Re: blogging using DocBook articles or so ...

[Jochen+oasis-open@Hayek.name]

>>>>> Dave Pawson <davep@dpawson.co.uk> writes:

> On Sun, 18 Jul 2010 18:01:37 +0200
> Jochen+oasis-open@Hayek.name wrote:

>> Just to make sure: there is no DocBook envolved at all?
>> I would love to see DocBook at the core.

> Not sure how it would help with Atom?
> Perhaps to create the inserted html?

I thought I could create the articles, i.e. the *content* using DocBook.

>> And you regard the file format, that "embraces" the articles
>> (DocBook?!?), as nice enough?

> The wrapper is atom, not docbook, which I publish directly.

ure. A standard. Fine. Isn't that OPML-related? Rings a bell somehow.

But usually (I thought at least) the wrapper (i.e. the feed) does not contain the entire articles,
but it only refers to the content.

But at least some feed providers restrict it to that, as far as I can say.

I would like it that way.

I don't like to slurp an entire newspaper into my iPhone and not via wifi,
and only read 2% of it.
Yes, it's a little more time consuming, to read the articles,
that you decide to go for from the feed "TOC".
But I find that more reasonable.

But maybe I got something fundamentally wrong.

>> I mean, I am sure, it must actually be nice enough, if it's Dave
>> Pawson's work.

>> Whatever grammar you are using,
>> I guess we can derive RNC from that,
>> which will enable my happy authoring in emacs with nxml-mode.

> I have some code which generates the 'wrapper'
> $new tag "Longer title" generates

> [...]

> Then I just type html into the div, as html.

But I could basically just wrap the URL for the article itself in fullblown HTML here, right?
I mean, some articles are not just 140 characters, right?!?

> I use emacs nxml-mode for authoring.... but nvdl for validation. 
> Then run a build script, using xslt.

Sounds great to me.

>> 
>> If I create an article (for the time being I asssume a DocBook
>> article in one HTML chunk), how much extra work is it around that?
>> do I have to manually keep redundant information like the article
>> title and its creation DateTime?

> As you can see, some info is required. 
> I use the date as the key 'hook'
> with tag and title coming from the command line. 

>> 
>> Can we get away from that sort of redundancy?

> From the above, you can see that the html header from docbook is all
> redundant?

Not from the above,
but if I want to keep the article separate, then yes.

>> 
>> 
>> I am only asking silly questions and creating cheaky requests.

> No, just strikes me as 'I have a hammer' 
> it all looks like  a nail? 

Well, I think we have different approaches,
but I think I can make use of your software for my approach.

> I don't really see how docbook can help with a blog distributed in atom?

Separate articles, each written in DocBook, atom being the wrapper pointing to articles living for themselves, one by one.

> regards 
> Dave Pawson
> XSLT XSL-FO FAQ.
> http://www.dpawson.co.uk

Kind regards,
Jochen hayek

http://Hayek.name/Jochen.html

P.S.
I am taking my knowledge from some O'Reilly "Short Cuts", namely:

$ pdfinfo 9780596528221.pdf
[бн]
Title:          Getting Acquainted with OPML
Author:         Amy Bellinger
[бн]

$ pdfinfo 9780596529383.pdf
[бн]
Title:          How to Build an RSS 2.0 Feed
Author:         Mark Woodman
[бн]

But maybe that literature does not really fit here.