psrc@mtunq.ATT.COM (Paul S. R. Chisholm) (08/20/90)
In article <1990Aug19.205508.16479@mthvax.cs.miami.edu> wb8foz@mthvax.cs.miami.edu (David Lesher) writes: >Does anyone else out there have a hard time sending replies to att-ites? > 1) Article is posted from "bell@cbnews.att.com (alex.g.bell)" > 2) I attempt to reply to that address > 3) My reply bounces, for various reasons, at cbnews. >All this comes about because, of course, Alex does not REALLY have an >account on cbnews, but instead on another att machine that lets cbnews >do all the postings. I'm surprised the reply doesn't work. Another way to answer Alex would be to send e-mail to alex.g.bell@att.com; Bell's name would be looked up in the AT&T internal e-mail directory, and forwarded to the e-mail address listed there. (It's impossible to look up information in the directory and have the result sent outside of AT&T; that's both a security and a privacy feature.) The middle name or initial is optional. Paul S. R. Chisholm, AT&T Bell Laboratories (paul.s.r.chisholm@att.com) att!mtunq!psrc, psrc@mtunq.att.com, AT&T Mail !psrchisholm I'm not speaking for the company, I'm just speaking my mind.
david@twg.com (David S. Herron) (08/23/90)
In article <1255@mtunq.ATT.COM> psrc@mtunq.ATT.COM (Paul S. R. Chisholm) writes: >In article <1990Aug19.205508.16479@mthvax.cs.miami.edu> wb8foz@mthvax.cs.miami.edu (David Lesher) writes: >>Does anyone else out there have a hard time sending replies to att-ites? >> 1) Article is posted from "bell@cbnews.att.com (alex.g.bell)" >> 2) I attempt to reply to that address >> 3) My reply bounces, for various reasons, at cbnews. >>All this comes about because, of course, Alex does not REALLY have an >>account on cbnews, but instead on another att machine that lets cbnews >>do all the postings. >I'm surprised the reply doesn't work. Another way to answer Alex would >be to send e-mail to alex.g.bell@att.com; Listen.. by putting out e-mail headers with addresses like that you are claiming to be able to handle mail addressed like so. ergo-cum-quid cbnews.att.com is *BROKEN* Saying that doing something unintuitive and specific-to-att.com is a very bad way to handle the situation. Mr. Lesher described the normal way in which people try to reply to mail. What he tried doing is the way it is supposed to work. This is why I say that cbnews.att.com is *BROKEN*. > Bell's name would be looked >up in the AT&T internal e-mail directory, and forwarded to the e-mail >address listed there. That's fine. That's wonderful. I don't like that attempt to toss all those people into one pool of names, but AT&T can do as they wish. Past experience with host names is that it is a very bad thing to try to use just one pool of names -- that seperating them into hierarchies is much better. My feeling is that the same holds true for the local-part of e-mail addresses. If the header were to read "From: alex.g.bell@att.com" then it would work. >(It's impossible to look up information in the >directory and have the result sent outside of AT&T; that's both a >security and a privacy feature.) The middle name or initial is >optional. I'm curious.. assumably your directory will eventually be attached to an X.500 server. One of the features which will likely be in X.400 user-agents is a hook into X.500 directories so that if you don't know specific details about somebody you can go querying around to get them so you can fill in attributes in the X.400 headers. With the sort of policy you claim AT&T has, how will this work? -- <- David Herron, an MMDF weenie, <david@twg.com> <- Formerly: David Herron -- NonResident E-Mail Hack <david@ms.uky.edu> <- <- Sign me up for one "I survived Jaka's Story" T-shirt!
anagram@desire.wright.edu ((For Mongo)) (08/24/90)
In article <7801@gollum.twg.com>, david@twg.com (David S. Herron) writes: > In article <1255@mtunq.ATT.COM> psrc@mtunq.ATT.COM (Paul S. R. Chisholm) writes: >>I'm surprised the reply doesn't work. Another way to answer Alex would >>be to send e-mail to alex.g.bell@att.com; > > Listen.. by putting out e-mail headers with addresses like that > you are claiming to be able to handle mail addressed like so. And so they can, usually. This sounds like a classic case of computer error. Even the computers make mistakes sometimes. > ergo-cum-quid cbnews.att.com is *BROKEN* > Not really. Gliched maybe, but not broken. I've replied to many that way, and never had a problem. A
mark@cbnewsu.att.com (Mark Horton) (08/31/90)
In article <7801@gollum.twg.com>, david@twg.com (David S. Herron) writes: > Listen.. by putting out e-mail headers with addresses like that > you are claiming to be able to handle mail addressed like so. > ergo-cum-quid cbnews.att.com is *BROKEN* > > Saying that doing something unintuitive and specific-to-att.com is > a very bad way to handle the situation. Mr. Lesher described the > normal way in which people try to reply to mail. What he tried > doing is the way it is supposed to work. This is why I say that > cbnews.att.com is *BROKEN*. If you reply to this article, I wager to say the reply will get through. cbnews has to deal with 200,000 AT&T employees in a company that is reorganizing at an amazing rate. It's inevitable that a few users are going to have some problems. The system as a whole is not broken, anymore than TWG would be broken if you suddenly left (and your account were deleted) or moved and then I tried to reply to something you posted right before you left. We have determined that the error rate recently went beyond our acceptable standards of quality and identified the cause. We have a fix designed and expect to install the fix next week after appropriate testing. > If the header were to read "From: alex.g.bell@att.com" then > it would work. Excellent idea, but it runs into trouble for all the John Smith's or David Jones in a 200,000 person company. We're looking into solutions for this, but we aren't there yet. > >(It's impossible to look up information in the > >directory and have the result sent outside of AT&T; that's both a > >security and a privacy feature.) The middle name or initial is > >optional. > > I'm curious.. assumably your directory will eventually be > attached to an X.500 server. One of the features which will > likely be in X.400 user-agents is a hook into X.500 directories > so that if you don't know specific details about somebody > you can go querying around to get them so you can fill in > attributes in the X.400 headers. > > With the sort of policy you claim AT&T has, how will this work? Simple. We will deliver mail, and even help fill out the form for a specific user. What we don't do is let you browse the directory to accumulate a list of employees. Mark
emv@math.lsa.umich.edu (Edward Vielmetti) (09/02/90)
There is a rumor in the Computer Underground Digest (alt.society.cu-digest, v2 n1,) from "Anonymous, somewhere in Texas" that AT&T provided Secret Service agents with access to 'attctc', nee 'killer', for the purpose of monitoring the hacker community. Read the whole digest (article <1537@chinacat.Unicom.COM>) for more information. Anyone from AT&T here prepared to reply to this one? --Ed Edward Vielmetti, U of Michigan math dept <emv@math.lsa.umich.edu>
david@twg.com (David S. Herron) (09/03/90)
In article <1990Aug30.183437.11916@cbnewsu.att.com> mark@cbnewsu.att.com (Mark Horton) writes: >In article <7801@gollum.twg.com>, david@twg.com (David S. Herron) writes: >> Listen.. by putting out e-mail headers with addresses like that >> you are claiming to be able to handle mail addressed like so. >> ergo-cum-quid cbnews.att.com is *BROKEN* [...] >We have determined that the error rate recently went beyond our >acceptable standards of quality and identified the cause. We have a >fix designed and expect to install the fix next week after appropriate >testing. Ok. Sorry. I overreacted. I'd seen a claim from somebody that a particular address didn't work. I've also seen signatures from cbnewsx.att.com users directing followups to some other host. And I've also had at least one problem mailing to a cbnewsx.att.com address. All of which led me to believe that things were "*BROKEN*". >> If the header were to read "From: alex.g.bell@att.com" then >> it would work. > >Excellent idea, but it runs into trouble for all the John Smith's >or David Jones in a 200,000 person company. We're looking into >solutions for this, but we aren't there yet. er.. yeah.. And you run into the same trouble with the "attmail" service you're already running. That there's only one pile of names and it's rather impossible to avoid collisions in a large enough pile of names. Leading to Harris's Lament: All the good ones are taken. Eh? :-) I see that problem as being very similar to host naming problems that we've seen on ARPAnet, UUCP and BITNET. Each one ran into problems when the host base reached the 3-5,000 range. [Long question about X.500 queries over network in order to fill out e-mail headers and the like] >Simple. We will deliver mail, and even help fill out the form for >a specific user. What we don't do is let you browse the directory >to accumulate a list of employees. > > Mark I read this as saying that particular non-AT&T people will have to be authorized in some way or another for particular queries and that you won't answer to open ended queries? -- <- David Herron, an MMDF & WIN/MHS guy, <david@twg.com> <- Formerly: David Herron -- NonResident E-Mail Hack <david@ms.uky.edu> <- <- Sign me up for one "I survived Jaka's Story" T-shirt!
psrc@cbnewsl.att.com (Paul S. R. Chisholm) (09/05/90)
What has come before: David S. Herron (david@twg.com, who posts from gollum.twg.com [which is sort of relevant to this discussion, if you think about it]) complained that replies to AT&T users posting from the cbnews systems didn't work. Both Mark Horton (mark@cbfsa.att.com, who posts from cbnewsu.att.com) and I (psrc@mtunq.att.com, who used to post from there, but now post from cbnewsl.att.com) recommended, as a workaround, sending e-mail to a human name at the logical AT&T gateway (e.g., alex.g.bell@att.com). The real solution is to get cbnews messages correctly forwarded. They work in general; Mark and the other members of the AT&T newnews team are working on fixing the remaining cases. My personal workaround is to add a Reply-To: line to my articles. In article <7801@gollum.twg.com>, david@twg.com (David S. Herron) writes: > If the header were to read "From: alex.g.bell@att.com" then it would > work. In article <1990Aug30.183437.11916@cbnewsu.att.com> mark@cbnewsu.att.com (Mark Horton) writes: > Excellent idea, but it runs into trouble for all the John Smith's > or David Jones in a 200,000 person company. We're looking into > solutions for this, but we aren't there yet. In article <7878@gollum.twg.com>, david@twg.com (David S. Herron) writes: > er.. yeah.. And you run into the same trouble with the "attmail" > service you're already running. That there's only one pile of names > and it's rather impossible to avoid collisions in a large enough > pile of names. Human name addressing is convenient . . . but it's not sufficient to handle all cases. If you send mail to attmail!john.smith, I think you'll get a complaint that your mail was undeliverable, a list of all the John Smith's, and your original message. I assume something similar works for john.smith@att.com. In either case, e-mail from John Smith should have a reasonably good From: line, with a "real" e-mail address, not one based on human names. (AT&T Mail will use a bangish address, e.g., From: attmail!jqsmith; it doesn't support DNS addresses. I've talked the the Appropriate People in AT&T Mail, and don't see DNS support any time in the immediate future.) > [Long question about X.500 queries over network in order to fill out > e-mail headers and the like] Mark: > Simple. We will deliver mail, and even help fill out the form for > a specific user. What we don't do is let you browse the directory > to accumulate a list of employees. David: > I read this as saying that particular non-AT&T people will have to be > authorized in some way or another for particular queries and that you > won't answer to open ended queries? Darned tootin', and I hope it stays that way. We *don't* want to give Joe Random a list of every member of the technical staff, with his or her title, address, phone number, and e-mail address. I can always tell when someone sells a (proprietary) AT&T directory to a headhunter or a stock broker; the phones start ringing in alphabetical order. It's a pain in the tush. Human name addressing, fine; invitations for harassment (or distraction from AT&T business) during working hours, no thanks! Paul S. R. Chisholm, AT&T Bell Laboratories att!mtunq!psrc, psrc@mtunq.att.com, AT&T Mail !psrchisholm I'm not speaking for the company, I'm just speaking my mind.