On Fri, 2 Nov 2001, J. J. Young wrote: > Steve wrote: > > >I've never had any complaints that my > >pages (which of course all contained the required DTD) > >didn't work with one type of browser or other. > > 'of course', 'required'... hmmmm.
In order to comply with the standard, the DTD is required. Of course, it's not "required" for pragmatic reasons, any more than it be "required" to use properly grammatical or validity wrds if they don't want two. ppl kan communicate irregardless of if they follow the rulez or naught. And browzers can render fawlty (to a degree) pages to. So, yes... for me, the DTD is "required." ;-) > > I even used to check my pages with either IE 3.x > >or 4.x > > 'even' -- if it wasn't too recently there was still > a large user-base and in any case IE3 is some kind of > baseline. NN2 is a step too far, for me (sorry Clarence). > Then there are the problems with IE on the Mac. If you write your code to the standard, then it's a problem with the browser if it's not rendered correctly. In order to accomodate as many browsers as possible, I write to 4.0 Transitional without CSS. I believe in the "any-browser" philosophy, and I believe that using a subset of that particular instruction set will enable the highest number of browsers to render it "somewhat as I intend." ;-) That said, I'm also coming round to the point that v3 browsers are almost non-existent, and that I really should start using CSS, which would leave the display for v3 browsers in simple black and grey. > Specific instance was to do with CSS. Problem solved > by removing the DTD. Sorry, I can't remember the location > of the article. Ah, ok... that would explain why I never got a complaint. > > Since M$ has a seat on the W3C, and since the > >standards are designed for the "real net," I don't > >see any conflict. > > I suppose not. The real Net is over-populated by users > of Win9x+ so nobody else counts? What I mean here is that M$ knows what the standards are before any of us do. If they CHOOSE to ignore the very standards they help create, then what's the point? The authors of NN6 were rained upon with complaints when their browser was more standards-compliant than any other browser to date. I still haven't downloaded NN6, but Mozilla .7 is pretty close I guess. I like the way pages are rendered in it... too bad it's such bloatware. I continue to use NS 3.04 as my default browser, though I have PINE set up to launch Arachne whenever I click a link from e-mail. > The majority are told > they should be happy. Then along comes an attack from > Ruritanians armed with flowerpots... I'm not familiar with the reference. > Real webpages are scattered with extensions like <nobr> > and kludges for v3 browsers that might validate as > HTML 4.01 Transitional but not as HTML 3.2 Right. Even such common usage as <font face> will not validate as HTML 3.2, yet all the v3 browsers will recognize it. It came out as a Netscape extension which was nearly universally embraced... which is probably the only reason it was added to 4.x Transitional, albeit as an already deprecated attribute. HTML was introduced as a formatting method which would keep content separate from presentation. HTML 3.2 crossed that line and introduced elements which merged the two. HTML 4.0, with CSS, headed back in the direction of the original intent of HTML, yet at the same time had to recognize the detour that had been taken by 3.2... and yes, that can be laid largely at the feet of Netscape. Of course, at that time, Bill Gates was still saying, "The internet is just a fad." Win'95 didn't even originally ship with ANY web browser! When he was hit with the epiphany that the internet was here to stay, a quick hack of Mosaic was done, and IE 1.0 was born, and released with a Win'95 "plus pack." Of course, Microsoft couldn't just play catch up to Netscape. They had to "embrace & extend," and then surpass the dominant browser. :-/ > The older > browsers couldn't care less about the lack of a DTD and > nor, yet, do the latest offerings. I've got code from > outside sources for e.g. hit-counters and content-rating > that won't validate under any DTD. At present the "need" for DTD seems to be limited to validators and some wireless appliances. Once XML (and perhaps XHTML) become more common, the DTD will likely become more necessary. > Personally, I'm stuck in a time-warp where I want pages > to be OK on v3 browsers with 640x480 256-color screens. > Those restrictions can cause enough hassle so I trust > in the more capable browsers being able to look after > themselves. Yup, me too. I write for 640x480, mostly stick to 256, and make it 4.0 Transitional... which means I can use font face and some of those presentation elements not possible with 4.0 Strict. I hate going to a page with NS 3.04, and not being able to read anything because the underlying HTML doesn't allow it, but CSS "fixes" everything for newer browsers. > New users in the UK are using a NetTV with > a browser identified as NN3 -- no CSS, so for static pages > I use HTML 4.01 Transitional with CSS for not much more > than justifying text and adding hover. I believe WebTV here also identifies itself as Mozilla 3. > My alternative pages are very basic (no tables > for layout or specified colors); the kind of non-design > I gravitate to, given the chance e.g. > http://news.bbc.co.uk/low/english/world/default.stm Shows very nicely in Linux Arachne 4.1.66, Lynx 2.8.3, Mozilla 0.7, NS 3.04, and NN 4.76. Amaya 4.3 is the only one in which I noticed anything odd. > Focusing on one aspect of the M$ involvement: the > widespread use of the Verdana font. This font is big > and IE's CSS implementation discouraged the use of > relative font sizing. Couple that with those thinking > it was acceptable to use <font face="verdana" size="1"> > for body text and you get squinty pages for anyone > without Verdana. "Default" font size is supposed to be 3. You're right. The widespread use of IE Verdana has made many web "authors" use font sizes which make the pages unreadable for many of the rest of us. (Linux, Mac, etc.) > Quoting from HTML: The Definitive Guide (Musciano & Kennedy, > O'Reilly & Associates, revised 1998) pp.46-47: > > "It's also unclear what doctype to use when including in > the HTML document the various tags that are not standards, > but are very popular features of a popular browser -- the > Netscape extensions, for instance, or even the deprecated > HTML 3.0 standard, for which a DTD was never released. > > Almost no one precedes their HTML documents with the SGML > doctype command. Because of the confusion of versions and > standards, we don't recommend that you include the prefix > with your HTML documents either." EOQ Argh! But then look at the grammar. "[O]ne" is the singular subject followed by the plural pronoun, "their." Is it any wonder that people without a grasp of grammar for their spoken language would likewise be deficient in the "grammar" of their markup language? I would probably just throw that book in the garbage! ;-) > Things have certainly changed but, like I said, I'm stuck > in a time-warp. I would love to use and write for a standards > compliant user agent for the OSen I'm somewhat comfortable > with, but I don't know of one. Don't write for any specific agent. Just write to the standard. Even that allows a wide latitude. I can't imagine any complainers, but if they do pop up occasionally, just point them to somewhere like http://amigaphil.planetinternet.be/bugsEN.html > Thanks for IPW, BTW. I certainly can't take any credit for it. All I did was install it and make some minor cosmetic changes to the web interface. - Steve
