Hold on a moment...
Only 116 people voted? Yet Firefox 3 just got 8 million downloads? I hardly think this speaks for even a reasonable percentage of the people who use browsers (IE, Opera, et al.), even if they are right on some requests.
There's been a late surge of interest in voting for a wishlist of features users want added to web browsers, after being ignored for months. With less than a day left before the OpenAjax Alliance closes its poll for most-wanted browser features 116 people have now voted, according to the project's summary. That's up from a …
The reason i did not vote on this even though i develop with AJAX almost daily, is that before i think about vector graphics or anything so nice, having the basics working would make life easier.
I no longer care what browser people want to use, but if they do insist on making IE the market share, then it would be nice if that browser didn't come to a grinding halt everytime you update the DOM in any significant way. In some cases it can take 10 x as long to render a AJAX'd drop down list than in Firefox.
My wishlist... fix IE so it complies with at least some standards.
As a commercial web developer, I'm likely only to see 'benefits' from this exercise in 5 or 10 years.
The browser release cycle and subsequent "uptake" by punters is so glacial, we're still stuck having to make sure websites work with ie6 which is now over 7 years old!
The reason we need to do this is because there is still a significant percentage of punters using this insecure old relic - over 20%.
So, should any of these suggestions be "adopted", from a commercial point of view, it's a moot point. We have to code for what the masses have, not what the geeks have.
Paris, because she knows Ajax is really only meant for cleaning bathrooms.
Developing for IE6 is hair pulling irritating - Please can MS just kill it off and force everyone onto IE7 - At least that supports XmlHttpRequest as is. Hopefully (although I'm not convinced) IE8 will at least be comparable to FF2 in terms of standards compliance.
And I have to agree with the previous comments - Doing something that should be simple just takes an age to complete in IE.
are for proper grown up desktop apps. Plenty of cross platform libraries.
Of course, that would require that the web 2.0 crowd grow the balls to install apps on the user's system, and develop the commensurate level of app-dev skills required to do so without disastarous results.
Or they could just use Flash like everyone else.
Will IE 7 (or FF 3 for that matter) run on Windows 95 or 98?
I know quite a few people using them (98 is the most recent windows I own, but I normally use Panther, Puppy or RISC OS 5.).
You could of course suggest they all buy a new computer, but I suspect that your customers would prefer site visitors to spend money on their products rather than a new computer.
What about sites degrading gracefully? If you don't have the latest and greatest, then you should get a simple non-scripted w3c compliant html site. I don't see any problem with IE 6 being below the threshold in such cases.
I usually run firefox with noscripts or netsurf (no javascript). If a site doesn't work, then I'll go to a different one. Companies lose my money, if their site doesn't work.
I think that the problem isn't just about which systems IE6 will run on. There are still sizeable numbers of users that insist on using IE6 because they have various applications that will not work happily with IE7. That and the fact that IE7 is broken in a few other ways which can sod up the rendering of some sites.
Personally, I prefer to use Firefox, mostly because of (in my case and, I suspect, in yours, especially with that Iyonix) it has the best cross-platform support.
As for degrading gracefully, there are too many idiots posing as webmasters these days to allow that to happen. Sadly, these tend to be the ones with the loudest voices when it comes to this sort of thing, hence the number of really badly thought out sites.
Native parsing does exist by means of eval() which is insecure, it allows for injection of functions into the retrieved object. A correct JSON parser ignores those and only presents the static values from numbers, strings and arrays. In the web 2.0 world of mashups and aggregations a (cough, spit) web developer must sometimes use untrusted sources for his data.
3rd party parsers exist but that is yet another script to link to the page. Aren't you fed up of waiting for 30+ .js files AFTER the page has seemingly loaded? I'm looking at you facebook!
MS would just give up and help those who wrote ActiveX / MS Jscript code port it to the wider world...
I don't do html / css / js very often and I know remember WHY the ugly hacks you have to use to get it to work with IE (document.all etc.) anyone!
But MS will do as MS likes and impliment their own b******ised version of it that will need a few more scripts to work roll on the AJAX-xbrowser.wrapper.js script.
I even tried to use VML once, still trying to forget that painful experience.
Evil Bill - well there is the MS WAY or the Highway...
It's ALL ugly hacks. Not just hacks to make different browsers all look the same and work the same, but to get them to do it at all. The web was never designed to do thin client or remote data/local interface applications, it was designed to display hyperlinked information. Scripting and vector graphics and offline working and data storage are things that need to be done in a managed and secure environment, not in browsers, developers of which once went to war over the syntax to make text blink, and everything else since.
Sometimes it is best to start again from new requirements, rather than make the existing solution work.