Wikipedia:Featured article candidates/Kernel (computer science)/archive1

From Wikipedia, the free encyclopedia

[edit] Kernel (computer science)

This is the second try (Wikipedia:Featured article candidates/Kernel (computer science)/Archive 1), and I've addressed the concerns, which were mainly about citations missing. Candamir 02:04, 19 August 2006 (UTC)

Comment I have a small wishlist for this article.

  • Make it conform to WP:FN (footnotes go after the punctuation)., thanks to User:Icey! :) Candamir 18:21, 19 August 2006 (UTC)
  • Spellcheck the whole thing, esp. references section., addressed by User:Danaman5 (thanks!) and me. Candamir 20:07, 19 August 2006 (UTC)
  • Summarise image contents in the caption (this is a common problem across Wikipedia, let's change it!), I hope it's better now, and you're right, many WP pages present this problem... Candamir 20:07, 19 August 2006 (UTC)

Feel free to cross them out as they're being addressed, and don't shoot me if I come up with further comments later on. :) - Samsara (talk • contribs) 11:23, 19 August 2006 (UTC)

Object. This article needs to have more history of kernel development. Right now, it doesn't contain a single mention of Microsoft Windows, the most prevalent modern operating system. Does Windows not use a kernel? I have copyedited the article, and fixed some wording issues. --Danaman5 16:35, 19 August 2006 (UTC)

Vote changed to Support, concerns addressed. --Danaman5 02:19, 20 August 2006 (UTC)

Response to objections. Thank you, User:Danaman5 for copyediting and spell-checking. I also added one more section to the history part, containing a description of the modern situation, mentioning Win, Mac and Unixoids as well as the osdev community. Please express any further concerns you might have. Candamir 20:07, 19 August 2006 (UTC)

Thank you for addressing my concerns. I have changed my above vote to support.--Danaman5 02:19, 20 August 2006 (UTC)
  • One more comment on the Windows issue. The article says:
The modern, NT-based line of Windows is generally considered a hybrid kernel. However, it is closed source and distributed under the terms of the Microsoft EULA, a proprietary license.
The last sentence leaves us in doubt as to what the license has to do with the kernel architecture. I assume the sentence is meant to hint that we don't know any details about the NT kernel. If, for instance, the architecture of the kernel considered a trade secret by Microsoft, this might be worth mentioning. Thanks again. - Samsara (talk • contribs) 22:30, 20 August 2006 (UTC)
  • Object "Modern Kernels" issues
  1. As of 2006, Microsoft Windows dominates the end-user market for operating systems with little space left for alternatives like Mac OS and various Unix-variants, such as Linux or *BSD.
    "with little space left"? This is awfully sensationalist and a mostly pointless aside at best, and at worst it is a biased take on how Microsoft is somehow forcing others out of the market. A suggestion would be shortening it to something like "The NT kernel, the current dominating kernel" with better english .
  2. The modern, NT-based line of Windows is generally considered a hybrid kernel. However, it is closed source and distributed under the terms of the Microsoft EULA, a proprietary license.
    "However"? What does this have to do with the previous sentence? Also, it assumes that kernels must be open source, which is a heavy POV!

Also, as a request it would be nice for it to elaborate a bit more rather then just labeling each kernel "hybrid", "monolithic" etc. as that doesn't seem that informative :). RN 08:21, 21 August 2006 (UTC)

    • Response to objections: I completely rewrote the Windows part of that section and slightly modified the other parts, hoping to achieve NPOV... Do you think I should further explain the architectures of those kernels instead of just declaring them as hybrid? Candamir 23:26, 21 August 2006 (UTC)
      • If at all possible - it would really add to it. IIRC MS originally claimed NT was a microkernel - maybe if possible try to sum up the relations in a sentence or two (difficult job but at least an overview of what major traits it takes from each structure would help). Anyway, thank you very much for clearing those things up! RN 21:29, 22 August 2006 (UTC)
        • Response to objections: Now I added a phrase describing a little bit of the NT kernel and also mentioned how XNU was originally created. Hope it meets your expectations. Candamir 22:39, 22 August 2006 (UTC)
      • Thanks! You are doing a great job! I read through the article a couple more times, I think I am starting to see a problem though - and that is that DOS, Early Windows, and Classic Mac arn't mentioned at all. For example, it delves deeply into Unix, but one is left with the question on whether DOS or [classic] Mac had a kernel at all or what kind, yet the article mostly assumes having a kernel is the norm (I'm deliberately being ignorant here :)). I don't know if each of the two needs three paragraphs on it, but I think people are going to have those questions and they are probably essential to the article. If it were me (and this is in no way a requirement - please use your own judgement :)) I would make a seperate section for early mac/windows there and just touch on it. Thank you :). RN 20:49, 23 August 2006 (UTC)
        • Response to objections:Added paragraphs on Mac OS and Windows... Candamir 17:16, 27 August 2006 (UTC)
      • Thanks. Not as deep as I thought but it is great as an overview! Thanks again :). RN 14:18, 3 September 2006 (UTC)

Comment I ran the auto peer review script and pasted the results on the talk page. Can you fix and or verfiy those and strike through when done? Also I know it's probably imposible to have a generic kernel structure picture but if one could be created for the upper right it would really help out. -Ravedave help name my baby 02:51, 22 August 2006 (UTC)

Oh and this may help make a slightly dry subject more intersting [1] -Ravedave (help name my baby) 03:08, 22 August 2006 (UTC)
I checked the link you provided, but IMHO it contains more information on how to make jokes in texts, and while I truly approve most of what he says in there, I don't think if it fits for an encyclopedia... Candamir 21:17, 22 August 2006 (UTC)
The history section could use more dates. Also working in apple's OS where it belong could help as it has been one of the most used OSes for quite a while. Also maybe the page should be re-arranged. The history section is very interesting and might do better at the top. Overall a good article though, I will definitly change to a support once some of these things have been addressed. -Ravedave (help name my baby) 03:42, 22 August 2006 (UTC)
Response to objections: The problem is that the history section is based on the knowledge of the kernel design approaches section (readers won't know what a microkernel is if history comes first). Thus, I ask you to reconsider your position about this issue, but of course I'll change it if you convincingly argumentate against my point. Candamir 22:39, 22 August 2006 (UTC)
Point takn. The Image adds alot, thanks for making that, overall its better. Everything seems to be addressed execpt for the lack of dates in the history section. There are 4 dates in 10 paragraphs. At the very minimum the beginning should talk about *when* the concept of a kernel started. Did it start in the 40's with the ENIAC? The paragraphs go in chronological order so just throw a "in the 19xxs...blah blah blah" -Ravedave (help name my baby) 04:37, 23 August 2006 (UTC)
  • Object. this article is not aware of the academic status of the subject. For instance it assumes the modernity of an architecture based on hierarchical privilege levels (as Supervisor mode), while academic research on computer architecture replaced this outdated technology many years ago (see for instance the bibliography of capability-based security), showing that it provides both poor protection and poor performance.--BMF81 00:31, 26 August 2006 (UTC)
  • Support. I made some changes so the article no longer suggests that all kernels use virtual addressing or pre-emptive (as opposed to co-operative) multitasking. I would like those to stay. Otherwise the article is a concise introduction to the subject. If it doesn't make featured status, I will happily promote it to good article status (just message me). Cedars 11:26, 27 August 2006 (UTC)
  • Comment. I've added a pending tasks list in the talk page.--BMF81 13:09, 27 August 2006 (UTC)
  • Object again. I've just had a look at the todo list on the talk page, and it is clear that this article has a lot of potential that isn't currently realised. If any of the points have already been addressed, maybe this can be indicated? I think it would be a shame to have this article go up on the front page before it's reached that potential. Best wishes, Samsara (talk • contribs) 18:08, 27 August 2006 (UTC)
    • MY GOODNESS that is a great todo list... Acedemically it isn't as deep as it could be and in referencing as well. However, it is reasonably comprehensive and sets out what it is trying to achieve in an article on kernels for a general audience. Taking those things into account one could very well be looking at a seperate article entirely - one more focused on being a reference work. While it is a very enticing proposition, the current article presents a readable overview on an otherwise difficult topic. RN 14:18, 3 September 2006 (UTC)
      • This is very technical subject, one of the core topics of CS; don't you think that we would give a poor image of wikipedia if we present as its best work something that is not academically relevant?--BMF81 03:13, 4 September 2006 (UTC)
  • Support. I commented last time on the large-scale and irritating use of "in order to". Just why you think it's necessary write three words when one will do is a mystery. Tony 13:28, 3 September 2006 (UTC)
    • My understanding is that this is sometimes done where the use of "to" – which can take various meanings as a preposition as well as preceding infinitives – is ambiguous. - Samsara (talk • contribs) 18:11, 3 September 2006 (UTC)
  • Object unfortunately. The referencing is still inadiquite in places (I put some fact tags a couple days ago...) and I agree with BMF81. If there was ever a case for good article, this would be a decent one :). It is such an important article which I encourage the editors to keep working on, because they are doing a really good job on a tough subject :). RN 05:53, 5 September 2006 (UTC)