[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [dita-comment] Classification domain and classification map: DITA TC is considering removing this from DITA 2.0
Metadata is extremely important. We have not used the classification domain so far. To me the use-case is not really clear. In a global, Git-based no-CMS environment fed by more than 20 technical writers it is difficult to identify topics for reuse. Editors like <oXygen> or FrameMaker don’t have (AFAIK) a
user-friendly faceted topic search function, at least not for technical writers who don’t know what XPath or XQuery is. The classification maps can be a good overview of the topics that you have in the repository. In a perfect world these maps should be automatically
generated by reading metadata from the topics. Modelling them by hand is not smart. If we understand them as faceted views on the content, they don’t need to be DITA at all. HTML5 would maybe make more sense if the use case is to offer a convenient way to
browse through the repository. If the use case is to flavor topics with metadata for publication I would recommend to remove the domain entirely, because injecting metadata on container map or publication map level
is IMHO no good practice at all and should be avoided. The metadata should be stored always on topic level. If this is really needed, that would be an edge case which should not blow up the core spec. Stefan Eike Phone +49 (0)2572 879 279 Cell
+49 (0) 172 8843770 Dometic Germany Holding GmbH – Betriebsstätte Emsdetten, Hollefeldstr. 63, 48282 Emsdetten Germany dometic.com
From: Shane Taylor <shane@taylortext.com>
FWIW, my team is currently building an implementation of the classification domain to support adding classification metadata to built HTML topics. I agree that the current specification is imperfect,
but I think there's enormous value is systematized metadata both for filtering and retrieval. The current lack of support might be more reflective of the difficulty of this challenge and the lack of a unified standard for outputting classification metadata than the lack of value. I would
hypothesize that were the DITA-OT to support a basic implementation—and the current proposals that reduce complexity of the classification domain would help with this—that others might be able to build support for specific implementations. If there's interest, we might be able to contribute a baseline of support back to the DITA-OT. Ideas to consider for spec changes/simplification:
Shane Taylor On Tue, Jan 4, 2022 at 1:39 PM Kristen James Eberlein <kris@eberleinconsulting.com> wrote:
|
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]