Re: Terminology for web servers and CGI?

["Steven J. Owens" <puff -at- NETCOM -dot- COM>]

Carl Stieren writes:

> When explaining how to use third-party products with your firm's software,
> things can get tricky. I'm in the midst of such a project and while I have
> good definitions for web server terms such as
>
> * server root
> * document root
>
> I need names for a few other terms whose concepts are known to me (and
> described in "Webserver in a Nutshell", another book on Microsoft MIIS, but
> whose names elude me:

     This specific question is off-topic,but I think I can make a more
general on-topic point, so I'll reply to the whole list.  The short
version is, try to do your own homework.  The web is an excellent
resource for questions just like this, the more so because this
question is *about* the web.

     Stuff for the web can usually be found at the W3 Consortium's
website, http://www.w3.org.  Start here, if you find the general topic
but it's a bit too dense for you to digest, look for a book about it.
Usually you can buy a "kinder, gentler" version of the information
from O'Reilly Associates, generally considered one of the best
publishers in the technical field(not to imply that they're ripping
off W3 docs, bu that their books stay pretty close to the standards in
terms of information content, while generally being more readable).

     Detailed information about Internet standards can be found in the
"Request For Comments" (RFC) document appropriate to each
standard. When a standard is proposed, a "working group" is set up by
the Internet Engineering Task Force (IETF) to develop the standard.
They draft something and then they distribute it as an RFC.

(The origin of the name came from the early days of the Internet;
the author of the first RFC has been quoted as saying something along
the lines of "Well, we were just grad students, we didn't feel like we
had the authority to say "This is the standard", so we called it a
"Request For Comments".")

     There are certainly many RFC archives some as FTP archives and
some as web sites; here's one that a quick Yahoo search turned up:

     http://www.cis.ohio-state.edu/hypertext/information/rfc.html

     When you're dealing with technology topics you should consider
checking into standards organizations like American National Standards
(ANSI) or the International Standards Organization (ISO).  I know the
ISO sells printed versions of their standards docs for large sums of
money, but you might check and see if a) there's a publically
accessible copy on the web, b) your company has paid for access to a
private web site with copies of the document, or c) your company has
purchased printed copies of documents pertinent to the question at
hand.

     Other organizations to check on are the Association for Computing
Machinery (ACM), which publishes several useful documents dealing with
the more abstract and high-falutin' topics of computer science, not to
mention their interesting and educational Communications publicaitons
(Communications of the ACM, or simply CACM), and the IEEE
(International Electrical Engineering or something, but really they
have stuff on all sorts of topics).

     For example, I know that the ACM publishes a couple of lexicons
of computer science terms that we found frequently useful backwhen I
worked for a UNIX RDBMS software company, and I've seen various
volumes of algorithms.

     (Speaking of which, for those dealing with really abstract
technical topics, Donald Knuth's three-volume set, recently reissued,
is considered canonical in the area of algorithms.  For example, when
I wrote a reference section for our implementation of the Soundex
algorithm, a system designed to take typed english words and find
words that "sound" like them according to the rules of english.

     Knuth is incredible; it seems like he's behind huge chunks of the
basic concepts in the field of programming and computer science.  If
you need to really, really understand the stuff, bring along a math
grad student when you do the reading.  If you can manage to work your
way through a significant chunk of this stuff on your own, you'll
score some major cool points with your programmers.)

     Also check at the Software Engineering Institute (SEI) web site,
they do a lot of research into process stuff but they also do hard
research, and by the way they house CERT, the Computer Emergency
Response Team that was formed by the government to act as a clearinghouse
of trustable information about computer security issues.

     There are also a variety of nifty online lexicons and dictionaries
out there for specific domains, for example the Free OnLine Dictionary
of Computing, FOLDOC:

     http://wombat.doc.ic.ac.uk/foldoc/index.html

     Find others at the Yahoo Computing Dictionaries category:

     http://dir.yahoo.com/Computers_and_Internet/Information_and_Documentation/Computing_Dictionaries/

> * "that part of the URL following the optional port number after the domain"
>    (I want to call this "Partial URL", but someone at my shop said there
>     is a real name for it.)
>
>    For example, suppose you have a URL entered in
>    a browser as "http://www.mycorp.com:80/widget-scripts/buy-widget-a"; and
>    you've mapped "/widget-scripts/" to the cgi-bin directory on your
>    web server computer and the  Partial Path "/widget-scripts/buy-widget-a"
>    maps to particular script, say "C:/webserver/cgi-bin/buya.pl"

     Hm, I know there's a specific name for that part of the URL in
the RFC, mainly because three or four years ago I was helping a friend
write a CGI script to verify URLs.  Mostly we just call that the
"arguments", specifically GET-style arguments (POST-style arguments
are passed to the web server separately from the URL). Also called a
"query string". Check the RFC.


> * "the handler for an extension" (not a CGI or web server term, but a more
>    generic one).
>
>    For example, suppose you map all your scripts ending in ".pl" on your
>    web server computer to "C:\Perl\perl.exe" and you map all your scripts
>    ending in ".exe" to "cmd.exe" (wherever that lives on Windows NT -
>
>    What do you call "handler for an extension", namely perl.exe and
>    cmd.exe in the example above? Maybe "executing program" (not a very
>    nice term for a Quaker lad, but maybe OK in the industry?)

     This is referred to as the "MIME type", where MIME originally
means "Multipurpose Internet Mail Extension" and is now used in a
broader context than Mail.

http://wombat.doc.ic.ac.uk/foldoc/foldoc.cgi?query=mime+type

     Netscape's web server can be configured to handle different MIME
types, for example to recognize a MIME type by extension and to react
by doing something other than simply grabbing the file and handing the
contents of it to the browser.  I don't remember the nomenclature
offhand, check the Netscape web server manual, or their web site.

Steven J. Owens
puff -at- netcom -dot- com


 From ??? -at- ??? Sun Jan 00 00:00:00 0000=