[kde-doc-english]update

[Éric Bischoff]

Le Friday 29 November 2002 13:36, Frederik Fouvry a écrit:
> | > If you update a document, please also adapt the FPI from 4.1.2 to
> | > 4.2.
> |
> | I suggest this is done in HEAD branch only.
>
> I kind of implicitly meant that, but that's of course not enough.
> Now it has been enounced clearly ;-)

;-) Better shout it again :

PLEASE USE THE 4.2 DTD ONLY IN HEAD, EVERYONE.

Thanks for listening ;-).


> | Could you make the default date more recent in the template too ?
>
> Done.

Thank you very much.

Now people who forget to set the date of their documents won't appear on my 
radar screens under the category "outdated docs" anymore ;-).

Are we sure that all the copies of the template are updated? It seems to me 
that there are several copie of it lying around on HTML servers, in various 
CVS modules, etc. Am I wrong, Lauri?


> You define per person two entities, e.g.
>
> (1)
> <!ENTITY Eric.Bischoff
> "<personname><firstname>Éric</firstname><surname>Bischoff</surname></per
>sonname>">
> <!ENTITY Eric.Bischoff.mail "<email>e.bischoff at noos.fr</email>">

> These can then be freely used in the documentation (both in
> othercredit elements, author elements, ... and in the running
> text).

Thanks for the explanation.

The problem with that method is that it does not help much normalizing the 
ROLES_OF_TRANSLATORS and CREDIT_FOR_TRANSLATORS messages.

In particular, it does not reduce sufficiently the complexity of the 
<othercredit> section. This complexity is a _major_ cause of compilation 
errors, amplified by the assymetry of markup between ROLES_OF_TRANSLATORS and 
CREDIT_FOR_TRANSLATORS.

In the French team, we started with your idea (two entities, one for the name 
and the other one for the email), but we quickly switched to the current 
method, which is perfectly satisfactory.

So I suggest you reconsider method (1), so every team uses the current French 
method.


> If you want, you can also add (like the Dutch and the French
> teams are doing) entities specific for certain contributors.

We do it systematically for all contributors of all types (translators, 
proofreaders, ...).


> <!ENTITY traducteurEricBischoff       '<othercredit
> role="translator">&Eric.Bischoff; &Eric.Bischoff.mail; <contrib>Traduction
> française</contrib></othercredit>'>
>
> obviously using the entities defined in (1).

I'm not sure that it brings much to do that in two steps (1) and (2). But this 
point can be discussed.

The important to me is that I would specially like to enforce the second (2) 
strategy, as it permits to have very simple constructs like :

=========================================================
msgid "ROLES_OF_TRANSLATORS"
msgstr ""
"&traducteurJoelleCornavin; &traducteurGerardDelafond; "
"&traducteurLudovicGrossard; &traducteurPierre-EmmanuelMuller; "
"&traducteurLaurentRathle;"

msgid "CREDIT_FOR_TRANSLATORS"
msgstr ""
"<para>Traduction française par &JoelleCornavin;, &GerardDelafond;, "
"&LudovicGrossard;, &Pierre-EmmanuelMuller; et &LaurentRathle;.</para>"
=========================================================


> If the approach as in (1) is taken, it certainly pays off to
> replace the author information (and adding <personname> in the
> process).  IMO, there's no point in doing it if the entities are
> not used.

They should be used ;-)


> I hope this is clearer?

Sure it is. Thank you.


-- 
Journalist: "What do you think about occidental civilization?"
Gandhi: "That would be an excellent idea."