Wikipedia:Articles for deletion/Ch interpreter (2nd nomination)

Ch interpreter

AfDs for this article:

• Articles for deletion/Ch interpreter
• Articles for deletion/Ch interpreter (2nd nomination)

Ch interpreter (edit | talk | history | protect | delete | links | watch | logs | views) – (View log)

(Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)

Relisting per Wikipedia:Deletion review/Log/2011 April 13. Primary argument for deletion appears to be lack of notability. Procedural nomination only, I am neutral. T. Canens (talk) 05:03, 1 May 2011 (UTC)

Note: This debate has been included in the list of Software-related deletion discussions. -- Cybercobra (talk) 07:26, 1 May 2011 (UTC)

Note: This debate has been included in the list of Computing-related deletion discussions. -- Cybercobra (talk) 07:26, 1 May 2011 (UTC)

• Delete. To establish notability, WP:GNG and WP:CORPDEPTH ask for multiple, reliable, independent, non-trivial sources. I don't think they're there. The requirement for independence eliminates the papers by Harry Cheng, the creator of Ch, plus those of Matt Campbell and Zhaoqing Wang, both of whom have co-authored papers with Harry Cheng. (In addition, Cheng was the director of the UC Davis Integration Engineering Laboratory where Campbell was a master's candidate.)

  Two more of the sources are also not useful because they're not actually commenting on Ch. Both the Glassborow and the Huber articles are actually book reviews of Cheng's textbook for teaching C. Yes, it apparently comes with a CD containing Ch, but I don't think that's what they were reviewing. I think they were a reviewing a book about how to teach C.

  The independent secondary sources we're left with are the Heller, Wilson and Gary Wang articles. They're weak.

  The Heller article is a First Impression web-only column article, not featured review and it says so. (Columnists have a lot of freedom to say anything they like; a featured review is one the editors stand behind.) A First Impression is just that; it just came out and here's what it looks like, who knows if it's going anywhere. Also, this was a web-only column, never in print. From The controversial end of Byte, the big print magazine most of us remember "ceased publication with the July 1998 issue, laid off all the staff and shut down Byte's rather large product-testing lab ... In 1999, CMP revived Byte as a web-only publication." The Wilson article is similar; it's a weak column mention. It made it into print, but in a much smaller circulation, special interest publication, the author isn't anywhere near as well-known as Heller and the column isn't actually about Ch. It's "looking instead at the Open-RJ open-source library (http://www.openrj.org/), along with its mapping to the Ch and C++.NET languages."

  The Gary Wang article is the weakest of the bunch. It's advertising being passed off as a legitimate review. It cannot plausibly have been peer-reviewed. Yes, it did appear in Spectrum but the whole point of asking for reliable sources should be more than just saying, okay, as long as the article appeared somewhere that usually exercises good editorial control, that's enough. This is an article, that if you simply read it cannot possibly be considered to be reliable and balanced. Characteristic of purely promotional writing, it fails to identify even a single shortcoming or any possible way in which the product might be improved or any purpose for which it wouldn't be absolutely fabulous. It's not surprising to see that SoftIntegration quotes Gary Wang in their press release predating the Spectrum article. What is surprising is that Spectrum would allow this to be printed.

  When I take these very weak sources together with the aggressive history of spamming Wikipedia (e.g., here and in the spammy external links) and Amazon with SPAs, the endless relitigating (we're now into the 4th week of debate on this page, prompting me to wonder if this what Spectrum also gave into!) and the unsavory canvassing in earlier rounds, I think the right answer is delete (but I respect consensus even when I'm on the wrong side of it.) Msnicki (talk) 07:43, 1 May 2011 (UTC)

Can you substantiate where papers authored by an interested party, but published by an editorially independent reliable source, are somehow inappropriate? Remember, our guidelines look at who paid for the publication of a paper, not who authored it. Jclemens (talk) 08:04, 1 May 2011 (UTC)

• WP:CORPDEPTH states that sources used to establish notability may not include "any material written by the organization, its members, or sources closely associated with it". Msnicki (talk) 14:26, 1 May 2011 (UTC)

Why are you applying Wikipedia:Notability (organizations and companies) to an article that is about neither an organization nor a company? —chaos5023 (talk) 14:58, 1 May 2011 (UTC)

• Because that's what it says to do. From WP:CORPDEPTH, Primary criteria: "A company, corporation, organization, school, team, religion, group, product, or service is notable if it has been the subject of significant coverage in secondary sources. Such sources must be reliable, and independent of the subject. A single independent source is almost never sufficient for demonstrating the notability of an organization." Msnicki (talk) 15:04, 1 May 2011 (UTC)

Ah, okay. Missed that; thanks! —chaos5023 (talk) 15:11, 1 May 2011 (UTC)

WP:CORPDEPTH Provides the following context for the above quote: "A primary test of notability is whether people independent of the subject itself (or its manufacturer, creator, or vendor) have actually considered the company, corporation, product or service notable enough that they have written and published non-trivial, non-routine works that focus upon it." Looking at the entire section and the nature opf all the other example given, it is clear what is being prohibited. Guy Macon (talk) 16:34, 1 May 2011 (UTC)

You keep repeating that Hubber's article is just a book review even though I have pointed that it is review of both software and book. Let me copy and paste the author's quote found from the vendor's website. "The bottom line for a post-use review is one question: Would I use the text again? In the case of C for Engineers and Scientists. An Interpretive Approach, the answer is yes. Students agreed with this assessment; in their post-use survey, when asked whether the book should be used in future course offerings, all students answered "yes." After using this textbook and the Ch interpreter, I have a new respect for the role of interpreters as an alternative to traditional compilers for introducing computer programming." One question: 100% students in the class like Harry Cheng's book when using Ch software for learning based on the review. Do you still think it is just a book review? Chuser (talk) — Chuser (talk • contribs) has made few or no other edits outside this topic.

• Comment: Save for Msnicki (talk · contribs), who has already participated in this AfD, I have notified the participants of Wikipedia:Deletion review/Log/2011 April 13#Ch interpreter and Wikipedia:Articles for deletion/Ch interpreter about the amended closure and the AfD. Cunard (talk) 07:54, 1 May 2011 (UTC)

• Keep The reason we prefer RS'es with independent editorial control is that it makes all this essentially moot. Dr. Dobbs, BYTE, and IEEE Spectrum are perfectly fine RS'es. The software has received multiple, non-trivial coverage in independent RS magazines, and thus meets the GNG. The associations of the software authors to these RS articles is not an issue, unless one wants to impeach the editorial independence of these publications. Jclemens (talk) 08:00, 1 May 2011 (UTC)

• WP:GNG and WP:CORPDEPTH both require that sources used to establish notability must be independent of the subject. They do not say independence is unnecessary just so long as the magazine that published it was independent. The whole thing, starting with the author, has to be independent. Msnicki (talk) 14:26, 1 May 2011 (UTC)

In my opinion, the above is a overly narrow and legalistic interpretation of WP:CORPDEPTH. The spirit of the policy is clearly explained in WP:CORPDEPTH itself: "Self-promotion and product placement are not routes to qualifying for an encyclopaedia article." Guy Macon (talk) 16:40, 1 May 2011 (UTC)

• Keep. There are in-depth articles in two top mass-market computing magazines: Byte Magazine (2001), Dr. Dobb's Journal (2004). There are also in-depth articles in less prominent magazines, as well as numerous less focused mentions in top magazines that together span a decade. They are all cited in the article. and together go well above the standard required by GNG.

It is true that peer reviewed research papers do not automatically establish notability if they only come from one small workgroup or a tiny set of collaborators. But we don't even have to examine whether this is the case here. Hans Adler 08:35, 1 May 2011 (UTC)

• Keep The Heller article is lengthy, substantial, and in a solid publication. The Wilson article also offers expansive coverage of the topic at-hand. Clearly, the Wang article, by nature of its inclusion in an edited publication, is valid. We decry Wiki articles for having peacock terms; we don't generally turn away articles for being based on reliable sources that happen to use what might, in an encyclopedia article, be called peacock terms. It is neither surprising nor relevant that somebody would quote a positive article about their product in their product's marketing materials. Keep. ɠǀɳ̩ςεΝɡbomb 09:32, 1 May 2011 (UTC)

• Keep The above mentioned sources are enough to establish notability. In addition, Mactech has a review, the Ch interpreter is very commonly embedded in LabVIEW, and is sometimes used to program Lego Mindstorms. Guy Macon (talk) 09:54, 1 May 2011 (UTC)

• The Mactech "review" is by Matt Campbell. He is not independent, having co-authored a paper on Ch with Harry Cheng and having received his Masters while studying under Cheng. (Read the bio at the end of the paper.) This is even more blatant than the Spectrum article; they've allowed Campbell to review a product he'd worked on himself! Undoubtedly,It seems likely this was not disclosed. Msnicki (talk) 14:26, 1 May 2011 (UTC)

I was about to write that you should not make serious accusations such as "Undoubtedly, this was not disclosed" without actual evidence of wrongdoing, but then I noticed that the Mactech review was published in 2003 while the paper was submitted in 2005 and published in 2006. Not having a time machine or reliable precognition, I think it safe to assume that he didn't disclose something that was several years in the future.

If I review a product on my own webpage, that does not show notability - pretty much every garage band has a website. If I review a product and Mactech or Byte publishes it, the fact that a well-know computer magazine with a large subscription base made the editorial decision to publish a review of the software is, in itself, evidence of prior notability (they don't review products unless they are notable) and also establishes notability (even if it wasn't notable before Mactech or Byte published the review, it became notable at that point). Evidence that the author has a conflict of interest calls the accuracy of the review into question, not the notability of the product reviewed. Unless you have evidence that the creator of the software bribed the editors of Mactech and Byte, you must presume that the decision to publish the review was not made by the author of the review. Guy Macon (talk) 16:24, 1 May 2011 (UTC)

• The association is not the paper in the future but that at the time he wrote the review, Campbell was working on Ch (as documented in that later paper co-authored with Cheng) while he pursued his masters degree under Cheng. Msnicki (talk) 16:42, 1 May 2011 (UTC)

• Keep Weak article on a perfectly horrible topic. I ask myself why such a thing existed post 2000, why it was still thought to be a good idea by then (we no longer had the grief of the 1980s), why it was based on C of all things and a comment like, "Students in disciplines other than computer science can just learn C" gives me the crawling horrors. In particular, we seem to have a long article here that still gives a poorer flavour of what it's really about and why it came to be, than the IEEE paper's abstract does.

That all said though, when a product gets in-depth coverage like this at the level of journals like the IEEE Spectrum (and I hope Wtshymanski will accept that Spectrum isn't obscure), then that's notability, as we define it. I don't like it, but I can't give reason to delete it. Andy Dingley (talk) 10:21, 1 May 2011 (UTC)

• Delete same reasons as before - WP:RS wants a knowledgeable source, which takes out the Huber source. Other editors appear to be arguing that Wang's close associates and even Cheng's papers constitute sources for the sake of notability. TEDickey (talk) 10:33, 1 May 2011 (UTC)

From my viewpoint, the Wang paper in IEEE Spectrum is a good ref for notability. Is there a reason why you would discount it? Even if there's a connection between Wang & Cheng, I trust the editors of IEEE Spectrum to publish an article free of any bias to a level that would be a problem for us. Andy Dingley (talk) 12:07, 1 May 2011 (UTC)

The point was already made in the previous round: a journal will print based on their perceived user-interest (whether novelty or just topicality), will peer-review full articles but not necessarily other content, and bias of the type we're discussing is not a factor in the choice whether to accept or reject. TEDickey (talk) 13:02, 1 May 2011 (UTC)

I don't believe we need full-blown peer review just to establish notability. I trust the editors of Spectrum as an adequate delegate for that. Andy Dingley (talk) 17:19, 1 May 2011 (UTC)

• Keep Seems a very clear-cut case at this point. Collect (talk) 12:59, 1 May 2011 (UTC)

• Keep - The articles about ch have appeared in respected journals. When evaluating articles about little software products we are often struggling to find any coverage at all from a reliable source; this is more than enough. EdJohnston (talk) 13:12, 1 May 2011 (UTC)

• Keep - Plenty of valid, notability-establishing cites. Wang's connection to Cheng is irrelevant; we're considering IEEE Spectrum, not Wang. The only time the author is the source is in a WP:SPS situation. We also need not rake source citation content over the coals for reasons to disqualify it, like speaking too positively of its topic, that are conjured out of nowhere rather than appearing in policy or guidelines; the question we're asking is whether it provides significant coverage, i.e. enough to provide a meaningful chunk of Wikipedia article content. These conditions are sufficiently met. —chaos5023 (talk) 14:04, 1 May 2011 (UTC)

• Comment I just added additional two independent sources in Ch Interpreter article. One is written by Buff Here which was mentioned in the previous AFD and another is written by Li-rong Wang from Tokyo Inst. of Technology here. The second article is a new article which talks about the application of mobile c based on ch.

In addition, I have two questions and appreciate anybody's comments. One is about Professor Tom Hubber's article in the IEEE Computing in Science and Engineering. TEDickey believes it should be removed from reliable source because the author is not knowledgeable and it was mentioned in every AFD. My comment is that I believe that it is IEEE editors's job to determine if he is knowledgeable about the subject to publish his article. WP:RS states: "Reliable sources may be published materials with a reliable publication process, authors who are regarded as authoritative in relation to the subject, or both." It is "or", not "and". In other words, RS can be established if one of the above two conditions is met. Also from WP:RS, "When available, academic and peer-reviewed publications, scholarly monographs, and textbooks are usually the most reliable sources". The IEEE Computing in Science and Engineering is one of the most authoritative academic and peer-reviewed publications. It has been repeated twice already. Can anybody comment if Hubber's article is suitable as RS?

Another question is about the link from Ch shell to C shell. Msnicki believes it is a spamming for such a link. C in C shell (csh) means C language. It talks C like language for shell programming. The link Ch shell (ch) talks the shell programming in C. It is C not something like C. Ch shell use the C language syntax for shell programming. Anybody think it is relevant and shares useful information in wikipedia with such a link? The reason I am asking is that Msnicki mentions it every time in AFD. Thanks. Chuser (talk) 05:47, 2 May 2011 (UTC) — Chuser (talk • contribs) has made few or no other edits outside this topic.

• What made your link spam was the sheer brazenness of putting such a tenuously-related commercial product at the very top of the list of C shell#External links, ahead of Bill Joy's own famous paper, An Introduction to the C shell. Msnicki (talk) 17:37, 2 May 2011 (UTC)

IEEE Computing is a reliable source. Our opinions about whether the author is knowledgeable are irrelevant.

The link is appropriate. The two are clearly related.

The remedy for an objection about it being on the top of the list is to move it lower. The above comment makes it sound as if your assessment of the link would have been different had it been placed lower. The placement has nothing to do with whether it belongs or not.

I would have waited for the result of the RFD, deleted the link to http://www.softintegration.com/docs/ch/shell/ from C shell if Ch interpreter was deleted. and considered replacing it with a See Also to Ch interpreter if Ch interpreter was retained. Guy Macon (talk) 18:20, 2 May 2011 (UTC)

• Unlike you, apparently, I actually downloaded and played with the product two months ago, so I know that no matter what's claimed here, Ch is a C language interpreter, not a C shell. (If you don't know the difference, maybe you aren't ready to prescribe what should be done to that article.) Msnicki (talk) 18:30, 2 May 2011 (UTC)

[sentence deleted] Nobody claimed that Ch is a C shell, and your WP:OR is irrelevant. Guy Macon (talk) 20:54, 2 May 2011 (UTC)

• And here's the link to the Wikiquette complaint you're obviously still upset about. Msnicki (talk) 21:02, 2 May 2011 (UTC)

Do you have a comment about the the proposed deletion of Ch interpreter article? That's what this page is for. Guy Macon (talk) 21:24, 2 May 2011 (UTC)

• I apologize. Allow me to restate. When I saw the external link added to the C shell article, I read the Ch interpreter article, which clearly identifies the product as a C language interpreter, not a C shell, and I downloaded the product to try it. There's a difference between inserting original research into WP and following the links provided to verify sources and claims. That's why we ask for sources, so that anyone can verify anything. I just tried to verify it. What I found was that Ch is an interactive C language interpreter that can also run commands and it's packaged with a lot of GNU utilities. But the actual language is C, not C shell (which is quite different despite the name), and that it was obvious this was not a relevant link per WP:ELNO. Msnicki (talk) 22:31, 2 May 2011 (UTC)

The above argument is compelling. Msnicki is right; there should not be any links (internal or external) from C shell to Ch interpreter or to any other implementation of the C/C++ language. The two are just not closely related enough other than having similar names, and even that is pretty much a historical accident. Can we add a link from C shell to Seashell? (just kidding.) Guy Macon (talk) 01:37, 3 May 2011 (UTC)

I saw that ch shell borrowed many C shell features from the link here. Many shell features such as filename wildcarding, piping, here documents, command substitution, history, login shell, command completion etc are supported in ch. There is some kinda of relationship but agree with you both that ch is a different language from csh. I am fine with no links. Thanks for the clarification. (note to Guymacon: welcome to modify my format and any other errors since I am new.) Chuser (talk) 06:49, 3 May 2011 (UTC) — Chuser (talk • contribs) has made few or no other edits outside this topic.

• This is so not true. For my testing, I used the 64-bit 6.3.0 version on Windows 7. It's a free download so anyone can check this and I hope you will; if you used a Unix shell (any of them, it doesn't matter) before, it'll take you maybe 15 or 20 minutes start to finish to verify what I'm telling you: Ch is not even close to any Unix shell.

  From pg 38 of the User Guide, if Ch can't otherwise recognize what's typed, it tries to run it as an ordinary external command. But the facilities available for running external commands are very limited. Think cmd.exe but brain-damaged.

  It can do piping or redirection of stdout only; if you redirect or pipe stdout, it loses stderr. The documentation claims (pg 71) that stderr can be redirected with constructs like 2>&1 but when I tried it, it was busted. It cannot do command substitution in an ordinary command (e.g., echo `date`); the backquoted string is taken as a literal. To get command substitution to work requires (see pg 68) that you first store it to a variable (e.g., string_t s = `date`) and then you can use it as $s (but you're supposed to free() the variable after you're done.) Here documents are not supported. Period. It claims (pg 73) to support background processes using & but this also is busted: It doesn't work on anything but a simple command; e.g., date | wc & doesn't work because the pipeline doesn't get set up.

  It has the usual C language constructs for iteration and condition testing, but they do not appear to work from the command line. (The examples on pg 141 do not work.)

  Ch claims (pg 66) to support wildcarding (it calls it "filename substitution") but actually, it depends (at least on Windows) on the child doing the wildcarding when an external command is invoked for everything except the tilde (home directory). (You can discover the behavior by typing echo *.) That's conventional on Windows (developers routinely link in a library routine setargv.obj to do the wildcarding prior to entry to main()) but it means there's no support for ranges (e.g., hello.[ch]) or multi-level wildcarding (e.g., */*.c) supported by a Unix shell (like Bash or tcsh, even on Windows.) History and filename and command completion appears to be done with Readline.

  Finally, the whole thing is very fragile (e.g., try typing for (int i = 0; i < 5; i++) at the Ch prompt and see what happens!) Msnicki (talk) 14:35, 3 May 2011 (UTC)

True or not, this has no relevance to notability (Oh, a claimed "shell" that's just a paltry sub-set - it's not as if we haven't seen plenty of those before).

If you want to delete this, you would need to address why each and every one of the sources is invalid, starting with the IEEE pieces. Andy Dingley (talk) 14:54, 3 May 2011 (UTC)

Andy, please see below. I take for granted the consensus this time will be to keep. I suggested we close early (over my own dissent) precisely because I think we're wasting our time discussing something that's pretty much a foregone conclusion. If you also think we're wasting our time, then support my suggestion for the speedy keep.

Otoh, as long as we're here to discuss this stuff, it seems reasonable to discuss the claims made in the article, including the claim, "Ch is a C-compatible shell similar to C-shell (csh)", especially when it's brought up here. I've already addressed the sources, including the IEEE article (above) but I presume you weren't convinced and it hardly seems useful to repeat the same argument; if you didn't agree last time, why would you agree now? But I will point out that the IEEE article also contains the testably false claim, "Another rewarding feature is that Ch is also a genuine C shell." This is consistent with my criticism earlier that this article is not reliable and that it's pure advertising masquerading as a review. Msnicki (talk) 15:02, 3 May 2011 (UTC)

To claim Ch is not close to any Unix shell, you should run Ch in Unix to verify, not from Windows. From the Unix machines I tried all commands such as command substitution in an ordinary command "echo `date`", and "date|wc &", multi wildcard */* etc. All works fine. In windows, it inherited the cmd.exe features as mentioned by Heller's article. I don't think anybody cares & to put the windows command to run in backend, though & works if not association with pipe as I tried. Let me put this way. If Ch is verified as a genuine C shell (C language shell), not C-like C shell (csh), will you agree to add the link under csh? C in C shell (csh) is not used by accident. It can be more easily adopted in early days when csh is promoted as a more C like language. We know csh is not C. It is interesting to see you want to distance C shell from C now. Let me walk you through some features you mentioned in Windows. 1) "for (int i=0; ....)" , that is C++ and C99 feature. The supported C99 features in Ch are listed here. in C90, the code should be "int i; for (i=0; ....)". 2) string_t is a Ch built-in string type for easy C/C++ scripting with auto memory allocation and deallocation. That is why free() is not needed. You can find more about it here. 3) I don't know how the docs says. To redirect both stdout and stderr in Windows, you need to use the following command in windows "a.exe 1>stdout_stderr.log 2>&1" to work in Ch, that is how I used ch most of the time. In Unix, you need to use "a.exe 2>&1 >stdout_stderr.log". 4) In windows from Ch shell, you can run "ls C:/Windows", it doesn't work the same as cygwin, Microsoft Services for UNIX or tcsh, etc. which uses "ls /C/Windows/". Unless there is a shell programming standard in Windows, you may argue that Ch shell doesn't conform to those Unix shells in Windows and that is correct. Ch shell is more like windows shell in windows. 5) The examples on pg 141 works fine from the script file and command line for me if the shell command mkdir is kept in a separate line. 7) in Unix, wildcard */* works fine. In Windows, if I replace those Unix utilities from mingw msys, */* works fine for those unix utilites in Ch shell. I am sure that you may have more questions or even find bugs when using Ch as I do, however, I think this might not be a right forum to discuss about it. Chuser (talk) 06:53, 4 May 2011 (UTC) — Chuser (talk • contribs) has made few or no other edits outside this topic.

How come Gary Wang says in his IEEE article that he ran Ch under Windows like I did and thinks it's a genuine C shell if you agree that's not possible there?

Regarding the rest: for (int i = 0; ...), you need to try it. I'm not complaining it doesn't support the feature, I'm complaining Ch is fragile: it locks up in a loop, spewing the same message over and over; it's unresponsive to Control-C, so you have to close the window. If you redirect stdout by piping or redirection, stderr is lost. The 2>&1 construct does not work. I was using Windows paths, not Cygwin paths. You make my case that Ch does NOT do wildcarding; you only get wildcarding if the child does the wildcarding, contrary to your claims here and the claims in the manual.

You're right that this is not the right place to discuss bugs, but I'm asking how one reconciles what Ch actually does with the claims of an article represented as independent and reliable. 08:19, 4 May 2011 (UTC)

1) I think you have the problem in understanding Gary Wang's IEEE article. Here is the quote from his article "Another rewarding feature is that Ch is also a genuine C shell. Beginners find it easy to use. If a students type "printf("Hello, World!")", Ch will print "Hello World!" on screen." What he meant is a shell follows the standard of the C programming language. In other words, it is a C language shell (ch), not C like shell (csh). What is genuine C shell in his mind? This is explained by the next statement "If a students type "printf("Hello, World!")", Ch will print "Hello World!" on screen." printf("Hello, World!") is a popular C statement, not csh statement. It has nothing to do with "C shell (csh)". You made a wrong assumption. 2) I tried yesterday the same code for (int i=0; ...) in linux and got syntax error. That is exactly expected and correct behavior in linux. Now I tried the code in windows, can repeat what you see. I believe it is a bug in windows. 3) for redirecting both stdout and stderr in Ch, did you try a.exe 1>stdout_stderr.log 2>&1 as I mentioned? it works fine for me to catch both stdout and stderr in the file stdout_stderr.log under ch shell. 4) if the user can run grep abc */*.c , Does it matter if it is child or parent does the wildcarding as long as it works fine as multi-wildcard under ch? Chuser (talk) 05:47, 5 May 2011 (UTC) — Chuser (talk • contribs) has made few or no other edits outside this topic.

• Comment. It seems likely pretty clear already that the consensus will be to keep. Only Tedickey and I dissent. I believe in consensus even when I'm on the wrong side of it. I propose this be closed early as a speedy keep per WP:SNOW unless others object. Msnicki (talk) 15:23, 2 May 2011 (UTC)

I don't think that would be appropriate. Regardless of the numerical weight of keep !voters, you've presented meaningful arguments for deletion and they should be given full consideration by whoever closes. —chaos5023 (talk) 15:27, 2 May 2011 (UTC)

I'm retracting my comment. Early on, it didn't look like there was going to be much real discussion, mostly just votes. I was premature; the debate did get serious. Msnicki (talk) 19:59, 8 May 2011 (UTC)

• Keep, per the exhaustive arguments I made at the very recent DRV that's linked in T. Canens' nomination statement.—S Marshall T/C 21:52, 2 May 2011 (UTC)

• Comment. I'm questioning the reliability of the Gary Wang review even though it appeared in IEEE Spectrum, a publication that we all agree is otherwise reliable. It looks to me like the guidelines are little ambiguous on this situation so I asked at Wikipedia:Reliable sources/Noticeboard#Reliable article versus reliable publication. Only one response so far (but if he's wrong, I expect there will be more) but that one response was, "There seems to be a consensus here that we look at the particulars rather than saying a specific publication or journal or newspaper can always be used as a reliable source." Msnicki (talk) 13:15, 4 May 2011 (UTC)

• Question for Chuser. On your talk page, I asked, "Do you have a relationship with the author of Ch, Harry Cheng? Are you working under his direction, perhaps as a student in his department or as an assistant or because he's asked you?" You haven't answered there. Will you answer here? Msnicki (talk) 15:21, 5 May 2011 (UTC)

Don't you mean, "Are you now, or have you ever been, ..." a Ch user? This question is entirely inappropriate. The justification of this article should be kept to objective judgements according to our strict policy. This kind of attack on an editor, not on the content, is right against the behaviour we're supposed to follow. Andy Dingley (talk) 15:46, 5 May 2011 (UTC)

I make mistakes all the time and it's possible I've done it again. It certainly isn't my intent to ask a question that constitutes personal attack. I am interested to understand if there's an undisclosed WP:COI. If you or anyone is able to shed light on how the guidelines treat this, I would appreciate your help. If I'm wrong, I'm an adult, I won't quibble, I'll retract immediately, apologize and promise not to do this again. Cheers, Msnicki (talk) 15:54, 5 May 2011 (UTC)

I have asked for advice here. Msnicki (talk) 16:15, 5 May 2011 (UTC)

What you asked isn't a personal attack, it's a legitimate question.   ArcAngel   (talk) ) 17:16, 5 May 2011 (UTC)

It's not a personal attack, but it's still not a question we should consider.

COI isn't an issue for notability. Notability isn't demonstrated by individual editors, with or without COI. The independent 3rd parties that do demonstrate it are, by their nature, off-wiki and so the issue just doesn't arise.

COI is an issue for content within an article, and appears as if it could be an issue for this article. However that doesn't change the notability or deletion issue either way. Despite recent digression, the quality of the shell interpreter just doesn't matter for the purpose of AfD. Andy Dingley (talk) 17:20, 5 May 2011 (UTC)

If we're now all agreed that my question is not a personal attack, then I would still like it answered. Chuser hasn't just contributed to the article itself; he's also chosen to participate here and I think that makes it relevant here. Msnicki (talk) 19:34, 5 May 2011 (UTC)

Sure- it's relevant. Chuser is the most supportive of the notion that Cheng's close associates should be the basis for notability of the topic TEDickey (talk) 23:37, 5 May 2011 (UTC)

I was supportive that articles written by Cheng's previous associates and published in reliable publications can be considered for notability. But, I don't take a position since the first AFD discussion. If you forget it, let me copy and paste my statement in the last AFD: "I have no comments regarding your next claim 'if they worked together, they are not independent'. It is wiki policy's call". You and Msnicki were arguing with S Marshall and others, not me. It is interesting to see you and Msnicki's positions are consistent from the beginning till the end --- no matter what: Delete. In the last AFD, gary questioned Msnicki's neutrality and COI, your comment is: "The page's author should not waste time and editor's patience by attacking the neutrality of other editors". Personally, I will be very glad to own friends like you and Msnicki. Friends are supposed to support each other all the time. Giving me 10 Msnicki and TEDickey, I believe I could control any wiki consensus no matter if it is AFD 1 or AFD 2. Since the AFD closer is supposed to follow consensus by policy, maybe I could even control wiki. You both have exhausted all of your methods and wiki's resources. From denying all reference articles, their authors, the contents of the articles, and even the software itself after downloading the software and reading the documentation in details. Don't get me wrong. I think it is a right way to do it. With the overwhelming consensus here to keep, what is the next step? You find the last target: me? I still don't know what is a driving force behind such a behavior. I have to go back and read the Hamilton C shell link [here] and [here] as gary mentioned in the previous AFD. @Msnicki, Can I ask you the following questions? You don't have to answer though. 1) "Do you have any relationship with the Hamilton C shell or its author? 2) Do you think Hamilton C shell competes with Ch shell in Windows? Chuser (talk) 08:50, 6 May 2011 (UTC) — Chuser (talk • contribs) has made few or no other edits outside this topic.

In brief, you make aspersions on other editors, claim friendship with them, and use that claim to make demands upon them TEDickey (talk) 09:42, 6 May 2011 (UTC)

Chuser, you're participating in the debate as an advocate. If you have WP:COI, you should declare it. (1) Yes, I have declared a COI with respect to Hamilton C shell here. (2) No, I do not think it competes with Ch; Ch is not a Unix shell and no Unix shell (not even the C shell) understands C language statements. (I imagine the real competitor must be gcc, the GNU C compiler; insofar as gcc is more in compliance with standards and includes a debugger, I imagine it must trounce Ch.) Now that I've answered your questions, will you answer mine?

The sources offered are weak. I don't believe the Wang article in Spectrum was anything but a plant. It's pure advertising and its hyperbolic claims are unreliable if not provably false. This was a complete failure of editorial control at Spectrum; they got spammed. We already know that the Campbell "review" was written while he studied under Cheng and while he worked on Ch; with that COI, that article also should never have been accepted for print.

All we have left is a web column First Impression (the weakest type of review possible) and a print column that's actually about Open-RJ, not Ch, in a small circulation, special interest magazine. This isn't enough under WP:CORPDEPTH. Msnicki (talk) 14:03, 6 May 2011 (UTC)

Msnicki, you just repeat what has been discussed. Your claim that "Ch is not a Unix shell and Hamilton C Shell doesn't compete with Ch shell" are incorrect. Based on the wiki Unix shell definition, "A Unix shell is a command-line interpreter or shell that provides a traditional user interface for the Unix operating system and for Unix-like systems.", Ch shell is of course a Unix shell. There is no such a restriction that a Unix shell should not understand C statements. The link here shows that Ch shell competes with C shell and it has a list of one-by-one feature comparisons. Hamilton C shell is just a C shell running in Windows. The Byte article here was written by Martin Heller -- a Mr. Computer language person at Byte. In the section titled "First Impressions and Comparisons", he said "If you already know C/C++, you'll find Ch a congenial development environment --- the fact that it's an interpreter and has a big library make a lot of things very simple. Ch also seems like it would be fabulous for beginning C/C++ programmers to use as a learning environment. As a shell-programming environment, I can see using it, with some reservations. On Windows in particular, I'm not sure whether I like Ch as well as the Hamilton C Shell or the MKS Toolkit, mostly because those packages have extensive collections of UNIX-like utilities, and more amenities in their shells than Ch currently supports." He clearly states that Ch is a shell programming environment. Apparently, Hamilton C Shell is the number one competitor coming to his mind from his "First Impressions and Comparisons". Hamilton C shell costs $350. Ch Standard Edition is free for commerical use. Your motivation on deleting the Ch article in wiki is very clear. Chuser (talk) 08:32, 7 May 2011 (UTC)

I placed a warning on Chuser's user page reminding him tp assume good faith when dealing with other editors. Guy Macon (talk) 11:06, 7 May 2011 (UTC)

thanks for the reminder TEDickey (talk) 11:25, 7 May 2011 (UTC)

Ok. I just deleted the last comment. To be frank, I am surprised to see that an anonymous editor can attack an IEEE Spectrum author and its editors as "shameless", etc. in wiki just because he doesn't like what is written, especially in light of the fact that IEEE Spectrum is the flagship publication of IEEE with the highest integrity and reputation in the field. This is the third round AFD, Msnicki repeatedly presents false statements and misleading claims. At the beginning of this AFD, he claimed that "Both the Glassborow and the Huber articles are actually book reviews of Cheng's textbook for teaching C". But, the fact is that there was no such a book for Glassborow (who is an active member of the ISO C and C++ Standard Committees and former Chair of the Association of C & C++ Users) to review in 2001. He has read the article multiple times and still gave such a misleading statement. Hubber's case has been pointed out in my posting already. He claimed that "The Wilson article is similar; it's a weak column mention, It made it into print, but in a much smaller circulation, special interest publication, the author isn't anywhere near as well-known as Heller and the column isn't actually about Ch." But, as mentioned at wiki link here for DDJ, "It covered topics aimed at computer programmers." DDJ is one of the most relevant and well known journals on the subject. Matthew Wilson is the author of "Open-RJ and Ch". and is a well-known columnist and contributing editor for C/C++ Users Journal and Dr. Dobb's Journal with multiple C/C++ books. It is interesting to see that Msnicki has already nominated to delete Matthew Wilson's wikipedia page here while his article about Ch is under discussion here. Chuser (talk) 04:33, 9 May 2011 (UTC)

Ch is not a Unix shell. Simply being able to run a command (minimally) does not make something a Unix shell. From the Unix shell article you've quoted, here's the part from the second paragraph you didn't quote: "The Bourne shell, sh, ... introduced the basic features common to all the Unix shells, including piping, here documents, command substitution, variables, control structures for condition-testing and looping and filename wildcarding." Ch does not support here documents. Piping and I/O redirection are busted. You've conceded that command substitution doesn't work under Windows. There are no control structures for iterating or conditionally executing external commands. (Ch has a for loop and an if statement, but they're usable only with C language statements, not external commands in the nested block.) Ch does not do wildcarding. Quite simply, Ch is not only not a Unix shell, it's not even a usable command processor comparable to cmd.exe.

What this demonstrates is a couple things. First, the Heller article is just what I said: it's a first impression and those are never in-depth, reliable product examinations. They're usually little more than, I opened the box and here's what I found in the first few minutes, and that's what the Heller column was.

Second, the Ch interpreter article itself and the intense lobbying we're seeing are consistent with my claim that the whole thing is spam. The article, your advocacy here and even the sources offered are filled with hyperbolic misrepresentations, dubious product claims and concealed conflict of interest. And to that, we can now add personal attacks. It's this kind of relentless and shameless lobbying that I think is likely the real explanation for what happened in the editorial failure at Spectrum. We should not allow that here. Msnicki (talk) 11:47, 7 May 2011 (UTC)

You are starting to sound very shrill here, devolving into personal attacks. Please reconsider your involvement in this AfD and article. Tijfo098 (talk) 15:41, 8 May 2011 (UTC)

On the other hand, Mcnicki has been somewhat more restrained than others on this discussion. No point in aggravating the situation by making your own accusations TEDickey (talk) 15:57, 8 May 2011 (UTC)

• Comment. Chuser claims that Ch and Hamilton C shell are competitors and that because I've declared a COI with regard to the latter, that my "motivation on deleting the Ch article in wiki is very clear." But Chuser concedes:

1. Ch "is a C language shell (ch), not C like shell (csh)."
2. "To claim Ch is not close to any Unix shell, you should run Ch in Unix to verify, not from Windows."
3. "Hamilton C shell is just a C shell running in Windows."

If the only way to get Ch to do anything (still not very) similar to what Hamilton C shell does is by running it on Unix, where Hamilton C shell doesn't run, I don't think they're competitors. I think this is an unhelpful deflection from some genuine concerns about COI raised by others (c.f., 1, 2) and which have nothing to do with me and from the dubious, promotional product claims (c.f., from the article, "Ch is a C-compatible shell similar to C-shell (csh).")

If I sound repetitious to Chuser or shrill to Tijfo098, my apologies. But this has been a bit of a Groundhog Day of debating the same article and re-examining the same sources over and over for the last couple months, so perhaps I'm getting cranky. Msnicki (talk) 18:07, 8 May 2011 (UTC)

• Good Lord, this debate's deteriorated enormously since I last looked at it. I do wish you would all cease these accusations of bad faith and these opinion statements, and get back to analysing the admissibility or otherwise of the sources.—S Marshall T/C 20:23, 8 May 2011 (UTC)

S Marshall, in the previous DRV you argued, "Notability is a tool for detecting and removing marketing spam."

Calling your attention to the Spectrum article by Gary Wang, the advice from the reliable sources noticeboard that the content of the article, not just where it was published matters in judging reliability and the points that have been raised here, I'm curious to know your opinion about that article. I'm asking if it's changed. Reading it again carefully, do you still find it to be reliable? If someone asked your opinion in real life, would you say, oh, yes, clearly reliable? Or would you have doubts? If yes, why would you have doubts? Msnicki (talk) 21:36, 8 May 2011 (UTC)

Msnick, I suggest that in the above post you change...

Wikipedia:Reliable sources/Noticeboard#Reliable article versus reliable publication

...to...

Wikipedia:Reliable_sources/Noticeboard/Archive 96#Reliable article versus reliable publication

Normally, I would have fixed this for you (making such a change is specifically allowed: see WP:TALKO: "Examples of appropriately editing others' comments ... Disambiguating or fixing links, if the linked-to page has moved, a talk page section has been archived, the link is simply broken by a typographical error, etc.") but you have expressed objections to such changes in the past, so I have to ask you to make the change yourself. Guy Macon (talk) 22:16, 8 May 2011 (UTC)

Alas, I was not able to comment at the reliable sources noticeboard prior to this being archivedbecause Msnicki did not inform anyone here about the discussion on the reliable sources noticeboard. He posted a link but I missed it. In my opinion, by using the wording "But right now, I'm just interested in understanding the reliability of the sources", Msnicki (inadvertently; I have no doubts about his good faith) steered the discussion in the direction of whether the source is reliable instead of whether it is evidence of notability even if unreliable.

If I understand the consensus, we should look at the particulars of an article when considering whether a claim in the article is from a reliable source, but for the purpose of establishing notability, the fact that the editors of a journal or newspaper that is a reliable source decided to publish an article on a particular subject is evidence of notability even if the actual article is unreliable. But that question was not asked. Guy Macon (talk) 22:41, 8 May 2011 (UTC)

Of course I posted a link to the reliable sources discussion. I feel like I've been begging for discussion. I haven't felt like I've gotten much beyond, well, if it's Spectrum, that's all you need to know. I appreciate the suggested correction to the link and I have made it, but I'd have gone with something like this.

The reliable source discussion mentioned in Msnicki's comment has been archived to Wikipedia:Reliable_sources/Noticeboard/Archive 96#Reliable article versus reliable publication. Msnicki (talk) 00:08, 9 May 2011 (UTC)

By particulars, I'm pretty sure he meant everything particular to the article including, yes, certainly, the content and whether you bought a word of it. But maybe we should try to revive the thread and see if he'll answer again. Believe it or not, I actually do care about consensus, not just winning. Msnicki (talk) 01:03, 9 May 2011 (UTC)

• Just to return to Msnicki's question, I think it's biased and POV through and through. I won't go so far as to call it "marketing spam" but I certainly wouldn't buy a product based on reviews like that. As far as that, I'm with you.

  But for me, this is why it's important that Wikipedia has a neutral article: because when someone reads that biased material in Spectrum, you see, they might turn to Wikipedia to start looking for the straight truth. Biased coverage in the sources is the initial trigger that gets end-users to turn to Wikipedia in the first place, which is why I take the view on notability that I do. When someone reads Wang's article in Spectrum and googles for the Ch interpreter, they ought to find Wikipedia at or near the top of the list, and Wikipedia ought to tell it like it is.

  It's okay for us to use biased sources, and most sources have bias or POV. What matters is that they're reliable, by Wikipedia's curious and twisty definition of the word. (Words on Wikipedia don't mean what they mean in real life. Something that's "notable" for Wikipedia is often "extremely trivial and boring" for me personally, and something that's "reliable" on Wikipedia is often in my personal "don't believe a word of it" zone.)—S Marshall T/C 00:01, 9 May 2011 (UTC)

That surprises me a little. I would have expected you to prefer a plain language interpretation of the term, reliable, i.e., if you don't believe it, you at least don't consider it reliable even if others think it is and that that would be your contribution to the consensus. But also, I would have expected you to agree with WP:CORPDEPTH where it asks that sources used to establish notability should not include "advertising and marketing materials by, about, or on behalf of the organization". It doesn't exclude only those sources created by the organization. It also excludes marketing materials about them, which asks for a judgment call: The only other way you could get marketing materials about them but not by them is if someone else created them but either way, they're excluded if that's what we think they are. In the case at hand, I'm asking, do you think this was an independent reliable review or was this a marketing material? I think it was the latter.

WP:CORPDEPTH seems to argue that in covering commercial topics as opposed to questions of whether we should have an article about your kid's middle school, that the bar has to be a little higher, that sources deserve closer inspection, that we should be more careful. I agree with that. Msnicki (talk) 01:03, 9 May 2011 (UTC)

WP:CORPDEPTH lists "advertising and marketing materials by, about, or on behalf of the organization" under the heading "Independence of sources" (emphasis in original). The clear meaning is that if we allow an organization to publish stuff about itself and then to point to that as evidence of notability, then anyone can do it and everything is notable. The IEEE Spectrum article clearly fails to match that description. It is an independent source. The decision by the editors of IEEE spectrum to publish material about the Ch interpreter was independent of the author they chose to write that material or the words that author chose. Nobody here is pointing to the content of those waords as evidence of notability. They are pointing to the decision to publish those words, arguing that IEEE spectrum would not have made that publishing decision if the Ch interpreter was non-notable. I conclude from this that the section of WP:CORPDEPTH you quote does not apply. Guy Macon (talk) 04:40, 9 May 2011 (UTC)

Note: All this sudden attention to my long-since deleted Hamilton C shell article has motivated me to request a DRV at Wikipedia:Deletion_review/Log/2011_May_8#Hamilton_C_shell. Msnicki (talk) 02:48, 9 May 2011 (UTC)

You might want to adjust the indent of the above. It is indented as a reply, but I cannot figure out which comment it is a reply to. Feel free to delete this if you wish, whether or not you fix the indentation. Guy Macon (talk) 04:40, 9 May 2011 (UTC)

I copied of the formatting of the notes offered for information, not for discussion, at the top of the section. There's probably a rule but I don't know what it is. Msnicki (talk) 04:52, 9 May 2011 (UTC)

• Comment. Msnicki, I am here to correct your false claims and misleading statements. You keep presenting false statements about Ch even after I corrected them in the previous post. Now, you claimed that "Ch does not support here documents." Another false statement. All listed Unix shell features piping, here document, command substitution, variables, control structures for condition-testing and looping and filename wildcarding, plus additional features you mentioned such as external commands in the nested block, and redirection work in Ch. Let me just go through all of them one by one with examples that are modified based on the ones presented in wiki.

1) here document. Since I am more familiar with Perl and Ch, for an illustrative purpose, I use Perl example listed in the above wiki link for comparison. Both have the same output.

Column-generating template families

The templates listed here are not interchangeable. For example, using {{col-float}} with {{col-end}} instead of {{col-float-end}} would leave a <div>...</div> open, potentially harming any subsequent formatting.

Column templates

Type	Family	Handles wiki table code?†	Responsive/ mobile suited	Start template	Column divider	End template
Float	"col-float"	Yes	Yes	{{col-float}}	{{col-float-break}}	{{col-float-end}}
	"columns-start"	Yes	Yes	{{columns-start}}	{{column}}	{{columns-end}}
Columns	"div col"	Yes	Yes	{{div col}}	–	{{div col end}}
	"columns-list"	No	Yes	{{columns-list}} (wraps div col)	–	–
Flexbox	"flex columns"	No	Yes	{{flex columns}}	–	–
Table	"col"	Yes	No	{{col-begin}}, {{col-begin-fixed}} or {{col-begin-small}}	{{col-break}} or {{col-2}} .. {{col-5}}	{{col-end}}

^† Can template handle the basic wiki markup {| | || |- |} used to create tables? If not, special templates that produce these elements (such as {{(!}}, {{!}}, {{!!}}, {{!-}}, {{!)}})—or HTML tags (<table>...</table>, <tr>...</tr>, etc.)—need to be used instead.

2) command substitution The following code copied and modified from the above wiki link works fine in Ch.

 #!/bin/ch
  vi `fgrep -l malloc *.c` 

3) variables interpolation. The following code copied and modified from the above wiki link works fine in Ch:

 
  char *name = "Alice";
  echo "${name} said Hello World to the crowd of people.";
  

4) control structures already covered C. All commands such as "ls *" can work inside if-statement and for-loop.

5) wildcarding works fine.

vi */*.c 

6) piping, the following command works in Ch.

  ls -l /proc/net |less 

In addition, the following code works fine in Windows.

  ls -l c:/windows/ |less 

7) External commands in the nested block example. Comparing with the examples copied from wiki C Shell.

Column-generating template families

The templates listed here are not interchangeable. For example, using {{col-float}} with {{col-end}} instead of {{col-float-end}} would leave a <div>...</div> open, potentially harming any subsequent formatting.

Column templates

Type	Family	Handles wiki table code?†	Responsive/ mobile suited	Start template	Column divider	End template
Float	"col-float"	Yes	Yes	{{col-float}}	{{col-float-break}}	{{col-float-end}}
	"columns-start"	Yes	Yes	{{columns-start}}	{{column}}	{{columns-end}}
Columns	"div col"	Yes	Yes	{{div col}}	–	{{div col end}}
	"columns-list"	No	Yes	{{columns-list}} (wraps div col)	–	–
Flexbox	"flex columns"	No	Yes	{{flex columns}}	–	–
Table	"col"	Yes	No	{{col-begin}}, {{col-begin-fixed}} or {{col-begin-small}}	{{col-break}} or {{col-2}} .. {{col-5}}	{{col-end}}

^† Can template handle the basic wiki markup {| | || |- |} used to create tables? If not, special templates that produce these elements (such as {{(!}}, {{!}}, {{!!}}, {{!-}}, {{!)}})—or HTML tags (<table>...</table>, <tr>...</tr>, etc.)—need to be used instead.

8) To redirect both stdout and stderr in Windows and Unix, stdout_stderr.log contains the output of both the stdout and stderr. I have pointed out that it works and you keep repeating there is a problem without providing the evidence.

  In windows, use: run.exe 1>stdout_stderr.log 2>&1
  In Unix,    use  ./run 2>&1 >stdout_stderr.log

I have just tested all above features in both Linux and Windows. They all work as expected except command substitution doesn't work in Windows. But you already know the alternative solution from your previous post: "store it to a variable (e.g., string_t s = `date`) and then you can use it as $s". In addition, Ch shell supports history, aliases, command/filename completion, etc in both Linux and Windows. One bug or one missing feature in Windows doesn't change the fact that Ch competes with Hamilton C shell in Windows. In summary, Ch is a Unix Shell. Professor Gary Wang@IEEE Spectrum and Martin Heller@Byte knew what they were talking about. Ch a geniune C shell. Hamilton C shell competes with Ch shell head on. You not only try to delete your competitor's product article and its related links in wiki, but also use the Wikipedia platform to attack your competitor's product with false and twisted facts. Chuser (talk) 06:16, 9 May 2011 (UTC)

I stand by my remarks. Anyone is free to download the product (or not) and decide what to believe. If someone actually wants me to respond to any specific point, I will, otherwise not. Msnicki (talk) 06:31, 9 May 2011 (UTC)