Talk:Postfix (software)

Security

Curiously, the word "security" never appears in article as per January 2018. Starting with Daniel J. Bernstein's comments at https://cr.yp.to/maildisasters/postfix.html, Postfix's security track record might be worth at least a brief mention. --Arto (talk) 06:08, 19 January 2018 (UTC)

License

The main reason Postfix has not seen more widespread use is because it is not licensed under the GPL or BSD license. In fact the official website includes no information about its licensing scheme (Oct. 2003) in recognition of this weakness.

Is there any evidence for this complex and worried-sounding claim? Postfix is distributed under the IBM Secure Mailer license, which is BSD-like. This license's chief distinctions from BSD are its more extensive warranty disclaimers, and its liability requirements for commercial distributors. There is no sense in which this license is a "weakness" as far as I can tell, and the chief reasons that it isn't posted loudly on the Web site are (1) Postfix isn't ideological about licenses, and (2) the Web site is disorderly. I'm going to remove this claim if it can't be backed up. --FOo 14:31, 23 Dec 2003 (UTC)

As promised, I have removed the claim. With Postfix distributed on all Macintosh sytsems (and with Apple having become one of the major Unix vendors) there isn't even any basis to say that Postfix isn't "widespread"! --FOo 14:26, 14 Jan 2004 (UTC)

It's also distributed with NetBSD as an alternative MTA to sendmail. Though sendmail is still the default. --B1ff 13:09, 17 July 2005 (UTC)

I just found recently that...

There was a very recent Wikipedia article that changed "postfixes" to "suffixes". Well, who would ever use "postfix"?? 66.245.110.129 01:45, 14 May 2004 (UTC)

Masquerade domain in postfix

I just merged in material from Masquerade domain in postfix. You might want to take it back out...it doesn't look hugely encyclopedic. NickelShoe 21:05, 12 January 2006 (UTC)

News

How come there is no news page detailing releases, not even a dedicated mailing list? 219.79.235.121 14:06, 29 October 2007 (UTC)

Buffer overflows

"One of the strengths of Postfix is its resilience against buffer overflows." Is there some specific reason for this in the way Postfix is implemented or is it just an empirical observation? As the article also says "Postfix's source code is often used as a famous example of good programming practice," it would be interesting to elaborate a bit. -- Coffee2theorems 11:28, 11 November 2007 (UTC)

Does the article still need more citations for verification?

The article has a warning at the top that says that the text needs more citations to reliable source for verification. In recent months I restructured the text and added a ton of links. At this point I wonder what is fundamentally missing at this point. Any discussion of the characteristics of the Postfix software, its strengths and weaknesses, etc., will ultimately have to be supported by empirical, accurate, first-hand, observations. What are your criteria for establishing the reliability of those sources? Wietse venema (talk) 00:59, 24 November 2012 (UTC)

SMTP

I would like to discuss recent changes by user 123.243.137.139, this user repeatedly removes SMTP compliance of Postfix without sources [1]. I must also mention his personal attacks in the comments. So I am opening a discussion about this here. Wolscmip (talk) 07:36, 17 May 2013 (UTC)

Source for the origin of the name

If anyone is interested in adding it back (not sure how relevant it is) with a source, https://marc.info/?l=postfix-users&m=164841848519002&w=2 is probably as authoritative as you'll ever get. The Crab Who Played With The Sea (talk) 23:46, 28 March 2022 (UTC)

Relevance of release history?

The Release history section of this article is something like 80% of its vertical length and 60% of its bytes (YMMV of course with screen size), and yet it contains almost no information – plenty of data, but no context for it. Wikipedia is not a changelog, and I don't see it belonging here. This isn't the place to record that version 1.1.8 was released on 2002-05-04, and 1.1.9 was released eight days later, or that versions 3.4.2 and 3.4.3 were released on the same day (2019-03-10), especially when the majority of the "significant changes" column is empty (implying that the changes between versions are insignificant – so why record them? ;)). This kind of versioning information belongs on Postfix's own changelog.

I'm going to go ahead and just remove it. If it turns out to have be beloved, we can discuss options, like maybe a much more summarized table with only x.y releases instead of every x.y.z release. oatco (talk) 23:16, 22 October 2023 (UTC)