Re: Programming project required

From: Gerhard Haak (gerhard_haak@hotmail.com)
Date: Sat Jan 20 2001 - 05:43:00 MST


> Don't choose languages/platforms/OSes etc based on what you like; choose
> what people are using, and are going to be using more in the future. In
real
> world coding, you very rarely get to choose your tools; they are part of
the
> initial parameter set.
>
> I highly recommend ASP, on IIS, with SQL Server backend, as a good
starting
> set of employable skills. I'm currently working using this set of
products,
> with the caveat that the same products must also support Oracle, and the
> caveat that javascript is used for ASP programming rather than VBScript, a
> very sound choice.
>
> Some of the purists will come out of the woodwork and attack me for this
> viewpoint. It does sound like a banal and colourless life. Don't be
fooled;
> the mainstream technology is not only rewarding in a financial sense, it's
> fun. It's fun because it keeps changing at a crazy pace. When you've got
> most of the coders, and most of the install base, you can afford to put a
> lot of effort into development. This means things are rarely ever the
same,
> which is damned hard, but always interesting.
>
> Note also; I've had a lot of jobs, and never used even similar sets of
> skills/languages from one job to the next; the primary language to use,
for
> instance, has always been entirely new to me. That's not uncommon. So
> there's no need to get cut up on becoming a specialist in language X or
> database Y. Just get really good at learning.
>

Beware of the hype, Emlyn, lest it become self-fulfilling to the blind
(which is exactly what MS, Oracle, etc. are counting on). Its due to
thinking like this that the software industry is in the mess it is (i.e..
cost over-runs, projects that fail to meet spec ...).

If given the chance, a software engineer/developer should consider the
project requirements and available resources, and use the tool-set most
appropriate for the job (unfortunately not always possible considering the
pro-MS bigotry present in relatively ignorant management circles) If all
you need is Access, then use Access. If you need something more meaty, then
consider using Oracle or SQL Server - but don't jump straight to SQL Server
just because its MS.

There's a good reason why Oracle has ~60% of the Enterprise database market,
and why Apache has ~60% of the web-server market (MS has ~20%). (See
www.netcraft.com for server info)

Last year we looked at moving across to the Sun devised J2EE platform, and
EJB's (Enterprise Java Beans) in particular. It was incredibly difficult to
separate the hype from the facts - it seemed that every consultant that was
in any way involved with these technologies was screaming at us to adopt it.
However, digging deeper revealed the hype to be exactly that - hype. At the
time, there was a complete absence of research that showed these
technologies (which are similar to the MS IIS etc. platform) reduce the
amount of effort one needed to expend to develop a product. In fact, the
small set of work we examined indicated that these platforms INCREASE the
amount effort required.

We'll be considering the change again this year, and am currently under the
impression that they've matured sufficiently to make the move worth our
while. I'm 90% sure that if we'd listened to the hype and made the move
last year, we'd be regretting it still.

Critical examination of the facts is a good habit to have in any endaveour,
whether Extropian or otherwise.



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:05:06 MST