koha-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Koha-devel] MARC Editor in 'stable' CVS Badly Broken


From: Paul POULAIN
Subject: Re: [Koha-devel] MARC Editor in 'stable' CVS Badly Broken
Date: Mon, 19 Jun 2006 10:49:26 +0200
User-agent: Thunderbird 1.5.0.4 (X11/20060516)

Joshua Ferraro a écrit :
Hi Paul et al,

Hi Joshua,


This morning I updated my working "stable" CVS repo to take a look at the recent MARC editor commits and I discovered that things are very badly broken:
After 6 hours of attempting to track down the cause of these and other
bugs I didn't take notes on, I have given up. I'd like to request that
we revert the stable CVS MARC editor to the previously working
versions:

I tried to install a 2.2.6RC1 at IPT, and they reported a few other problems. They are back to 2.2.5

Biblio.pm 1.115.2.54
addbiblio.pl 1.52.2.40
addbiblio.tmpl (not sure, many unrelated changes committed at the
same time)

I am OK with this, although i'm quite unhappy because I spent many hours on the modifs.

I will repeat that MARC::File::XML can handle iso-8859-1 encoding
if you specify that you're working with non-MARC21 data (and if
you have 8859 encoding, you don't have MARC21 data) IF YOUR
XML PARSER CAN HANDLE 8859 ENCODING. Someone with non-MARC21 data
should take a look at the latest Sourceforge version of MARC::File::XML to see how this is handled and modify Koha's editor
accordingly (I may have a chance to do this at some point, but I'm
a bit burned out on this at the moment).

I could not succeed to use MARC::File::XML
However, this is not really a problem : the hack to use HTML2XML then xml2marc or directly html2marc is 4 lines long, and totally harmless when TemplateEncoding=UTF-8.

All your problems are related i'm afraid to the work I made to have a 100% javascript CloneTag. The 1st one did not work well, I tried to fix it, and had to make too many changes after... The MARC editor has so many limits, that I'm afraid we (I) tried to do too many things with it : we have no chance to succeed.

So I think we can :
* go back to previous version
* apply TemplateEncoding patch
* apply the tmpl_loop BIG_LOOP patch, that makes the templates much much more easy to fix. * reintroduce the management of authorities in npl (they were not working, a small javascript was missing)

I'm away for 3 days in a few hours, I won't do that before departure, but feel free to do the revert, i'll take care of re-applying patches on thursday/friday.


How/what can we do to use this failure as a lesson, to be sure we won't reproduce it another time ?
--
Paul POULAIN et Henri Damien LAURENT
Consultants indépendants
en logiciels libres et bibliothéconomie (http://www.koha-fr.org)
Tel : 04 91 31 45 19




reply via email to

[Prev in Thread] Current Thread [Next in Thread]