WebJunction

A number of themes have emerged in this week's discussions that lay some good groundwork for libraries ready to embark on using Open Source software. One is that OSS provides greater control and flexibility than commercial software. If you are unhappy with one or more areas of the library's computing landscape, then the added control and flexibility that comes with OSS might make these good starting points for an Open Source evaluation. We have talked about Eric Raymond's notion of a developer's itch, and that this can result in a good application because the developer can directly identify with the benefits. The same approach can be used in selecting potential OSS projects. Identifying applications, or functions of an application, which drive you crazy, may provide fertile ground for planting the seeds of OSS.

John has provided a link to the [url http://europa.eu.int/ISPO/ida/jsps/index.jsp?fuseAction=showDocument&parent=news&documentID=1647 ]IDA Open Source Migration Guidelines[/url] and it cautions against what it calls Big Bang Migrations, where everything is changed at once. The guidelines provide a number of scenarios for introducing OSS in conjunction with other systems, such as adding Linux servers to an existing NT domain. Starting with a well defined project, such as whether Open Office could be used to replace Microsoft Office, or whether Linux should power the library's web server, is a good way to gauge how well OSS can fit into the library's day-to-day operations.

OSS is not just about software. See [url http://www.wired.com/wired/archive/11.11/opensource_pr.html ]this article in Wired[/url] about the use of OSS to foster mass innovation in other areas, such as [url http://www.wikipedia.org/ ]Wikipedia[/url]. OSS participation is much like barn building, everyone can have a role in its success. I suspect that libraries are rarely vehemently opposed to OSS, but the greater challenge is probably overcoming inertia and finding an easy onramp to the Open Source world. The good news is that the communities and developers behind OSS projects are generally accessible and purposely want to be close to their real and potential users. For those not using Open Source software in their libraries beyond some default underlying OSS-based Internet layers, what would help make OSS a real option in more areas of your library? What OSS options would you like to see that don't exist now?

Whoops, our firewall hiccupped during my post and I didn't see the reply about other OSS filters. It sounds like there are great options out there. With CIPA, is there some sort of language where you can open up content if you present some sort of disclaimer or get some indication of acceptance of non-liability from the patron? That would be the only technical advantage of having something on the desktop but maybe it can already be done with the squid-based tools.

I don't think that there are concerns with squidGuard and DansGuardian in terms of CIPA compliance, but with ease of use. They're not exactly WYSIWYG applications, and I imagine that their installation and management would be quite overwhelming for the average librarian--not to disparage librarians at all, mind you, but most librarians don't have the time it takes to learn all that's necessary in order to do so, especially in small libraries where there are only one or two librarians on staff.

I am shooting at this blindly but I think Squid or Squidproxy server can be managed via a GUI, web-based interface called
Webmin

[url http://www.webmin.com/standard.html ]http://www.webmin.com/standard.html[/url]

Webmin can be installed during the installation of Mandrake 9.x
and I use it in conjunction with the standard config files.

As an aside, I really think that some librarians (I am one) should consider training themselves in this area.

Webmin allows one to make changes in many of the text files used to configure Apache, Squid, Samba and many other apps.

Many linux/BSD heads do not use any type of GUI interface because it is an inefficient use of processor power. Things like Webmin, KDE, Gnome, and Xandros are "plug ins" to sit upon Linux.

I guess to answer your question. Webmin is available if you want to use it and install it. The issue is choice and empowering the end user to make the decision to use or not use any aspect of Linux. But most cool apps are available with the Linux distros upon installation

Ed's right, a lot of true Linux geeks tend to look at GUIs as unnecessary baggage. But I think that a GUI interface (like Webmin?) is almost a pre-requisite if a piece of software is going to be used by any significant portion of the computing population.

So far in this week's conversations, Chrystie has consistently been the voice of the above-average librarian who has an interest in open source but is not a techie. (Sorry if I just insulted you, Chrystie.) And since those folks are the ones who will ultimately be willing to try open source, it would be a mistake to advocate omitting a GUI in the interest of purity.

Which means, of course, that many open source apps still have a long way to go before they're ready for the general library audience.

Shedges

I totally agree that a GUI is vital in this day and age. I use Mandrake rather than Debian because of Mandrake's management interface (GUI).

Although it is not my favorite distro, I would even suggest that libraries consider Lindows because it is cheap and easy to install programs. Lindows educational pricing is quite reasonable.

CRH, there is much to be said about ease of installation and use. But at the same time, I think that it is important that we understand the tools we are using. It is kinda like when as reference librarian we gently scold teenagers and college students when they use tools like Google to do their research. They use it because it is easy and they get lots of stuff but are they getting the type of results that would get if they used Lexis/Nexis or Proquest.

Freedom is not "free"

Do people even care about the operating system? I doubt it, they just want their applications to work. I appreciate the quality of OpenOffice and the rest, but we really need some library-specific software a lot more. Walt Crawford, and some people on the oss4lib list have talked about an OSS-based, CIPA filter, libraries that are dealing with CIPA must be paying through the nose for apps like NetNanny. The secure public desktop seems to be coming along, we need WIBS so if that can keep going, then it can be a big savings as well. The other area that is ripe for OSS is print management. What about applications that are not in libraries now, there's some wild ideas [url http://webvoy.uwindsor.ca:8087/projects/flow ]

here[/url], it would be great if OSS staked a claim on some new territory as well. (cloaking, now back to lurking...)

I agree with our (lurking again) friend bookerT about developing OSS applications that libraries need and don't have. I'm reminded of sylvar's excellent example earlier (in the week - on Tuesday?).

It seems to me, however, that Mozilla and OpenOffice are easy entry-points into the world of OSS. If you have people trying those, buying into the community idea around them, then perhaps you have more advocates in what djsarge refers to as the 'rank and file' and then you can move on from there.

Ah ha! I made you come out. All good points, indeed.

So, who's in? And what to work on first?

CIPA-compliant filtering software?

Something else?