More storage format remarks
bo at klaralvdalens-datakonsult.se
Wed Jun 16 11:56:10 CEST 2004
On Tuesday 15 June 2004 20:21, David Faure wrote:
> <recurrence> : <month> is the full month name (june in the example).
> Wouldn't it be simpler for everyone to use month numbers instead?
A bit, yes. But by putting the text in here, it gets a bit more readable
to humans. Granted, month numbers are used so often that people would be
likely to match it instantly, but by writing the text, we make sure the
tag is understandable. I mean, if it just says <month>6</month> that
could mean some other things - like every 6 months after the initial
date. When it says june, there is no misinterpretation possible.
> 4.2.3: the first example says type="None", instead of "none" I suppose.
> The whole notion of ranges isn't very well defined. "none" means all
> the time, "number" is the number of times the recurrence happens, and
> "date" is... ah the ending date. Found from the examples, but a line
> about it before, would be a good idea :)
> 4.2: it says <exclusion> but the examples say <exception>. I guess it's
> the latter?
No, actually it is <exclusion>. I fixed the examples.
> Is the version attribute (for notes, events etc.) mandatory or
Don't know, really. I would say mandatory. Objections?
> All types (note, event etc.) refer to <attachment> at the end of the
> list of fields that are common to all types, but that field isn't
> defined in 4.1 (nor anywhere else?)
> 4.1: says last-modification-date. But Chapter 5 and following say
> "last-modified". 4.1 is in fact not really normative, I noticed that
> the default values etc. are specified in the definition of each of the
> type. I suspect it might be clearer to fully define the common fields
> in 4.1. (e.g. default value for sensitivity, etc.)
> Chapter 6: under the xml, it says "in a contact, the body is the note".
> Which note? Copy-n-paste error, or does this mean some random text
> associated with the contact?
Fixed the description.
> Chapter 6: in the XML, there's <email> twice, at two levels. This is
> error prone and hard to model in e.g. DTDs (although it's no problem in
> RelaxNG). So if needed we can keep it, but it looks a bit unclean. Not
> sure what to suggest... raw-email for the inner one?
I changed it to <address>.
> Note done yet with the schema, but it's in good progress. Will commit
> it, as well as test files and a script to run the test, to the "doc"
> cvs module.
I see you already checked it in - very nice :-)
Bo Thorsen | Praestevejen 4
Senior Software Engineer | 5290 Marslev
Klarälvdalens Datakonsult | Denmark
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: not available
More information about the format