On Sun, 2013-04-07 at 18:50 +0200, Peter Boot wrote:
> Thanks. I'm afraid I can't comment on the project. My main reason for
> asking is that the approach is so different from what I've done myself
> and from what I have heard that others do.
Hard to tell what motivates others.
For its part, the main application for the Buddhist Translators
Workbench project (of which I am the director of software development)
is being developed in a way similar to what your developer proposed: the
server converts the document into a format that the client can use, the
client works in a large part independently from the server, and
eventually the edited data is sent back to the server for storage. For
some specific projects that may be specific reasons to not develop an
editing solution this way but generally speaking this is how I would do
> I always thought of json as
> mostly meant for data structures, and not so much for texts with spans
> and further embedded spans having various (rendering) properties and
> hyperlinks attached to them. But that may be my lack of familiarity with
> client-side programming.
I've briefly experimented with XML-as-json structures. I do no remember
running into any theoretical limitation to what could be converted to
json. Some of the implementations I tried were limited but that was just
a problem with these implementations, not with the general principle.
Ultimately though I've stopped using json. The solution I've settled on
is to have the server convert a TEI document into an XHTML fragment
(ultimately embedded into an editor page) which is structured so as to
be convertible back to TEI without loss.