Re: New document (rules) / collisions with infrastructure schema [message #1045 is a reply to message #1044] |
Tue, 28 September 2004 16:12 |
Matthias Hengartner
Messages: 57 Registered: August 2003
|
Member |
|
|
Hello,
it's fine to have some guidelines for creating railML schemes.
But I detected some collisions between these rules and the current version
of the infrastructure schema (regarding the rules for the naming of
designators).
[I already posted this in the infrastructure-newsgroup.]
In the current version of "Bildungsvorschrift für railML-Teilschemen"
(Creating Instructions for railML partial
schemes) from Nils Poldrack
(www.railml.org/documents/general/creatinginstructions_ger_R C1.pdf), section
3.5.6 "Bezeichner"/"Abkürzungen" ("designators"/"abbreviations"), there
is(among other) the following rule:
* Abbreviations or contractions may not be used as part of designator names
---> this rule collides with different designators in the infrastructure
schema (e.g. infraAttrGroup, ocp***, dir, posOnNet, absPos, .......)
It's not my opinion that we have to rename these designators, but perhaps we
should change the wording of this rule ("should not" instead of "may not").
Best regards,
Matthias Hengartner
-----------------------------
Matthias Hengartner
IVT ETH Zürich
++ 41 1 633 31 09
hengartner(at)ivtbaugethzch
-----------------------------
"Nils Poldrack" <poldrack(at)ivifhgde> wrote in message
news:cjb6tg$quh$1(at)sifaivifhgde...
> Hello,
>
> a new document was released: Creating Instructions for railML(R) partial
> schemes, Release Candidate 1. Available at
> www.railml.org/documents/general/creatinginstructions_ger_RC 1.pdf
>
> The English version will be released within a few days.
>
> Here is a short summary of the content:
>
> - railML(R) is no data modell but a data exchange modell;
> - schemes are written in English, according to W3C's XML Scheme
Definition;
> - comments within the scheme (annotations and element history);
> - names conventions for elements and attributes;
> - format of date and time;
>
> Not included is the never-ending story about Globally Unique Identifiers
> (See discussion articles in group railml.misc for further information.)
>
> Please, send me your comments and ideas via news group. A discussion is
> possible only this way. Thank you
>
> Nils Poldrack
>
>
>
|
|
|