mic_none

Wikipedia:Articles for deletion/System bus model Source: en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/System_bus_model

The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.

The result was no consensus. There is no clear consensus for deletion, keeping or merging (the latter not being helped by a suitable target which has reached consensus view). Incidentally, although a merge with von Neumann architecture was indeed proposed on the article's talk page, there is no clear agreement with it there, but that is a matter for further discussion at the talk page - a note on the Wikipedia talk:WikiProject Computing mentioning the merge proposal, and asking for more input would not be a bad idea. PhantomSteve/talk|contribs\ 20:32, 22 May 2011 (UTC)[reply]

System bus model

[edit]
System bus model (edit | talk | history | protect | delete | links | watch | logs | views) – (View log)
(Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)

I PROD'ed this article last October, but it was contested. The reason given was: rm PROD - this is a variant on the Von Neumann architecture, and I found alot of books with it, not just two - it describes the standard configuration of an x86/68k/etc machine. A discussion subsequently took place at the article's talk page; Talk:System bus model. A merge to von Neumann architecture was proposed, but has not been carried out.

My rationale for deleting this article is basically the same as it was last year. There is no indication that its topic, the system bus model, is notable. My original rationale, is still as relevant today as it was then. I am of the opinion that the rebuttal of the editor who contested the PROD inadequatly addressed the concerns I outlined in the PROD rationale in two ways. Firstly, it was asserted that there was coverage; not shown that there was; either by stating how sources were found so it could be independently verified (in the PROD, I detailed how I looked for coverage), or by listing the sources themselves. Secondly, I get the impression that because the topic is associated with the von Neumann architecture, those opposing deletion are doing so on the basis that its notability is presumed and/or inherited. For the latter case, WP:NOTINHERITED sums up why it is not a good idea.

As previously mentioned, the article was proposed for merging. In practice, problems with articles that could be fixed by normal editing (which includes merging) should not be nominated for deletion. In my opinion, this article cannot be fixed by merging because its the coverage its topic has is trivial compared to that of the von Neumann architecture, which would result in undue weight if merged.

Lastly, the article is referenced to sources that are reliable, but does coverage of the article's topic in these sources meet WP:N's requirement of non-trivial coverage? I can only assess the first source, a book, since it is the only one that can be previewed at Google Books. In that book, coverage begins at the bottom of page 31, and resumes half-way down page 32 and ends shortly after. The amount of text on the topic is around one typically-sized paragraph. It is clear that the first source is not significant coverage of the topic. For something which claims to be an evolution of the von Neumann architecture, I expect there to be substaintial coverage. This was raised at the article's talk page, but no direct response was given. Rilak (talk) 07:33, 30 April 2011 (UTC)[reply]

You've put a lot of work and careful thought into this AfD. I wish that energy and intelligence could somehow be funneled into the articles in question. I apologize for dropping the ball on the merge. I still beleive this can be done sucessfully and at least one other editor has concurred. I will try to find time to do it. I'm confident that the WP:UNDUE and WP:N issues can be dealt with by the editors of Von Neumann architecture. As far as the AfD is concerned, I favor whatever approach most expeditiously gets rid of this article while salvaging anything of value. --Kvng (talk) 15:12, 30 April 2011 (UTC)[reply]
Note: This debate has been included in the list of Computing-related deletion discussions. -- • Gene93k (talk) 17:17, 30 April 2011 (UTC)[reply]
  • Comment Delete Without further researching the subject on other sources, let us first read the relevant sections of the references. The second reference of the article ([1]) starts its 1.4 The System Bus Model section, just after the 1.3 The Von Neumann Model section with the (page 5):"Although the von Neumann model prevails in modern computers, it has been streamlined.", which suggests that the so-called System bus model is the modern version of the Von Neumann model. The image on the article page is also a re-drawn copy of the "Figure 1-3 The system bus model of a computer system. [Contributed by Donald Chiarulli, Univ. Pittsburgh.]", page 5 from the same source. The term System bus model is only mentioned once in the first reference ([2]), inside the 1.7 Von Neumann Model section and before 1.8 Non-Von Neumann Model section, with "This architecture (ed. Von Neumann) has also been streamlined into what is currently called the system bus model" (page 32). This reference is published six years later than the other. We can conclude that these two references, both of which are published books about computer architecture, describes system bus model as a modern version of the Von Neumann model. Thus, merging the article into Von Neumann Model is plausible if it doesnt fail WP:WEIGHT.

    When we check the web for other sources, we can see that the majority of the sources are direct copies from either of these two books (mostly the first) or from the course pages of the universities that use these books as the course textbooks. Some articles mention it (like [3]) but they either don't discuss it deeply or they're not directly related to the subject. This is (if I understand correctly) the main concern of the nominator per WP:GNG.

    This lack of sources except these books (that are mainly used as textbooks) suggests that the term is in fact coined to address a general modern version of the Von Neumann model. The main reference describes the model further by (page 5-6): ""Most important to the system bus model, the communications among the components are by means of a shared pathway called the system bus, which is made up of the(..) the system bus is actually a group of individual busses classified by their function." The system bus page redirects to front-side bus, probably because of the sources like ([4]). There is an old discussion on the talk page which address the difference between two terms (Talk:Front-side_bus#System_Bus).

    I think the article strongly needs {{expert-subject}}. Nimuaq (talk) 01:58, 1 May 2011 (UTC)[reply]
To clarify, I never challenged the existence of the system bus model. I challenge its notability and whether it would be due weight to merge it into von Neumann architecture. After all, the claim that the von Neumann architecture has been streamlined, and that the resulting streamlined version of it is called the system bus model, has only been shown to be supported (so far) by the two textbooks mentioned as references in the article, and their accompanying materials such as lecture presentations. Now, consider the amount of coverage the von Neumann architecture has received: two textbooks (including one whose coverage of the topic is a single paragraph) versus the 6,210 results Google Books found containing the term "von Neumann architecture" but without "system bus model".
Finally, regarding the IEEE Design and Test of Computers article, the term "system bus model" is referring to a behavioral model of the system bus. It is not claiming that it is an evolution of the von Neumann architecture. As I mentioned in my original PROD notice, there are far more instances where the term has been used to refer to a behavioral or electrical model of a system bus than what the article claims. Rilak (talk) 04:00, 1 May 2011 (UTC)[reply]
I know you haven't challenged the existence, but rather challenged the notability, as I already stated above. I can also easily verify your finding of "more instances where the term has been used to refer to a behavioral or electrical model of a system bus" compared to the two sources (published books) above. I searched for the system bus (not the model) this time together with the Von Neumann architecture and found out that the majority of the book sources discuss the system bus, usually under Von Neumann architecture, with the same definition and the same structure of which the source of the article discusses the system bus model (one of the recent ones of the several examples here) Thus, I think the authors of the book ([5]) used the system bus model term to explain the system bus and the modern interpretation of the Von Neumann model in one section.
More than ten years after the books first edition, this model is still not widely discussed under that particular name, so I think you're right when you say it fails WP:GNG. For the merging articles, I -now- think that since all of the content the article offers is discussed under the system bus in numerous sources, rather than a merger, the relevant parts should either be moved to the article System bus or the page should be created with other particular sources, where a link can be created for the front-side bus article, for it might also refer to it. Nimuaq (talk) 10:14, 1 May 2011 (UTC)[reply]
Comment I think that would be fair. I've seen the image on a number of other books too, some of which are published before the main source of the System bus model. But I'm not sure about if its a common diagram (which can be attributed to common knowledge) or specially drawn since the source its taken says "Contributed by Donald Chiarulli, Univ. Pittsburgh.", as I explained above. The diagram is either first drawn by Donald Chiarulli or the authors did not have such a diagram and they couldn't draw it themselves. If its the latter, it is a bit odd since the authors of the book first used the term System bus model yet they cant draw that basic diagram and still need someone to contribute it for them. Nimuaq (talk) 08:36, 5 May 2011 (UTC)[reply]
The diagram is not up for deletion; it's not part of the article but an entry in the file namespace: [6]. The diagram is also incorrect. The CPU should be shown to contain an ALU, control unit, and registers. I agree that system bus should be turned into an article, but I don't think merging content from this article into there is a good idea since the content is very specific to the system bus model. Rilak (talk) 08:56, 5 May 2011 (UTC)[reply]
Only the introduction is specific to system bus model, the content in Communications section discusses the system bus like other sources: [7], [8]. I think those relevant sections can be easily merged into a separate article for the system bus. Nimuaq (talk) 21:19, 5 May 2011 (UTC)[reply]
As far as I am concerned, the entire section in question is specific to the system bus model since the source it cites has that coverage for the purpose of discussing the system bus model. Quite frankly, I don't see any of redeeming qualities that you and Nowicki see in that section. The act of merging the content might be easy, but fixing up the content is not. The section contradicts itself. It says that the I/O bus is part of the system bus, and then it claims more sophisticated architectures have separate I/O buses. It also makes some claims which are quite questionable; such as the power bus being part of the system bus; and that the address bus can be used by the receiver to determine the transmitter (the command bus would be used instead for this purpose). I think starting anew; without the structural baggage of the section in question (the section was intended to talk about the system bus model; not system buses in general; and is structured according) would be preferable in regard to editing effort and the resulting quality. Rilak (talk) 08:10, 6 May 2011 (UTC)[reply]
I took a second look to the article, the main source ([9]) does not state the I/O bus and the power bus being a part of the system bus. It just discusses those separately inside the "System bus model" section. Thanks for pointing out, I agree that starting a new article without a merge is a better solution. Nimuaq (talk) 13:07, 9 May 2011 (UTC)[reply]
Relisted to generate a more thorough discussion so a clearer consensus may be reached.
Please add new comments below this notice. Thanks,  Sandstein  07:02, 8 May 2011 (UTC)[reply]
Relisted to generate a more thorough discussion so a clearer consensus may be reached.
Please add new comments below this notice. Thanks, Spartaz Humbug! 15:21, 8 May 2011 (UTC)[reply]
The article partly describes the system bus in a general sense; and it only does so to permit its primary topic, which is the system bus model, to be discussed. I think the amount and depth of discussion regarding this article's worthiness for inclusion deserves a better keep or merge argument; specifically responses to the concerns regarding the notability of the topic; and the quality and utility of the article's content. Rilak (talk) 09:07, 10 May 2011 (UTC)[reply]
Either— reading it again, I would agree just deleting this one would also be reasonable. My main problem is that it seems to confuse an "architecture" like the von Neumann, with an implementation style, which I would say a system bus is. Not a "model"? That is, a system bus was one popular way of organizing a von Neumann architecture, as computers went from multiple racks to a single box with modular printed circuit booards. The article probably should be in past tense, since it generally popular from, say the 1970s through the 1990s. And then the more modern front-side bus and HyperTransport etc. with separate memory and I/O busses as the follow-ons need to be clearly stated. W Nowicki (talk) 17:50, 10 May 2011 (UTC)[reply]
Keep and take over redirect of System bus, as IMHO System bus is something significantly more specific than generic Bus (computing), specific enough to warrant a separate article. Ipsign (talk) 11:52, 14 May 2011 (UTC)[reply]
Last year, it was keep, because the article is about the von Neumann architecture (which it is not). This year, it's keep, because the article is about the system bus (which it is not). Sigh. Rilak (talk) 02:31, 15 May 2011 (UTC)[reply]
WP:ATD: "If the page can be improved, this should be solved through regular editing, rather than deletion". Ipsign (talk) 13:44, 16 May 2011 (UTC)[reply]
Yes, if a page can be improved through regular editing then it should not be nominated for deletion. Thank you for stating something that I knew before you even made your first "argument" in this discussion. Now, instead of using policy as a thought-terminating cliche and insinuating that I haven't got the slightest clue as to what I am doing; how about you make an argument as to why you think this page can be solved through regular editing, with responses to the existing arguments that say it is not possible and not desirable? Rilak (talk) 05:03, 17 May 2011 (UTC)[reply]
Personal attacks (saying that I am "insinuating" IMHO easily qualifies on this account, in particular violating WP:AVOIDYOU) will lead us nowhere. Why the article can be improved - this article contains information (in particular, a diagram) which doesn't seem to exist anywhere else on Wikipedia, and I think this information clearly belongs to Wikipedia. In addition, "System bus" is an obviously notable concept (have been in widespread use for about 20 years - references can be easily found), and is not really covered anywhere else. Argument saying something like "it is not an article about the system bus", is not a valid one per quote from WP:ATD which I've provided (if it is not an article on system bus, let's improve it and make it an article on system bus). While providing further arguments is possible, it looks pretty much pointless, as I don't see how discussion in this AfD can come even close to "rough consensus" required for deletion. Ipsign (talk) 08:12, 17 May 2011 (UTC)[reply]
It has been well established that the diagram is not up for deletion; that the "valuable" diagram is wrong (last year at Talk:System bus model!); and that the "unique" content is of poor quality (and is deemed to be not salvageable). My opposition of your argument that the article should be kept because it is about the system bus is perfectly valid because the article is not about the system bus (or the von Neumann architecture); it is about the system bus model. If you want coverage about the system bus, then correct place to add it is system bus. WP:ATD says nothing about rewriting an article about one topic to another under an incorrect title.
Furthermore, I must add that the mergers and keepers seem to be opposing deletion of this page because they want an article on the system bus. Why they are commenting here, in this AfD, escapes my comprehension, since the amount of effort they have expended on discussion here could have been spent turning system bus into an article. It seems that they view me as deleting coverage of the system bus from Wikipedia. That cannot be further from the truth. I want coverage of the system bus model deleted, which only happens to contains tangential coverage of the system bus. Also, the system bus model and system bus pages are independent of each other; content can be added to the latter page at any time and be safe from whatever the outcome of this AfD. The logic of some of the opposing arguments can be compared to fighting tooth and nail for keeping an article on about an non-notable expansion bus because the article, in addition to specific coverage on the bus, also happens to include a basic definition of expansion buses in general for the purposes of accessibility and context. Rilak (talk) 08:58, 17 May 2011 (UTC)[reply]
If there would have been an article about the system bus which this could have been redirected to, I would have voted "redirect". I you really want this gone, perhaps try remedying this first? —Ruud 12:06, 16 May 2011 (UTC)[reply]
I have no obligation to create articles that you desire. AfD is not a game where editors compel each other to perform favors in return for support and votes. Although I know of no policy or guideline that states this (since I have no interest in such games), I believe that my position on what AfD is not is the community norm. Rilak (talk) 05:03, 17 May 2011 (UTC)[reply]
Disturbing. —Ruud 22:51, 17 May 2011 (UTC)[reply]
  • Merge with Bus (computing). Except for a couple of obscure textbooks, nobody treats this as a topic of major importance, or sufficiently distinct from that of bus interconnect. There's no great theoretical concept here, just some pedantic wankery from a couple of obscure academic authors. Tijfo098 (talk) 10:55, 18 May 2011 (UTC)[reply]
    • I've looked some more through the textbook that was used to write most of this article, and the problem is not that it's fringe, but that it's simply too introductory. There's only a brief treatment of buses in real systems for instance on pp. 319-320. Tijfo098 (talk) 16:44, 18 May 2011 (UTC)[reply]
The system bus model is claimed to be an evolution of the von Neumann architecture (the 1st and 5th sentences in the article) whose sub-units (CPU, memory, and I/O) are connected by a one central bus, the system bus (the 2nd sentence). The system bus model is not about the system bus or any other bus (although the article does discuss buses, but only in the context of the system bus model), so why should the article be merged into Bus (computing)? Rilak (talk) 05:44, 19 May 2011 (UTC)[reply]
Actually, the system bus model is a fringe topic because the claim that there is an evolved form of the von Neumann architecture called the system bus model (which uses a system bus to connect the CPU, memory and I/O) is completely different to the claim that some von Neumann architecture machines use a system bus to connect the CPU, memory and I/O. Rilak (talk) 05:58, 19 May 2011 (UTC)[reply]
The problem is that "von Neumann architecture" is a fuzzily defined topic. So, depending how you choose to define that, "System bus model" is or isn't an extension. In fact, nothing in this interconnect model implies that the program is stored in the memory. There are some textbooks that don't present it as an extension of VN, but only as a simple (or simplistic) overview of bus-based computers (this one also has a DMA controller in the pic), (this has a ROM too). It's not really an encyclopedia topic by itself. Tijfo098 (talk) 11:17, 19 May 2011 (UTC)[reply]
Whether the system bus model is or is not an extension of the VNA, depending on how one wishes to define it, or interpret its definition based on the context of how VNA is used in any context, is irrelevant. What is relevant is how reliable sources discuss the system bus model in the context of the VNA; and the sources in the article define it as an extension of the VNA in terms of how the subunits are connected. Therefore, we must discuss whether to include coverage of the system bus model in Wikipedia or not as defined by the sources in the article, and not what we deem are reasonable viewpoints because that would be improper synthesis and original research. Additionally, the textbooks that describe a VNA computer using a system bus cannot be portrayed as coverage of the system bus model because no source has called such an implementation a system bus model; nor has any source stated that system bus and system bus model are synonymous terms. Rilak (talk) 01:17, 21 May 2011 (UTC)[reply]
  • Delete I have been out of town for a week but am back now. An introductory book would be fine for a basic article if it were up to date. But from the author's web site the 1999 book has been replaced with a 2007 book already. I still have a basic problem calling it a "model". It is not an "architectural model" in the narrow sense usually used in computer science. For example, the PDP-8 had some implementations that used a system bus (the PDP-8/E) while others (like the PDP-8/I) did not, even though they were binary compatible and therefore the same "architectural model". Early PCs also used a single system bus, while later ones evolved into dual independent busses, and then modern architectures with even more busses. Here is a proposal that I might have time to start today:
  1. redo the diagram in svg without the "ALU, control, register" confusion
  2. expand system bus to an article about the use of a single bus that was popular in the 1970s and 80s.
  3. narrow front-side bus a bit to apply only to the Intel-compatible design of the 1990s-2000s.
  4. add links from other related articles to the system bus article, putting it into both historical and architectural context
  5. delete this one, after replacing the one link to it

W Nowicki (talk) 17:25, 20 May 2011 (UTC)[reply]

Well, I ran into a problem in step 1. I drew a diagram in Inkscape, but uploading it to commons as File:Computer system bus.svg results in big black blotches. Some incompatibility between Inkscape and WIkimedia renderer? I also found File:Computer buses.svg which might work in the interim. It is a bit misleading since it shows data going into ROM, while it generally only comes out (that is what makes it read-only!). But perhaps a decent start. W Nowicki (talk) 20:23, 20 May 2011 (UTC)[reply]
The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.