The Future vs Your Archives

Time to stop and think

[Crescendo]

I've lost track of the times I've rewritten parts of the code on my websites. To improve them, to add accessibility, or to implement new code such as the latest DTDs from the W3C.

It struck me recently, that with the brand new tags like < line > and < h >, we will never be free from rewriting pages! There will always be new code, new devices and screen sizes to cater for, etc.

So how can we hope to archive our pages? If we do, they will not work in the future without rewriting the code every few years. For instance, I put in "target=_blank" in every link on my site. Then I found it was not right for accessibility, so I had to edit every page and take it out.

Then the XHTML DTD changed, so again, every page had to be modified. What happens if it's changed again? (My solution would be to make every page a PHP page, and load the DTD from a string. That way you could just change the one file with the string in it!)

And will I soon have to replace all < br > tags with < line >? I mean, authors don't rewrite their books over time do they? So why should we rewrite our pages?

Perhaps a solution might be to define everything as XML. Then process it into HTML, so all you'd have to do is change the process. (Eg: a newline swaps from becoming a < br > tag to being encased in the < line > tag instead.) Or is that just too extreme?

[rewboss]

My guess is that, for a while at least, browsers will retain the ability to render pages using old DTDs. You can still use HTML 3 if you want, you just won't be able to rely on certain features introduced with HTML 4.

XHTML 2 will not, of course, be backwards compatible with HTML or XHTML 1. What that means is that if you use one of those markups, XHTML 2 browsers might elect to attempt to transform your code, or to switch to "HTML mode" (like MSIE 6 currently has a "quirks mode") or perhaps refuse to display it. Any browser that refuses to display HTML or XHTML 1 is unlikely to sell, though, until the vast majority of websites are using the new standards. This will take a very long time, because this will require everyone to update not just their sites, but also their Microsoft Frontpage/Adobe GoLive/Dreamweaver/whatever -- and that will take years.

Some enterprising software-type bloke might come up with a piece of software that will automatically transform valid HTML into valid XHTML 2.

[Crescendo]

Another point to consider is that of broken links. If you move pages around, or delete them altogether, you get a problem when people try the old links.

When moving servers, I've put in a redirect file (against W3C recommendations) to point to my new site.

Tim Berners-Lee comes up with some interesting thoughts on this topic in his article Cool URIs don't change [w3.org]. One idea he suggests is:

Make a database which maps document URN to current filename, and let the web server use that to actually retrieve files.

So you could have a file that worked thus:

old filename / new location
index2.html -> /archives/sept00/index2.html

Great! But with 1000s of pages, would it slow down your site?

Thinking about the web as a whole also, in 5 or 10 years time, won't half the links no longer work?! If you've ever found an old site, you just know it's a lottery whether the links all work still or not.

How do we deal with this problem globally? Should everything be permanently archived in fixed addresses? (I'm aware of one site that is attempting to do this, going back to the Arpanet!)

Otherwise I fear we shall be lost in a sea of broken links, leading to loss of information.

[ergophobe]

My guess is that, for a while at least, browsers will retain the ability to render pages using old DTDs.

My guess is that all future web browsers with any significant market share will understand the standard DTD for HTML. Thus in the future you may need to add a DOCTYPE statement clearly indicating that your pages are in HTML3.2 so that even if the browser defaults to LGWT 7.6 (Latest Greatest Web Technology 7.6), it will still render HTML 3.2 pages correctly.

Of course, your pages may not look as intended on my ocular implants, but they will still be readable on the platform for which they were written.

Tom

PS - I'm a historian, so while what I say about the past is *sometimes* well-founded, what I say about the future never is!

[Filipe]

Off topic:

I mean, authors don't rewrite their books over time do they?

Many do. That's what 2nd, 3rd, 4th, nth editions are for.

[rewboss]

Usually, new editions have small typographical corrections. In rare cases, a word or two may be altered to eliminate some kind of continuity error. But that doesn't qualify as a rewrite: it's an edit. Hence the word "edition."

[ergophobe]

I mean, authors don't rewrite their books over time do they?

Actually, that's mostly because of overhead. If I could update my books every time I found an error or changed my mind about something, I would be updating them weekly.

This also doesn't work for academic books like mine because if you muck with the page count, people can't cite them accurately. If the texts had some dynamic, electronic indexing that wouldn't be upset by rewriting, again, I would do it constantly.

Tom