----- Original Message ----- From: "Hector Santos" <hsantos@xxxxxxxx> To: <ietf@xxxxxxxx> Sent: Saturday, April 18, 2015 5:10 PM > On 4/18/2015 3:59 AM, t.p. wrote: > > I find it worse than that. Just over a year ago, Martin Rex said, about > > an upgrade to a different IETF server, > > > > "Using IE6 to test a web page is perfectly reasonable. > > > > If it doesn't work in IE6, it uses more complex/esoteric features > > than what is necessary and should be used in the first place. > > I understand, and once upon a time that was very true, but I think by > this point in browser history time, its closer to IE9 as a minimum for > the windows world of end user support. That is why I go by for our > intranet Web products (we supplied canned dynamic web pages) and I > have to test all the browser out there. Its was a difficult job. > > I think a good way to lay this out is to have some web-support level > terms such as: > > Web 1.0 Pure HTML/CSS, no Javascript, no client side offloading. > Web 2.0 with Javascript, Ajax, Client side offloading, etc. > Web 3.0 with HTML5, "Web Sockets" > > The trend is to offload mode of the client side work to the end-user > devices. Well, trend has always been cyclic; centralized, > decentralized, we are in a hybrid mode now, of centralization with > smart offloaded fatter devices aka apps. Most of the top browsers > support HTML5, but javascript should be a given. It would be nice > (not as fancy) to keep Web 1.0 web site, but the IETF decision should > be made if WEB 2.0 is the minimum device expectation for the IETF. > > > The server obviously fails the golden IETF rule "be conservative > > in what you send out" (as HTML) and the server ought to be fixed" > > Its hard to support all the browsers, especially all the devices. But > its all possible of course. It is expensive. We made that move > ourselves around 2006 from Web 1.0 to Web 2.0 before requiring > Javascript where it was needed, otherwise it was optional. I expect > within a year or so, more Web 3.0 technology to be used for our > intranet package and expecting MOST of the interested end users > (targeted customers) to be ready. Testing will tell us. :) > > > I have performed the test. It fails, comprehensively. > > Well, when simple Javascript-disabled testing is obviously missed, it > says people working on it probably have all the latest gadgets, more > current/modern hardware, speeds, etc, and they don't do full end-user > testing and/or they presumed the interested end users will be ready > for it as well with latest hardware and browsers, etc. Its just a > matter of testing. > > So the question is, does IETF want to continue with WEB 1.0 or at > least WEB 2.0 for now? If WEB 1.0, then perhaps IE6 browsers should > be included in the test suite. IE6 does have some major issues with > Javascript so, I would expect a IE9 level of support. Hector Your analysis is much more professional than mine. I was struck by the comment in your initial post about Javascript. I do enable it but notice the impact on scrolling in return for which I see little or no improvemeent in function. Even IE8 is noticeably slower in paging down and as to why someone thought it a good idea to put all the links to the other bits of the IETF nexus (IANA, IAB, IRTF, ...) that I use so often, 20 page downs away (at least on http://datatracker.ietf.org/wg/ which used to be my home page) well, I am clearly on a different planet:-( Ditto whoever thought it a good idea to take a away the search function for I-Ds from that page. What I often do is disable images since there are so many of them on websites and they take so long to download, so another bugbear of mine is when the replacement text is meaningless. In the worst case, you get the same text in place of forward, back, cancel and submit icons!. So meaningful text for anchors is something I would regard as a requirement. In passing, the TLS list regularly carries figures, which I can never find when I want to, for the prevalence of such as XP, with its latest level of IE8, and I seem to recall providers there saying that XP still had to be supported, it had too many users to be discarded. So I would suggest IE8 still needs to be supported. Tom Petch > -- > HLS