forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nicola Ken Barozzi <>
Subject Re: cvs commit: xml-forrest/src/resources/stylesheets text2document.xsl
Date Tue, 22 Jul 2003 06:49:17 GMT

Juan Jose Pablos wrote, On 21/07/2003 20.46:

> Dave,
>>> As far as I am aware you only crawled files that are include on the 
>>> book.xml / site.xml files.
>> All links on a page are crawled, unless specified in filterlinks.xsl.
> The way the links got on a page is defined either on site.xml or a local 
>  book.xml file.

Not necessarily. The links *can* be there, or can oven not be there.
For example, CSS styles can use images inside, but Cocoon does *not* 
crawl CSS files!

Or if we want to include some html as-is, Cocoon will not crawl them, 
and the links that they have will not be copied.

>> I have seen a lot of different ideas and suggestions about Forrest's 
>> directories
>> over just the few weeks I've been on the list.  I suggest we get a 
>> wiki set up,
>> post all the different ideas and allow for comments.  After some time and
>> perhaps IRC meetings, we could establish a firm vision of what the 
>> directories
>> will look like in 0.5, 0.6, 1.0 or whatever.
> If there is a lot of ideas on this is because not many people like that 
> set up.
> I do not think we are going to have any change on this release, but 
> having a direction to follow would helps.

There has already been a discussion on this, and you have partecipated. 
I regard that there was consensus, but if it's not the case please 
explain why WRT that discussion.
I will post a summary in a next mail titled "[SUMMARY] Cleaning Forrest 
source directory madness".

>> This would only be possible in a webapp, not a static generated 
>> website.  Even
>> then, most browsers can handle most formats.  It is normally the end 
>> user who
>> chooses if they want plain text (eg, to copy to a document), html (eg, to
>> browse), pdf (eg, to print).
> 1) I can not see the technical reason of why "This would only be 
> possible in a webapp". Please explain a bit futher.

Content negotiation is dynamic. The CLI makes *one* request, not 
different ones.

> 2) I do not talks just on formats, It could be diferent language.
> 3) Not all the user agents are end users. A search engine bot who wants 
> just the content in spanish.
> It is not me who has these clever ideas (I wish!) , it is on the 
> protocol stuff:

Do you know Cocoon? Look at it, there are all the things you need.
But remember that static versions of the site will always be less 
dynamic, of course.

Nicola Ken Barozzi         
             - verba volant, scripta manent -
    (discussions get forgotten, just code remains)

View raw message