jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stefan Guggisberg <stefan.guggisb...@gmail.com>
Subject Re: Use CND for builtin nodetypes
Date Wed, 08 Apr 2009 09:48:58 GMT
On Wed, Apr 8, 2009 at 11:24 AM, Tobias Bocanegra <tripod@day.com> wrote:
> On Wed, Apr 8, 2009 at 11:05 AM, Jukka Zitting <jukka.zitting@gmail.com> wrote:
>> Hi,
>> On Wed, Apr 8, 2009 at 10:53 AM, Tobias Bocanegra
>> <tobias.bocanegra@day.com> wrote:
>>> i would like to drop the builtin .xml file and read the builtin node types
>>> only from the .cnd file and additionally add support for a "custom_nodetypes.cnd".
>>> this will make the node types file much more readable and offers users a simpler
>>> way of providing their node types.
>> I'm a bit afraid that making the file more readable will just
>> encourage more people to modify it directly. :-(
> well, but that's security by obscurity :-) but it certainly helps the
> developers. furthermore, all the node types in jsr283 are now speced
> in CND, and converting them to XML is a pain and error prone.
>> Instead, couldn't we store the node type information as actual nodes
>> in the same persistence manager we currently use for versioning
>> information? This way we could get rid of the current virtual item
>> states we use for exposing the node types under /jcr:system. Storing
>> the type information as content would also simplify clustering,
>> backups, etc.
> hmm, i think you're talking about something else :-). storing the
> registered (builtin and custom) nodetypes in the persistence manager
> (instead of the custom.xml) would certainly be a big advantage. the
> 'versioning' workspace should then be the general "rep:system" shared
> tree.
> i was talking about installing the initial set of nodetypes, like:
> nt:base :-) those are currently stored as resource in
> builtin_nodetypes.xml, and i would like to replace it with the
> (already present) .cnd aequivalent.


WRT custom_nodetypes.cnd:
-0, i do share jukka's concerns. with hindsight i consider it a
mistake storing the registered custom node type definitions
in a human readable format ;)

> regards, toby

View raw message