On 1/9/98 12:45 AM, Dromio <[log in to unmask]> wrote...

>On Thu, 8 Jan 1998 22:20:39 -0600, in message:
><[log in to unmask]>, Adam Bailey
><[log in to unmask]> wrote:
>
>> On 1/8/98 4:54 PM, Jacky Foo <[log in to unmask]> wrote...
>>
>> >----------
>> >From:   Pauline M. Marshall
>> >>For example, you can add two people with different cases ([log in to unmask]
>> >>and [log in to unmask] are two different subscribers), but you can't delete
>> >>them singly (if you ask to delete mmouse, both mmouse and MMouse get
>> >>unsubed), but, as shown below, not only does it not care about case, it
>> >>doesn't care about domain/machine names).  Aghghgh.
>> >
>> >If LISTSERV could automatically convert all user-ids to the lower case (as
>> >a standard), it would solve the problem with double subscriptions.
>>
>> No, no, no. That would be wrong. RFCs state that usernames don't have to
>> be case insensitive. On some systems, usernames ARE case sensitive.
>>
>> That's where this whole argument started.
>
>        Thank you, I think that it is WONDERFUL that listserv makes
>the domain-name insensitive (converts it to all upper case),

That's because RFCs state that the domain name is ALWAYS case
insensitive. Anyone who has a case-sensitive domain is wrong.

>but is
>this problem that has been stated actually there?  I had two aliases
>to this account ([log in to unmask], and [log in to unmask]), but
>there is nothing to say that those two aliases DO point to one
>account, those could be two entirely diffrent people, just as
>[log in to unmask] could be a diffrent person than [log in to unmask]

As Eric explained, LISTSERV (1.8d, which I've yet to have the pleasure to
work with) only does fuzzy domain matching with domains that have been
pre-registered by the ISP. For example, MSN was the first case this was
discovered for, because [log in to unmask] is the same as [log in to unmask]
So, since Microsoft informed L-Soft of this, LISTSERV now considers
[log in to unmask] and [log in to unmask] to be the same person, and will
accept commands thusly.

>Is the case sensitivity issue ignored as Pauline is sugesting, and if so
>how quickly could it be fixed?

Case sensitivity has nothing to do with domain matching. It's just that
both are parts of Pauline Marshall's argument in favor of standardization.

Right now, the whole case sensitivity thing is a minor annoyance, at
best. I'm certainly in favor of adding a switch that would specify
whether you want the address in your command to be case sensitive or not,
but it's not the first thing I'd do if I were lead developer at L-Soft.
:)


--
Adam Bailey      | Chicago, Illinois
-=-=-=-=-=-=-=-=-| "Do not take life too seriously;
[log in to unmask] | you will never get out of it alive."
[log in to unmask]   |  - Elbert Hubbard
Finger for PGP   | http://www.tezcat.com/~adamb