The Trip
So I finally had my stuff packed together and was ready to leave Aachen for the Plone Archipelago Sprint in
Norway. It was Sunday morning, 9am, a warm, sunny day when I was leaving first for the Cologne-Bonn
Airport and then flying to Oslo. Shortly before landing I did a peek out of the window and what did I see?
Snow! So finally that was gone at home, I was in fact going back to where it’s sorta cold. Thank you ;-)
(well, snow wasn’t all over, just some spots here and there. Weather actually seems more greyish with
some rain here and there but not really that cold).
So at Oslo and on the train to Tonsberg, from where the boat was actually going to the island, I’ve met
Balazs Ree from Hungary, David Convent, Godefroid Chapelle, Jean-Francois Roche, Raphael Ritz and
Martin Aspeli. Of course the sprint started immediately after leaving Oslo with Balazs getting his laptop
out and Godefroid joining him in discussing kukit (I think).
In Tonsberg we met many other folks joining us in the boat (and loads of cans of beer actually). After
putting these and ourselves on the boat we had an hour tour to the island Bolaerne where we did arrive
at 9pm actually (and thus 12 hours later after leaving home. But I shouldn’t complain as Geoff Davis
just came in telling me about his 48 hour trip).
About the location
The location is actually quite nice. It’s an island which was once used by the military reaching
back to WorldWar II times. You can see the remnants of that and the cold war still and in some of
these we’re actually sleeping and sprinting. It’s the old officer’s building and it’s sort of
comfortable. Unfortunately it’s a bit away from the restaurant which is not so bad because of
the way you have to go to get fodd but due to the fact that the place with the restaurant
is actually the area with internet. Thus for breakfast, lunch and dinner times you usually
see the most laptops out there.
This situation (having a sprinting area without internet) makes it actually a bit difficult.
Some effort has been made to copy all things over, get a copy of the subversion repository
and make sure it also will merge in the live one again. Thanks to Godefroid Chapelle, Andreas
Zeidler, David Convent and others this seems to be working now.
Additionally we have the tower with internet access. It’s a bit remote and at the restaurant
area and does not have space for all of us so it’s mainly used for setting up the SVN server
and so on.
Monday
So monday started with me being the bad guy. Just because I was triggering the fire alarm
by having a shower… So why’s that, you might ask :-) Geir told us the day before that
we should keep the door to the shower closed when having one as the steam would otherwise
trigger the alarm and 40 laptops will be watered. In fact I was having the door shut but
had it slightly open 10 minutes afterwards. This seems to be enought for the alarm to
be triggered. But now we know at least that there’s no sprinkler installed and everybody
was at least awake :-)
So actual work just started after breakfast and so this is the time to get a bit more
technically now.
So it started by some introduction to how sprints works (especially those organized by limi)
and was is going to happen here. Alexander presented us a big list of topics (and PLIPs) to be
worked on. In a further post I will post that list with annotation of how far what went.
The new PLIPs we generate here will probably copied over by Wednesday evening by Martin Aspeli.
So as you might know the main area of work is being done on the UI side of things which will be
the main story for Plone3.0. In Plone 2.5 there has been laid some foundation for that work.
The main goals here are to improve the (perceived) UI speed, get infrastructure speed and make
using Plone even more fun as it is already :-)
The JavaScript discussion
So when speaking of improving on the perceived speed the first thing which comes to mind is
of course AJAX. Right now we do use only a little bit of that in Plone (in form of the LiveSearch)
but there was already lots of work done within the community to create even more. I want to mention
Bling by Benjamin Saller and kukit by Godefroid Chapelle and others like Florian Schulze and Balazs Ree
(if you’re interested in that you can look up my interview with Godefroid about it from the Belgium
Plone Meeting back in December 2005).
Related to this was the big discussion which JavaScript library to use. There is basically MochiKit
and Prototype in the discussion and both have been used by people already to do things and apparently
both have advantages and disadvantages.
For MochiKit speaks that it’s written by Python guys and thus is sort of pythonic and has a more
functional style while Prototype does not have that (so it’s merely a question of what you do like
more) but has the broader userbase as it’s used by Ruby On Rails. Thus there are lots of modules
already available for use in Plone.
So this was discussed and this discussion was closed by some demonstrations by Ben Saller and
Florian Schulze.
Ben Saller was showing his Bling product which basically uses one-liners in TAL for inserting
JavaScript code into templates which then e.g. transform the
Title field of a view into an editable component (via AJAX). More examples have been shown.
Godefroid and Florian showed us kukit, which uses MochiKit (but only marginally). It injects JavaScript
by not changing the source template but instead dynamically without touching it. It does that with
an XML syntax which defines which part of the template will get which behaviour attached (so it
works basically like Behaviour but knows more about it’s Zope/Plone context).
After that we just defined who’s working on what and made up some bigger groups, like the UI group, the
AJAX group and so on. The day ended with some fire in the fireplace, some guitar music with trashcan drums
by Martin and Martijn and (as rumours tell) some drunken people :-)
Tuesday
This was the actual first day of real work. People split up into their teams and started working
usually on first creating a PLIP. I actually needed to fix somem client bugs so I stayed at the tower
around noon (and I also needed to upload the episode 1 video of course). I then went back to the location
to work with Tom a bit on the markup support. What we want to do is basically allow for Wicked and Markdown
to be used in Plone and enabling to edit the defaults without changing the code (right now allowed content
types are defined within the schema of an object and we want to change that at least for objects which
do not define that and thus use the default). First thing to do was to ask all the people who know more
about it and we sort of have a plan now (implement the actual support in PortalTransforms and create
a Z3 utility to give the defaults).
At the end of the day we wrapped up what has been done. I am editing a video right now (which is also
the reason why I got to bed only at 3am) and I will produce a written report, too, which will be available
on plone.org soon. That’s also the reason why I am not going into detail here. We will do more reports
as progress is being made. So all of you will be uptodate.
So Wednesday has started now and we already had 3 new fake firealarms (not me this time ;-) and we’re close
to heading back to keep on saving the Plone world.
For more information make sure to check out my archipelago sprint videos also being published in this stream or use My Video Blog (also available via iTunes).