Hi Dennis
On 26/08/11 16:28, Dennis Sosnoski wrote:
Hi Sergey,
On 08/26/2011 09:37 PM, Sergey Beryozkin wrote:
On Fri, Aug 26, 2011 at 7:04 AM, Dennis Sosnoski<[email protected]> wrote:
...
The principle of being strict in what you send, open in what you expect,
is what led to the browser wars on the web. Each browser implemented its
own way of handling things which weren't correct, so web sites might
look good with one browser but not with another. We'd all have been much
better off if the browser builders had all agreed to just reject
anything which was not correct.
All browsers will ignore unrecognized HTML tags which is the
underlying idea behind the forward
compatibility. I think what you are referring to has mostly to do with
non-portable scripts being deployed
at individual sites, ex, IE and Firefox will probably show the same
dynamic content slightly differently...
Ok, it's true that Postel's principle didn't really lead to the browser
wars on the web. The people who were producing content were violating
the principle by being sloppy in what they produced, and the browsers
had a competition to try to work with the sloppiest HTML.
There was a lot of this, at least back in the '90s. People were creating
HTML pages with missing required start or end tags, misspelled tags, and
so on. The browsers would interpret this malformed HTML in different
ways, resulting in web pages that would look fine in one browser (the
one the developers of the page had tested on, of course) but not in others.
Sure, attempting to cope with a broken content is one issue. This is why
XHTML is now used.
But I was referring mainly to the browsers (consumers) ignoring
unrecognized tags [1]. Generally speaking is not 100% safe to ignore
unrecognized content, the classical example is a new tag informing how
many pills or something a patient should take from now on or similar,
this is why CXF takes it to the next level :-), disabling the validation
is one option but is somewhat unsafe too, thus is where a transform
feature comes in - where insignificant unrecognized elements can be
dropped selectively for the validation to succeed while still protecting
the validity of the incoming payload...
Cheers, Sergey
[1] http://www.w3.org/DesignIssues/Extensible.html#Introduction
- Dennis
