowly, slowly, most software on my website will be hosted on Sourceforge or Bioinformatics.net.

P.S. Obviously I chose the time where they are in the middle of a large scale site changes and upgrades so nothing is available for now (except the screenshot).

About file formats accepted by BioMed Central

BioMed Central is one of the main Open Access publishers in the world of Science, Technology and Medicine. On a side note, that’s where I published my two articles (in Proteome Science and the Journal of Circadian Rhythms). One might think that, given their support to Open Access, they would also support Open Source software and Open Format documents.

For the software side, it’s not very clear. Although they ask authors to consider releasing software described in publications under a free (or at least open source) license, they also support and advertise for a bunch of proprietary software. While it’s not a bad thing per se (it enlarges the number of potential authors), it’s sad to see they don’t cite popular free software like OpenOffice.org (to write your article), Gimp (to edit your figures) or Zotero (for reference management). These are the three main software in each category but the free software world has many more of them!

I decided to write this post because I recently received an e-mail from BioMed Central stating that BMC Bioinformatics, one of their flagship publications, accepts a variety of different file formats in the submission process. This was already true when I submitted my articles. I wanted to know how they improved their submission process in this respect and if they now added open document formats (in a broad acceptance, not only the OpenDocument format somehow linked with OpenOffice.org).

E-mail from BMC Bioinformatics with file formats accepted for submission

My first comment is that the list of accepted file formats usually applies to all BioMed Central journals, not just BMC Bioinformatics, since they share the same publication platform. In the Instructions for Authors, the following file formats are accepted: Word, RTF and LaTeX (with the BMC template) for text, EPS, PDF, TIFF as well as PNG, Word (sic), PowerPoint (re-sic), JPEG and BMP for figures. In addition, they list CDX and TGF to represent chemical molecules. How disappointed am I!

I’m disappointed because some interesting open formats have been left out. And I can’t find interesting links stating that BioMed Central will support them soon.

With some stating that OpenOffice secured more than 15% of the business office suite market as of 2004 and despite an ISO standardisation (ISO/IEC 26300:2006), the OpenDocument formats are still absent. Many young scientists now use OpenOffice.org because it’s free (mainly free like in free beer, though), because labs can’t afford MS-Office licenses prices, even educational ones but also because it allow them to do everything they want. I agree that you can easily convert your ODF, ODS or ODP documents into their respective proprietary DOC, XLS and PPT. But it would have been nice to directly have the ODx documents. On the technical side, ODx documents are “just” XML files: tools exist to automatically parse them and transform them in the journal final format (I didn’t write it’s easy but it should be more easy than reverse-engineering closed, proprietary file formats).

I’m also disappointed because although the PowerPoint format if there, SVG is not. I guess it’s just because they only use bitmap versions of the PowerPoint files. All vector graphic editors supporting SVG (and all of them support SVG: Adobe Illustrator, Inkscape, Dia, …) have conversion functions to bitmap equivalent of your drawings. So it may have little impact. But it would have been better if BMC support for SVG was direct.

In conclusion, I’m hoping the extraordinary work done by BioMed Central in the publication area will extend to the formats they accept for submission. A partial example could come from PLoS submission guidelines (here for PLoS Computational Biology, especially for figures) where they explain a lot of technical as well as license aspects and cite free software as reference.

Fedora 11 is out (since a week or so)

This week, I nearly emptied my internet quota by downloading and seeding the new Fedora Core 11. For those who don’t know yet, Fedora is a Linux-based operating system that showcases the latest in free and open source software. What I particularly like in this GNU/Linux distribution is that its developers prefer to make changes to the original software instead of applying fixes specifically for Fedora ; in this way, all the other distributions may also take advantages of the improved software. You’ll find many reviews and “tours” on the web about this new Fedora. In the next paragraphs, I’ll just highlight some of the most interesting points I saw until now. Coming from a Fedora 9 also helps to pinpoint the major improvements (mainly from a user point-of-view).

Fedora 11 screenshot

The first nice thing I noticed was the reduced time to boot my ageing laptop (ok, 2 years is not so old) and its simplicity (just a progress bar in the bottom of the screen). I won’t give you numbers. Anyway, numbers would have been specific to my computer. But I can tell you it’s quite faster than with the Fedora Core 9. And anyway faster than Vista. In general, I think it’s a big improvement for the user experience: after all, most users just boot their machine to use it as a tool, not to wait with amazement that everything is loaded.

Another nice thing is to finally get new versions of some software. OpenOffice is in version 3.1, Netbeans is in version 6.5, Python 2.6, … One thing I was waiting since a long time is Istanbul, a screen recorder. Otherwise, all the other main software I use are also present: Gimp, Inkscape, vim, gthumb, xmms, etc. The only thing a little bit tricky was to add the ability to read two proprietary formats: mp3 and flash. For mp3, you have to add the fusion nonfree repository and it’s clearly explained here. For Flash, I was surprised that Adobe offers an installation procedure just identical to any other software on Fedora, based on yum. I was also pleased to see projects like bioperl and biopython already available at the installation step (along with XDrawChem and JMol which will allow me to read old molecules I drew a few years ago with proprietary software like Chemdraw). One last thing a was eager to see in action is the ext4 filesystem (Fedora being the first distribution to include it by default). To be honest, I don’t see the difference.

XDrawChem and JMol in Fedora 11

Finally, if I have to summarize the experience with this Fedora Core 11 so far, it will be: smooth installation, up-to-date software, no surprise, ready to work after 1 day (I had to play with nearly all the new toys inside). I think GNU/Linux is really ready for the desktop 🙂

Implication of Oracle buying Sun on Open Source projects?

Oracle and Sun announced a few days ago that Oracle will buy Sun. Others are more apt than me to comment on the financial and strategic impacts of this move (for example, in the Guardian, the New York Times, the Wall Street Journal or on Slashdot). I’m more interested in the potential implications this move could have on some Open Source projects which were backed by Sun. I indeed believe Oracle will continue the development of his contributions to Open Source software, whether they are notable (Btrfs or Oracle Enterprise Linux) or less visible.

In the last few years, Sun opened or started to open some of its (key) software like OpenOffice.org, Netbeans, OpenSolaris, Java, … Sometimes these moves were considered as a last hope to see them used (and developed) at a lower cost for Sun. Very often, these moves were criticised because the “opening” was only partial (non-free licenses, stranglehold on the development processes, …) or just announced (Java still needs to be fully opened). However, the openings of OpenOffice.org and Netbeans can be seen as successes: OpenOffice.org is a more and more used office suite and Netbeans fairly competes with another open development source-editor-cum-development-platform, Eclipse. In the beginning of 2008, Sun acquired MySQL AB, the company behind the probably most used database system for website development, MySQL. Unfortunately, rumors spread that Sun will close some of the MySQL features, leading to forks like Maria(DB) (rumors where later dismissed). Anyway, these software are (nearly) free. But they may not be in Oracle strategic plannings.

Oracle now owns 2 database management systems: Oracle and MySQL. Although they maybe do not compete at the same level and although I don’t see Oracle dumping one RDBMS (because of their respective user base), it could become expensive to maintain 2 code bases for the same goal.

Oracle now owns 2 operating systems too: Oracle Enterprise Linux and (Open)Solaris. And here, they compete at the same level: on enterprise desktops and servers. The beauty of Open Source is that OpenSolaris may survive thanks to its community if it would be abandoned by Oracle.

Oracle has now the lead on the development of an IDE, Netbeans, while it extensively uses and promotes its rival, Eclipse. Fortunately for Netbeans, it has a strong community behind … I guess it’s approximately the same for Sun virtualisation software, VirtualBox (no immediate use for Oracle) but I’m not really following these technologies so I won’t bet anything on this.

Oracle now also has the lead on the development of Java, a programming language cherished by a lot of companies around the world (some say Java is the COBOL of the 1990s …). Oracle also uses Java for its tools so I guess Oracle will continue its development. Whether the opening of Java will continue and if it does, at what speed, one can assume it will depend on the financial and/or fame benefits Oracle can gain from it.

Oracle owns now an office suite. I don’t really see how it fits into Oracle software portfolio unless Oracle really pushes hard its adoption in companies where Microsoft Office has a monopoly. Or Oracle intends to beat Microsoft by offering a complete solution, from corporate servers (with Oracle DB, Enterprise Linux, BEA/Tomcat application servers and Sun hardware) to corporate desktops (with OpenSolaris (?) and OpenOffice.org), Oracle’s CEO Larry Ellison being known to forecast the end of Microsoft. By providing top-to-toe-solutions, this would make Oracle the next IBM but this is another subject.

So, except for Java (and maybe OpenOffice.org), I’m rather pessimistic on the future of these Open Source / free software projects. Does this mean that they will not survive? I don’t think so. They users/fans base is sometimes huge. And similar high-quality Open/Free projects live very well without one big corporation behind them ; think of PostgreSQL, Linux, Eclipse, Python/Ruby, etc.

Ryan Paul wrote an article in ArsTechnica on the same topic, for those who are interested.