Re: Moving forward

["Daniel E. Markle" <syntax@ashtech.pa.net>]

On Mon, Jun 05, 2000 at 08:44:49AM -0400, Robert Myers wrote:
> > The reason for trashing the current site and code boils down to one basic
> > problem: scaleability
> 	Isn't this a problem that could be solved by re-coding the site at
> a later date?

Aaron didn't get across to you just how bad the scalability was.  With every
page load the system had to run everything.  The design was horrible, our box
was dieing when only 2 or more people were accessing our site at the same
time.

> 	I talked with Jay about this a little bit, he said that this was a
> result of the comments section for the articles.  Wouldn't it have been

This section made it even worse, but removing it only made the site marginally
useable.

> waisted.  The program logic could easily be moved to another
> language.  This is really all a part of a program lifecycle, as new

I would agree with this for most of the code, but the underlying design itself
was so flawed that it warrants a re-write.  We will re-use what we can, of
course.

> that.  It seemed as though the last time I worked with Jay on the code,
> that you guys were so close to having something, albeit not the grandious

Yeah, what we had was something that worked fine on a single user box with
single accesses.  Heck, even in that environment it was hammering more modest
boxes.  It was simply not useable in a multi-user context.

I should also mention the searching feature, a core part of our site, was
incredibly flakey and ridiculously system intensive.  Searching did a htdig
search of our very system intensive site, and with only a number of documents
in the low double digits, it would hammer the box for many minutes.  Not to
mention it broke horribly every time a considerable change was made to our
base code.

What we were facing was a site that was so system intensive and so poorly
designed that it would even start returning errors if only a very few people
were on line at once.  Had we actually gone live with this code in a larger 
scale, it would have inevitably been a very major flop that would have caused
more of a setback in bad press and enraged contributors than the time delay a
re-code will cause.

Really, this fiasco was caused by our inexperience and our failure to attract
other, more experienced, developers to the project to help us out.  These are
all problems that we are determined not to repeat, even if it means the
project is delayed by a few more months.

-- 
-------------------
Daniel E. Markle
syntax@ashtech.net
http://www.ashtech.net/syntax
-------------------