OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   RE: RE: [xml-dev] XML Technologies: Progress via Simplification or Co mp

[ Lists Home | Date Index | Thread Index ]

Just taking it as presented.  As I said, I know that's not 
what he meant but that bit gets repeated too often and 
actually, one should look at it.  Duct tape is what we 
use for broken technologies.  (I've spent more of my 
life on stage than in a computer lab but enough time 
in both to know where the comparisons fall apart.)

The vision of the gleaming technology is older than 
that.  Go back to the thirties and even earlier for 
the stainless steel buildings, art deco, etc.  The 
Empire State building is still standing.  A lot of the 
ticky tackys of the 50s aren't.

Raised floors hide the chaos.  Facades hide the boilerplate. 
A good designer enables upgrades.  That is another reason 
not to duct tape cables.  It only works if the setup 
isn't changed (ever have a drummer show up with a different 
kit, a new drum, or the bass player and lead guitarist have 
a fight in the dressing room and now want to be on opposite 
sides of the stage?)  Late bind the moving parts if you can. 

And yes, that is why scripting is invaluable.  The real 
difference in what we do now and what we did in the 80s 
is that objects are scripted in a thin client.  Interpreters 
have found their natural and effective role.  On the 
other hand, web clients never come up to the complexity, 
reliability, and real workflow efficiency of thick 
clients build with languages that are optimized for the 
task at hand.  Foxpro and its language are a good example 
of application environments that refuse yield simply 
because writing a SCAN statement is fast and hard to 
screw up when compared to looping over a recordset 
and keeping up with position.

It never completely settles down on a new paradigm 
except where the environment is very very simple. 
Tim still doesn't understand what information 
ecosystem concepts really mean.  It is all about 
the predictability of environments and what to do 
when they no longer are.


From: Adam Turoff [mailto:adam.turoff@gmail.com]

> http://tim.oreilly.com/opensource/paradigmshift_0504.html
> Only amateurs and desperate roadies duct tape mic cables.

I think you're reading too much into the duct tape metaphor.

On the one hand, there's a vision of an precision engineered system
where everything is well-designed up front, full of smooth polished
chrome and nary a stray line or unasthetically curved surface.  It's a
vision of the future we had since at least the 1950s.

Then there's the hard reality, where things never work as flawlessly as
originally designed.  This is the where the "duct tape" is visible, and
hacking is necessary to make things run in the real world.

Then there's the mix of the two approaches.  In the real world, you
cannot avoid all of the chaos.  So instead of ignoring it and dealing
with it at the last minute, you can support chaotic behavior by
installing raised floors and running all of the cable out of sight.

Perl, and the metaphor of the "duct tape of the internet" isn't about
amateurish use of immature technology.  It's about accepting that shiny
gleaming COM/CORBA/SOAP/whatever components never work as flawlessly as
originally designed, and more often than not, someone needs to hack on a
system to make it work in the real world.  


News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS