William H. Magill on 13 May 2005 19:55:46 -0000

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] POP and leave-on-server

On 12 May, 2005, at 22:09, Stephen Gran wrote:
On Thu, May 12, 2005 at 09:48:31PM -0400, Jeff Abrahamson said:

In a private discussion, a flame war erupted concerning the "leave on
server" option that many POP clients offer.  The other party
maintained that mail should initially be left on the server in case
something happens to the client.  He concluded that this was the
reason for the feature.

I opined that the feature was likely for use by those who ran more
than one POP client but only wanted one personal mail store.  I was
not able to justify this assertion, but suggested (with link to the
rfc) that, since the protocol is stateful, it is unlikely to suffer
from reliability problems that would cause one to loose mail by
issuing DELE's after RETR's.

Does anyone here know more of the history? I imagine experience with
earlier versions of POP might elucidate this, as web mail clients were
not popular in the old days before the web.

The references I can find immediately on google suggest to me that you are both correct, and IMAP came into regular use for both of these reasons. I see references that claim that mail is best handled on a central server (backups, security, etc) and other references that are talking about the problems of off-line management of mail for multiple users/machines and a single account.

POP is unlikely to suffer from data loss, judging from the protocol
specification, so you are correct there.  I think it's mostly that POP
was around first, and there are a number of shortcomings in the
protocol, including but not limited to the ones you two mention, and
these shortcomings lead to IMAP.

This is the basic history, (and how we implemented it at Penn many, many years ago).

Back in the NSFnet days, before the Internet was invented by Al Gore in 1995, POP was around first. It was first proposed for standard net- wide use circa 1989 by Tim Burners-Lee.

[RFC 918 defining the protocol was published in 1984. POP V2, RFC 937 in 1985, and RFC 1081 in 1988. "RFC 1081 was obsoleted by, in turn, RFCs 1225, 1460, 1725 and 1939. Despite the large number of revisions, the protocol itself has not changed a great deal since 1988; these RFCs contain only relatively minor tweaks to the original description of the protocol. RFC 1939 was published in 1996 and POP3 has not been revised since that time, though a few subsequent RFCs define optional extensions and additions to the basic protocol, such as alternative authentication mechanisms."]

POP was developed in reaction to the "client server" model of computing, popular in the 1980s, in reaction to the "host - terminal" model of the past.

While many people now had actual "Personal Computers" on their desks at work in the 1980s,
there were still very few people who had machines at home. The explosive growth of the Home PC was occurring in the early 1990s. ... "Portable" computers were still a joke, "Laptops," did not yet exist. The individuals "Personal Computer" was the client, the central main- frame was the "server."

In the mid 1980s, some of us in the education community DID have multiple machines, in the office, in the Lab, at home ... and POP's downloading mail ton ONE only was considered a MAJOR problem. The IMAP protocol was defined at Stanford in the mid 1990s to address the issue of "mobile computing." ["Webmail" has since been developed as "yet another solution" to the problem of "mobile computing" ... by going back to the original host-terminal model of computing, but with a browser based GUI instead of a CURSES based character cell device.]

The first version of IMAP formally documented as an Internet standard was IMAP version 2 (IMAP2) in RFC 1064, published July 1988.

Implementation of the IMAP Protocol was "spotty" at best. Mark Crispin at the University of Washington wrote what became the "standard" implementation somewhere in the late 1980s. While it was implemented at the University of Washington, it took several years before it became widely adopted by others, POP still being the "preferred" method. Penn didn't implement IMAP until the mid 1990s, and even then, POP was still the "standard" protocol as the IMAP server was anything but robust at that time.

William H. Magill
# Beige G3 [Rev A motherboard - 300 MHz 768 Meg] OS X 10.2.8
# Flat-panel iMac (2.1) [800MHz - Super Drive - 768 Meg] OS X 10.3.8
# PWS433a [Alpha 21164 Rev 7.2 (EV56)- 64 Meg] Tru64 5.1a
# XP1000  [Alpha 21264-3 (EV6) - 256 meg] FreeBSD 5.3
# XP1000  [Alpha 21264-A (EV 6.7) - 384 meg] FreeBSD 5.3

___________________________________________________________________________ Philadelphia Linux Users Group -- http://www.phillylinux.org Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce General Discussion -- http://lists.phillylinux.org/mailman/listinfo/plug