ADV: XML to CSV Converter

[Fairlight]

>From inside the gravity well of a singularity, Brian K. White shouted:
> Unfortunately price was the main issue for us.
> $1k is no big deal, but we are more and more moving towards many small 
> unimportant servers where:
> 
> $2500 in hardware, "free" os and much of the software, where we did invest 

I think I can safely snip there and say that your keyword is "free" when it
comes to software, based on the rest of your paragraph(s).  We're back to
the PuTTY vs Anzio argument in just a different form.

As for the cost, you're citing $1k against each machine.  The question
then becomes, "Who owns the machines?"  If they're all owned by Aljex,
someone can feel free to talk to me about a site license.  If they're
all owned by individual companies, well then saying $1k for part of a
24/7 enterprise-centric solution (this kind of processing is becoming
increasingly central to the operations of a lot of businesses) is a small
price for each company to pay once per major version in order to run it
24x7x365xyears_between_major_releases, nonstop and multiply concurrently if
necessary.

By your argument, fP's price should be an issue for you.  Well, it would be
if you were using development kits at $1500 a pop.  The runtime is a
different story, admittedly, if you're using that.  

I'm looking at it this way, same as I do with OneGate:  It's software
designed to be central to a decent percentage of doing business in one
arena or another.  If it's that important, then it's enterprise-level, and
then $1k isn't unreasonable because a decent amount of their business is
being handled by it.  In fact, it's actually cheap.  Sure, nobody likes
to spend money on software, including me--but if you want to do certain
things, it's a fact of life.

> So, similarly to anzio's webpo, which would also be pretty great for us, far 
> more valuable for our particular operations than an xml/csv converter 
> actually, $1k per box for just one little feature of the many that each box 
> needs, that doesn't work out to be worth it.

"One little feature," is -not- how I'm viewing it.  I'm viewing it in
the context of a business that has 200+ XML documents come in a day
from a single source, and they have multiple sources.  How much money
would that be for an office temp to key in manually?  Don't answer, it's
rhetorical--as a former temp myself, I know exactly how much.  The cost of
this program is less than a data entry temp's fees for a week (at the
employer end, which is about 3x the cost of what the temp actually gets,
last time I talked to someone about it).

> It makes more sense to pay a developer to do some contract work for a couple 
> days and have something you can use however and as much as you want.

That has to be the most short-sighted statement I've seen all day.  Okay,
it comes in second after Jose's comment about "aren't other such things
available".

Sure, you can pay someone (me included) to do custom.  It's usable
-once- per format, unless you're asking for a custom solution that does
generic conversion.  In which case, that's the product I just wrote, and
I guarantee you it cost more than $1k at my hourly rates.  I dunno what
you're thinking, but you either code a hell of a lot faster and better than
me, or you're engaging mouth before brain.

> I'm sorry this is not helpful and that I don't have any constructive 
> suggestions.

It's very helpful.  It says you're not interested unless you can get it
free, or at least very cheaply.

> I'm just saying what the things are that one potential customer is weighing 
> so that you can see what is their situation and try to think of ways you 
> might make yourself into an attractive part of that situation. 

Believe it or not, I do at least some market research.  More than one
fP developer told me it was worth the price when asked directly.

> Right now you are very "full service" which is great for end users. In
> which case maybe we were simply never part of your intended market in the
> first place.

And there's never been a bug in any application that you've bought which
you had to wait on the developer for a few weeks to a few months to get
fixed.  My longest time from report/discovery->fix is 72hrs, and it was
really a huge logic issue brought out by someone's request for a new
feature.  Thinking time, (70hrs - sleep).  Coding and testing, 2hrs.  Most
of my patches are < 1hr from notification usually.  Less than 24hrs in
99.9% of cases.  I can only think of that one exception.

I've been doing business for just over 12 years now.  I can't think of an
example when someone -didn't- want "full service".

mark->