[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Status update: IMC CMS and short-term plans for Samizdat

From: boud
Subject: Re: Status update: IMC CMS and short-term plans for Samizdat
Date: Wed, 13 Sep 2006 02:10:42 +0200 (CEST)

hi dmitry, all,

On Wed, 13 Sep 2006, Dmitry Borodaenko wrote:


The IMC techs once again decided to do something about the situation
with Indymedia codebases. The results of the previous attempt in 2002
inspired me to start development of Samizdat. This time, this push for
better codebase gives me incentive to move Samizdat forward.

cool :)

Samizdat has come a long way since 2002, here is a quick sketch of
where it stands now with regards to the features needed for new
Indymedia CMS (see details at the above link).

While Samizdat covers a lot, many important features a missing, and
most of them can be added without big changes to the architecture. So
here is the roadmap for the next few weeks:

- bug: fix for current libgettext-ruby
- install: setup running default site
- security: block external images (via xhtml.yaml)
- doc: update docs, unit tests

I plan to finish this today or in the worst case tomorrow. Once done,
I will finally release the long-overdue Samizdat 0.6.0, and upload a
package of it to Debian/unstable.

OK. i'll try installing/testing once it's uploaded/mirrored.

- security: notify moderator button
- security: captcha
- security: spam filter
- bug: supplement published date with modified date
- ui: inform guest users that they can't edit even open-for-all messages
- ui: show file size in non-inline messages

This stuff can be done within this week (including weekend). What will
help me do that is 1) pointers to some ready captcha (or the like)
modules and web-ready spam filters for Ruby; and 2) help with testing,
pointing out bugs, and keeping translations up to date.

spam filters: no idea about ruby, but for twiki, i've found BlackListPlugin to work fairly well:

It functions using a mix of several different strategies against wiki
spam which attempts to get high google rankings.

i don't think this helps much except for inspiration...

- design: get rid of remains of built-in focuses
- feature: wysiwig HTML editor
- feature: image manipulation (thumbnailing, galleries)
- feature: user notifications (research existing mailers)
* > - bug: fix translations (?)
- speed: mark-and-sweep in cache
- ui: show nested comments
* > - feature: import RSS feeds (filtered syndication of blogosphere)
- feature: pod/vodcasting
- install: research PostgreSQL clustering
- install: restart Apache in postinst
- bug: catch and report db connect timeout error
- bug: reproduce and fix boud's Sanitize issue

All this is a more difficult or lower priority work, I'm not going to
start on it until I finish with the previous two batches. I would be
really appreciate if someone could pick up one or two of these tasks,
the ones marked as features and ui are more easy marks as they are
relatively straight-forward and can be done without deep knowledge of
Samizdat internals.

i'm going to try:
* > - feature: import RSS feeds (filtered syndication of blogosphere)

i've started RTFMing a bit about basic ruby (/me has no OO experience :P)
and i'm likely to start experimenting in the next few days.

dmitry - i looked for the standard ruby library package you mentioned
in /usr/share/doc/libruby1.8/README  and i guess you mean REXML ??

i also found this discussion:

which points to what seem to be at least 4 different ruby packages/routines:

And here's a 10-line hack:

i'll also do: * > - bug: fix translations (?)

pl - there are some pl translations missing
fr - totally absent at the moment - i should hopefully be able to get this done this week

BTW: does anyone know if we are ok with BIDI (bidirectional text - like
hebrew and farsi/arabic) ?

Now let's see how much I can before Sunday and how much help I can get.

/me will do my best...

reply via email to

[Prev in Thread] Current Thread [Next in Thread]