On Tue, Feb 05, 2002 at 05:32:44PM -0800, Jacqueline Craig wrote:
> Tom,
>
> Here are just a couple of early comments.
>
> I have some concerns about the subjective comments in the opening
> paragraphs: "Most universities..." Some disagree with this claim.
> How about "Many universitites..."?
I would be reluctant to weaken this statement. I have looked at directories
for at least 30 other major universities, including most of the large
research institutions comparable to Berkeley. So far I have found three,
outside of Berkeley, who don't appear to provide top-level addresses; UCSC
(which uses "cats.ucsc.edu"), UTexas ("mail.utexas.edu"), and UPenn, which
appears factionalized, and probably is, due to its organizational structure
which is more modeled on the loose-knit English universities than the newer
American campus.
All of the others--Columbia, Cornell, UNC, UMich, UVA, Penn State, Harvard,
Yale, Stanford, USC, CalTech, MIT, and all the UC campuses except Berkeley
and Santa Cruz, provide at least some addresses at the top level, and most
of them have the bulk of their addresses at the top level.
(Thinking about UPenn got me looking at Oxford and Cambridge--Oxford,
not surprisingly, has addresses at the college level, but Cambridge has
cam.ac.uk for everyone).
I think the debate, if there is to be one, should be between "Most
universities..." and "Nearly all universities..."
(My information was gathered by going to the schools' online directories
and looking up common names).
> "There is an implication of factionalism." -- I disagree -- I'm not
> comfortable with this sentence. I think the argument that @berkele.edu
> simplifies email address is a sufficiently good reason without reaching
> for other subjective claims in support of this service.
I'd be willing to remove this if people think the support for the idea
is still strong enough without it. We should talk about it on Monday.
> Under "Name retention" -- shouldn't we recommend a minimum length of
> time that names would be retained before another can use them? I
> understand that LDAP entries have a termination date some 6 months after
> the individual's information is no longer in employee/student data.
> Would this be the length of retention? Also, what is the intended
> meaning of the term "deprecated" in this context?
I probably should expand that point.
The idea is that if the name doesn't exist from a mail perspective
(if you get a standard "user unknown" bounce message when you mail to it),
it should be considered available in the namespace if someone else wants it.
But we might not want to expire names as rapidly as CalNet does, because
what happens when, for example, jlee@berkeley.edu graduates? If we shut
off his forwarding after 6 months and then make the name available, someone
will snap it up and then start getting a lot of mail for the old jlee.
So, we might want to have a period of time where the name is
"deprecated"; maybe instead of forwarding mail, we send back a bounce
message that says "this person is no longer affiliated with the university
and his last known address is foo@hotmail.com", or maybe a pointer to the
alumni community for students.
The crux of the recommendation is that if we're not providing any service to
the name, the name should be available to campus users. I'll see what I can
do to explain it better without being too long-winded.
> The current status of the student email account policy is that all
> students should have an email account (uclink or other is acceptable),
> that they must provide the campus with the email address and to keep
> it current in the campus online directory.
>
> This isn't an Undergraduate Affairs proposal, but one that is coming
> out of various committees -- it remains to be seen how it will resolve,
> but I think you should change the bullet "Undergraduate Affairs is
> considering..." to something like:
> Current policy discussions are proposing that all students should have
> an email account (uclink or other is acceptable) and that students must
> maintain their current email address in the campus online directory.
> if you think it necessary at all to include this information.
If the extent of the proposal is that students must maintain a valid e-mail
address, I don't think we need to include a recommendation about it; it will
fit into our scheme the same way as any other account assignment scheme.
It's only if student e-mail accounts are auto-assigned (and therefore
unfriendly) that I think we have anything to say about it. My understanding
was that this is one of the proposals on the table--is that no longer true?
> In the section under "Ongoing workload" --
> I ask that you remove the phrase "... complaints about inappropriate
> names." This is a non-issue now with uclink, and I wouldn't expect it
> to become one with this new service -- it's not worth mentioning. Also,
> I would certainly hope that UAS would *not* experience a 10-20% increase
> in workload. I would put the figure at less than 5%.
Changed.
> Typos: unfriendlyness > unfriendliness
> publicise > publicize
British spelling!
> Thanks for pulling together a comprehensive report!
And thank you for your comments.
-- Tom Holub (tom_holub@LS.Berkeley.EDU, 510-642-9069) College of Letters & Science 249 Campbell Hall
This archive was generated by hypermail 2b29 : Tue Feb 05 2002 - 18:32:43 PST