Fwd: [IMC-Tech] Re: [Seattletech] Proposal: CVS source control for city websites

Tue Jan 15 14:44:32 CST 2002

Is any of this info useful for us?

>Delivered-To: mediageek-org-listgeek at mediageek.org
>Delivered-To: imc-tech at lists.indymedia.org
>From: Micah Anderson <micah at indymedia.org>
>To: Jeremy Kahn <jeremy_kahn at yahoo.com>
>Cc: seattletech at lists.indymedia.org, imc-tech at lists.indymedia.org,
>         webcoders at cat.org.au
>User-Agent: Mutt/1.3.23i
>Subject: [IMC-Tech] Re: [Seattletech] Proposal: CVS source control for 
>city websites
>Sender: imc-tech-admin at lists.indymedia.org
>X-BeenThere: imc-tech at lists.indymedia.org
>X-Mailman-Version: 2.0.6
>List-Help: <mailto:imc-tech-request at lists.indymedia.org?subject=help>
>List-Post: <mailto:imc-tech at lists.indymedia.org>
>List-Subscribe: <http://lists.indymedia.org/mailman/listinfo/imc-tech>,
>         <mailto:imc-tech-request at lists.indymedia.org?subject=subscribe>
>List-Id: IMC Tech Team <imc-tech.lists.indymedia.org>
>List-Unsubscribe: <http://lists.indymedia.org/mailman/listinfo/imc-tech>,
>         <mailto:imc-tech-request at lists.indymedia.org?subject=unsubscribe>
>List-Archive: <http://lists.indymedia.org/mailman/public/imc-tech/>
>Date: Tue, 15 Jan 2002 12:33:45 -0800
>
>Note:I have added webcoders to the CC list, and removed editorial (until
>the technical issues are worked out).
>
>Jeremy and Jonathan, excellent proposal! You have clearly defined the
>terminology so that people can understand the technical terms involved
>and have described the problem with an acute understanding of the
>issues which the content producers of IMCs face. This is commendable
>considering the complexity involved with the various pieces. It is
>without a doubt that various aspects of the design have been done for
>technical reasons, often leaving the important aspects of how to
>manage the content in the dust. It is for this reason that your
>proposal is important.
>
>What follows is:
>* a note on formatting
>* a couple clarifications
>* an ammedment to the proposal
>* suggested further reading
>* the advertisement tacked to your email removed
>
>
>--------------
>Note on format:
>
>I will quote your proposal in-line using the standard ">" symbol to
>denote text from your copy. In an effort to reduce email overload, and
>conserve space, I have remove parts which I did not consider necessary
>for my comments. Where necessary, I have added editorial comments in
>standard bracketted format [like so] to clarify the text out of context.
>
>The place where the advertisement at the end of the original email was
>removed to save us all.
>
>
>--------------
>Clarifications
>
>* Under the PROBLEM DECLARATION:
>
> >[IMCs have only part of their pages under source control, which is an
> >advantage], in some ways -- the shared code, controlled under the
> >Active CVS server, is only a single copy, while some dozens of IMCs
> >operate off the one copy.
>
>I do not have the appropriate understanding of Active's CVS structure
>to properly comment on the reasoning behind the current source code
>setup. The only detail I know which is relevant is that each city gets
>a *complete* copy of the Active code, and its relevant CVS directories
>when it is created, in otherwords, each IMC actually operates off of
>their own code, instead of off of one copy. The logic behind this (as
>detailed from http://tech.indymedia.org/cvs_info/cvs_18oct_log.html)
>was that each city would then have the opportunity to hack on the code
>independently, as if a developer.
>
>After a while, it was decided we didn't want to give cvs write
>access to everyone who sets up a new city, so the new-city setup
>script was re-written to do an anonymous checkout of the code, and an
>export copy of the local folder containing the config files.
>
>
>* Under the REQUIREMENTS:
>
> > * a box for the CVS server and DNS registration.  It
> > really needs to run Linux or some other Unix-like OS,
> > and definitely needs a stable internet connection and
> > enough disk space for about three times the size of
> > all the code to be checked into it. (not much disk
> > space, these days).
>
>It should be clarified that there is actually a server that was
>budgetted for, purchased, and assembled partly for this purpose. This
>is the development server, also known as "kropotkin". It is housed in
>L.A., was assembled by Chris B. and has freenet running on it (set up
>by the freenet folks themselves). The intent of this server was to be
>an encouragement of involvement of the IMC in the freenet project, and
>vice versa; a place for code development to happen, outside of the
>production environment, a sandbox for site development whereby access
>restrictions would be less than those in the production environment; a
>cvs mirror of the CAT CVS tree, a place to setup residual services
>which do not belong on a dedicated server (such as a request tracker).
>
>The server was to be a "cloudy mirror" of the production environment
>(currently stallman), meaning that it would as closely as possible
>maintain the same distribution and revision of stallman so that
>development can happen on the same platform so thta when changes are
>propogated to the production server there would be no unnecessary
>surprises due to version differences between the two servers. "Cloudy"
>because when you are doing development you are, by definition,
>changing how things are.
>
>This server was officially announced a few meetings back on IRC, and
>was included in the tech summary (see *further readings below).
>Stefani volunteered to set people up with accounts on request.
>
>* Also under REQUIREMENTS:
>
> > * an admin for that server.  I hereby volunteer to
> > do this.  I have three years of experience in
> > administering CVS servers.
>
>This is excellent, we've had some work on kropotkin, but not enough. I
>am not even sure if it can properly serve a single IMC site yet. The
>last I knew about this is that there is a database and a webserver
>running, but active still needs to be tweaked to get working there yet.
>
>Currently the debian packaged "cvs" is installed, there are other cvs
>packages available (apt-cache search cvs, apt-get install <package
>name>) which might be useful for this project, or you may decide that
>removing the packaged version and installing a built version may be
>the better way to go, that would be up to you.
>
>On a personal note, I can vouch for Jeremy as a dedicated IMC
>volunteer who would be a perfect person for this. :)
>
>-------------------------
>Ammendment to proposal
>
>* With what has been said so far, I think it is only logical to ammend
>the proposal to include kropotkin as the CVS server, enable Jeremy to
>make it happen by giving him the access to do so.
>
>* I think that Jeremy should be able to create accounts as needed, so
>that stefani isn't the sole person responsible for this.
>
>* The only piece which I found missing in the proposal was
>documentation and training of other IMCs in the process of how to
>utilize this setup to their advantage. I'd like to ammend the proposal
>so that it includes this, so that once everything has been setup
>properly there is a concerted and dedicated effort to make the system
>viable to non-technical people.
>
>
>-------------------------
>Suggested further reading
>
>I would suggest looking through these documents for further reading:
>http://tech.indymedia.org/cvs_info/index_toc.html
>
>Tech summary where the development server appeared (it was actually
>available much earlier than this):
>http://lists.indymedia.org/mailman/public/imc-summaries/2001-November/000028.html
>
>http://reflect.cat.org.au/lists/webcoders/msg01983.html
>
> >
>_______________________________________________
>imc-tech mailing list
>imc-tech at lists.indymedia.org
>http://lists.indymedia.org/mailman/listinfo/imc-tech