[Web4lib] RE: Open source support models

[Tim Spalding]

Thanks for the correction on Evergreen.

I wouldn't say I needed correcting on VuFind, though. Saying that
VuFind isn't mostly PHP because there are chunks of shell scripting
and HTML is like saying that a car isn't made out of steel and glass
because it also has leather seats and pockets of air. Much the same is
true of SQL and, depending on what you're doing with it, JavaScript.
And if you're going to count lines HTML against lines of code, you
should also factor in all the GIF and JPEG images, not to mention
Cascading Style Sheets. This way lies madness.

Tim

> Ha! Fact-checking is readily available for a few of these projects at
> http://www.ohloh.net/projects/evergreen/analyses/latest (for
> Evergreen) and http://www.ohloh.net/projects/10977/analyses/latest
> (for VUFind).
>
> So a correction: the business logic of Evergreen is written primarily
> in Perl and SQL, with a few optimized sections rewritten in C. The
> current catalogue interface is primarily JavaScript with XHTML, and
> the staff client user interface is written in Mozilla XUL (XML +
> JavaScript). The user interface for most new staff client
> functionality is being built with the Dojo JavaScript framework.
> Python is used for the internationalization build infrastructure and
> for the new EDI piece.
>
> The choice of language in the project largely comes down to using the
> most appropriate tool for the job. That's one of the advantages you
> get from building an application using a service-oriented
> architecture.
>
> --
> Dan Scott
> Laurentian University
>



-- 
Check out my library at http://www.librarything.com/profile/timspalding