Re: [xsl] [ANN] Saxon-CE - XSLT 2.0 on the browser

[Olivier Jeulin <olivier.jeulin@xxxxxxxxx>]

Le 28/03/2011 18:31, Dimitre Novatchev a C)crit :
> Having the same XSLT 2.0 processor in all five major browsers is a
> page turning event.
>
> While we are now closer to the day when sending an XML response to a
> "page request" will be mass practice, I see one often cited problem
> still not addressed:
>
> How would search engines find data? It seems that they also would need
> to perform the specific XSLT transformation on the XML (which they
> don't do currently, AFAIK, and probably would refuse to do so due to
> the additional costs involved and to the web-xmlization still not a
> predominant trend -- seems like a wicked closed circle, doesn't it).
In this case, you have at least 2 choices (there are probably others):
1) let search engine use the content of the page and of 
<noscript>...</noscript>
2) detect the type of browser and redirect it to a static HTML page 
created by the server

The problem arises also for people who doesn't use javascript in their 
browser (links, elink, ... used for example by visually impaired 
people). In this case, the 2nd solution has the advantage that you can 
still use XSLT2 or XSLT3 to create the HTML page.
It is still a big step ahead: the server can be a lot less powerful to 
process the few percent of non-javascript browsers, while most of the 
users will use their computer (PC, phone, ...) to render the page.

To me, the problem comes from the technologies used nowadays (and in the 
futur): how does this kind of transformation work with existing web 
frameworks and architectures? Would a Ruby on rails developer use XSLT2 
in the browser (choose Python, PHP, Java if you prefer)? (hint: the Ruby 
community invented YAML because they were sick of XML)
Sure, for XSLT aficionados, this is a great news, but we are not alone :)

--
Regards,
Olivier Jeulin