From david+informl-talk at folklogic.com  Mon Dec  3 02:37:59 2007
From: david+informl-talk at folklogic.com (David Anderson)
Date: Mon, 3 Dec 2007 08:37:59 +0100
Subject: [Informl-talk] What we can learn from other Wikis - 0
In-Reply-To: <474CDDE5.6090200@onghu.com>
References: <4748569B.4080300@si.on.ca> <474CDDE5.6090200@onghu.com>
Message-ID: <c4da0b700712022337m524a38d9p9d3fe7027dee6c0f@mail.gmail.com>

Yes indeed. I'm looking forward to further discussion too.

Besides TWiki, there are a few other tools that have influenced our
thinking about Informl. I'm thinking of JotSpot, openrecord, and
DabbleDB. JotSpot seems to have disappeared since they were acquired
by Google, but they certainly had interesting ambitions. With informl
I would hope to enable user creation of "applications" with less
"programming" than jotspot required. DabbleDB is also proprietary, and
not really in the wiki genre, but as a tool for helping users impose
some structure on their information, it is very interesting. Of these
three, openrecord.org is perhaps the most comparable to informl, but
makes some different choices -- such as keeping the database schema
loose, allowing free text such as "next Friday" in a date field.

Perhaps in a future post I can say more about some recent
architectural changes we've made to informl, and our (tentative)
thinking about where that's going.

--david anderson

> Thanks for the very informative and interesting post.  I look forward to
> the subsequent messages.
>
> Cheers,
> Mohit.

From david+informl-talk at folklogic.com  Mon Dec  3 05:31:24 2007
From: david+informl-talk at folklogic.com (David Anderson)
Date: Mon, 3 Dec 2007 11:31:24 +0100
Subject: [Informl-talk] problems with latest revision
Message-ID: <c4da0b700712030231i6f2d03d1r28560145a2a367d7@mail.gmail.com>

Anton,

I'm not sure what's wrong here. Perhaps you can provide more details?
And when it comes to "Is this really really really critical about
using an old revision?" I'm not sure what you mean -- and old revision
of what? Ruby-debug, or informl, or a wiki page?

I'm not 100% certain if it's failing in a migration, or before
starting to execute the migrations. Do you know? If it's getting to
migration 26, that migration is particularly complex. We've decided to
change the persistent representation of versions of wiki pages -- from
wiki markup to HTML. So 026_convert_version_bodies_to_dbhtml.rb is
doing some tricky stuff, but we decided it would be better to get this
DB conversion done now, once and for all.

Do you see a series of ". . . . . . . ." on the screen? The migration
prints a dot after converting each wiki page version. How much content
do you have -- ie how many pages/versions are being converted? Perhaps
there's a page with something unusual that our testing didn't uncover.

If it's not getting that far, perhaps the issue has to do with
informl's use of ruby-debug, and the fact that rails recently includes
it as well. We haven't done any testing on Rails 2.0rc-x, and very
little testing on versions newer than 1.2.3 -- what version of Rails
are you using?

Probably best to take this offline until we resolve the issue, and
then we can report back to the list.

--david

> ---------- Forwarded message ----------
> From: Anton Aylward <anton at si.on.ca>
> To: informl-talk at rubyforge.org
> Date: Sun, 25 Nov 2007 13:09:08 -0500
> Subject: problems with latest revision
> I have a problem running rake after the latest svn update
>
> [anton::~]$ cd ~/Ruby/informl
> [anton::informl]$ svn up
>
> .....
>
> Updated to revision 616.
> [anton::informl]$ rake db:migrate
> (in /home/anton/Ruby/informl)
> rake aborted!
> can't activate ruby-debug (= 0.7.5), already activated ruby-debug-0.9.3]
>
> (See full trace by running task with --trace)
> [anton:informl]$
>
>
> The trace doesn't seem to help.
>
> Is this really really really critical about using an old revision?
>
> --
> Do we, holding that the gods exist, deceive ourselves with insubstantial
> dreams and lies, while random careless chance and change alone control the
> world?
>     --Euripides, Hecuba
>
>
>