Subject: Re: CSS + behavior vs. XSL (was: EcmaScript, gone?) From: Daniel Glazman <Daniel.Glazman@xxxxxxxxxx> Date: Wed, 09 Sep 1998 15:05:54 +0100 |
Chris Lilley wrote: > > As you and I both said: you can probably already do almost everything that > > you will be able to do in three years, but today you use gobs of scripting > > code and the DOM and in three years you will use more simple, robust, > > declarative languages like XSL. > > Actually, at the moment Daniel uses a simple and declarative language > called STTS which is I think his point (ie you are both arguing for a > declarative approach) but I will let Daniel describe further. Well, it seems I have the floor, isn't it ?-) This discussion is a bit away from the main subject of this mailing-list but it gives an opinion on XSL, so... Ok, STTS stands for Simple Tree Transformation Sheets. Its version 2 is a submission of Electricité de France to the World Wide Web Consortium you can find in [1]. The basis of STTS is the CSS syntax. Transformations are attached to a selector or a group of selectors. STTS 2 is based on a CSS 2 working draft but CSS 3, we are planning to submit soon, will be based on CSS2 selectors and will even propose to extend them (:last-of-type and other things like that !-) I have been discussing that with Sharon and SteveZ at the Geneva W3C AC meeting. If you read [1], please consider it as an very ugly draft of the future version. Subtree transformations are based on the use of selectors on the right side of a declaration. To be clearer, a transformation is a declaration made of a property name and a value for this property. This value is a CSS selector. Selectors are used in STTS for the description of a condition before the curly braces and for the description of fragments on the right side of a declaration. No scripting at all, no DOM at all, only declarations... You can find a simple tool implementing STTS 2 at [2]. There are binaries for Solaris 2.5 and Win32bits. We are planning to release a new version of this tool conformant to STTS 3. It allows to manipulate the document's tree, regardless to the DTD for the moment, inserting and removing structural data. It is also able to insert CSS rules in the document. We have already implemented STTS in a http server so we are able to serve HTML 4.0 documents to a client *after* removal of all deprecated elements/attributes in favor of CSS for instance... With STTS3, our server translates XML docs into HTML for instance. We clearly see some major advantages to such an approach : a) it is coherent with CSS, which is not a future solution but an *existing* solution. No script, no DOM. b) a very complex transformation is simple to write and simple to read in a single glance ; it is NOTEPADable (don't laugh ; the web is what it is now because it was notepadable). c) it is coherent with Netscape Action Sheets (in fact with CAS) which is a terrific submission to w3c. d) STTS and Action Sheets will co-evolve with CSS syntax. The three of them can easily be agregated. e) very very low-cost... f) no problem with XML. [1] http://www.w3.org/Submission/1997/16/ [2] http://www.edf.fr/der/html/produits/publications/w3c/CSSize/cssize.en.htm I am leaving my office for a week and a half and won't be able to read my mail all the time, so if you send comments on this message, please be indulgent if I do not answer immediately. </Daniel>, Electricité de France rep. @ CSS+FP WG XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: CSS + behavior vs. XSL (was: E, Chris Lilley | Thread | Who can have qualifiers in XSL?, Steve Dahl |
ANN: second release of the Koala XS, Jeremy CALLES | Date | RE: second release of the Koala XSL, Smith, Brooke |
Month |