W3C home > Mailing lists > Public > spec-prod@w3.org > October to December 2011

Re: References Re: What are the requirements/problems? Re: Working on New Styles for W3C Specifications

From: Marcos Caceres <w3c@marcosc.com>
Date: Wed, 14 Dec 2011 10:36:46 +0000
To: ""Martin J. Dürst"" <duerst@it.aoyama.ac.jp>, Dominique Hazaël-Massieux <dom@w3.org>
Cc: Julian Reschke <julian.reschke@gmx.de>, Charles McCathieNevile <chaals@opera.com>, "spec-prod@w3.org" <spec-prod@w3.org>
Message-ID: <A53105F6A2E5465FA34294135730FF33@marcosc.com>

On Wednesday, December 14, 2011 at 10:00 AM, "Martin J. Dürst" wrote:

> > I haven't touched (or encountered it) in in about 10 years, so I wouldn't know any more.
> No problem. But then please don't call it weird and obscure.
Will refrain from doing so in the future. It was my personal opinion - maybe I will take another look.   
> > > > > though
> > > > > looking at /TR/ right now, there is not even a link there anymore to
> > > > > the RDF file. Much more helpful would be JSON or plain XML so it can
> > > > > actually be parsed easily by off the shelf tools like Node.js and JQuery.
> > > >  
> > >  
> >  
> > > I think giving not only a specific programming language (JavaScript) as
> > > a requirement, but on top of that requiring specific libraries is going
> > > a bit too far.
> >  
> > I think you misunderstood. I said "like" because I wanted to give those two as examples:
> Ok, sorry.
> > though ignoring the dominance of Javascript and JQuery seems counter productive.
> For writing Web applications, very much so. With respect to tools for  
> writing specs, I don't think JavaScript is at all dominant. If I have  
> missed some really important recent trends that moved all spec writing  
> to JavaScript, I'd appreciate some pointers.

Robin can probably give you a good indication of how many specs are using Respec.js.   

The tools that I know of are:

1. Respec.js (JS) http://dev.w3.org/2009/dap/ReSpec.js/
2. CSS 3 post processor (??) https://www.w3.org/Style/Group/css3-src/bin/postprocess  
3. Anolis (Python) https://bitbucket.org/ms2ger/anolis/

I'm wondering, what other tools do people use?  
> > Regardless, seems kinda sad that the format that the specs is being delivered in is not one that is supported by Web Browsers or easy for Web developers to work with.
> I haven't written any of the W3C or IETF specs that I wrote in a Web  
> Browser. Given that, I don't understand why support by Web Browsers  
> would be relevant. Maybe you can explain.

My idea was basically to XHR the /TR/biblio file and just generate the bibliography dynamically for W3C specs. That way, I don't have to do any maintenance and I always get the freshest information from /TR/. Tools like Respec.js could do the same also (it's own bibliography file is constantly, and unnecessarily, falling out of date).   
> Also, the RDF stuff was developed by people with interest in RDF (and  
> maybe even writing RDF-related specs, of which there are quite a few),  
> and at a time where JSON was not the big thing it is now, if it existed  
> at all. It's not that the W3C decided to have this data in RDF only and  
> never produce anything else.

Dom, can we have a JSON version of the /TR/ specs please?  
> > > Anyway, if you are okay to lower the requirement to just JavaScript, and
> > > are ready to do a search of two, you would quickly have come up with
> > > quite a few choices.
> >  
> Maybe I have to repeat myself, but why don't you check yourself? Google  
> (and other search engines) are your friend. Maybe you will be pleasantly  
> surprised :-). When I checked, I found quite a few RDF parsers for  
> JavaScript, but I haven't used any of them, so I can't recommend any one  
> of them.

Like most people, I'm lazy and don't have time to learn new things (would have to learn a new lib and potentially the RDF data model). I'd rather things just work with the tools provided by browsers (e.g., JSON).  
> > Again, I used JavaScript as an example - the requirement is "lots of people can use it on the Web without special tools".
> I'm still not understanding why I should write a spec "on the Web". If  
> you are writing your specs "on the Web", I'd like to know how that works.

It's pretty cool: http://dev.w3.org/2009/dap/ReSpec.js/documentation.html   
(though it's not really on the Web, just uses the Web platform to spit out a document… and fetches some resources from the Web… so it's a mix)

Kind regards,
Received on Wednesday, 14 December 2011 10:37:28 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:42:19 UTC