Wikipedia:Featured article candidates/OpenBSD

OpenBSD

In the previous attempt to feature this article, only three opinions were ever given - during that time the complaints were taken back and one even converted to a support, however this was not enough to get the article a passing grade. So here it is again for another attempt, here s the old peer review and previous FAC attempt. Janizary 08:23, 6 January 2006 (UTC)

• Support. Article contains a great deal of information, well-referenced, seems quite well-written. The order of later sections may need a slight tweaking, but besides that, I can't think of any other wide-scale suggestions for improvement. RyanGerbil10 14:41, 6 January 2006 (UTC)

• Support. Same as above, and it seems to have a good balance of history and technical information. Rst, Fri Jan 6 14:48:36 UTC 2006.

• Neutral. Much better after the rewrite. I'm concerned about meeting the stability criteria after changing that fast, but what is there is good. Object for now. Sorry, as I know you guys are really trying to get this one featured, but the writing is simply not very good. There's lots of short paragraphs and choppy sentences throughout the article that makes the text flow poorly. I know that's a hard objection to handle because it's pretty fundamental, but good writing is clearly fundamental to a FA too. I can pick out examples if you want, but there are literally so many, that I'm not sure listing them would be fruitful. I'll see if I can set aside some time to work on fixing what I see, but honestly I'm not sure I'll be able to. 2) There's really no mention of the downsides of OpenBSD. For one, it's really not terribly usable by someone without significant technical skills. That of course is a downside to some and not to others, but it is an important facet. Saying a desktop can be run kind of glosses over that important bit. Performance. Pretty much all head to head tests put it way behind the other BSD's and Linux for most general tasks. The developers will generally admit this explaining performance is secondary to correctness and security. But the fact it is way behind is also ignored in the article. Another tidbit: is DJB's software still out of the ports tree? The article doesn't seem to know unless I missed it. So the article suffers from lack of balance. While spending great detail (about 4 full paragraphs) on the split from NetBSD, it hrdly mentions some of the most commonly cited advantages of OpenBSD, the integrated cryptography (which is not even linked to in the article, much less discussed well, and one sentence on the high quality of the documentation. So there's a lack of coverage of the important advantages and disadvantages. Overall it suffers from being edited by committee. - Taxman ^Talk 15:00, 6 January 2006 (UTC)

* No apologies needed, being critical is the core of these reviews - it's how we separate the wheat from the chaff. I can't deal with your points just now, but I will be going over them in like 6 hours or so. As far as usability goes, I completely disagree - I went from using Windows 2000 to using OpenBSD with nothing but the man pages and a friend telling me to keep trying and to read the man pages, there is no special technical skill required to use OpenBSD, just the willingness to read the documentation. Performance, yes, we could even mention the lack of a UBC. Janizary 17:30, 6 January 2006 (UTC)

Hehe. I was going to suggest UBC as an example too :-). NicM 17:45, 6 January 2006 (UTC).

• Thanks for your comments; I think your points are fair, in general. Restructuring the article layout and text so that it flows better has been on my list for a while, although I make no comment on whether it is too bad for a FA, just that there is scope for improvement. Unfortunately, it is quite a big job that needs to be done carefully and I can't seem to find the time. I'm not convinced that the text itself is as bad as you say, but perhaps we have different ideas of style; it may help if you would be good enough to list some of the most egregious examples.

I had also thought about the lack of detail about cryptography and documentation but a) as the article already covers the technical security topics in some length I was reluctant to add more, b) given the current structure, I wasn't sure just where to put them, particularly stuff on documentation, c) to my mind, they are less interesting and important than the unique security features OpenBSD has developed, but I'm not really interested in crypto so I would think that :-) d) as I'm not interested in the topic I don't really know what to say about crypto, so I was hoping someone else would do it :-).

DJB's software is still gone and will probably never reappear; the article doesn't say this, but nor does it say there was a reconciliation and it was put back in—it may indeed be better if it said this explicitly, I'll probably fix that now. The split from NetBSD was perhaps the most important event in OpenBSD's history (and a significant one in NetBSD's) so I don't think four paragraphs is too much. A section about criticism/problems is a good idea, notably, as you say, performance and usability; however, I don't think there is practically that much criticism of OpenBSD in these terms, most people (both users and non-users, and developers) seem to understand that speed and ease of use are not top priority, so I'm not sure how long such a section would be :-)... I did try to gather material for a section on criticism of OpenBSD's security record but I found it impossible to seperate real allegations and evidence from trolling :-/.

The biggest problem at the moment is probably the flow and structure. I think I'll start a discussion on the talk page to try to get a layout sorted out before trying to fix it. NicM 17:45, 6 January 2006 (UTC).

Oh, never mind about pointing out examples in the text, I see what you mean now. NicM 18:48, 6 January 2006 (UTC).

• Neutral (was: Oppose.) I think the fact the article contains a sentence like "The only available records of these events are an incomplete set of emails, published by Theo de Raadt on his personal site" is a rather concerning problem. The article indicates that this is a rather critical matter, but seems to concede the source used is unreliable. This wasn't something added just now. While I hate to be picky, frankly given recent extra attention/concern lately for verifiability and reliable sources at Wikipedia, I think it would be a mistake to endorse this article, and thereby accept it's sourcing standards. I think we want to feature articles that cite reputable published sources. This is especially true when saying something that reflects on a person's character (which the relevant paragraph does). I see other notes/references that are apparently from mailing list discussions, which have never been published (as far as I know) outside the mailing list. Now admittedly, this article is better sourced than most Wiki articles, there are many worse articles, and maybe there's good justification for everything. But, we shouldn't use it as an example of what we think is best. --Rob 17:51, 6 January 2006 (UTC)
  • Unfortunately, most of the OpenBSD discussion and development takes place online, so almost all of the available historical sources are mailing lists and online articles. I understand what you are saying, but I do not think this criticism can be realistically met. Without expressing an opinion on whether or not they are good enough for a FA, I think the article's standards are as good as they can be—the section in question does explain the source is not backed up and provided by one side of the events. NicM 18:20, 6 January 2006 (UTC).

• So you oppose it based on the fact that we cite an archive of e-mails? That archive used to be complete, but someone edited it, back when Theo still ran NetBSD. Hmmm, perhaps we could add the bit about some NetBSD developer cracking Theo de Raadt's webserver and removing some of the mail from that archive, not the nicest bit of history - but it was one of the reasons there is such bad blood between NetBSD and OpenBSD and is part of what got OpenBSD going into security. I don't think it the best thing to go into because that brings a bit more POV than I am able to neutralize. And I'd not say the source is unreliable, that it has been tampered with is a bit closer to the truth. It's a dump of e-mails that someone with a grudge took parts out of and it's hosted by the guy who was on one side of the argument - that there is noone saying another side of the argument more lends to the validity of the mails than anything for me. Anyways, as NicM says, everything done for projects like this is done in e-mail, not press releases unfortunately, so we can only deal with what we have. As it stands there is one source of information regarding it all, and that is the one from Theo - not citing it would probably be worse than not covering the matter. Janizary 18:25, 6 January 2006 (UTC)
  • When it comes to negative characteriztions, that can harm a person's professional reputation, and be deemed defamatory, email archives, are even less adequate then normal. Now, if a investigative journalist for reliable publication were to examine them, analysis them, and publish a report, supporting what's said here, then that would be an adequate source. But that hasn't happened, and we can't do original research, which is what would be required. I've also put a note on the talk page, as I do now feel, that no only should this not be FAc, but much of it should be removed. --Rob 19:12, 6 January 2006 (UTC)

• Neutral (was: Oppose.) I gave a rather complete peer review months ago, which didn't go over well with them, so I'm not going to go into many specifics this time. I can on request take a closer look if they desire.

1. POV/comprehensiveness issues - and I'm particularily dissapointed with the history section as I really wanted to see more here. A lot of the conspiracy theories are gone, which is great, but for such a pivotal event it is really kind of bare... also, IIRC the DARPA grant was revoked because Theo was working in Canada and grants were techinically supposed to fund US-only endeavors.
2. writing issues - i.e. "The reasons for this event have never been fully and publicly explained"
3. flow problems (esp. in the "Highlights" subsection).

WhiteNight ^{T | @ | C} 20:42, 8 January 2006 (UTC)

• History was a lot less bare about the fork but was reduced because of repeating damaging claims with lack of sources. I've made a lot of structural and flow changes, and efforts to reduce the length, at User:NicM/OpenBSD that I'm going to use if nobody objects within the next few days, detailed comments on it would be appreciated if you have time, on my talk page for preference. NicM 21:05, 8 January 2006 (UTC).

• GREAT! The beginning history is a lot better in terms of comprehensiveness but the writing isn't too good though :\. I'll try to change it myself if I have some spare time. WhiteNight ^{T | @ | C} 21:15, 8 January 2006 (UTC)

Thanks, any improvements are welcome. I do my best with the writing but I know it can end up a bit clunky :-). NicM 21:19, 8 January 2006 (UTC).

It looks a lot better now... WhiteNight ^{T | @ | C} 18:47, 19 January 2006 (UTC)

• Neutral (was: Oppose.) The article is too focused on the social circles and personalities involved in the early history of OpenBSD, and not enough on its technical qualities or its architecture. OpenBSD is not a social club; it's an operating system. I've made a first shot at cleaning up and summarizing some of the history, but this article needs a lot of refocusing before it's featured. --FOo 05:04, 10 January 2006 (UTC)
  • The personalities, particularly that of Theo de Raadt, is a very important and widely known aspect of the OpenBSD project, and most people do think (with some justification) that the fork was due to social issues (the core NetBSD world was small and tightly integrated at the time, and Theo was a skilled and very active contributor). In any case, I've already made a lot of alternative changes here, which I intended to merge this evening, please let me know what you think. I'll take a look at your modifications and see if I can include them, I do prefer parts of your simpler text. NicM 08:30, 10 January 2006 (UTC).
  • On another note, OpenBSD may not be a social club, but Wikipedia is equally not a technical manual, it is important to discuss social aspects where they are relevent. Particularly as so much of OpenBSD, and open source in general, is social. It is analogous to Wikipedia: the goal is to produce an encylopedia or operating system, but since it is being done by people, often in their spare time and often based on their beliefs, it can develop a large social component and personalities and relationships can become very important. NicM 08:48, 10 January 2006 (UTC).
    • I keep rereading my comments above and I'm not convinced I have really made my point: forgive me if I have and seem to be trying to hammer it home, I'm not. The thing is, OpenBSD is a unix-like and a BSD, all of which have basically the same architecture; I think this article should focus on the specific history of OpenBSD and what makes it different and noteworthy, namely license purity and security. This is not to say I'm not interested in seeing technical material added (although the emphasis in my restructuring was to remove material, not add it), I just don't think the current focus (my new version at least) is too far off the mark. NicM 13:10, 10 January 2006 (UTC).
      • I agree that Wikipedia is not a technical manual, and the history of OpenBSD as a project is interesting. However, the personality politics of Theo de Raadt are not nearly as interesting as the details of what OpenBSD is and how it differs from other BSDs and Unix systems. I think your recent changes are a definite improvement, and I withdraw my opposition. --FOo 02:14, 11 January 2006 (UTC)
        • Thanks for the comments and withdrawing your objection. Its slightly difficult because, as far as opinions go, Theo de Raadt is the project, so to explain the project's, especially the early history, you need to explain his, but I take your point, somewhat. I was just going to write a list here of how there weren't many differences and most of what there are, are things that are missing in OpenBSD, but I think I've just convinced myself to write a short section on technical differences: things, aside from security enhancements, OpenBSD has (pf, bioctl, there must be more), things it hasn't (a UBC, rc.ng, dynamic /dev & no *devsw, UFS2, SMP on non-i386) and things that are different (native threads, encrypted disks, supported platforms (<NBSD >FBSD)). Although, the reason for most of these being missing seems to almost always be one or more of the same three: no convincing reason to change existing solution, lack of developer time, or dissatisfaction with the solution of the other BSDs. I'll have to think see if I can come up with any more and where to put it in the article. NicM 09:11, 11 January 2006 (UTC).
• Comment Alright, the article has been restructured. This is the second time we've done one of them, hopefully it leads to a better overall article. Please everyone who's voted take another look to make sure you still agree with your previous opinion. Those interested in that further discussion and are looking for a quick-link, see here. Rob's change is marked in the history page, so I changed the image to make it easier to follow at a glance. Janizary 08:26, 11 January 2006 (UTC)

• Detailed and informative. Support. Dysprosia 09:45, 11 January 2006 (UTC)

• Support The article is to the point and informative. Overall this is a good article and a great reference. I also enjoyed reading the "history" part. Cmihai 07:20, 12 January 2006 (UTC)
• Support Article is informative, good work. --Terence Ong 14:22, 16 January 2006 (UTC)
• Support Lots of references and information. Gflores ^Talk 23:03, 17 January 2006 (UTC)
• Comment -- I would determine what other articles an OS novice would have to read to understand this article and use the {{subst:preq}} prerequisites template. In addition, I would suppose that even if someone knows enough about operating systems to understand "BSD," "kernel" and similar terms, he or she might not know the background of NetBSD. For that reason, I would remove the mention of that OS from the lead paragraph and instead briefly explain it at the beginning of the history section. ("Theo de Raadt had been working on NetBSD, another open source operating system, when...") You also might want to spell out Berkeley Software Distribution on first reference just to be safe. -- Mwalcoff 02:32, 20 January 2006 (UTC)
  • Thanks for your comment. I've made a few changes to the first paragraph, is this any better? I'll have a think about a list of prerequisites, have you got any particular suggestions? NicM 08:18, 20 January 2006 (UTC).
    • Okay, I've added a prerequisites box. Not entirely happy with its contents or position. Is it possible to make it appear beside the infobox? NicM 08:57, 20 January 2006 (UTC).
      • I've removed it again as the combination of it and the infobox is broken in IE6 and I don't know how to make it work. NicM 10:56, 20 January 2006 (UTC).
        • It does work, I had an extra blank line was screwing it up. NicM 11:05, 20 January 2006 (UTC).
          • Yes, that was the kind of thing I was talking about. Unfortunately, I don't know enough about the subject matter to cast a vote, but it looks good to me. -- Mwalcoff 00:16, 22 January 2006 (UTC)
  • I actually added to the history of NetBSD for that purpose, but really I thought the articles linked in OpenBSD gave most of the information needed to know what's going on in it. I'll see if I can convince a few random people on #wikipedia-en to read the article and see what they didn't understand to see if there is anything to add to the prerequisites. Janizary 15:00, 20 January 2006 (UTC)
    • Yeah, of the three people I convinced to read the article, none seemed to think it was too over-the-top or complex, one didn't think the prereq box was needed, since those were linked in the article - with the exception of computer insecurity. Janizary 21:41, 20 January 2006 (UTC)
      • Computer insecurity is linked in the article. It is about what you and I would call computer security (and is linked as such in the article), but that term has more specific meanings that are covered in the computer security article and mostly don't apply to OpenBSD (they do apply to the ACLs and MAC features in, eg, FreeBSD but not to most of OpenBSD's features). NicM 22:04, 20 January 2006 (UTC).
        • I think the prereq box is a good idea. Someone who doesn't know anything about operating systems would wind up on a wild link chase trying to understand the article without it. -- Mwalcoff 03:19, 22 January 2006 (UTC)