Talk:Ajax (programming)/Archive 2

From Wikipedia, the free encyclopedia
Jump to: navigation, search
← Archive 1 Archive 2 Archive 3 →

Request for external link addition

  • I would like to make a formal request to add to the external links section of this page. It is my own website, so I wouldn't like to add it myself, but I would prefer that someone else review the website, and make the decision to add it. Feather Ajax is a 1KB Ajax library intended to teach users how to program a small (but extremely effective) ajax library using Javascript and any server-side script they prefer (the examples are written in PHP). --Sugarskane 04:03, 20 July 2006 (UTC)


The article states that the term AJAX was coined in 2005 - is there a reference for this somewhere? -Andreas

The first known use of the term in public was by Jesse James Garrett in his February 2005 article Ajax: A New Approach to Web Applications. --Sleepyhead 13:57, 22 May 2006 (UTC)

I updated the data today include that the term was thought of when Jesse James Garrett was in the shower. Additionally, Jesse thought of the term while consulting for a large financial services firm. Jesses was trying to present how Ajax could improve the financial services firm's web application. The term Ajax was officially first published in an internal presentation to that financial services firm on February 11, 2005. Post that presentation Jesse wrote his article that officially termed Asynchronous Java XML: Ajax --Jacecole 22:47, 14 June 2006 (UTC)

Lack of development tools? Also: Hummingbird and links

I'm not sure the "Lack of development tools" critisism holds true - while it's true that there are no specific Ajax development tools (unless you count the various frameworks) that's largely because AJAX is a composite of several different server and clientside technologies, all of which do have development tools.

OK, maybe I should have worded it more precisely... There are development tools that allow one to develop Ajax applictions. Some of them even help a little in the process. Then there are frameworks and/or librares that make specific tasks easier. However, there are no full-encompassing IDEs comparable to those for desktop applications.
I am not a VB fan, but let's use it as a well-known example. Developer starts it up, drags and drops UI components, ties handlers to events, writes some code and voila - there you have an app. Although this is theoretically possible even for RIA/Ajax apps (Hummingbird has entertained the idea but that is a whole different story), there is nothing even remotely comparable to what VB IDE offers. Arguably, the ease of development of VB apps in VB IDE is the spark that started the explosion of many (Windows) desktop applications - it became easy to write them. We can argue that most of them are of arguable quality, but the same idea exists in other IDEs as well. We have yet to see that for RIA/Ajax.
--Aleksandar Šušnjar 20:07, 13 January 2006 (UTC)

Hmm. I'm still severely dubious of several premises of this section: that having some kind of drag and drop development environment is the be-all-and-end-all of programming, that current frameworks don't count towards that, the implication that existing web development practices don't work for sites using AJAX, that a site using AJAX technologies has to be some kind of equivalent to a desktop app.

BTW What's the relevance of the Hummingbird link? It doesn't seem to mention AJAX on the site at all ("Your search - ajax - did not match any documents.")

--Artw 21:49, 13 January 2006 (UTC)

Well, put it this way. Imagine if you had as many ready-made UI controls as you can find for desktop apps. Imagine if the IDE did all the linking between the code and instances for you, as pure JavaScript and HTML usually do not. Imagine if you, in fact, did not have to worry about any communication with the server at all - all you do is call methods of smart objects. Creating "Ajax" apps would then be a breeze. I am not implying that existing web dev practices don't work for sites using Ajax nor that Ajax apps have to be equivalent to a desktop app - but that is the direction they can be taken to and that is how they differ from "traditional web apps". Lack of tools is one of the reasons why there aren't many (I know of only one) desktop equivalent app.
Relevance of Hummingbird link is there for people who built their solutions on Hummingbird platform. You won't find Ajax there because that is just a very new term. Hummingbird calls its architecture different terms, incl. "Webtop", "Web UI Objects", "Constructs", etc. but is essentially the most powerful Ajax app (and platform) I've seen so far, and it was done in 2001 - long before "Ajax" is coined. On top of that it is flexible to use any (incl. custom) transport protocol, completely abstracted from both client and the server, so either XML, JSON or other "serialization type" can be used, via separate (de)serialization streams. Other important parts woth mentioning were the environment controller (sort of a rudimentary client-side operating system - resource manager), HTML element to proper object event distribution, library to aid/simulate classes in JavaScript along with inheritance, a library (class hierarchy) of ready-made UI controls, etc.
I also added "", albeit without a link. I think those people deserve to be mentioned. I remember them being alive somwhere between 1999 and 2001 - not sure. They made, essentially, a simulator of a typical desktop - all with applications, program managers / start menus, window management, etc - your desktop. You could log in from anywhere and deal with your apps. To top it all they did it in IE4 and announced NN4 support. Bad part: took ages to log in, while it loaded all the stuff it needed. Generally very nice idea, maybe ahead of its time, which is why the domain now seems to belong to someone else (or at least there is no mention of what it used to be). Maybe someone can bring some light about
I thought it makes sense to put the link to people who actually did something about it when there are links to those who just talked about it or did comparatively very little (Macromedia, Microsoft, Google).
--Aleksandar Šušnjar 22:39, 13 January 2006 (UTC)
If it worked in IE4 and NN4 it probably wasn't AJAX, as the XMLHTTP object wasn;t around until later versions of that browser. It sound slike what you're describing is some clever ASP/DHTML interaction that works in some other way. The links should be removed as they are confusing and not helpful to anyone wanting to know about AJAX. --Artw 23:09, 13 January 2006 (UTC)
I was interested in at the time and saw what it did. If you limit AJAX to XML than they may not qualify. However, neither is "Ajax" any more limited to XML - see article itself (e.g. quote: "XML is commonly used, although any format will work, including preformatted HTML, plain text, JSON and even EBML"). On the other hand it had more verbose and powerful client-side implementation than most todays players and is comparable in complexity to Hummingbird's. I am not describing "clever ASP/DHTML" interaction in any other way. I had my hands "dirty" on both.
As for helpfulnes, this article also presents history. I beleve the credit for starting the idea should go to people who did something notable and either or Hummingbird alone probably outweigh all others mentioned in the article combined.
--Aleksandar Šušnjar 23:25, 13 January 2006 (UTC)
Without wanting to get into an edit war, I really don't see much evidence that they're part of the history of AJAX, or even use AJAX at all - in fact all i can see is a dead link and a link to a rather impenentrable product information page. Are they browser based applications, in HTML and Jabvascript, which interact with the server without reloading the page? If so then they'd be AJAX-like, but without using XMLHttp (note: not necessarly XML) they fall outside of most definitions of AJAX. And if they don't use XMLHttp by what method do they work?
--Artw 23:44, 13 January 2006 (UTC)
It is hard to find evidence for as it no longer exists as a company. Domain is currently owned by some other company. I tried my best and created the stub article for it, in hopes that people involved in developing it will notice and provide some more information. For now have a look at it and the links at the bottom - "online traces of existence" I've found.
As for Hummingbird solutions I was quite involved in developing and architecting (for) it. Unfortunately I haven't found public information - documentation seems to only be available to customers and partners, but I will probe further. Humminbird has a framework that can (but does not have to) use XMLHttp (different ones for different browsers). For example, for JSON communication it uses IFRAMEs. You are the first person restricting Ajax to XML. I grant that XML is in its name and that others are not but you have to acknowledge what the article already states and what most people think of Ajax - XML is common, but not required. If you want to discount all but XML then the article has to be seriously rewritten and many other references and links removed. Last but not least, AJAX does not require XMLHttp. There are ways of transferring and parsing XML otherwise, although not as efficiently.
I wrote a "toy" component for Hummingbird Webtop in early 2002, that relied on client-side Javascript, CSS and XMLHttpRequest. It updated a static page once per second, displaying a dynamic (moving) realtime graph of certain server statistics. The page itself was loaded once only; only the data was refreshed by the once-a-second XMLHttpRequest. For what it's worth, anyone hoping to defend a patent action in connection with Ajax might want to contact me! However I'm not at all impressed by the claim that Webtop itself is an example of Ajax. I'm closely familiar with the architecture of Webtop, having worked on the Webtop kernel for several years; and I think this claim is simply wrong. MrDemeanour 12:23, 15 May 2006 (UTC)
Last but not least, I am honestly not trying to put ads here. I removed the links, but I believe all those people deserve the credit for their achievements, especially when compared to other listed. Don't get me wrong - I don't want them removed. Although rather simplistic (when compared to the other companies mentioned (e.g. Google) are very popular and influential. But they are not the (only/true) pioneers.
Instead of spending time explaining all of these things here in the talk page, I'll set up the article for Hummingbird DM Webtop and put everything there. I just have to confirm with Hummingbird as to how much can be made public.
I'm also asking you for a little patience. The fact that you're not personally aware of some facts does not mean that they should be removed. After all most of us know very little and rely on others to provide information. I am myself very interested in Ajax/RIA/whatever-you-want-to-call-it architecture as I really never fully believed in "traditional web apps". They all too much smell of classic "dumb terminal and mainframe" paradigm, only with more colourful and less interactive terminals. Furthermore, while the servers are truly powerful, they are not as many times faster than a typical user workstation as many users they have to handle - therefore too much client-side power is wasted... If only we had compiled client-side JavaScript, for example, thing would start looking really great...
--Aleksandar Šušnjar 04:28, 14 January 2006 (UTC)
A lack of dedicated development tools seems like a very empty criticism to me. Tools will turn up if there is a definite requirement among developers for them. Since Ajax is simply well-known technologies that we've been developing in for many years, brought together, I don't see why — necessarily — development tools on the level of those designed for desktop application development are required.
Secondly, the addition of all of this Hummingbird etc. "historical" data has made the first sections of this article disastrously complicated. Aleksandar, you are talking about parts of the history of remote scripting, not of Ajax as we know it today. Please clean this section up. Rufous 16:40, 14 January 2006 (UTC)
Hmmm... I agree with you that "History" needs to reorganized and cleaned. But I disagree that I am talking about "remote scripting" (as proposed by Microsoft, e.g. via Java applets). Hummingbird approach is true Ajax in every aspect although it does take it even further (than anyone else) - underlying protocol and data formats are pluggable and not restricted to any one. What we call Ajax today is essentially a JavaScript-powered client (browser) side app technology where code is delivered on demand from the server and in which the client-side code can asynchronously communicate data (not visual HTML pages or snippets). XML is commonly used transport format but not really required. Hummingbird architecture, for example, fits the entire description and extends it further and was created years before the "Ajax" term has even been coined, before other players came up with their little Ajax "enhacements" to their existing products (Hummingbird app is "Ajax" in its entirety - it is not just used as an enhacement). I think that qualifies Hummingbird as a "pioneer" more so than others listed. is another example.
If anything was "hastily" added it is the mentions of everyone else. What has Microsoft done? Talked about it and implemented proprietary and rudimentary enhancements here-and-there. What has Google done? Came later and made some (although important) enhancements to their services. However, due to their popularity their actions became immediatelly visible. What has Macromedia done? Came really late, talked about it and wanted to do it with Flash (it is worth mentioning). So, if those companies are to be present in the article as late-coming all-talk-no-action (with some exceptions) then actual performers and pioneers such as and Hummingbird (and possibly others I would like to hear about) absolutely MUST be credited.
Last but not least about the lack of tools being "empty" criticism. It worries me as well, as I don't like including information to be dated in encyclopaedic content. However, in this case we have no choice. Pretty much all pros and cons are dated. Current fact is that even when tools do exist they support advance Ajax only to the level of "checkmarking" it - rudimentary. I am personally eager to see the time when that stops being the fact and will want to see the guys making those tools credited for their efforts. History is about what happened, explanations and important people. In this case and Hummingbird are at least as important as anyone else listed.
--Aleksandar Šušnjar 17:37, 14 January 2006 (UTC)
I;ve just made a number of edits to the article, all of wehich seem to originate witgh yourself, and all of which seems to be intended to tilt the meaning of the term AJAX away from it;s common usage and (I suspect) towards wahtever and Hummingbird are. sounds interesting, but I don't think it's as big a part of internet history as you imply it is, and is tertiaqry to AJAX history at best. This article will become meaningless if we start inclusing everyone who was using an iframe hack or similar to transfer data before AJAX. Perhaps a general article on remote scripting techniques would be a good idea?
Your claims that Hummingbird is an important part of AJAX history and are the top rpoponents of AJAX development at this time are completely unsubstantiated by anything Ican find - as far as I can tell they make a document management system with a nice but unremarkable interface. Until you can back it up it should not be included, and certainly not linked to in multiple places throughout the article as you would have.
Also I've deleted the "development tools" section as it seems to have devolved into gibberish. Free free to rewrite it in a clearer manner, and without a surfiet of crufty links, but please don't simply restore it.
--Artw 18:06, 14 January 2006 (UTC)
I tried the search myself ... did not find much on-line but here's one mentioning the customization course: 411: Customizing DM Webtop.
--Aleksandar Šušnjar 22:52, 13 January 2006 (UTC) Reference

Ajax Patterns keeps being added (sometimes by me, sometimes by others), and removed again, usually indiscriminately with several other references. What's everyone's take on it? It's basically the complete draft text to a book on Ajax, with topics ranging from basic Javascript/DOM stuff to advanced architectural concepts, as well as details on 100+ famework and library details. With all content under CC. Sems fairly appropriate as an external reference, I'd have thought. Mahemoff

I suppose all this stuff is already dispatched in various pages at Wikipedia, else let me know... Spankman 12:14, 12 April 2006 (UTC)
As someone who watches this topic closely, I can attest that Ajax Patterns is the most complete and authoritative single point of reference available on the topic of Ajax. If there were to be only one external site reference included in the Wikipedia Ajax entry, this is the link I would recommend. Brentashley 13:25, 12 April 2006 (UTC)
There are billions of external links but no useful information related to Ajax for me. Spankman 06:11, 13 April 2006 (UTC)
Where are these billions of external links? Most links are to original content inside the website. Mahemoff 11:12, 14 April 2006 (UTC)
According to the guidelines one must avoid to link to "Any site that contains factually inaccurate material or unverified original research". Spankman 10:46, 17 April 2006 (UTC)
Well, you've got me there. If you're not able to point me to the billions of external links, could you help me locate the factually inaccurate material or unverified original research? --Mahemoff 10:20, 18 April 2006 (UTC)
You are funny. Since this is a wiki, the whole site is made by no matter who and it is clear there is not at Ajaxpattern enough people to verify all the pages and all the links. The bookmark sections require a code to enter and this is also forbidden by the guidelines. Spankman 12:11, 18 April 2006 (UTC)
It is "clear" because...? The site gets thousands of eyeballs a day. As it happens, many pages are read-only (although I don't see how an openly editable wiki is actually a downside). I'm not aware of any bookmark sections requiring a code, the entire content is publicly available. --Mahemoff 13:40, 5 May 2006 (UTC)
Seriously, I have watched this website closely, and found nothing valuable about Ajax. There are articles that are out of topic and seem taken from other websites, and links, links, links, but no real work. Alcalazar 11:21, 13 April 2006 (UTC)
Alcalazar, can you please provide an example an article that is "out of topic" on Ajax Patterns? Also, did you look at the 70 chapter-length entries of original writing, or just the Ajax Links page, which is the only page I'm aware of that is essentially just only links. Mahemoff 11:12, 14 April 2006 (UTC)
For example, "Page rearangement", is an article on Dom, not on Ajax. Alcalazar 10:58, 20 April 2006 (UTC)
DOM has nothing to do with Ajax? Please refer to JJG's original article (which I quote below). At least you now seem to accept the content is original.--Mahemoff 13:40, 5 May 2006 (UTC) is one of the most comprehensive collection of resources about Ajax on the web that I have seen. I do think it is very useful for Ajax developers. But so are many other sites and portals about Ajax. The problem we have with this Ajax article is that people keep adding links to their own site - useful or not. So as we have seen in the past, the link list keep growing and growing. There are two issues with keeping this link. First of all the link section is divided in two: articles and tutorials. is a collection of both but the link is to the main site while all the other links are to specific articles. I am not saying that this sectioning is good - in fact it makes it makes the job of cutting the links down harder as there are many good Ajax articles and tutorials. Second; there are many other portals and sites with a collection of tutorials/articles. Shall we add this as well? --Sleepyhead 11:45, 14 April 2006 (UTC)
I believe we should keep here links to articles closely related to Ajax (XMLHttpRequest +JavaScript all together) because we can except lot of developments in the future using Ajax, and the size of this page will become virtually unlimited. Maybe we should create other pages to cover all subjects. Ajaxpatterns is not a tutorial actually and if a link is added here, put it at root of "External links" instead. Booles 12:28, 14 April 2006 (UTC)
There's been a lot of misinformation and irrational logic in this thread, which I've tried to address, but I can see it going around in circles from here, so we're going to have to agree to disagree. I do want to point one specific technical point out though: Ajax does not mean, and was never intended to mean "What can we do with XMLHttpRequest". Please read JJG's original Ajax article and the FAQ below, which states, "Ajax is our name for the overall approach described in the article, which relies not only on XMLHttpRequest, but on CSS, DOM, and other technologies.". Although some people here do feel the site is link-worthy, there seems to be a general resistance against changing any of the links or their structure, which to me goes against the principles of wikipedia, but I guess it's understandable given the disputes that might arise. Until any structural changes happen to those links, I'm going to jump out of this conversation. --Mahemoff 13:40, 5 May 2006 (UTC)

In general I'm against adding any new links, but for what it's worth I'm in favour of adding AJAX Patterns to external links - it's a solid resource. --Artw 22:27, 5 May 2006 (UTC)

If there is a very good page on the site, it may be linked. The site if a farmlink for me. Spankman 13:51, 16 May 2006 (UTC)

I don't see why people must be fanatical against As someone deeply into AJAX and lead developer of an AJAX open source project (, I consider the most authoritative source right now. Not wikipedia. So it's funny some guys here don't think ajaxpatterns is "good enough" (oh?). Is this a cat fight? Live and let live guys. You can both exist, and one thing is for sure: this site is a sure bookmarker for anyone into AJAX, by blocking it because of your personal issues, you deny users this valuable information. Wikipedia is for its users, not its editors. My vote goes to addition to the external links section. --Lennert 19:30, 28 May 2006 (UTC)

Huh? There is no vote and you create a sockpuppet to vote? Spankman 13:14, 29 May 2006 (UTC)
They are now hidding backlinks into the body of the text. Spankman 12:29, 1 June 2006 (UTC)
Spankman, once again, you've made it seem like you have some agenda by making a wild, derogatory, claim with *zero evidence*. Now the possibility you have an agenda here is I'm sure just silly - after all, this is Wikipedia, and you are just an impartial (if anonymous) contributor hoping to maintain a high quality article. So far, you've called AjaxPatterns a "farmlink", "factually incorrect or unverified original research" (by implication), "no useful information related to Ajax", and you've made some inane comment about the "bookmark area" requiring a code, which you never explained. Which might lead some people to question your motives, but not me. So please enlighten us all on this, your latest effort to improve the quality of this article...Who are "They" are where are these hidden backlinks? --Mahemoff 12:30, 7 June 2006 (UTC)

Lists of browsers

Would someone please explain why are there lists of browsers that support/doesn't support Ajax in the article? I don't see much value in both lists because whether an ajax application works properly is more down to the application itself than the browser. Every case is different and such an over-generalisation won't help the reader understand much more, if not outright mis-leading. Case in point: people would tend to say "Google Calendar does not support Safari" rather than the other way round, ie "Safari does not support Google Calendar". --John Seward 15:24, 28 April 2006 (UTC)

  • I beleive it's more correctly a list of browsers that support both XMLHttp and a decent level of DOM Scripting, which is pretty much the purist definition of AJAX. --Artw 15:58, 28 April 2006 (UTC)
A list of browsers that support Ajax is less and less useful with time. Should be removed very soon. Booles 09:48, 30 April 2006 (UTC)
I have removed the two sections for the reasons given above. --John Seward 17:53, 13 May 2006 (UTC)
Now that section is removed I would suggets a list of browser characteristics required to support an ajax application (Javascript is esential, lack of xmlhttp would leave you relying on iframe hacks, lack of proper DOM support makes it difficult to do anything useful with any data retreived). Wether or not this should get into a discussion of individual browsers is another question- if it does then the we'll end up with a list identical to the one we just removed. --Artw 19:00, 13 May 2006 (UTC)
Perhaps a matrix of selected browsers' support of the various composition technologies of the AJAX umbrella would help. --John Seward 14:33, 18 May 2006 (UTC)
Excellent suggestion! Er... how do we make one in wiki? --Artw 14:53, 18 May 2006 (UTC)
How about this:
Support of the Ajax composition technologies across popular browsers
JavaScript XMLHttpRequest DOM ...
Microsoft Internet Explorer
IE7 Green check.png Green check.png Green check.png
IE6 Green check.png, but... Green check.png Green check.png
IE5 10px 10px 10px
Mozilla family and other Gecko-based
Firefox 1.5 Green check.png Green check.png Green check.png
Safari 1.3 Green check.png Green check.png Green check.png
--John Seward 17:25, 23 May 2006 (UTC)
That's totally awesome. We probably need to footnote it with various qualifiers as well, but I think this is totally the way to go. --Artw 17:30, 23 May 2006 (UTC)

Multiple links

Several links to a same page, in several articles at Wikipedia is SPAM. Please choose the best location, and link your page only once. Alcalazar 12:33, 18 May 2006 (UTC)

When the links are relevant to both articles, I disagree, and I have not found any mention of this on WP:EL. The link has been removed from XMLHttpRequest. Rufous 16:17, 18 May 2006 (UTC)
I think the "See also" section is due a major cull. Creating an Ajax frameworks page and putting some of them there might be a good idea, though obviously such a page would be at risk of becoming a link farm. --Artw 15:27, 18 May 2006 (UTC)
Should remove also the link to Mozilla. It is linked twice. Spankman 13:44, 19 May 2006 (UTC)

Back Button Support

I deleted the following section as it is such a horrible solution I seriously doubt anybody uses it.

An alternative to this method could be to open up an application in a fullscreen window, with no buttons or menu whatsoever, thus eliminating the need for (or even existince of) a back button. Once the new window has opened there will be no back state available.

Feel free to restore it if i am in the wrong, though I'd quite frankly be horrified to discover anyone actually doing this.

Anyhow, I thought everyone was using the fragment identifier technique these days? --Artw 18:37, 26 May 2006 (UTC)

To my knowledge, using a pop-up with toolbar and buttons disabled by JavaScript was a common trick employed by applications ever since web applications exist to avoid unexpected behaviour. For instance, the HSBC online banking system would show your accounts in a pop-up with no buttons. How much of this trick is specific to Ajax, however, is something more difficult to ascertain. --Pkchan 18:42, 26 May 2006 (UTC)

IBM Developerworks link

Do we need to delete that? It's a pretty good set of tutorials and, in one form or another, they have been part of this article for a long time. Does anyone have any background on this "spamlist" and how they came to be on it? --Artw 18:54, 28 May 2006 (UTC)

It should absolutely be blocked. Somebody at Alphaworks was involved in a concentrated (and infuriating) spam assault on various Ajax-related pages on WP. See for example, this user. See especially this comment. I can't abide that. Rufous 14:50, 29 May 2006 (UTC)
First off, your use of the term SPAM is not entirely correct in this instance. Secondly, I had read two out of the five tutorials at the IBM site and they were not selling anything (I intend to read the other three soon). Thirdly, I sent a link for this wikipedia page to colleagues telling them to explore the IBM links at the bottom and then was told by those people that they could not find the links. These links were in an external area. There is nothing wrong with doing this. I will be rolling back your changes (so stop being mental). --Neilrieck 00:44, 30 May 2006 (UTC)
AlphaWorks/DeveloperWorks are in the spam blacklist. No need to mental introspection. They must be deleted without discussion. So I do. Spankman
I still resent the fact that you (or perhaps it was someone else) decided to treat this "External Link" as SPAM. I personally found the articles very helpful and I'm sure others would too. So just for my own personal information, just who will decide when to stop considering links to AlphaWorks/DeveloperWorks as SPAM? When will they be removed from the so-called "balck list"? --Neilrieck 15:50, 30 May 2006 (UTC)
While I agree that the articles were quite good, they were added to the page aggressively and continually by a single user over the course of a number of months, despite resistance from other editors. The term spam does not have to connote lack of quality — the articles are on that blacklist because of the behaviour of its authors. Rufous 17:12, 30 May 2006 (UTC)
Fare enough but please answer my question. Who is the king pin in this virtual world that will determine when someone, or their URL, will be removed from the black list? How do any users of this system (like me) know that the people controlling this "black list" designation aren't working for an IBM competitor like Microsoft? --Neilrieck 11:32, 31 May 2006 (UTC)
Even if they are working for an IBM competitor, that doesn't change IBM's sleazy spam tactics.
If you follow the link from the first revert, you'll see that both AlphaWorks and DeveloperWorks were intended to be added to the blacklist, but for some reason, only AlphaWorks was added. If either of you disagree with this, I suggest that you bring it up on that page or with Amgine. In the meantime, I think it's fair to say that AlphaWorks links should not be added but DeveloperWorks links are okay, but Neilrieck, please bear in mind, it's possible that if you add DeveloperWorks links, then they'll simply be added to the blacklist like originally intended, and the links will be removed again.
On a more long-term note, if you resent these websites being on the blacklist, you can either convince everybody to change Wikipedia's style guide regarding external links or you can convince IBM to change their practices.
--Bogtha 14:08, 31 May 2006 (UTC)
Lots of companies contain the occasional over-zealous employee. I don't think IBM is any more sleazy than any other (although I personally have more respect for newer & younger companies like RIM + Google). That said, no one has answered my question: who decides when the IBM urls will come off the black-list? Is is an official Wikipedia person or a self-appointed Wikipedia monitor? What I'm worried about is some Wikipedia pages being monitored a little more closely than others. BTW, I've just read Wikipedia's style guide regarding external links which seems to contain a little more wiggle room than you have suggested. In the interim, what would you think about me moving the IBM links to my personal web site the add an external link to my page at the bottom of this page?
I think nobody is answering your question because the information is right there at the top of the spam blacklist Talk page. Any meta admins can edit the spam list. Please, if you have a complaint, at least do the minimum to follow it up instead of demanding that the people you disagree with help you complain about their actions.
I don't think linking to your personal homepage just to get around a spam block is such a good idea. If you can't get consensus on blacklist removal, you should respect that.
Saying that this might be the work of a rogue employee is not applicable. IBM were given the chance to respond and they did not. That indicates policy rather than an individual causing trouble for their employer. The comment Rufous points to backs this up; saying "We are trying to determine the traffic and interest of each one, to help us determine future commitment by IBM for this type of content." You really can't call that an over-zealous employee, can you?
--Bogtha 09:15, 1 June 2006 (UTC)
"I agree that the articles were quite good". Sorry, for me it is not a valid link: it looks impressive at first, but once I have read the 5 parts, I have found nothing that is not already in WP and more specially in the XMLHttpRequest specification. It is only more verbose and annoying.
Please, perform a search on developerWorks keyword, for backlinks from WP to their website, look at the number at top and tell me if I am in dream? Alcalazar 10:21, 1 June 2006 (UTC)
Another comment: read carefully the guidelines: "Generally, it is best to avoid using URL redirection sites in external links". Links to developerWorks are redirections actually. Alcalazar 10:29, 1 June 2006 (UTC)

I'm not trying to be difficult, but all WP links provided so far seem to be very subjective. No where have I seen an objective definition what WP considers a SPAM attack. Is it one perceived event per day? Four per day? One per hour? In the real world our actions are governed by objective rules, in the form of laws, which may only be interpreted by a judical system. In this virtual world I don't see any anything (yet) which could be considered police or a judicial system. There are, however, many volunteer monitors and I supect that some may be a little more agressive than others. --Neilrieck 11:38, 4 June 2006 (UTC)

History (2)

Since it's noted that there are alternatives to XmlHttpRequest, might it be relevant to include: One alternative is to employ a persistent connection for asynchronous exchange between browser and service [1]. -- 21:43, 8 June 2006 (UTC)

A link to this page has been repeatedly added to the history section. It's apparently a "a javascript module for Firefox 1.5 " and as far as I can see it isn't particularly notable, isn't particularly relevant to the history of AJAX and since it's Firefox specific may not even be relevant to AJAX. So far I've been removing it as link spam, but it keeps coming back. Artw 21:51, 8 June 2006 (UTC)

The idea was to point out a product that implements a persistent-asynch method. That page may not be highly notable, but surely the concept is? -- 22:00, 8 June 2006 (UTC)

Semiprotection 2 - the return of semiprotection?

Would all you regular editors trying to keep the article clean appreciate the return of semiprotection? I'm not sure I'm seeing much in the way of reasonable contributions by the anons, and as the unprotecting admin I feel kind of like I've left you swinging in the wind. Syrthiss 12:46, 9 June 2006 (UTC)

  • Against - although the page attracts an above average number of edits that have to be reverted to maintain the status quo they;re usually well-intentioned - and some of them are actually keepers. I think there's probably enough people watching this page with an idea of the difference to keep things running.
In general I'm against semiprotecting for long periods of time too, and it looks like you're reasonably on top of things. Syrthiss 15:51, 9 June 2006 (UTC)

Clean seperation between application logic and user interface

I just want to share a few of my personal AJAX expirences. As a team leader in a company that has been struggeling with the AJAX concepts for years, I have seen both power and trouble in this model.

I do think that by far the most fruitfull features of AJAX model is its ability to cleanly seperate user interface details from the actual functionality of the application. XML is powerfull enough to meet almost any requirements on complexity of data to be exchanged, while at the same time it allows us to limit the number of different forms that this data can take - for example by means of an XML schema serving as a contract (design by contract) between the two.

This kind of seperation is a fundamental design principle that provides transparency and stability and thus, maintainability, as any experienced "hacker" knows. An obvious practical benefit is that you can invoke the AJAX services from the command line shell, which will ideally give access to exactly the same functionality as through the browser. This in turn can easily be utilized in automated testing.

Now, this is where I whish the story ended, but unfortunately not. In the beginning the DOM API / JavaScript combination seems like a treasure of oppertunities, but all too soon you are forced to make compromises. Genrally you will only be able to acheive almost what you want, and generally only be able to support a few of the popular browsers. It is always a fierce, troublesome and extremely time consuming effort which is needed. So one problem is that the DOM API is too low-level and another that browser need a lot of quirks to do what you expect. I belive it is rather unique to this platform that by far the most time is spent on finding workarounds for API quirks.

Although JavaScript is a pretty powerfull and expressive language (when you really get to know it) I feel it does add to the problems of the AJAX model. The problem is that as you build more and bigger application components you will have to start thinking about things like reusability and stability of you JavaScript code. While JavaScript does support object orientation and various forms class derivation, it does lack the concept of encapsulation of implementation withing class hierarchies. You can use JavaScrips concept of closures for safe encapsulation, but it does not fit constructivly into the concept of class hierarchies.

One prticular problem with JavaScript is that a member of a base class cannot be made private. Thus whenever you want to add new properties to a derived class you must take care not to reuse names of already used properties in the entire chain of base classes. While most other issues can be dealt with reasonable well, the lack of encapsulation is a serious problem and really does limit the scalability of your code base.

Hope some of this input can be used in the main article.

Kristian 08:28, 11 June 2006 (UTC)

Your points might be better off as part of a discussion of DHTML or Javascript in general, since they apply to a broader field than just AJAX -Artw 17:46, 12 June 2006 (UTC)

Googling ?


I wonder how Ajax will influence google indexation ?

Does anyone know?

That would depend entirely on how the non-JS version of your site or web-app works, since Google ignores Javascript --Artw 17:45, 12 June 2006 (UTC).

Not an Acronym?

How is Ajax not an acronym? -- 05:20, 18 June 2006 (UTC)

That was a hobbyhorse of User See when I gave up arguing about it. --Nigelj 11:11, 18 June 2006 (UTC)

According to Wikipedia's acronym article, acronyms are "abbreviations... written as the initial letter or letters of words, and pronounced on the basis of this abbreviated written form." Hmmm... let's see Asynchronous JavaScript with XML. Seems to fit the definition. --Munchkinguy 01:51, 19 June 2006 (UTC)

Actually the article says Asynchronous JavaScript and XML. Note the "and", not the "with". And that's not an acronym.....? --JCipriani 01:57, 09 July 2006 (EDT)

Regardless, this should be fixed. The comment is irrelavent (does it really matter wheter or not it is an acronym?), so it should either be deleted, or followed up with a sentance explaining that it is an acronym. --Munchkinguy 01:54, 19 June 2006 (UTC)

Seems weird to me too...of course it's technically an acronym. I've heard JJG talk several times and never heard him say that it's not an acronym. Sure, he's said it's "Ajax" not "AJAX" and he's emphasised that Ajax is more than "Asynchronous Javascript + XMLHttpRequest", but it's obviously derived from an acronym to begin with. --Mahemoff 08:33, 21 June 2006 (UTC)

Double Buffering

Double buffering is a similar solution to a similar programming problem, as applied to computer graphics, i.e. video writes. The idea is to build each video page in the background then flush to presentation for the most seamless appearance, avoiding flicker, distortions, etc. I added a reference to this technique, which was promptly removed by a regular editor without explanation. It seems to me that web developers are done a disservice by ignoring the tried and tested techniques in other development arenas, and would like to see this reference reincorporated.
CleffedUp 14:19, 20 June 2006 (UTC)

It was probably removed as it's not similar enough, or relevant enough, to warrant a mention. I'd be against any reincorporation. Artw 16:42, 20 June 2006 (UTC)
Irritatingly enough, you're the one who removed it. If you don't know why you removed it in the first place, than I'd hardly think you're in a position to judge whether it belongs. CleffedUp 05:20, 9 July 2006 (UTC)
And I've removed it again. They are two seperate subjects, without any kind of link except perhaps as a weak metaphor. Artw 15:49, 9 July 2006 (UTC)
You miss the point: There are lessons to be learned in other disciplines of programming and, in fact, other professions. The classic example of the latter being the broken windows study. What do actual broken windows have to do with programming? Nothing, but there are lessons to be learned nonetheless. My interest here is in relating current practice to established theory to encourage exploration. Surely there's value in that. I would even speculate that a web developer would have conceived the Ajax framework sooner if more spent some time out of their specialty. On top of this, I simply can't understand your vehement objection to an innocuous reference to a programming metaphor. CleffedUp 14:59, 10 July 2006 (UTC)
Wikipedia isn't a software engineering learning tool, it's a wikified encyclopedia. Including similar but historically unrelated concepts doesn't add any value to the wikipedia entry. Encourage exploration in media designed for it -- there are plenty of design pattern wiki's on the web. Anon


One of the disadvantages of using Ajax is that there are better alternatives in some circumstances. Other articles report better alternatives in their disadvantages sections. For example, the Java applet article says in its Disadvantages of applets section:

  • "Though not strictly a disadvantage of Java applets, alternative technologies exist (for example, Ajax and Flash) that satisfy much of the scope of what is possible with an applet."

So, out of symmetry, it makes sense to mention in the Ajax article that Java applets exists as an alternative. As this change was reverted, I'd like to open the floor to a discussion. Stephen B Streater 11:43, 24 June 2006 (UTC)

Hmm. I might actually pop over there and modify that - I think they more properly mean DHTML. And to be honest a comparison between DHTML and flash is apples and oranges, though the comparison between Flash and Java-applets holds up a little better.
And I'd argue that if we do include it then IMHO it doesn't belong in the disadvantages section. Artw 14:52, 24 June 2006 (UTC)
I agree the disadvantages section is not really appropriate - and it probably should be moved around in the Java article too. Perhaps we could have an article to compare different Web technologies with advantages and disadvantages. It would be good to be able to see all the options laid out in one article. Stephen B Streater 18:11, 24 June 2006 (UTC)


Anyone heard of this site called AjaxProjects ( before? It seems that an anoymous user has been systematically adding links to this site and I am not sure whether the links are genuinely useful and relevant or we should revert them all as link spam. --Pkchan 18:58, 26 June 2006 (UTC)

The site looks useful and relevant, and isn't obviously a link farm (though it does feature a prominent adsense ad) but the blanket linking IS definately suspect, and the quality of the links should be checked. Artw 19:16, 26 June 2006 (UTC) links appear frequently at the end of Ajax-related blog posts (e.g. on and they are usually a generic, one-sentence, link to the site with absolutely no reference to the article. The site's content itself does seem to be original AFAICT. (Disclaimer: The site content overlaps in some respects to the site I run, --Mahemoff 19:31, 3 July 2006 (UTC)


A page seems to have been created for AJAJ, or "Asynchronious Javascript with JSON". I see no reason for such a page to exist seperatly, but should it be merged here or to JSON? Or just dropped entirely? Artw 22:36, 27 June 2006 (UTC)

Asynchronous server query is another troublesome page that may need some attention and possible merging. Artw 22:39, 27 June 2006 (UTC)
OK, I've gone ahead and redirected AJAJ to here... Artw 17:48, 29 June 2006 (UTC)


What's the fuss about the infobox now apprearing at the beginning of the article? It doesn't appear to be a general infobox, so we don't even know what that infobox is trying to describe. And the image doesn't appear to bear any official recoginiton at all (not to mention the aesthetics). I would strongly suggest to take it down. --John Seward 06:02, 9 July 2006 (UTC)

Iwas a bit confused by that. Ajax has a logo now? It probably should be taken down. Artw 15:47, 9 July 2006 (UTC)
Great. --John Seward 19:13, 11 July 2006 (UTC)