Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uwm.edu!spool.mu.edu!snorkelwacker.mit.edu!bloom-beacon!dont-send-mail-to-path-lines
From: rws@expo.lcs.mit.EDU (Bob Scheifler)
Newsgroups: comp.windows.x
Subject: Re: xstuff: could be improved ?
Message-ID: <9103272008.AA07951@expire.lcs.mit.edu>
Date: 27 Mar 91 20:08:31 GMT
References: <1991Mar26.082611.9783@siesoft.co.uk>
Sender: tytso@athena.mit.edu (Theodore Y. Ts'o)
Organization: The Internet
Lines: 48


    I have had the same problems with xstuff, I get no response for days and
    then finally I get a bounced message with something like "expo refused
    connection for 3 days".

I don't believe that's the same problem at all, actually.

    Another whinge is the out-of-date (or incomplete) nature of the indexes
    that xstuff provides, or at least of the extremely limited number of files
    which it tells me that it has available.

I don't think xstuff is out of date.  It is extremely limited, basically
limited to public MIT fixes and Consortium public review documents.

    How much
    stuff can I actually get hold of through the xstuff server ?

Very little.

    I would
    really like to be able to get hold of things like a totally current
    ICCCM,

The R4 distribution has a "totally current ICCCM".

    I don't want to clog the mail system
    with huge volumes of source code, just keep informed on developments.

If by "developments" you mean Consortium-published documents, the xstuff
server does provide public review documents.

    I realize that xstuff is a public service facility, and is provided as
    a concession and not a right.

It is provided because the software already existed, and it was relatively
painless to get set up.

    But it would be much more useful to
    non-ftp'ers like myself either if more information was available on
    its facilities, or if it was able to provide me with a more comprehensive
    base of retrievable information.

Of that I have no doubt.  You are welcome to encourage some site that shadows
our ftp tree to make the information available by mail.  You are welcome to
write a bunch of new software to do what you desire, debug it for a while on
a real database, and make sure it is secure, that it is reasonably robust,
and that it is designed with a built-in load limiter.  Make the software
freely available, and we'll see about using it.