CERN Accelerating science

Previous:		The Microsoft Mail Migration in CERN	(See printing version)
Next:		Dublin Core Metadata

What happened so far

We can now move towards a stable and consistent set of URLs.

Which server does what

wwwinfo.cern.ch

is the server hosting the Web pages describing the work of the IT groups individually. The IT top page leads you to the groups' top pages from which you can learn about their mandate, their staff and the detailed descriptions of their activities and projects. (In about half the cases the URLs are of the form wwwinfo.cern.ch/groupname).

-> wwwinfo.cern.ch is oriented towards groups

A comment on its name: the original name of wwwinfo was wwwcn. This name had to be given up two years ago when the name of our division changed. Having chosen the rather "pale" name of wwwinfo we are confident that we can keep it stable.

consult.cern.ch

The main purpose of consult is to facilitate access to relevant information independent of its origin. The user does not have to know which person or group (later on even division) is responsible for a given service or information.

consult does not necessarily host (but can do so at the request of the responsible) all the information about a given subject, but helps the user where to start.

The following may make it clearer what to expect from the consult server:

• Its top page shows which areas it covers, explicitly.
• The quality of what you can access via consult is kept (hopefully) high by a very restrictive mechanism of submission. An exception is news where only the rules for the news group itself apply.
• In a world of sometimes very complicated URLs, those served by consult are easy to remember and sometimes even "guessable", at least for the first two levels: URLs like consult/service/mail or consult/writeup/garfield can even be transmitted by phone! You may also observe that the URLs are always lower case.
• Finding items is made easy by the two search engines xwho (people and services) and xfind (cnls, question&answers, writeups etc.) as well as via structured lists.

-> consult.cern.ch is oriented towards users

A comment on its name: consult was always called consult, but as it could not be distinguished from wwwinfo under the old server-software it had a very low visibility in the past, and was even neglected. Now, with its independent existence, it will be maintained increasingly well. Look out for an article in the next CNL, describing its marvels in more detail, and above all what makes it different from a normal search engine.

home.cern.ch

serves what are called personal home pages. For this server also, which could not have its own identity, major improvements are foreseen. Look at its home page to get a description on how to set up your home page.

-> home.cern.ch is oriented towards individuals

What will happen during January 1999, and afterwards

Phasing out the old server names

The server name wwwcn, as its usage is still relatively high (15% of all requests), will be given six months more, i.e. will "survive" until end of June 1999. Two measures will be taken nevertheless:

1. All requests for wwwcn are redirected to wwwinfo (which delays the requests and puts more work onto the server)
2. The server code for a wwwcn request is artificially set to 404, i.e. a program (robot) will think that it did not get the page)

Hardware Upgrade

For matters related to this article please contact the author.