Dmitry Borodaenko
2/12/2009 8:24:00 PM
On Thu, Feb 12, 2009 at 9:18 PM, Pit Capitain <pit.capitain@gmail.com> wrote:
> Hi Dmitry, nice to read from you again. We once met in Karlsruhe a
> couple of years ago, where you talked about Samizdat at EuRuKo. I
> always liked your reasoning for the name Samizdat, and now also the
> dedication of the new release. Unfortunately I don't have anything to
> do with RDF, so can't say more to Samizdat itself.
Hi Pit!
Thanks for your support, much appreciated! I remember you from Karlsuhe,
too :) It was a nice time, a shame actually that we couldn't establish
EuRuKo as a regular event...
You know, over the last years focus of Samizdat development shifted away
from RDF and more towards open publishing. It still does RDF, and RDF
code has become more advanced (and, by necessity, much more
well-structured), but it has taken a back seat to things that matter in
real world: security, usability, multimedia support, etc. If the
prominent place RDF takes in Samizdat's description scares people off
like that, I should probably rephrase it: Samizdat is no longer an RDF
system that incidentally does open publishing, it's an open publishing
system that happens to have RDF under the hood.
I think that other standalone modules in Samizdat library, while small
and single-purpose, deserve more attention than that scary RDF storage.
For example, the story of the Sync#try_lock bug kind of proves that
Samizdat Cache module is the most advanced Ruby object cache
implementation out there (unless I'm missing something and there is a
way to do thread-safe two-level locking without that API ;)
I'm also quite proud of the Sanitize module: it might be comparatively
slow due to invoking both Tidy and REXML, but it is small and elegant,
making it easy to understand and to keep secure. With tools like that
out there, who would need another HTML parser?
The point is, Samizdat is more than an RDF engine. It has other bells
and whistles that may just as interesting!
--
Dmitry Borodaenko