need of a web based message board

Fri Jun 29 07:10:00 PDT 2007

Only Jay Ashworth would say something like:
> Yeah.  MW is at it's best for Distributed Knowledge Capture.

I looked at the Wiki entry in Wikipedia again this morning.  Came close to
starting an fP Wiki.  Decided I have more pressing matters ATM.

> > I don't get caught up in the "web 2.0" buzzword scene.  There -are- some
> > cool site interfaces out there.  I've run into one that's just beautiful.
> > It has video delivery, but you basically can choose from arbitrary ranges
> > between any two timelines as deliniated by thumbnails of the start of a
> > segment.  So you may have 36 thumbnails, but you can choose to use 2-5,
> > 2-30, 11-15, whatever it looks like you want.  Totally drag-n-drop.  A
> > -lot- of time and love went into that interface.  And it's worth it--all
> > parties save on bandwidth, you spend less time and disk space on stuff you
> > don't want and more time on stuff you do want.
> 
> Where's that?  Or is it subscrition?

I'll contact you about this privately.

> Indeed; there's a Flash animation of how a Chrysler A604 automatic
> transmission works that's floating around out there that *almost* makes
> up for all the Entirely-Implemented-In-Flash site I want to burn.

I -hate- sites like that...  *shakes head sternly*

> *My* personal problem with Zope is that's it's too much of amoving
> target, a problem I've always had with WebGUI.  They both have somewhat
> of an impedance mismatch problem as well.

*coughsputter* PHP *coughsputter*

And I'd rather recompile and replace all of X11 than PHP, to be perfectly
honest.  Hardcore linux kernel devs have told me they'd rather just
reinstall a new OS than install a new PHP.  Having done it several times, I
can see why.  The -last- time 4.x needed to be patched, they had an
arbitrary code execution issue.  It took 18hrs to mesh it with SuSE's
bloody spec file from rpm's like...well it was 4.3.3+patches into which I
rolled 4.4.6.  You don't want to think about how broken that
was--specifically their use of uhm...(*thinks*) libtool, that's the little
bastage.  Horrid, horrid breakage.

THEN...they rolled a -bad- patch.  So less than a week later there was
4.4.7.  You know what they did?  They figured that as long as they had to
patch their patch, they'd toss in a new parser version while they were at
it, in a third revision level bugfix patch meant to fix a patch.  How
-bloody- stupid does one have to be to do something like that.  That one
took me very little time to fix, but they -did- break the makefile
assumptions doing what they did between the prior patch and that re-patch
by virtue of changing the parser version.  I was pissed, to say the least.

> I haven't seen the video, and I'll go look.  What'd they compress it
> with?

Stupidly, MPEG4 with AAC-22.050KHz mono.  They should have gone with H.264
instead of MPEG4 for what they were doing.  Which again tells me they don't
know their arse from a hole in the ground.  Anyone with brains making a
.mov like that should be using H.264.

> > That holds up until you want the apps to actually -do- something beyond
> > simple input/output--and even then, sometimese.  The
> > second you want even data validation or the tiniest thing, you need lines
> > of code.  I'm just not seeing the comparison.  Add to that the fact that he
> > had massive lines of code even in the smallest example for something that
> > didn't need more than "<html><head /><body>Hello World.</body</html>"
> > and...well, I can't say much for his logic, either in terms of example
> > choice or design choices.  At least one of the two was a bad idea.
> 
> Sure.  But it's really hard to demonstrate a web-app-builder framework
> reasonably in less than about 3 hours...

It should have at -least- taken input, or been somehow -dynamic-.  Showing
an example of static content wrapped in needless dynamic fluff is...just
plain dumb, I'm sorry.  If he didn't have the time to do it right, he
shouldn't have done it at all, IMHO.

> There are always LOC.  The question is just:
> 
> Who has to *write* them.
> 
> And don't forget Component Software Hell.
> 
> If you base your code on other people's... then you have to track and
> test as they upgrade, or supply your own copies; each has pitfalls.

That's why you use reasonably mature technologies like perl or even
python.  I would put python in that category from what I've read and been
told.  If Bill Campbell thought it was mature enough to toss over perl for
it, that says something to me.  I'm not up for the change myself, but at
least I can feel it's reasonable to consider if needed.

I do -not- consider PHP mature, nor do I consider it even sanely designed.
These others (Ruby, plone/zope, et al) may be sexy, but as you say...moving
target syndrome.

The irony is that people have told me that perl is a moving target.
*laugh*  That's rich.  Sure, I get behind versions, but overall you can
bet on it being relatively stagnant compared to other things out there.

Seems like the newer it is, the faster it moves, too. :( And unfortunately,
many of these "designers" don't know one solid principle of solid
engineering:  It's not complete when there's nothing left to add, it's
complete when there's nothing left to take away.

mark->
-- 
No matter what your problems, modern medicine can help!
http://members.iglou.com/fairlite/fixital/