I'm glad to announce that Pliant release 99 is available for download, at:
This is a very important release, for several reasons:
For the ones that missed previous stories, from my point of view, the
framework of the final Pliant has been led down with release 97 through adding
the new UI user interface and the storage subsystem, both on top of PML
Now, with release 99, I'm convienced that all of this is now production quality
(please don't mistake production quality with feature complete) because the
main Pliant server I'm administrating is behaving smoother than ever.
Lastly, I've spent a fiew days off thinking how to properly drive the futur
of Pliant, and decided the details that will follow.
So, first I'm sorry to have missed number 100 for such a major mistake, but
nobody is perfect,
then I want to thanks some of the main contributors that helped me achieve
the present sresult:
first of all, my wife Helene for helping me not get crazy under the load for
more than ten years now ... and I hope for a lot more,
then Patrice Ossona de mendez for organizing the first ASPIC (Pliant meatings),
Marcus Santos for beeing the first brave to teach Pliant before any decent
and lastly my boss Didier Dumas for trusting me so that earning my leaving
has always mean freely putting Pliant to real work, so sorting out nice
looking ideas from really good ones.
Now, starting from release 99, the development model of Pliant will change
In very fiew words, priority number one is documentation.
Since it might just look at a good resolution with no practicle changes, here
are more details about the changes that will help me achieve it:
. first of all, providing a decent documentation will require making more
user friendly (and providing a satisfying documentation from the intelectual
point of view would require to adjust some details in the design of) many
peaces of Pliant. My first huge decision is I will not do these changes,
at least for now. The reason is that Pliant is a large projet so polishing
all peaces (as opposed to making these stable which has already been
achieved) will take years (just the same as it took to Unix operating
system). So, what I will do is document as it is.
. then each time I've started to think or write Pliant documentation, I've
tried to do it top down, starting through explaining the great concepts
that Pliant relies on because I wanted to justify and proove that Pliant
relies on great new concepts as opposed to beeing just a new instance of
well known receipes. The problem is that I have not yet the writing skills
to do so, and I have no existing materials to rely on so that it's a
daughting task. As a result, I will now do it bottom up: the new
documentation will be a large set of small poorly written articles
describing how to do something with Pliant or how one part of Pliant is
working with no overview or design choices explainations. I'm now assuming
that the reader either already decided to use Pliant for unknown (maybe
strange) reasons and I have to help him or her do so, or that the reader
wants to start hacking Pliant code so that I have to help him or her
understand the existing code.
. lastly, to write some nice looking documentation, I would need to master
publishing tools to make nice looking graphics and screenshots, and these
are not ready in Pliant desktop tools relying on the UI, so I will start
with words only.
So, as a summary, this will be documentation for the braves (braves does not
mean already skilled programmers, just highly motivated), not at all mainsteam
documentation, but it will be effective documentation so that any guy studying
it carrefully will be abble to continue Pliant without my help.
There will be room for a second layer of better polished documentation later
written by myself or by more skilled writters, but it should just reduce the
required braveness and a little bit time for mastering Pliant.
Sunday, november the 4th of 2007.