pyjamas 0.8alpha1 release

[lkcl]

On May 18, 6:29 am, harrismh777 <harrismh... at charter.net> wrote:
> Terry Reedy wrote:
>
> > No, because I think you are exaggerating.  That said, I think core
> > Python is pretty close to 'complete' and I would not mind further syntax
> > freezes like the one for 3.2.
>
> I am exaggerating only to the extent that someone can imagine folks
> becoming just annoyed with PEP progress to drop the whole thing... I am
> exaggerating only to the extent that we define 'it' as language death...
> if the user base narrows, python's future is held tentative... on the
> other hand, if the user base grows and campers are happy, then python's
> future is more stable... I don't think this is an exaggeration...

 one thing that people are not aware of - because it normally simply
does not make its way out into the public world: you're forgetting
those people who "just use" python.  they don't get onto public
mailing lists, they don't develop free software projects.

 i've mentioned this story before, but it's worth repeating in this
context.  i worked in a military environment (NC3A) back in
2006-2007.  the version of python that they were using was http://python
two... point... ONE.  six years after its release.  why???

 well, it went something like this.  someone got the idea that doing a
portal would be good.  so they looked around, and found Zope.  so,
they evaluated the latest version somewhere around ooo april to june
of 2001.  ok they _started_ evaluating it.  so, some four months
later, after doing some coding examples, we're now up to august 2001,
a decision has to be made by the internal client.  they say "yep, go
for it", but that took another four months (dec 2002).  now we do 18
months of software development (july 2003) to produce a base package.

 now the code now has to be handed over to a team who perform security
evaluations.  this has to be paid for.  another six months go by, and
the security accreditation is received (dec 2004).  but this was just
for the "base" code: now we have deployment and actual product /
portal development, and a maintenance cycle of 2 years (2006).  now
i'm called in to help with that maintenance and development cycle
(2007).

 and throughout this time there is *no way* that they can upgrade from
python 2.1, because it would cost another $EUR 10,000 to get the
accreditation certificate.

 it's now 2011. for all i know, some TEN YEARS after python 2.1 was
released, they're still using it.

 you ... _just_ don't normally hear about these kinds of deployments
of free software, but it illustrates that a particular version can
hang around for a hell of a long time.

 l.