Wouldn't it be nice ...

The deadline for proposing ideas for OSCON talks was Monday at 11:59 p.m. Naturally I clicked Submit for the final time not long before then, at 11:56 p.m. Note I said "final time," implying more than one—I decided to go all-out and submit three ideas, in hopes that at least one will slip through. For the curious, they are

• Design and publish beautifully and professionally: How to create a professional-looking document, using Scribus and Inkscape. Many people who read this probably don't realize I spent four years working as a page designer and copy editor at newspapers. I continue that interest today as editor of a newsletter about open-source activity for the Open Source Educational Laboratory at Oregon State.


• Community dynamics in a large open-source project: Problems, solutions and conundrums in Gentoo. The funniest (and saddest) part is when the same things begin to repeat themselves, and nobody else remembers last time it happened.


• Open-source software in the biosciences: Where we are and what we learned on the voyage: The cruel joke of scientists as programmers. As a grad student in biochemistry at OSU, I've had to deal with more ugly software than many of you can imagine. But at least more of it is open source now, right?

Anyone else submit proposals? I'd be interested to hear what they're about.

For anyone who's interested, here are the projects I've got going right now. Many of them could use some help, so take a look and let me know if you're interested in any. Roughly in my order of interest:

• Add the Sugar desktop environment for OLPC — it's in my overlay, but it segfaults on startup of sugar-emulator somewhere in sugar-shell code. Try it out and see whether you can come up with a fix.


• Port LTSP to Gentoo — pioto, straaken and perhaps another person or two are working with me on this. This involves changes to the client-building plugins, init scripts, and adding some ebuilds. Also, probably creating Seeds for the client and server.


• Get the rest of the system-config-* GUI tools from Red Hat working — some remain masked. Would appreciate testing and fixing on any that remain masked.


• Add virt-manager into the main tree from my overlay — haven't got a Xen instance to test it with yet. If anyone would like to test this and let me know, that'd be great.


• Fix our X init scripts to be more like upstream intended, then fixing upstream to be current. havner is taking the lead on this, and I look forward to seeing his work.


• Add some new science packages, including KiNG and friends from the Richardson lab, CCP4MG, CCTBX and more.


• The infamous bug #44132 — make multiple MPI implementations simultaneously installable.


• Resume my occasional series of blog posts on Gentoo in the enterprise, embedded, cluster etc environments. One post I want to make is how to use the Gentoo installer's CLI frontend to make large, automated installations easy.

And of course these are beyond the usual ongoing maintenance of X, science packages and cluster packages.

We're beginning to put an academic paper together, and of course I'd like to do this using open-source software if I can. My PI (principal investigator, the head of the lab) uses Word — so whatever ends up getting used, it needs some capability to at least export to .doc or .rtf. A critical aspect of any solid academic paper is citing your reference in a bibliography. OpenOffice.org does have some basic bibliography capabilities, but that's what they are: basic. Work is underway to fix that, but it's not expected to get anywhere for a year or so.

After some research, I've come across a few promising packages:

Zotero: A Firefox 2.0 extension, public beta started less than 2 weeks ago. No integration with word processors yet, but you can copy and paste a formatted bibliography across, and export and import the actual data. "It lives right where you do your work — in the web browser itself." As a result, adding references from online searches such as PubMed is as simple as a single click. Every other package needs explicit support added for online searches.
Bibus: Uses OpenOffice.org's Python functionality, also integrates with MS Word. The build system sucks — it should use distutils, but instead it's got some custom Makefile and weird shell scripts and configuration files. Its functionality comes highly recommended, however. Will do PubMed and eTBLAST queries.
Pybliographer: The development version (1.3) integrates into OO.o and LyX. The 1.3-series GUI is alpha-quality and just had its first release. Will do PubMed, Web of Science and CrossRef queries.
bibutils: Command-line filters to convert between a variety of formats, including EndNote (which is currently in use under MS Word). Also handles RIS and BibTex, so that provides for OO.o import and export as well.

Update: As of now, bibutils and Bibus are both available in Portage. Try 'em out.

I finished my graph figure around 6 p.m. today. Since then, I've probably spent an hour pondering it. Two weeks or more went into working on the information in and presentation of that one little figure with three graphs in it. It's really humbling when you think about how small the material is, and how much effort it took to produce. When it's printed, it will be something like 6 cm wide and 12 cm tall.

If you really have an urge to see it, drop me a line and I'll let you know once the paper's been submitted and accepted. Till then, I'm gonna keep it on the D-L. =)

Yesterday was an enlightening day. At work, I'm trying to compare two protein structures — one is a higher-resolution version of the other. I'm plotting a certain characteristic of each against the other and finding the slope of the line. Great news! There's a really cool trend between pairs of structures. My boss asked me to switch the X and Y axes around, since one typically puts the newer data on the Y, and the data to which it's being compared on the X.

So I did, and guess what? The slopes don't match! That's right, switching the axes around doesn't necessarily result in an inverse slope when you're doing a linear regression. Why, you ask? Because the technique only minimizes for the Y direction. Suddenly my really cool trend isn't a trend at all, and all the slopes are equal within error.

What we've decided to do is put the more accurate data in the X axis, to better account for the larger error on the Y axis using linear regression.

If any of you know much statistics, I'd like to hear a more accurate, better way to come up with a slope that's robust to flipping the axes (perhaps by minimizing both X and Y distances or residuals?). This method needs to already be implemented in some open-source program and fairly trivial to learn.

It's been a while since I posted. I've discovered that the more time I spend blogging, the less actual work I get done. And I don't have any bosses or employees who I need to convince I've been working.

Lots of work has gone into improving the Gentoo scientific computing experience. I've written new, more flexible eselect modules for BLAS and LAPACK, a couple of very commonly used linear algebra libraries. They're in my overlay now, and they will enter the Portage tree soon.

I'm trying to make Gentoo the distribution of choice for my day job -- protein X-ray crystallography. This means packaging all the weird Fortran stuff with even weirder, broken build systems. I'm waiting on a patch to go into Portage 2.1.1 to fix RESTRICT=stricter, then I'll be able to add the remaining apps chilling in my overlay -- primarily CCP4 and Coot. I've recently added packages for CNS, SHELX and ABINIT. Ghemical-2's ability to run GAMESS-US also helps quite a bit, and I hope to find some time to work on the WebMO package that's hard-masked in Portage.

The entrance of GCC 4.1 into testing has made a lot possible, but also requires a lot of fixing. This is particularly the case in scientific computing, where there's an entirely new Fortran compiler -- gfortran. Please note that gfortran is NOT a renamed g77. It accepts different options, and it compiles different code. Only gfortran will compile Fortran90 and Fortran95 code. Only g77 will accept common options such as -fno-globals. Do not assume they are the same.

I've also gotten back into trying to do some work on the cluster project, prompted by a couple of active cluster builders in the #gentoo-cluster IRC channel. I hope to be able to set up a test cluster this summer and rework the high-performance computing documentation. It's currently a bit of a hack, outdated and incomplete but better than nothing. I'd like to create a clean cluster setup using diskless nodes with UnionFS, PXELinux, etc. Also on this list is getting OpenMPI into the tree and creating some sort of 'eselect mpi' module that will allow not just system-level switching but also changes on a per-user level.

Also, X.Org 7.1 has entered testing a bit under a month ago. Look to see this stable well before the 2006.1 release, which is in August. Now that Portage 2.1 is stable, the last major blocker appears to be gone. Perhaps we can get it stabilized around the beginning of July. That would be a nice Independence Day present.