case-sensitive e-mail addresses
Hello All,
Please, could someone advise me if there is still an issue with using case-sensitive e-mail addresses with mailman? ie. john.doe@email.com vs. John.Doe@email.com would both get delivered? Or would the case-incorrect version bounce?
Thanks!
*Larry*
"S" == SillyGod <sillygod@hotmail.com> writes:
S> Please, could someone advise me if there is still an issue with
S> using case-sensitive e-mail addresses with mailman?
S> ie. john.doe@email.com vs. John.Doe@email.com would both get
S> delivered? Or would the case-incorrect version bounce?Mailman is case-preserving case-insensitive w.r.t. email addresses.
What this means is that you cannot subscribe both john.doe@email.com and John.Doe@email.com; they are considered the same address for memebership purposes.
However, Mailman will case-preserve the username part of the address, so that if you subscribe John.Doe@email.com, messages will be sent to John.Doe@email.com and not john.doe@email.com.
-Barry
On Wed, May 02, 2001 at 10:17:37AM -0400, Barry A. Warsaw wrote:
"S" == SillyGod <sillygod@hotmail.com> writes: S> Please, could someone advise me if there is still an issue with S> using case-sensitive e-mail addresses with mailman? S> ie. john.doe@email.com vs. John.Doe@email.com would both get S> delivered? Or would the case-incorrect version bounce?
Mailman is case-preserving case-insensitive w.r.t. email addresses.
What this means is that you cannot subscribe both john.doe@email.com and John.Doe@email.com; they are considered the same address for memebership purposes.
Ok... since I know that LHS's are supposed to be treated as if they *might* be case-sensitive, I dug further.
The controlling paragraphs are 2.4 and 4.1.2 in RFC 2821, both of which specify that a local-part may be case sensitive.
IE: oops. :-}
Cheers, -- jra
Jay R. Ashworth jra@baylink.com Member of the Technical Staff Baylink The Suncoast Freenet The Things I Think Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
"JRA" == Jay R Ashworth <jra@baylink.com> writes:
JRA> The controlling paragraphs are 2.4 and 4.1.2 in RFC 2821,
JRA> both of which specify that a local-part may be case
JRA> sensitive.
JRA> IE: oops. :-}No oops on Mailman's part! We case preserve for the benefit of the receiving smtpds, but I think it's perfectly valid for Mailman to case-fold subscription keys.
-Barry
On Wed, May 02, 2001 at 11:02:38AM -0400, Barry A. Warsaw wrote:
"JRA" == Jay R Ashworth <jra@baylink.com> writes: JRA> The controlling paragraphs are 2.4 and 4.1.2 in RFC 2821, JRA> both of which specify that a local-part may be case JRA> sensitive.
JRA> IE: oops. :-}No oops on Mailman's part! We case preserve for the benefit of the receiving smtpds, but I think it's perfectly valid for Mailman to case-fold subscription keys.
I'm afraid I must disagree: that section of 2821 implies that
John.Doe@example.com and john.doe@example.com
*may be separate mailboxes*. Admittedly, it's a corner case, and very unlikely ever to bite us, but doing it that was *does* violate the standard.
Cheers, -- jra
Jay R. Ashworth jra@baylink.com Member of the Technical Staff Baylink The Suncoast Freenet The Things I Think Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
"JRA" == Jay R Ashworth <jra@baylink.com> writes:
JRA> I'm afraid I must disagree: that section of 2821 implies that
| John.Doe@example.com
| and
| john.doe@example.com
JRA> *may be separate mailboxes*.I understand that.
JRA> Admittedly, it's a corner case, and very unlikely ever to
JRA> bite us, but doing it that was *does* violate the standard.Except that Mailman isn't an SMTP server, so it doesn't have to play by those rules. For a list server, I claim that allowing keys to differ only by case is confusing and needless generalization. In practice, I don't remember it ever biting us.
-Barry
On Wed, May 02, 2001 at 11:10:31AM -0400, Barry A. Warsaw wrote:
"JRA" == Jay R Ashworth <jra@baylink.com> writes: JRA> I'm afraid I must disagree: that section of 2821 implies that
| John.Doe@example.com | and | john.doe@example.com JRA> *may be separate mailboxes*.I understand that.
JRA> Admittedly, it's a corner case, and very unlikely ever to JRA> bite us, but doing it that was *does* violate the standard.Except that Mailman isn't an SMTP server, so it doesn't have to play by those rules. For a list server, I claim that allowing keys to differ only by case is confusing and needless generalization. In practice, I don't remember it ever biting us.
No, but Mailman *is dealing with RFC[2]82{1,2} addresses, and ought to play by their rules. As I noted, it *is* a corner case; anyone who actually *does* that on a machine ought to be shot...
and I wouldn't expect us to be bitten by it; being a pedant is just my *job*, damnit.
Cheers, -- jr 'actually, it's an adventure' a
Jay R. Ashworth jra@baylink.com Member of the Technical Staff Baylink The Suncoast Freenet The Things I Think Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
"JRA" == Jay R Ashworth <jra@baylink.com> writes:
JRA> No, but Mailman *is dealing with RFC[2]82{1,2} addresses, and
JRA> ought to play by their rules. As I noted, it *is* a corner
JRA> case; anyone who actually *does* that on a machine ought to
JRA> be shot...
JRA> and I wouldn't expect us to be bitten by it; being a pedant
JRA> is just my *job*, damnit.:)
On Wed, May 02, 2001 at 11:32:51AM -0400, Barry A. Warsaw wrote:
JRA> and I wouldn't expect us to be bitten by it; being a pedant JRA> is just my *job*, damnit.:)
That's an excellent attitude, Barry. I think we'll get along just fine...
:-)
Cheers, -- jra
Jay R. Ashworth jra@baylink.com Member of the Technical Staff Baylink The Suncoast Freenet The Things I Think Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
On Wed, May 02, 2001 at 11:13:25AM -0400, Jay R. Ashworth wrote:
and I wouldn't expect us to be bitten by it; being a pedant is just my *job*, damnit.
Don't worry, you're not alone. I remember some pretty frustrating pedantic discussions where I had trouble convincing people I really *was* a nice guy, but I still stood by my pedantic point. Some of them were on python-dev, even :)
-- Thomas Wouters <thomas@xs4all.net>
Hi! I'm a .signature virus! copy me into your .signature file to help me spread!
On 5/2/01 8:05 AM, "Jay R. Ashworth" <jra@baylink.com> wrote:
I'm afraid I must disagree: that section of 2821 implies that
John.Doe@example.com and john.doe@example.com
*may be separate mailboxes*. Admittedly, it's a corner case, and very unlikely ever to bite us, but doing it that was *does* violate the standard.
Sorry, I disagree. "may be" means the standard allows you to consider this as optional. That's good, because 821 made case sensistivity mandatory (and the switch shows the clear intent of the standards folks, which is moving to full case insensitivity, but they must have felt completely removing it was too big a step -- but they're telling anyone who might still have a case sensitive mailer it's time to fix it)
In practice, Barry's approach is the best one. The only thing that happens if you practice case sensitivity on the user part is that users end up with multiple subscriptions, multiple copies and no clue how to fix it or what's wrong (and it's your fault, your mailer is screwed up).
The rewrite of the standard makes it clear this kind of situation (John.Doe and john.doe being different people) is now optional -- and I don't know that any MLM would work cleanly if it happens (and any admin that allows it should be shot...)
On Wed, May 02, 2001 at 08:26:59AM -0700, Chuq Von Rospach wrote:
On 5/2/01 8:05 AM, "Jay R. Ashworth" <jra@baylink.com> wrote:
I'm afraid I must disagree: that section of 2821 implies that
John.Doe@example.com and john.doe@example.com
*may be separate mailboxes*. Admittedly, it's a corner case, and very unlikely ever to bite us, but doing it that was *does* violate the standard.
Sorry, I disagree. "may be" means the standard allows you to consider this as optional. That's good, because 821 made case sensistivity mandatory (and the switch shows the clear intent of the standards folks, which is moving to full case insensitivity, but they must have felt completely removing it was too big a step -- but they're telling anyone who might still have a case sensitive mailer it's time to fix it)
Hmmm... Ok. If *you* think that, Chuq, I'll re-evaulate my opinion. I hadn't compared it with 821. I'm wondering why, though, they didn't make that explicit in the rewrite.
In practice, Barry's approach is the best one. The only thing that happens if you practice case sensitivity on the user part is that users end up with multiple subscriptions, multiple copies and no clue how to fix it or what's wrong (and it's your fault, your mailer is screwed up).
Got it.
The rewrite of the standard makes it clear this kind of situation (John.Doe and john.doe being different people) is now optional -- and I don't know that any MLM would work cleanly if it happens (and any admin that allows it should be shot...)
I believe I *did* evince *that* opinion, at least... :-)
Cheers, -- jra
Jay R. Ashworth jra@baylink.com Member of the Technical Staff Baylink The Suncoast Freenet The Things I Think Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
On 5/2/01 8:30 AM, "Jay R. Ashworth" <jra@baylink.com> wrote:
Hmmm... Ok. If *you* think that, Chuq, I'll re-evaulate my opinion. I hadn't compared it with 821. I'm wondering why, though, they didn't make that explicit in the rewrite.
I can think of two reasons:
they're really codifying existing practices -- nobody was paying attention to the case sensitivity issue anyway, but in a nod to the previous standard, they didn't want to throw it out completely. It allows them to get it out of the way while still claiming (for the most part) backwards compatibility.
they wanted it in the standard and out the door before the fight started... (grin)
In my big muther custom server, I squash everything to LC (and in fact, since my backend data store is mysql, making things case sensitive would require some significant work). I've found no cases (not few -- none) where this has caused a problem, and in fact the only time case issues come up at all is when an AOL user can't find their subscription and emails me for help and says something like "if you can't find juser@aol.com, try Juser..."
I probably shouldn't squash case, really, but it simplifies dealing with this stuff somewhat, and I've found from watching the search strings that most users don't try to maintain case in their lookups anyway... Except from AOL, where it's cosmetic...
I haven't really pawed through the new standard in detail, but from what I've seen, I think the intent is to say that case sensitivity is optional ("may be", not "must be"), and I think that implies that if you want to write stuff that's case sensitive on your local system, that's fine, but you can no longer assume it'll work once it leaves your control. Which is, in practice, how it's been for many years...
On Wed, May 02, 2001 at 08:43:17AM -0700, Chuq Von Rospach wrote:
Hmmm... Ok. If *you* think that, Chuq, I'll re-evaulate my opinion. I hadn't compared it with 821. I'm wondering why, though, they didn't make that explicit in the rewrite.
I can think of two reasons:
- they're really codifying existing practices -- nobody was paying attention to the case sensitivity issue anyway, but in a nod to the previous standard, they didn't want to throw it out completely. It allows them to get it out of the way while still claiming (for the most part) backwards compatibility.
But is this actually true? I haven't tried it, but I thought sendmail still treated LHS's as case-sensitive...
- they wanted it in the standard and out the door before the fight started... (grin)
<chuckle>
In my big muther custom server, I squash everything to LC (and in fact, since my backend data store is mysql, making things case sensitive would require some significant work). I've found no cases (not few -- none) where this has caused a problem, and in fact the only time case issues come up at all is when an AOL user can't find their subscription and emails me for help and says something like "if you can't find juser@aol.com, try Juser..."
Hmmm...
I probably shouldn't squash case, really, but it simplifies dealing with this stuff somewhat, and I've found from watching the search strings that most users don't try to maintain case in their lookups anyway... Except from AOL, where it's cosmetic...
Indeed.
I haven't really pawed through the new standard in detail, but from what I've seen, I think the intent is to say that case sensitivity is optional ("may be", not "must be"), and I think that implies that if you want to write stuff that's case sensitive on your local system, that's fine, but you can no longer assume it'll work once it leaves your control. Which is, in practice, how it's been for many years...
Hmmm... *I* interpreted it to mean "someone out there might be treating this as case-sensitive, so you ought to be careful not to break their shit."
But I can see your interpretation as well..
Cheers, -- jr 'be ye not overly annoying... nor too easily annoyed' a
Jay R. Ashworth jra@baylink.com Member of the Technical Staff Baylink The Suncoast Freenet The Things I Think Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
On 5/2/01 8:50 AM, "Jay R. Ashworth" <jra@baylink.com> wrote:
But is this actually true? I haven't tried it, but I thought sendmail still treated LHS's as case-sensitive...
Not to my knowledge, and I just tested it on my system and it's not, and my config should be standard for this. A quick grep of the docs doesn't show me any documentation at all on case sensitivity, so it looks like they removed it completely at some point -- I know at one point there was an option to turn it on; that seems gone.
Hmmm... *I* interpreted it to mean "someone out there might be treating this as case-sensitive, so you ought to be careful not to break their shit."
Not unreasonable, but compare it to the old definition. It used to be case sensitive, even though we all ignored it. Now, it may be case sensitive. The intent is clearly to make it non-mandatory. That means you can't depend on it being allowed unless you control the e-mail completely. And that means while I shouldn't arbitrarily do weird things to it simply because I can, you can no longer assume I WON'T do weird things to it, because the standard no longer tells me not to...
On Wed, May 02, 2001 at 07:59:20AM -0400, SillyGod wrote:
Please, could someone advise me if there is still an issue with using case-sensitive e-mail addresses with mailman? ie. john.doe@email.com vs. John.Doe@email.com would both get delivered? Or would the case-incorrect version bounce?
The *right hand side* of a mailbox name (which is what we're actually talking about here) is case-insensitive because it's a DNS name, and *DNS*, in turn, is case-insensitive.
The *left hand side* (the "username"), OTOH, must be treated as case-sensitive, because whether it actually *is*, or not, is system-dependent... and the "system" in question is the final target mail server -- in general, I believe that Unix machines are case-sensitive and Windows ones are not, as is usually the case, but I don't know Exchange, nor non-Sendmail Unix MTA's, well enough to guarantee that.
Interestingly, RFC 2822 appears silent on this point, making the *third* thing this week it's been silent on to my surprise.
Hmmm...
Cheers, -- jra
Jay R. Ashworth jra@baylink.com Member of the Technical Staff Baylink The Suncoast Freenet The Things I Think Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
participants (5)
-
barry＠digicool.com -
Chuq Von Rospach -
Jay R. Ashworth -
SillyGod -
Thomas Wouters