This post originated from an RSS feed registered with Web Buzz
by Christian Machmeier.
Original Post: Features done in meetings
Feed Title: .redSPLASH - Blog
Feed URL: http://businesslogs.com/WebLog/RSS.xml
Feed Description: Share what you know, learn what you don't
I don't know about the vaunted 10x productivity of Rails, but that's
only because I haven't done much web-work otherwise. All I can tell you
is that for Pimki I added two new features (and a bug fix)... while I was sitting in a meeting and came across the need for them!
Embedded GraphViz graphs
As
I was taking notes in a lecture (using Pimki on a brand shiny new
laptop), I needed to copy a diagram from the whiteboard. Laptop are not
notoriously easy in that aspect, but the diagram was simple enough and
I know that dot language reasonably well. So as the meeting went I just
typed the graph into a page. During a short break I added a new chunk
that looked for such embedded graphs, runs GraphViz and replaced the
text with the image result. (Well, the short break wasn't quite enough
and I didspend that evening fixing this, but the point, man, the point!).
So if you're using Pimki2
and update to the latest CVS snapshot (be warned - still alpha), you
can embed a GraphViz diagram in a page with the help of
<graph></graph> tags. Just read the help section.
Google Searches
In
another meeting, I wanted to take a note to search google on something.
What better way that to extend the existing external links with a
Google search? Just embed a [g[query string]] in the page and you're laughing.
Of
course, just adding more such links is not the full solution I'm after.
But until I get a chance to do the full macros, it'll do quite nicely.
Not bad for a couple of meetings in which I also managed to learn a few things :-)
As
for the bug-fix, in the upgrade to Pimki2 I managed to miss the ToDo
text expansion - the rendering of the text after the todo as part of
the page itself. That is not fixed and links to pages, glossary items
and the like are displayed correctly.
The next major item on the
list is still the backend format. This is also the major holdback in
development - almost everything is dependent on how the information is
stored. (So if you use Pimki, please fill the user survey at: http://rubyforge.org/survey/survey.php?group_id=447&survey_id=25!)
Currently there are two contenders: SQLite and KirbyBase.
SQLite comes with Instiki-AR, but I think KB may have other advantages
that merit consideration: Import from Instiki can still be acheived
through export/import (and that may be the case anyway with
Instiki-AR); It is a pure-ruby solution, so nothing extra to install;
It uses plain-text to store the data; I like the ruby-blocks format of
querying data. The major disadvantages is of ActiveRecord: KirbeyBase
is not an SQL format and there is no AR adapter for it. This means I
can't learn-by-example from Instiki and that I miss out on all the
existing goodness like the #acts_as_taggble mixin. I will
also lose on changes from Instiki, but since I'm not sure about how
continued integration will work (expect for export/import) anyway, and
that may turn out well in that it will force me to better review each
change in Instiki and get a better understanding of what goes in where.
What
I need is mainly time to play with both options. Not a thing I have in
bounds, so I'm asking questions on ruby-talk and here. If you have any
experience with these issues, please feel free to contact me directly
or on the pimki-users mailing list.
Previous Topic
