Talk:Comet (programming)

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Computing  
WikiProject icon This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
 ???  This article has not yet received a rating on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.
 

touch —Preceding unsigned comment added by R3tr0 (talkcontribs) 04:08, 21 September 2008 (UTC)

External links to comet[edit]

I think its a good idea to have links to some Comet servers. Some 78.232.434.22 guy added a link to StreamHub. Some other guy who hasn't contributed to the page removed it, but I thought it was useful so I added it back in. If you know of any other Comet servers that have free or open source editions like StreamHub please add - I think it is useful so a lot of people don't have to re-invent the Comet wheel from scratch in PHP - trust me I tried! :). Please discuss. Danke. :)

CometGuru (talk) 00:25, 24 July 2009 (UTC)

It's a commercial product. We generally don't put links to sites promoting commercial products on Wikipedia. - MrOllie (talk) 00:34, 24 July 2009 (UTC)

There is a community edition that is free. I'd waste all week referencing wikipedia articles that link commercial products. MySQL is a quick example, it's a commercial product but it's linked from Wikipedia.Woods01 (talk) 23:11, 24 October 2009 (UTC)

2006 is a bit late[edit]

Wasn't CGI:IRC doing COMET in the early 2000s?

68.224.90.49 (talk) 03:55, 25 July 2009 (UTC)

Probably, since "comet" is just a made-up word for a (ill-defined) collection of existing techniques. --Damiens.rf 12:29, 2 September 2009 (UTC)

Indeed. I was using Comet style updates (we called them 'slow load') in 2003 for the Russound Media Server ( http://www.russound.com/product_detail.php?i=3241 ) and more primitive versions as far back as 2001 for IntelliNet RS1000 (http://www.twice.com/article/249540-IntelliNet_Merges_Audio_Home_Control_Systems.php ) KevinSeghetti (talk) 20:39, 30 September 2011 (UTC)

Persistent connection?[edit]

These words should not be used because I feel it creates confusion with what an actual HTTP persistent connection is, which from what I am reading on the sources, Comet does not use. I am reading that Comet sends one HTTP request and the HTTP response is slowly fed back to the client through emulating lag on the server through some sleep method. A persistent connection is an actual connection that is kept alive after the initial HTTP request and response for sending additional HTTP requests and responses. How does Comet let me send additional HTTP requests without creating an additional connection? It doesn't from what I am reading, unless Java or a similar technology is used for the request, but this article mostly talks about XMLHttpRequest and IFRAME, which do not allow this as far as I've seen. I feel this needs to be reworded to "long-lived" under implementations or sources are needed that state an HTTP persistent connection is used. --Quilokos (talk) 06:51, 1 August 2009 (UTC)


Applets don't have security restrictions?[edit]

Since when? Both a Java applet and the Flash swf can only communicate with the host they were served from. They also cannot write files, or generally do anything the browser wouldn't be allowed to do. —Preceding unsigned comment added by 216.119.215.193 (talk) 20:17, 8 September 2009 (UTC)

This is indeed false and I removed the claim.--85.77.166.16 (talk) 09:49, 9 November 2009 (UTC)

The difference is that many comet transports do not allow connecting to a server other than the one the containing page was served from. This is a restriction not shared by embedded Java applets or Flash. You're right that the text as written did not carefully enough explain this, and was therefore incorrect, a sloppy oversight. —jacobolus (t) 04:58, 30 November 2009 (UTC)

Multiple issues[edit]

This article is tagged with multiple issues, but I think some of them can be cleared.

I have been working with ajax/javascript/html, and I approve the *technical content* in this article, except for the following sections (I can't tell with precision if their content is exact, but none of it sounds false to me) : the header, the historic part on netscape in XMLHttpRequest, the details on cross-site scripting (I know it's disallowed by browsers in some situations, but I can't confirm which), History and Alternatives.

These sections have a sufficient number of external references to my mind. For the part about BOSH, some say it does use the "Comet" technique (so it might not be an alternative) : http://metajack.wordpress.com/2008/07/02/xmpp-is-better-with-bosh/ .

As for advertizing, I can't see any. There are some product/company names in the "First Comet applications", but they are very different companies, and I think the major companies using this technology have been included.

The main problem I think is that this term is a neologism that I haven't seen anywhere before. But if it's the only problem, I suggest keeping this article (only renaming it, or merging it with another article, such as [AJAX] (but it may be too technical for this merging)), because it's a good review of techniques that can be used to have a stream of events going from a server to a web browser.

Jahvascriptmaniac (talk) 17:42, 27 November 2009 (UTC)

A previous version of this article had many more sources, and was generally more thorough, but one zealous editor decided that it didn't meet his personal standards, and since he was willing to devote more time to the ensuing revert war, he successfully removed much of the content and sources, and left behind, unsightly detritus, those notices at the top of the page. As for the “neologism” angle.... well, everything starts out as a new word. But “Comet”, whatever it is called, is a genuinely useful and notable concept, and thus clearly warrants an article. Feel free to remove the notices, but be prepared for the aforementioned editor to fight you vigorously. Unless you have some time on your hands for a tedious argument, it may not be worth the trouble. Cheers. —jacobolus (t) 04:55, 30 November 2009 (UTC)
Jahvascriptmaniac, here's a link to the previous version mentioned above so that you can judge by yourself (and understand where those tags came from). Also, the talk archive reveals the drastic changes were not exactly the work of just "one zealous editor". --Damiens.rf 00:54, 1 December 2009 (UTC)
Thanks to Damiens.rf provided the information about the previous version of this wiki page which is exactly the info I needed, the current version seems much clear though, but a lot of vaulable information has been lost in comparsion of the previous one. Techlivezheng (talk) 17:15, 19 May 2012 (UTC)

HTML 5 draft specification is outdated[edit]

The article links a "HTML 5 draft specification produced by the Web Hypertext Application Technology Working Group (WHATWG)" which is a very outdated draft document. The current version (http://www.whatwg.org/specs/web-apps/current-work/multipage/, 2010-07-16) does not mention the concepts of server-side events, I guess they have been removed from the specification. Maybe this shouldn't be listed as an alternative then if it didn't make it into the draft standard? —Preceding unsigned comment added by 93.111.233.185 (talk) 10:19, 18 July 2010 (UTC)

The Server-Sent Events spec was turned into a separate document. It is located at http://dev.w3.org/html5/eventsource/. I’d change the page, but I gave up any desire to work on it after Damiens.rf’s campaign of unilateral destruction a couple years ago. –jacobolus (t) 16:05, 18 July 2010 (UTC)

Web-based chat to Gmail[edit]

The web-based chat in Gmail (GTalk) it's a Jabber client, and it's based on BOSH, defined in XMPP Protocol. —Preceding unsigned comment added by 200.55.140.181 (talk) 15:13, 28 July 2010 (UTC)

What? That is simply false. Gmail web chat is different from its jabber chat, and has nothing to do with BOSH. –jacobolus (t) 17:30, 28 July 2010 (UTC)

protected[edit]

I've protected this article for a couple of weeks from editing. Perhaps those interested need to decide if they're writing a collaborative encyclopedia or playing tennis - Peripitus (Talk) 08:38, 9 June 2012 (UTC)

Thanks. –jacobolus (t) 19:07, 9 June 2012 (UTC)
Don't consider the fact that protection happened when your version was live any form of endorsement. The edit-revert-revert rubbish over the last few weeks is silly and my action is simply to stop this - Peripitus (Talk) 00:46, 10 June 2012 (UTC)
I agree. I have tried several times in the past to engage with Damiens in discussion, without any success. I don’t typically try to bring in bigger heavy-hitting process, but then most of the conflicts I’ve been involved in can be resolved amicably by editors willing to compromise. It probably would have been good to escalate the problems w/ this Comet page dispute to some higher process, but I haven’t had the time and motivation to make up lists of grievances and even think through exactly what I think should be done. A few years ago, I wrote a comprehensive article here about Comet. Damiens decimated the article, and tried to delete it repeatedly. I didn’t have enough free time or interest at that point to really make a fight out of it. Anyway, thanks for protecting the page. Maybe it’ll force some kind of attempt at real engagement. Cheers. –jacobolus (t) 02:43, 10 June 2012 (UTC)

My recommendation is to restore most of the content of the article as of this version before Damiens.rf went on his deletion spree in 2008. Some of his criticisms at that point were legitimate, so I recommend that some third party, ideally someone with some knowledge of Comet or back-end Web programming, make such changes as to fix whatever problems there are with that version. –jacobolus (t) 23:20, 12 June 2012 (UTC)

Edition decisions[edit]

Quotation marks[edit]

Should the article use as a mark of quotation:

  1. the character “ , like in this revision; or
  1. the character " , like in this revision;

--damiens.rf 17:32, 12 June 2012 (UTC)

Discussion/!Voting
  • This article had curly quotes between 2006 or 2007 and 2012, with no complaints. Wikipedia policy is to leave formatting of articles internally consistent, but not necessarily mandate such petty stylistic changes across the full wiki. Beyond that, I can’t believe that Damiens cares strongly about this issue: it’s just along for the ride. Distracting other editors with this trivial side issue is a petty waste of time. –jacobolus (t) 23:34, 12 June 2012 (UTC)

Formatting of references[edit]

Should the citation #13 (Rob Butler, et al. ) be formatted

  1. Without the {{cite web}} template: <ref>Rob Butler, et al. (June 2006). “[http://bugs.webkit.org/show_bug.cgi?id=14392 Bug 14392: Add support for <code>multipart/x-mixed-replace</code> to XMLHttpRequest]”. Webkit bug tracking. Retrieved 29 November 2007.</ref>, as in this revision; or
  2. With the {{cite web}} template: <ref>{{Cite web |author=Rob Butler, et al. |month=June |year=2006 |url=http://bugs.webkit.org/show_bug.cgi?id=14392 |title=Bug 14392: Add support for <code>multipart/x-mixed-replace</code> to XMLHttpRequest |publisher=Webkit bug tracking |accessdate=29 November 2007}}</ref>, as in this revision

--damiens.rf 17:32, 12 June 2012 (UTC)

Discussion/!Voting
  • I personally prefer 2, since the use of templates across the wiki promote consistency and eases maintenance. --damiens.rf 17:50, 12 June 2012 (UTC)

External link to Comet blog[edit]

Should the article link to "Comet Daily""?

  1. Yes, use it this external link.
  2. No, don't link.
Discussion/!Voting
  • It's a advertisement based collection of posting by self-appointed Comet experts, described in the site as " leading software engineers". Despite its name, the site the site is not updated "Daily": since it's foundation in 2008 it has less than five posts a year. --damiens.rf 17:32, 12 June 2012 (UTC)
  • No. I say an external link to this site does more good to them than to us. I don't find the site listed (prominently or not) on Google searches for "comet", "comet programming", "comet web programming", "comet programming technique", "comet ajax". When searching for "comet articles web" we see "Comet Daily" appearing mentioned in the snippet for the entry for this Wikipedia article, but not as a search result. When searching for "information about Comet techniques'", that is exactly how the website describes itself, the website appears in 5th place. --damiens.rf 17:50, 12 June 2012 (UTC)
  • The website was an attempt to make a blog centered on Comet, the topic of this article, in 2007, with posts by implementors of Comet servers – i.e. experts on the topic. It contained a wide variety of posts about Comet, including a great deal of relevant technical information, but mostly went dormant after 2008. It remains one of the most useful resources about the topic, and is much more accessible than dead-tree books. It meets the typical criteria for external links: The site’s content is accessible to the reader. The site’s content is proper in the context of the article (useful, tasteful, informative, factual, etc.). The link is functioning and likely to remain functional. –jacobolus (t) 23:29, 12 June 2012 (UTC)


Folks, just my 2c. (1) straight quotes vs curly. The difference is trivial and barely discernible (on my monitor) and the WP:MOS thing is a recommendation. It's hard to care about this and impossible to see why it's worth an edit war over. (2) I use the cite web (easier and more maintainable) but prefer the first as it's a bit less messy in the edit window, given that both look mostly the same to readers it's simply not worth the effort of caring over, especially as the text of the article is far from a polished complete article. This , however, is beside the most important points. A bugzilla link is hardly a reliable source, and linking to a conversation to assert that such is (or was) hapenning is verging on original research—where is the third-party source for this sentence ?. Personally I'd remove the text this relates to AND the reference. As for the external link, sorry but my flabber is to gasted by the quotation mark thing to comment. 100 edits since January has resulted in no substantive change to the article. Perhaps it's time to take the article off watchlists, and move on ? Peripitus (Talk) 12:12, 26 June 2012 (UTC)

The oddness of the bugzilla "source" has been pointed out since the first time it was proposed, but it was not taken seriously by the proposing editor and end up slipping into the article. --damiens.rf 12:39, 26 June 2012 (UTC)
The reference supports the claim that there was a discussion about adding it to webkit, by linking directly to the discussion in question. I really don’t understand why that’s controversial... In any event, that was much more relevant 4 or 5 years ago, before WebSocket got enough momentum to pretty much supplant multipart/x-mixed-replace. There has been enough change in the last few years that this article should be substantially re-written. I don’t have the stomach for it, though, because every time I’ve tried to add constructively here Damiens.rf has interjected to keep the article limited and confusing. It became not worth the effort. –jacobolus (t) 00:58, 27 June 2012 (UTC)
You don't seem to really understand why that's original research.
If I edit the article to use the MOS suggested straight quotes, will you have the stomach to revert me once more, or did it became, after Peripitus input above, not worth the effort? --damiens.rf 06:11, 27 June 2012 (UTC)
Why do you so fervently care that the quotation marks be one way? From my perspective, it looks like you want them changed only for the sake of fighting. –jacobolus (t) 21:59, 27 June 2012 (UTC)
As for “original research”, I think you misunderstand. “Original research” is developing some new synthesis or analysis. Simple statements of fact are perfectly reasonable to back up with direct sources. But it hardly matters, because as I mentioned the subject that the source was backing up there is now outdated. This article has made no progress since 2008 or whatever, and that is entirely because of your personal choices and interaction, stalling any efforts to update or improve the text, and chasing away would-be contributors who know anything about the subject. Here’s a question: what do you think the articles about Comet, WebSocket, “server push”, and so on should say? As far as I can tell, your preferred state is to have them say nothing, but settled on an equilibrium where they say a few things, because trying to delete them altogether caused too much of a ruckus. Right now, all of these articles are unhelpful and out of date, but it’s not worth trying to constructively add to them if any effort will be reverted and rejected. You haven’t added any constructive text, and it’s not even clear to me that you understand enough about these technologies to explain them. So again, what is your purpose, and what do you imagine these articles looking like in your ideal Wiki? –jacobolus (t) 22:06, 27 June 2012 (UTC)
If the qoutations look the same to you, then why do you fight for them? Don't pretend you were the first one to revert the quotation fix and that it was implemented by another user.
You refused to answer a very simple and direct question I made, and presented me with a very loaded and confusing question nevertheless. I answer you anyways.
Those article should explain those technologies for what they are, avoiding a commercial hype tone, avoiding the mystification and glorification of not only the technology itself but also of the people using the technology. You should know that at this point, since some years ago a number of editors pointed you about how wrong was to single-handedly write an article that sounded like a magazine brochure and promoted yourself and your workmates and the blog you guys worked in as the vanguard of a stupendous technology. You should not use Wikipedia to promote your name, Jacob Russ. Neither the name of your friends or the Blog you write for. The version of the article you wrote was unacceptable for many reasons. Stop whining about it.
And here I repeat the question you ignored: If I edit the article to use the MOS suggested straight quotes, will you have the stomach to revert me once more, or did it became, after Peripitus input above, not worth the effort? --damiens.rf 16:51, 4 July 2012 (UTC)
The point is not to promote anything: neither I nor anyone I can think of stands to gain from this “promotion” you’re so worried about. The point is to explain how the thing works, because it requires background explanation for a lay person. As this article (and WebSocket, and push technology) stands right now, a non-technical reader will be left mystified and confused, entirely unclear about how these technologies work or what the point of them is. As for the quotation marks: “Don't pretend you were the first one to revert the quotation fix” – I think your grammar here makes this say the opposite of what you intend? But anyway, yup, I reverted the quotation mark change, because using proper typography was just fine for years and there’s no reason to use improper typography instead. The MOS doesn’t have any hard requirements on that point, but only loose guidelines, and the important part is for articles to be internally consistent. But as I said, I don’t think you care about the quotation marks, and are only using that as a pretense for a fight, and I don’t understand why. The question still remains, and you haven’t answered it: what is your vision for this/these articles, and why haven’t you bothered trying to constructively add text expressing that vision, instead trying to slash and burn them down to uselessness? –jacobolus (t) 23:07, 4 July 2012 (UTC)
  • "The question still remains, and you haven’t answered it..."
Yes, I did.
  • "...what is your vision for this/these articles...
"Those article should explain those technologies for what they are, avoiding a commercial hype tone, avoiding the mystification and glorification of not only the technology itself but also of the people using the technology. " (copied from may last edit above)
  • "and why haven’t you bothered trying to constructively add text expressing that vision"
Because "adding text" is not the only valid and useful way to contribute to Wikipedia. Fixing, applying MOS, tagging mistakes, removing improper content is also contributing. In this subject, as in some others, I feel more comfortable with these kind of contributions, and I do not feel like a second-class wikicitizen for that.
  • "...instead trying to slash and burn them down to uselessness?"
That was childish and unnecessary. --damiens.rf 17:08, 6 July 2012 (UTC)
The “number of editors” was actually mainly just you. When it was noticed (not by me, but I believe by Simon Willison) that you’d been trying to delete the Comet article, there was a huge furor on the programming reddit page and a few other places about how absurd that was, and you backed off slightly. I was told by 6 or 8 different people, spontaneously and without solicitation, that the Comet article as it stood before was the best explanation around, and very helpful for engineers trying to use Comet in their web applications, and the article was widely linked from web technologist blogs for the same reason. Not because it was a load of self promotion, but because it was legitimately helpful, as an explanation. Now it’s not. And you personally are the sole cause of that. And it’s not just that the previous content was deleted, but your negativity and protectiveness over these articles has kept other would-be contributors away, and so all of these articles are almost entirely stagnant, and now woefully out of date. So again, what do you want them to be? From your action, I can only assume you want them out-of-date, overly technical and useless to non-experts, and entirely stagnant. –jacobolus (t) 23:12, 4 July 2012 (UTC)
Great revisionism. The "huge furor on the programming reddit page" event was just a bunch of redditors coming to Wikipedia as single-purpose-accounts to vandalized my user page (and that of other users you disagreed with) and revert-war on this page (like [1], [2] or [3]). They were all following a call to arms made by some influential programmer in readit after you personally asked him offwiki for "help" when your article was deeemed improper. The article had to be protected due to the amount of vandalism by clueless newcommers.
It's really annoying that you try to reinvent all that happened in this article. You wrote a terrible article all by yourself and was full of pride for it, but when the problems were noticed, the article as it were could not survive. The article you wrote put yourself, Jacob Russ, as one of the leading experts in an astonishingly new and innovative technology! That is promotion. That is commercial hype for personal gain. You even used your own blog posts as sources! Stop defending that version of the article!
I suggest you re-read the archived talk page to refresh your memories of the events that led to the clean up of the garbage that was once on this article. --damiens.rf 17:08, 6 July 2012 (UTC)
Yeah, I looked, and people’s anger still seems pretty legitimate. Basically, a bunch of prominent web experts (for instance, Armin Ronacher, whose comments you just linked to, and Arve Bersvendsen, whose “call to arms” it was, and Simon Willison) found it absurd that a useful resource was wiped out for no good reason by a random anonymous guy on the internet with no special knowledge of anything but a bone to pick (namely, you). “Slash and burn” is IMO an entirely fair description. But maybe “burn and then salt the earth” would be a better metaphor: the atmosphere you generated on these pages was so acrimonious that no one with anything productive to contribute was willing to stay around. For most people, there are much more fulfilling things to work on than fighting aggressive internet trolls who have time on their hands. “Because "adding text" is not the only valid and useful way to contribute to Wikipedia” – unfortunately, if all the text gets deleted, and every good faith effort to contribute is met with scorn and contempt, then these other functions are about as useful as enforcing the law as sheriff of a ghost town. “commercial hype for personal gain.” – what’s the personal gain? I’ve never made a single cent either from work on Comet servers, or from work on Wikipedia. I don’t give a damn about personal gain here, whether monetary or any other sort; the goal has always been to actually explain what’s going on. It so happens that many of the best explanations around are by people who themselves implemented Comet servers (what a shock!), and therefore it’s worth linking to those explanations if the point is to direct readers at the best explanations. If personal gain was my motivation, I’ve been colossally unsuccessful. Do I personally think people should use Comet, the technology (whether using WebSocket or via some other transport)? Of course; otherwise I wouldn’t have sunk so much time into trying to explain it. Does that benefit me personally? Not really. “personally asked him offwiki for "help"” – now you’re just making stuff up. “Those article should explain those technologies for what they are” – okay, well who do you propose will write that material? Currently, all of these articles are awful, full of inaccuracies, several years out of date, highly technical and full of jargon, entirely confusing for newcomers and experts alike, and almost completely stagnant. If these articles are going to improve, someone has to do real research, real writing, and real consensus building. But you seem to have zero interest in any of that, so who do you suggest will do it? Or are you satisfied with the current state of affairs? I defy you to find a single person with even a passing knowledge of web technology who thinks these articles are technically excellent, or a single non-expert who thinks they provide a good explanation for the layman. –jacobolus (t) 03:54, 7 July 2012 (UTC)

Comet support in IE10[edit]

XMLHttpRequest enhancements: Comet streaming support (which links here). Yappy2bhere (talk) 22:47, 7 August 2012 (UTC)