Recently, I've taken a position as a network administrator for a local tech company. Although this isn't directly related to science, there are some really interesting insights that have come out of it. Some background: the internal net uses Linux to provide internal services to a Win32-based network. Although I tend to spend at most 10-20% of work time on the servers, I find it very ironic that supporting the workstations that tends to consume most work time. The issues tend to revolve around three general themes: user issues ('my foo won't work' or 'the network doesn't work'), standards issues (getting different-language versions of the same software to work properly together - accented text and other issues relating to i18n text, or my favourite, getting different-language versions of Outlook to play nicely together), and finally machine issues, ether updates or actual workstation rebuilding (which I do everytime a machine is alloted to a different user).

Interestingly, looking at this in lab or departmental context, the open source approach allows quite a bit of flexibility to resolve these three general issues. First, although user education is particularly important, few institutions can commit the resources to ensure that new users obtain the information they require. That said, putting relevant information online locally (perhaps as web pages) is a great way to habituate users to searching for info when they have questions. The irony I find in many academic and research organizations is the realization that although one of our greatest tasks is to contribute to the written knowledge base (academic journals and the like), much of our day to day knowledge is passed on in the oral tradition. This has the unfortunate tendency of allowing labs to lose significant practical know-how when post-docs leave or when students graduate.

I'm curious to know how others are dealing with this problem, and as this volume of information grows over time, how the info is maintained to ensure its utility.

I should also mention that in addition to information pertaining to scientific software, I would be very interested to learn about projects, papers or other documents which have relied on Linux as the primary research platform

Here are some more links and updates for some scientific software that we've found...

Chemsuite - http://chemsuite.seul.org/
Richardo Stefani recently announced an update to his Chemsuite application to the SEUL/sci list.

Hello Folks,

I updated Chemsuite home page. The new version is 0.0.39 and it's still unstable. Now I'll try to put new versions weekly or biweekly on the site. It now compiles with wxGTK and GTK+ The chem2D drawer is unstable and dump some core. The molecular weight calculator is ready.

Grace - http://plasma-gate.weizmann.ac.il/Grace/
One of the most important tasks for scientists is to create clear, understandable graphs to explain or present results. These programs tend to be rather expensive. Grace is a very flexible WYSIWYG graphing program. Among other features, it offers graph exports into numerous high-resolution graph formats (ps, eps, pdf, mif, svg, pnm, png, jpg) and excellent control over graph axes and other graph elements. The program can be called interactively via GUI as well as via script. A user's guide also is available. The latest release of Grace is 5.1.2. (GPL)

SciGraphica - http://scigraphica.sourceforge.net/
Another WYSIWYG graphing program, SciGraphica purports to be a Microcal Origin clone and my initial exposure to it thus far reminds me of SigmaPlot. I'm still getting up to speed with it, but it seems that between SciGraphica and Grace, Linux users have access to two great graphing packages that will make them the envy of their Windows and Mac colleagues. Latest release 0.6.1. (GPL)

FREEGIS - http://freegis.org/
Last but far from least, I recieved an email from Bernhard Reiter about the Free GIS portal...

You might consider linking to the freegis.org project, which basically tries to be a portal for free software (as in the FSF sence) and Geographic Information Systems (GIS).

Though not necessarily limited to free software on GNU/Linux, a lot software runs on GNU/Linux. This is also the operating system of our choice.

Kind Regards,
Bernhard

-- Pete St. Onge (pete@seul.org)