Talk:Qmail

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Linux (Rated Start-class)
WikiProject icon This article is within the scope of WikiProject Linux, a collaborative effort to improve the coverage of Linux on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.
 
WikiProject Software / Computing  (Rated Start-class)
WikiProject icon This article is within the scope of WikiProject Software, a collaborative effort to improve the coverage of software on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.
Taskforce icon
This article is supported by WikiProject Computing.
 

Wrong Title[edit]

The "wrong title" template should be removed because, grammatically, even words that are not normally capitalized are capitalized when they begin a sentence, paragraph or article. Thus, what is usually spelled "qmail" should be "Qmail" at the top of the article and whenever it begins a sentence.

qmail is a proper name, and is formally spelled all lowercase. The title of the article is incorrect, the software package name is qmail, never Qmail. The 'wrong title' template must remain. Anastrophe 07:55, 15 November 2005 (UTC)

Licensing[edit]

"As a consequence, some Linux distributions will not install..." This is also true on other open source operating systems, such as OpenBSD (see Re: Why were all DJB's ports removed? No more qmail?). Perhaps replace "Linux distributions" with "open source operating systems" or something similar? --204.239.167.96 18:25, 20 March 2007 (UTC)

Since it is definitely something you can cite, go for it! — RevRagnarok Talk Contrib 18:34, 20 March 2007 (UTC)
Hopefully I've made a decent edit, my only concern is the four citations that could be trimmed down. I will spend some time reading up on style before trying anything major. 204.239.167.96 18:52, 3 April 2007 (UTC)

"Controversy" section to be removed[edit]

it's been tagged - appropriately - for quite some time. it's loaded with weasel words. there isn't a single citation of source within it. if someone desires to improve the section by finding some reliable sources, and de-weaseling it - have at it. otherwise, it's going to be removed. Anastrophe 17:22, 13 August 2007 (UTC)

The information presented in the section is basically correct (iirc) - so the correct way is to tag it with some {{cn}} tags for information that you believe dubious. I agree that it misses references - but if we just deleted everything on wikipedia because of that - there wouldn't be much left. This article is old enough to have been around before the requirements for verifiability. Tag it - and wait for corrections. --Kim D. Petersen 09:49, 25 August 2007 (UTC)
Burning the article down is the right way to go. The deleted paragraphs do not belong in an encyclopedia. MonsterShouter 16:44, 25 August 2007 (UTC)
agreed. it's all weasel-word-loaded dreck. yes, there are bits and drips of accurate information in it, but it's POV to the hilt, and it's been unsourced for one month shy of three years. so, are we to understand that crap, if it survives long enough, become fact, by virtue of longevity? that's not a an encyclopedia is built. but i'll humor kimdabelsteinpetersen and tag the living shit out of it, as there's no other way to tag it. Anastrophe 17:01, 25 August 2007 (UTC)
Anastrophe, despite your attempts at sarcasm - my guess is that people haven't sourced it because Qmail has mostly gone "out of fashion" since the article was written - which also was before things had to be referenced to the extent it is today. I have nothing against a rewrite btw. since i have no attachment to the subject at all - i've personally always implemented sendmail/postfix solutions. But like most other sysadms with a couple of decades of experience - i've become familiar with the controversy that DJB and his software raises in the community. --Kim D. Petersen 17:27, 25 August 2007 (UTC)
'out of fashion'? care to provide a citation for that bon mot? didn't think so. i'm also a sysadmin with a couple of decades of experience. most of the 'controversy' is little more than people hating djb's software because of djb's personality. some of the shortcomings of qmail cited are legitimate; the only 'controversy' is that trumped up by those with an axe to grind in that regard. sysadmins work around shortcomings of sendmail, postfix, communigate, exchange, etc etc every single day; it's simply the blind-spot of 'i chose this mta, therefore it's the best' at work. the difference is that sendmail, postfix, etc don't have an author that rubs people the wrong way. but i digress. the section is unencyclopedic. it is given undo weight by having it sit there full of weasel words and lacking sources.Anastrophe 17:36, 25 August 2007 (UTC)
As for out of fashion - see this compared to this. --Kim D. Petersen 18:17, 25 August 2007 (UTC)
smtp server surveys tend to be unreliable, for reasons you should be familiar with as a sysadmin. [1] gives a completely different response from only a year earlier - further suggesting unreliability. the high ranking of 'imail' defies logical explanatio, frankly. Anastrophe 18:56, 25 August 2007 (UTC)
Yep they have large uncertainties applied - especially now where your SMTP server is typically frontend'ed by some sort of firewall relayer - or by people mangling their welcome message. But such survey's do give an indication - or do you think that Bernstein did his 2000 and 2001 survey's for fun - or because he wanted to debunk the 50-75% sendmail claims? (here's a rather new one [2]) --Kim D. Petersen 19:49, 25 August 2007 (UTC)
Yes a large part of the controversy around Bernstein's products stem from personality clashes. But the reason that i for instance have run mostly on sendmail is that it was the best and most complete tool for the task. A hassle in learning the syntax - but you can make it dance, even when you have to support UUCP at the same time. But while such things are rather interesting to discuss in general - this is about qmail... And while some of the controversy may stem from personality clashes - it doesn't make the controversy less correct or notable. --Kim D. Petersen 19:49, 25 August 2007 (UTC)

congratulations to kimdabelsteinpetersen for replacing one set of weasel words with one of the weakest 'citations' i think i've ever read. one person states that qmail violates the standards - and we have a citation of controversy! kind of fits with the undo weight i was talking about....Anastrophe 17:53, 25 August 2007 (UTC)

Anastrophe - be civil please. Attack the source - not the editor. The reference is to an IETF meeting, where all members are amongst the very top people in MTA and MTA protocol. That Qmail gets referenced and commented at this level - is indicative of quite alot more. --Kim D. Petersen 18:17, 25 August 2007 (UTC)
you seem to have a misconception of what the IETF is. the qualifications to join and participate in IETF meetings are precisely as follows: none. no reliability of the source you've cited is conferred by his or her off-the-cuff comment in the logs of an IETF meeting. i'm tempted to tag it as {{dubious}}. and a reminder: the program is named qmail, not Qmail. Anastrophe 18:56, 25 August 2007 (UTC)
Possibly correct - but the commenter is John C. Klensin - chair of the Internet Architecture Board, and author of several RFC's [3]. --Kim D. Petersen 20:02, 25 August 2007 (UTC)
except that what klensin says is benign and non-controversial; it's what cnewman says that suggests controversy. either way, it's about three lines of off-the-cuff banter, which really doesn't strike me as reliable. which is not to say that there aren't reliable sources regarding some of the controversy, i just think this particular citation is very weak in support of the contention. Anastrophe 21:01, 25 August 2007 (UTC)
Oh - you mean Chris Newman [4], another author of several RFC's [5]? Yes - he is another chair of a IETF group (and author of SMTP RFC's) - someone i wouldn't be scared of calling an expert on SMTP ;-) --Kim D. Petersen 21:14, 25 August 2007 (UTC)
we have one person saying 'gross violations of the standards', the other saying simply 'bends' the standards. i suppose taken together we have a controversy, since they're at odds. it's an awful lot of kruft to plow through to get to that however (3 lines out of 863...)Anastrophe 21:38, 25 August 2007 (UTC)
Thats two of the worlds top-most experts on SMTP protocol issues (experts here as in they have written part of the specifications) that are commenting that qmail doesn't do exactly what the SMTP protocols are specified as. One of which is stating that its gross violations and the other that it is "bending" them. I dare say that that should be a sufficient reference for "There is some controversy among mail system operators over whether qmail is as standards-compliant as its author claims" - and not a weak reference at all. --Kim D. Petersen 23:23, 25 August 2007 (UTC)
Kim, you are required to cite reliable sources, and mailing list posts --- by anyone --- explicitly do not qualify. WP is not a venue to amplify or resolve Klensin or Newman's issues with Bernstein. In contrast to your claim, I rebut that this is simply a personality conflict spilling out into an informal mailing list. The issue remains: this is a weak, unencyclopedic, unbalanced section, and unless you can track down something better than allusions in mailing lists, it's going to be removed. --- tqbf 17:04, 30 November 2007 (UTC)
Which is entirely incorrect since most protocol work, standard work, discussion amongst the experts etc. on this particular subject is happening on mailing lists. Had this been something that was duplicated in other media, then your argument would have been correct. But like it or not this is the way that the SMTP (and IETF) people work. --Kim D. Petersen 10:47, 1 December 2007 (UTC)

Mailing lists are not reliable sources. It does not matter whose mailing lists they are. Find reliable sources. --- tqbf 18:58, 1 December 2007 (UTC)

Find me the place in WP:RS that says that mailing lists that are part of standardization processes, and are the official mailing lists of the standard boards aren't reliable. Until then - please let the information rest. --Kim D. Petersen 19:29, 1 December 2007 (UTC)
You don't seem to have read the edit you're reverting. The substance of the criticisms in that section weren't removed; they were simply condensed to balance them against the rest of the article. But that's neither here nor there: anybody can post to an IETF mailing list. There is no fact checking or or editorial oversight on them. Please stop citing them as sources. --- tqbf 20:08, 1 December 2007 (UTC)
Tgbf - I've read the edits that i've reverted, and i suggest that you look abit at who the persons involved here are. They aren't just "anybody". You also deleted official IETF meeting data. So far you haven't provided anything that even remotely indicates that this isn't WP:RS. --Kim D. Petersen 21:04, 1 December 2007 (UTC)
it's further worth noting that in fact neither of the cites reference mailing list posts. the first is a log of a chat session. the latter is the minutes of a "Telechat" meeting of the IESG. i needn't comment on the quality of those as WP:RS, need i? Anastrophe 21:02, 1 December 2007 (UTC)
And why is that? Can you state any other reliable reference to IETF standardization work? How do you think the IETF and IESG conduct most of their meetings? --Kim D. Petersen 21:42, 1 December 2007 (UTC)
your questions constitute a non-sequitur. informal commentary in a chat log is not a reliable source. the comments are off-the-cuff, and not directly responsive to any specific matter under formal discussion or on the agenda of the meeting. had the agenda included "Discussion of qmail's violations of standards" then it would have probative value. this isn't an article about how the IETF conducts its work - if it were, the citation would indeed be a reliable source. Anastrophe 22:55, 1 December 2007 (UTC)
You make a serious error here. Stating that its "informal commentary in a chat log" is ignoring that this is an official standards committee meeting. Again, that this happens on IRC is a choice of convenience - and no less reliable than a transcript of any meeting. Had this been "informal" then the transcript wouldn't be available.
Try turning it around - does the choice of technology to conduct a meeting, make the meeting, or the minutes of that meeting less reliable?
What we have is a bunch of experts on the subject of MTA's, who are meeting for standardization work on MTA's, talking about a specific MTA. Which is something that WP:SPS specifically states is a reliable source. --Kim D. Petersen 00:33, 2 December 2007 (UTC)
i'm sorry, but i do not believe i'm making any error. i am characterizing the comments you have focused upon as informal because within the context of the transcript, they are informal comments. reread the agenda for the meeting. read the rest of the discussions in the transcript. qmail, it's standards compliance or lack thereof, is not on the agenda, and there is no formal discussion of same. the entirety of the commentary extracted from this transcript that pertains to qmail is as follows (please forgive the crappy formatting):
  • [09:31:35] <alexeymelnikov> The demo implementation doesn't do mailing list related examples, but they are working on it.
  • [09:31:40] <xiaodong> CNNIC's is based on Qmail
  • [09:39:23] <alexeymelnikov> John suggested to skip the demo, as it is difficult to follow.
  • [09:39:23] <cnewman> John is far too polite about qmail's gross violations of the standards.
  • [09:41:21] <cnewman> As a data point, the presentation is of no interest to me, and I have not seen it before. Of course, I'm not seeing it now, either. But I'm having fun imagining which qmail ideosynchrasies are causing John's blood pressure to rise. :-)
  • [09:44:53] <klensin-ietf> Chris, nothing serious -- just oddities in how trace fields are constructed. Qmail bends the standard in ways that would be almost unnoticable in 2821/2822 mail, but makes it hard to tell what changes are due to these extensions and which to qmail basic behavior. My blood pressure is not rising over that.
  • [09:47:47] <pguenther> ah, I see: eaipara is indeed a parameter; their demo works because qmail doesn't require the brackets around the address and the client isn't inserting them!
this amounts to little more than cnewman making a sarcastic comment about qmail, and klensin pointedly not rising to the bait. the only actual substantive content is from klensin, and the tenor is that the slight variance from standards is not a big deal to him. this handful of informal comments during a formal meeting about other matters isn't reliable. Anastrophe 01:45, 2 December 2007 (UTC)

qmail.com.hk[edit]

This site and company, a Hong Kong based hosting reseller, uses the qmail name and logo. Is this officially endorsed by DJB? They appear to abide by the license restrictions, only remote or on-site installation is sold, no binary CD packages, although that doesn't necessarily mean they compile and build qmail for their customers. StevenMcCoy 14:14, 12 September 2007 (UTC)

"Controversy" section[edit]

Two things are going to happen to this section, and I'm giving a heads-up:

  1. Any critique of qmail that isn't sourced to a reliable secondary source --- IE, Rick Moen's personal website does not qualify --- are going to be stripped out.
  2. This section is going to be balanced with the rest of the article, so as not to give WP:UNDUE weight to a factional POV in an article about a piece of software.

The result of both of these changes is that I expect a number of pet criticisms to be removed. Source what you can, please.

--- tqbf 16:13, 30 November 2007 (UTC)

Adjusted. You cannot have an article on a piece of software with 7 grafs of description and 8 grafs of "controversy". --- tqbf 05:23, 1 December 2007 (UTC)
agreed. i plowed through the old history diffs a while back, and as i noted way above, the huge chunk of uncited criticism had been there for three years. that's an intolerable violation of basic WP policy. editors are obviously welcome to add back any notable criticism they believe belongs here - so long as it has a WP:RS to go with it. Anastrophe 05:45, 1 December 2007 (UTC)

You know, qmail is published on one person's personal site (DJB's). As such, it does seem that the same standards should apply to statements published about it on sources of a similar reliability. --FOo 08:12, 1 December 2007 (UTC)

Uh... OK? Have you read WP:SPS? Noncontroversial SPS are acceptable when they are writing about themselves. Which citation are you looking to challenge? --- tqbf 10:20, 1 December 2007 (UTC)

"talk is out"?[edit]

Kim used the following edit summary while reverting: "rv for the same reasons as previous. Talk it out..". Could you please explain what that means? Anastrophe 21:18, 1 December 2007 (UTC)

I mean that i've tried to restore the referenced section - but the references seem to be continually deleted, despite that there has been no argumentation (imo) on their reliability. I can't argue against the deletion of the rest - since that has had a reasonable amount of time, for other editors to reference.
That the references are to mailing lists and minutia from the IETF, is merely a reflection of how the protocol and standardization work are being done. While this is unconventional, it is the way things are done. In general i'd object to these kinds of sources, but in this particular case, it is the way things are done. --Kim D. Petersen 21:31, 1 December 2007 (UTC)
Kim: I am going to remove these two paragraphs, because they are (a) poorly sourced and (b) duplicate points made in the paragraph above them. If you hadn't noticed, that paragraph addresses your critique, and doesn't even attempt to rebut it, even though I think it's a ridiculous critique of qmail. We can go round and round on this, but you should not expect this to settle itself where you seem to want it to. The qmail article on Wikipedia is not the place to resolve all your concerns about Dan Bernstein. --- tqbf 22:03, 1 December 2007 (UTC)
i'd like to find a middle ground on this matter. i believe the problem informally deltas with the difference between malum in prohibitem and malum in se, so to speak. qmail violates standards. that's citable, and referenceable. the problem is, so does sendmail, and so do all MTAs. sendmail sort of gets the catbird seat, because many of the standards at issue were simply back-formations from sendmail's default behavior. Most RFC violations are minor and tend to fall into the 'quirk' category - every MTA has them. Major RFC violations generally don't show up in production MTA software as they prevent the MTA from Playing Well With Others. those with an axe to grind, for whatever reason, tend to exaggerate the significant of minor standards violations. no, i'm not accusing Kim of that, for the record. i've had to work around the 'shortcoming' of qmail's proclivity for acting as a joe-job bounce reflector in my own mailservers. it's frustrating. it's not a deal-breaker for me though, balanced against the reliability and security of the base software, and the ease of administration. but i digress. in fact i don't even remember where i was going with this. sigh. Anastrophe 23:08, 1 December 2007 (UTC)
The middle ground I propose is:
  • Retain the critiques
  • Do not rebut them explicitly, thus making them noncontroversial
  • Condense them to 1 graf to balance them with the rest of the article
  • Move on.
Does this make sense? The article as of this afternoon no longer looks anything like the article from yesterday. --- tqbf 23:11, 1 December 2007 (UTC)

Reliable sources[edit]

I can't tell people how to spend their time (and it always pisses me off when other editors try to do it for me), but let me voice my opinion:

It is a waste of time arguing about whether IETF artifacts are reliable sources:

  1. Per WP:V, it is unlikely that any of these will be acceptable sources. People are not reliable sources, writing is. Reliable sources are writings published in venues with a reputation for fact-checking and accuracy. Specifically excluded from this definition are venues composed primarily of opinions and personal beliefs.
    Per WP:SPS and WP:RS, since these are experts (published) - who are talking specifically within their area of expertise - your argument is wrong. --Kim D. Petersen 00:24, 2 December 2007 (UTC)
    That depends on the mailing list; the mailing lists you're citing are unmoderated and unauthenticated, and you're extracting snippets of conversations without context to suit your arguments. --- tqbf 00:37, 2 December 2007 (UTC)
  2. Mailing lists and chat logs are inherently WP:SPS, which are always sources of last resort, and only suitable for non-controversial topics (not that we're arguing over the controversy section of this article.
    Yes, and these are experts, talking within their field of expertise - something which WP:SPS makes specific exceptions for. The "chat logs" are the official transcript of the IETF meeting. That this happens on IRC is a choice of convenience - since the members of the IETF task force are spread across the globe. It doesn't make the meeting log less of a WP:RS, than a webtranscription of a face-to-face meeting. --Kim D. Petersen 00:24, 2 December 2007 (UTC)
    WP:SPS are the references of last resort. You're not addressing the argument. --- tqbf 00:36, 2 December 2007 (UTC)
    Erm, sorry? Should that be an argument? The trouble here is that all documentation of standardization, standardization work, and thus controversy over this, happens in this media. The end results are the RFC's. --Kim D. Petersen 01:18, 2 December 2007 (UTC)
    That's not at all true. The IETF, its processes, and its outputs are documented in numerous reliable secondary sources. --- tqbf 01
    33, 2 December 2007 (UTC)
  3. When "controversies" are factored out of the remainder of this article, the resulting article is mostly controversy, and like it or not, WP:WEIGHT applies: an article about a popular piece of software cannot consist mostly of debate about the software.
    This particular "popular piece" of software (and its creator) has been subject of lots of controversy. The solution to the controversy issue - is not to delete the controversy - but instead to increase the amount of writing on the software. --Kim D. Petersen 00:24, 2 December 2007 (UTC)
    All you've demonstrated as this controversy is important to you. --- tqbf 00:36, 2 December 2007 (UTC)
    Actually i couldn't care less. Qmail and the debate over this, is something that i've followed on the sidelines. I agree in a large part with Bernstein on security issues - but i also acknowledge that Bernsteins personality and his way to react to criticism has been a hurdle. --Kim D. Petersen 01:18, 2 December 2007 (UTC)
    How do you propose to improve this article? --- tqbf 01:33, 2 December 2007 (UTC)
  4. This debate is occurring almost entirely over the "Standards compliance" critique of qmail, which is not the most notable critique (licensing and security are). This is easily established by the balance of indisputably reliable sources. We cannot spend 4 grafs on an NN critique while allocating only 4 grafs to all the rest of the notable critiques and a comparable amount of time to the software itself.
    Within the are of mailservers (and other internet communication software), standard compliance is a major issue. So is security, something where Qmail apparently excels. --Kim D. Petersen 00:24, 2 December 2007 (UTC)
    Things do not become notable simply because you say so. Do you have any non-crappy references that say standards-compliance is more notable than the current graf alludes to? If not, what are you arguing about? --- tqbf 00:36, 2 December 2007 (UTC)
    Are you of the opinion that IETF standardization work and the issues raised, within the context of MTA's is non-notable? --Kim D. Petersen 01:18, 2 December 2007 (UTC)
    Oh, come off it. You aren't talking about the IETF --- you're talking about two people who happen to contribute to the IETF, and what they happen to think of Daniel Bernstein and qmail. Critiques are notable by dint of people writing about them in reliable sources. --- tqbf 01:33, 2 December 2007 (UTC)

I recommend: give up. Let's move on. I've refactored the article, which was unencyclopedic, hard to read, and confusing; this issue no longer makes any sense, and is a moot point.

--- tqbf 23:07, 1 December 2007 (UTC)

Security vulnerability[edit]

The rewording, "No qualifying vulnerability has been published" introduces an unsupportable POV: it is not resolved that Guninski's finding, which is in fact a textbook integer overflow (despite the extraordinary difficulty involved in exploiting it), does not qualify, only that Bernstein has opted not to pay for it. --- tqbf 04:31, 2 December 2007 (UTC)

but that's POV as well. the reward challenge has always stated that he is the final arbiter as to whether an exploit qualifies for the reward. therefore, if his decision is that an exploit does not qualify - then it does not qualify. whether one likes that or not is immaterial. based upon the specific qualifications outlined in the reward challenge, "No qualifying vulnerability has been published" is 100% accurate. Anastrophe 17:03, 2 December 2007 (UTC)
Strong disagree. The word "qualifying" is a clear judgement on the finding. The only fact we can report is that Bernstein has declined to pay. Consider that Bernstein could unfairly disqualify, say, an uninitialized stack variable vuln (perhaps claiming "nobody knew about these in 1998, unfair to fault me for them"), and by the same logic you just used, their finding would not "qualify". We're simply debating the word "qualify"; let's find a more neutral word. Also: the award is quite disputed, and I say this as huge Bernstein fan and a security practitioner. --- tqbf 17:05, 2 December 2007 (UTC)
i have no disagreement that the award is disputed; however, characterizing it as simply 'bernstein declines to pay' is incorrect. pretend it's a creative writing sweepstakes. buried deep in the fine print, it says "entries that fail to capitalize the first letter of the first word of sentences are invalid, and will be disqualified". captain lazy-caps (that's me) submits an extraordinary first person account of my encounter with aliens. in all respects, it's perfect, but i fail to capitalize the first letter of the first word of sentences. i'm disqualified. i can bitch about it all i want, but the fact is, according to the terms of the sweepstakes, my entry did not qualify, so if i don't win, i can't say that the sweepstakes sponsor 'declined to pay' me. dammit, too much coffee this morning. Anastrophe 17:24, 2 December 2007 (UTC)
(edit conflict) The difference between a sweepstakes and this contest is that the arbitrator published the criteria for qualification, and many (if not most) security practitioners agree that Guninski's finding qualifies. But that's how I'm characterizing on Talk, not in the article. I have no problem with rewording what's there now, but I have a real problem with wording it in a way that casts Guninski's finding in a negative light. It's a real finding, and a real mistake in the code. The only one, to my knowledge, ever found in a mainstream DJB package. It's a Big Deal. I've publicly offered to pay Guninski the $500 (thankfully, he hasn't taken me up on it). And again: I am a huge fan of Bernstein's. --- tqbf 17:35, 2 December 2007 (UTC)
i (or the coffee) decided to be WP:BOLD and integrate the reward info into the body as part of the guninski controversy. this should adequately mitigate the problems with it in the lede, since the wording can be eliminated as it's implicit. Anastrophe 17:34, 2 December 2007 (UTC)
Well played, sir! --- tqbf 17:36, 2 December 2007 (UTC)

Full-featured[edit]

What's up with "qmail is the only full-featured MTA in the public domain."? How is "full-featured" defined and who gets to decide? —Preceding unsigned comment added by 67.166.3.14 (talk) 22:32, 19 January 2008 (UTC)

please sign your comments when you post them by typing four tildes after it (this is noted at the top of the 'editing' page). full-featured is a reasonable term of art that a reasonable person can understand. full-featured means it's not just a single purpose script that can only inject a single message into another mailserver, for example. qmail provides all of the standard functionality a "full-featured" MTA would be expected to provide.
but there's an easy way to turn the question around: can you provide the name of another MTA that is in the public domain? then we can compare and see if it passes the 'reasonable person' test as an issue. Anastrophe (talk) 22:37, 19 January 2008 (UTC)
Pretty sure that's not how Wikipedia works, Anastrophe. The question isn't "prove that statement isn't correct." The question is "What's the citation for that statement?" 67.166.3.14 (talk) 00:14, 20 January 2008 (UTC)
Consensus wins disputes here, and right now, you don't seem to have it. What wording do you propose besides "full featured" to describe the qmail feature set? Note that Anastrophe did not use "remarkably flexible", unlike Sendmail, or "a famous example of good programming practice, unlike Postfix, or "highly configurable with features lacking in other MTASs", unlike Exim. --- tqbf 03:13, 20 January 2008 (UTC)
Actually i'm interested in what "full-featured" means here as well. Are we referring to a complete conformity to the ESMTP protocol? All features that users expect? That it can be used for all mail purposes?
The reason i'm asking is that i've dismissed it as usable specifically because it lacked a feature i needed (UUCP). If we are going to use this - then full-featured should be cited or defined. --Kim D. Petersen (talk) 04:11, 20 January 2008 (UTC)
perhaps then we should just elide "full-featured" and call it what it is - the only public domain MTA, if we are to go by Comparison of mail servers. the problem of course is that WP can't be used as a reference for WP. but i'm not sure if there's any other reliable source with such comparisons. i still maintain that a 'reasonable person' understanding of what 'full featured' means applies. while we're discussing this - what unix/linux/xBSD system doesn't include UUCP? Anastrophe (talk) 04:31, 20 January 2008 (UTC)

i've modified the claim in the article. it should be a reasonable compromise, but we shall see. Anastrophe (talk) 04:40, 20 January 2008 (UTC)

Portal:Free software: qmail is now the selected article[edit]

Just to let you know. The purpose of selecting an article is both to point readers to the article and to highlight it to potential contributors. It will remain on the portal for a week or so. The previous selected article was Frets on Fire - a guitar playing game which has be the subject of software patent discussion in recent weeks.

For other interesting free software articles, you can take a look at the archive of PF's selectees. --Gronky (talk) 23:16, 7 April 2008 (UTC)

Things moved on, as usual. The new selectee is Wireshark - a packet sniffer which people concerned about security might be interested in. --Gronky (talk) 11:55, 14 April 2008 (UTC)

requested edit[edit]

What is wrong with this edit ? Regards, NonvocalScream (talk) 22:43, 25 April 2008 (UTC)

did you read the edit summary? it isn't an improvement, particularly if you take the time to read more than just the first sentence. what are you trying to accomplish? Anastrophe (talk) 22:55, 25 April 2008 (UTC)
Sentence structure that appears correct, and this can be accomplished by simply restructuring the sentence without changing the meaning. Are you averse to that? NonvocalScream (talk) 23:31, 25 April 2008 (UTC)
This actually gave me an idea, what do you think about a small section regarding why the author decided qmail as opposed to Qmail? Do you think that would add to the article? NonvocalScream (talk) 00:17, 26 April 2008 (UTC)
yes, it would be good to have it in the article. i'm sure it's discussed somewhere tho' offhand i don't know where. Anastrophe (talk) 00:52, 26 April 2008 (UTC)
If the author of the software states it in a source, I'll find it. Once I do, I'll add it, and we can see if it fits. If not, we can remove it. NonvocalScream (talk) 01:22, 26 April 2008 (UTC)

No longer very popular[edit]

Here are three surveys of mail server software:

None of them show qmail as having above a 5.3% market share or a rank higher than sixth place across all mail software ... and that survey, the O'Reilly one, is a survey of companies rather than mail servers:

Rather than surveying all the domains in the world--which would have been a fun project--we chose instead to survey only those domains that have a real company behind them. We partnered with an old-school company data firm to get a list of 400,000 companies worldwide as our source material. It's not the perfect solution, but it suits our needs, and we can survey all 400,000 in a couple of hours without so much as raising a single abuse complaint.

The falkotimme.com survey was based on IP scanning and found 0.03% of mail servers running qmail, compared to 29% for Sendmail, 19% for IMail, and 20% for Postfix.

The securityspace.com survey was based on MX records. It found qmail running on 0.09% of mail servers, compared to 35% Sendmail, 20% Microsoft, 15% Exim, and 11% Postfix.

So, it is *not* clear that Qmail remains one of the most popular pieces of mail server software. --FOo (talk) 16:43, 24 December 2008 (UTC)

It seems like this claim has been without a cite for over a year. Now we have contradicting sources. I'm going to tag it dubious to motivate the issue.
On a side note the phrasing of the claim is going to make this a recurring problem. Once we have a cite, the claim should probably be phrased like "As of X, 200X, qmail is one of the top five ...". Reskusic (talk) 02:43, 18 April 2009 (UTC)
I have removed this claim, I seriously doubt that it is a popular system today and without a citation from a reliable source on the subject it should not be here. Chillum 23:46, 1 May 2009 (UTC)

Differences[edit]

Can anyone comment (find a citeable source) as to the most popular MTA to date for a unix system? Qmail? Postfix? Maybe have a comparison for use of a programmer making a decision as to which MTA to go forward with. Security is mentioned but what about performance, usability...?

Snowmaninthesun (talk) 23:11, 10 February 2009 (UTC)

Qmail-distributions[edit]

Anastrophe:

I'm not happy with your reasons for removal:

"Same rationale for removal as before. the license status is already covered; the stuff about the google code base reads like an ad."

1. There's minimal overlap about license status. In the "Problem" section, I'm simply going into more detail about the headaches that the old license gave to patch developers.

2. Don't just delete the whole thing because you think it reads like an ad. Suggest how I can improve it.

Until you demonstrate that you will do something other than delete my entire contribution, I will undo your removal.

Let's work together for Make Benefit Glorious Nation of Qmail!

Yonas yonas (talk) 00:08, 24 April 2009 (UTC)

you may not be happy with my rationale, but my rationale is based on wikipedia policy. based on your edit history, this appears to be a single purpose account, with the single purpose of driving eyeballs to the distribution you wish to promote. wikipedia is not a vehicle for promotion, whether commercial in nature or not. you addition is not about qmail itself (please note, qmail is always spelled with a lowercase 'q') - it only indirectly pertains to the topic of this article. on that basis alone, the added section is in conflict with undue weight - this particular distribution is not particularly notable, and certainly not notable enough to warrant significant inclusion in this article. notable software associated with qmail is easily found via the link to qmail.org found in the external links. that's entirely sufficient coverage. Anastrophe (talk) 05:34, 24 April 2009 (UTC)
  • SPA: I've edited Wikipedia a lot, but not under an account (anonymous). My intentions are not related to my account anyway......
  • Vehicle for promotion: I'm not promoting, I'm including something that is relavant to people that want to learn about qmail, specifically, about new open source efforts that came out of the new license.
  • Not about qmail itself: Well, I first started a separate article "qmail-distribution", but people kept deleting it, saying it should be part of qmail article.
  • qmail is spelled lowercase: true dat.
  • qmail-distributions is not a notable distro: it's not a distro at all! It's an effort to combine patches into your own easy to download and deploy distro, which wasn't possible before.
Let the people learn about qmail-distributions in this article. Period. We can edit the amount of information in its paragraph, but don't delete everything. 13:55, 24 April 2009 (UTC)
As long as no secondary reliable sources have commented on your project - its out. Sorry - but that is wikipedia policy. We do need outside sources to weight whether it is a notable project or not. Wikipedia is not a directory of possibly useful projects, interesting developments or of what people might consider relevant. We leave that completely and utterly in the hands of reliable sources. (and qmail-distribution isn't) If you want to promote your project, use Freshmeat or other software repositories to make it available to a broader audience. --Kim D. Petersen (talk) 15:10, 24 April 2009 (UTC)
Alright, I'll find a reliable source. I'll be baaaaaack!!! Bew haHahaHAaha Yonas yonas (talk) 16:30, 24 April 2009 (UTC)
Don't take my Bew haha out either. Yonas yonas (talk) 16:30, 24 April 2009 (UTC)

Secret ingredient[edit]

What is it that makes qmail safe to stand the test of time? The answers at http://cr.yp.to/qmail/guarantee.html do not suffice to make a program safe. Have any formal methods been used to implement qmail? --Abdull (talk) 21:59, 10 September 2010 (UTC)

Backscatter[edit]

Qmail as released is seriously flawed. It accepts mail for non-existent recipients. Once it has determined the recipient doesn't exist, it sends a mail to the apparent sender. Spammers may therefore use an un-patched qmail essentially as a spam relay. http://en.wikipedia.org/wiki/Backscatter_%28e-mail%29 . One would expect the standard distribution of Qmail to include patches which various people have written to combat the problem, however, it appears not: http://mirror.tje.me.uk/pub/mirrors/www.qmail.org/top.html#patches . I suggest the article should have a criticisms section, outlining qmail's backscatter problem. More information can be found here to solve the problem: http://http.netdevice.com:9080/qmail/rcptck/ . I suggest it is reasonable to strongly recommend against using standard qmail (without patches), Even the current version 1.06. Nick R Hill (talk) 21:14, 5 February 2011 (UTC)

IndiMail[edit]

Wouldn't it be a good idea to add IndiMail here? It's an updated and extended package, and after netqmail the second resurrection of qmail. I don't think it helps visitors/users to show netqmail as the most recent version, as it's hopelessly outdated and very hard (virtually impossible) to patch into an up-to-date system. — Preceding unsigned comment added by Offerman (talkcontribs) 23:25, 21 September 2013 (UTC)

Abandoned?![edit]

QMail rather looks like a completely abandoned project, now an actual risk because of a number of severe security problems. At least the wording should reflect that. — Preceding unsigned comment added by 2A02:908:EB20:1AC0:4032:50D3:22DC:A761 (talk) 02:55, 11 February 2015 (UTC)