On 11/24/2011 12:44 PM, Lana Brindley wrote:
On 11/24/2011 11:56 AM, David O'Brien wrote:
On 11/23/2011 04:00 PM, Jeff Fearn wrote:
On 11/15/2011 05:38 PM, Raphael Hertzog wrote:
Hello,
I don't have the habit to discuss decisions of maintainers but in this
particular case I would like the feedback of the wider community
on this specific bug report that got closed as NOTABUG despite
feedback from the upstream community proving that publican
was doing the weird thing here.
https://bugzilla.redhat.com/show_bug.cgi?id=752851
The upstream docbook-xsl stylesheets do preserve whitespaces and
linebreaks in inlines elements within verbatim environment.
Do any of you have any document where you need the current Publican
behaviour of converting any sequence of whitespace/newlines to a single
space in inline tags within verbatim environments?
To me this behaviour just ensures that I'll never be able to use
any inline element within a verbatim tag because I'll lose
the formatting of my<screens> <programlisting> and so on.
It's really counter-productive IMO and I can't really see who
would like to lose his formatting...
Cheers,
For me this raises a larger issue, should Publican be interfering with
the source XML at all?
XmlClean exists for two historical reasons, neither of which still
exist. Migrating old html/sgml content to XML, enforcing certain
standards are met.
The old content was migrated years ago and that process required a bunch
of scripts that never made it in to Publican ... they were horrific
people, horrific!
We backed away from enforcing standards sometime ago, see the death of
STRICT mode.
Given that neither of the reasons the code exists are of any relevance
any more, maybe we should just remove XmlClean completely. This would
see the death of clead_ids as well, but that was specifically for
migrating content and has never been supported for general use.
Comments?
Cheers, Jeff.
_______________________________________________
publican-list mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/publican-list
Wiki: https://fedorahosted.org/publican
I would mourn the loss of clean_ids.
Perhaps it was created for a particular purpose that may no longer (or
rarely) exist, but clean_ids has proven to be very useful for other
purposes. I know I'm not the only one who uses it; I also know several
people who are against using it. Those sorts of debates will always
exist, and are probably healthy.
Is it not possible to remove only those aspects of xmlClean that are not
used, and retain clean_ids? If I were a coder I would happily get
involved in this exercise, but I'm an end user who sees the potential
loss of a very useful tool.
+1
Used deliberately and with caution, clean_ids is a great tool.
L
Since the white space munging code in XmlClean is causing yet another
issue [1], for 2.9 I'm killing off the white space munging, so clean_ids
won't reformat your XML in 2.9. It will still change all your ID's when
and how it sees fit and it might even update the PO files to some extent
... maybe even correctly ... you roll the dice and take your changes
with it :)
Cheers, Jeff.
1: https://bugzilla.redhat.com/show_bug.cgi?id=756756
_______________________________________________
publican-list mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/publican-list
Wiki: https://fedorahosted.org/publican
