|
|
| GeoCommunity Mailing List |
| |
| Mailing List Archives |
| Subject: | RE: GISList: Automatic data recording |
| Date: |
07/01/2003 03:50:01 PM |
| From: |
Dimitri Rotow |
|
|
> >
> ------
> Everyone involved in cgi programming/server-side services actualy knows
> a serious drawback pf asp - it is still poorly suitable for heavily used
> services. If you have hundreds of hits per second on your site then your
> asp starts "idling". Actualy speed difference for the identical
> algorithm written in asp and some other "older" server programming
> languages like PHP or Perl is usualy about ~60 times (asp is slower).
Nonsense. If anything, CGI scripting is usually, but not always, slower
than .asp.
> And if you are going to use it under windows environment, you are going
> to have serious problems with it. Not to mention windows and IIS related
> security problems. Windows world is nice until you smash into serious
> efficiency and security issues.
More nonsense. Windows from 2000 onwards and *especially* recent releases
like Windows Server 2003 are *extremely* efficient, reliable and highly
secure. Yes, it is true that there are more security reports for Windows
than for Linux, but that's because more people use Windows, by a factor of
100 to 1. For web sites, many more commercial users utilize Windows and
Windows has vastly more web applications running under it. Note that
although Linux advocates are fond of stating that very many websites run on
Linux, what they don't say is that most of those are irrelevant, "one
person" websites involving trivial HTTP service that receive nearly zero
visitors and for which security does not matter. The great majority of
websites on which money changes hands run on Windows and IIS, as do most
serious commercial websites for which security is a factor.
> While learning some php, perl (perl_mod incl.) is not too hard if you
> are already involved in GIS or any kind of professional computing (many
I agree it is not difficult. But there are many things that are not so
difficult to learn, such as making one's one shoes, that people who place a
reasonable economic value on their time do not undertake. Instead, they buy
their shoes pre-made. Note also that most people don't hire a consultant to
make their shoes for them when they can easily buy them off the shelf.
I agree that if you have a hobby interest in programming, you might enjoy
that even though it is not an economically sensible activity, just as some
people who have a hobby interest in leatherwork might make their own sandals
and find that more gratifying than buying off the shelf. But again, that is
not most people.
To repeat, most people running sophisticated software want to use pre-built
tools to get the job done without programming themselves or hiring
developers to write custom applications. Suggesting that it is economically
sensible to program one's own application solutions using "free" software
(when "paid" software can do the job without programming) only makes
economic sense if the programmer's time is nearly worthless [which raises
the important managerial question... who in their right minds thinks it is a
wise management practise to populate an organization with nearly worthless
programmers?].
If you can acquire a sophisticated, inexpensive application that does what
you need without programming that makes a lot more sense. I agree that if
you *cannot* do what you need to do with a pre-built application than, of
course, it makes sense to look at all possibilities. After all, I work for
an organization that makes a lot of money when people hire them to do
programming. I am simply arguing that casually stating that one can easily
learn to program or that one should hire somebody to do the work is in most
cases a wildly uneconomical way to proceed, and that minimizing the vast
cost of custom programming to justify "free" Linux is, in most cases,
fundamentally deceptive and foolish.
> have accomplished this easily) and hiring somebody to do the work in
> Unix/Apache or Linux/Apache environment is actualy a more proffessional
> approach in this sence. One day you may be surprised to find out how an
> "old outdated" Pentium II/Linux/Apache/Perl_mod system outperformes
The PII strategy can be very good or very bad depending on who the user is
and what their economic situation is.
What's relevant in third world cases is not what is the best strategy for
people whose time is worth more than pennies per hour. If your time is as
valuable as the cost of food you eat every day, you won't be fooling around
with Pentium II's. They're just not adequate to run modern Microsoft Office
suites of applications that you'll be using, nor are they able to handle
those web applications of interest to most people who are have something so
unique that they are not hosting it on their ISP.
I grant you that a PII with Linux and Apache would be fine for a
low-bandwidth FTP site or
|
|
Sponsored by:
For information
regarding
advertising rates
Click Here!
|
