OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

xdi message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [xdi] Full XDI Graph Model ABNF inspired by Markus


I love the idea and how we're making progress, but I'm still having problems with it.

For example, now with this latest version again I don't like that the "member" rule is part of the "collection" rule.

E.g. if you have =markus$(+email)$!(!1), the ABNF would tell me I have 2 subgraphs, but I really have 3.

Yes let's talk more about it, I'm available.

Markus

On Wed, Feb 20, 2013 at 6:07 AM, Drummond Reed <drummond.reed@xdi.org> wrote:
Markus and Joseph and I had a good talk about the "long form ABNF" on the https://wiki.oasis-open.org/xdi/XdiAbnf/Discussion page today. First Joseph posted an update with some fixes, including the one that was causing the APG ABNF parser generator to throw an error. Then we agreed on several other small revisions for which I just posted an update version of the long form ABNF at the top of the https://wiki.oasis-open.org/xdi/XdiAbnf/Discussion.

Please look this over and feel free to post any comments - Joseph and Markus and I hope to have another discussion about it tomorrow.

=Drummond 


On Tue, Feb 19, 2013 at 9:50 AM, Drummond Reed <drummond.reed@xdi.org> wrote:
Markus, good comments. I noticed the "member" issue last night - a member can't actually be a standalone subgraph outside of a collection. So we need to fix that.

Yes, let's talk later today. I should free up about 11AM PT. Ping me on Skype.

Thanks,

=Drummond  


On Tue, Feb 19, 2013 at 8:43 AM, Markus Sabadello <markus.sabadello@xdi.org> wrote:
Just posted a reply to that discussion page. 

Wanna talk about this some time later today?

Markus


On Sun, Feb 17, 2013 at 7:29 AM, Drummond Reed <drummond.reed@xdi.org> wrote:

Per the minutes of the last meeting that I just sent out, after yesterday's call Markus started working on an ABNF that would incorporate the rules for the full XDI Graph model and not just the basic statement structure and legal characters that it has expressed ever since we first published it.

He posted as far as he got to https://wiki.oasis-open.org/xdi/XdiAbnf/Discussion last night, noting that it was not complete, but that it did look very promising. 

That inspired me to start working on it, and in fact I spent about 8 hours between last night and today (my wife is out of town ;-).

The reason I became so engaged is that I quickly became convinced Markus was right - we owe it to ourselves and implementers to capture every key semantic concept and rule about the XDI graph model in the ABNF if we can. And it should be a big red flag if there's a rule we find we can't express in the ABNF.

I'm very excited about the result. Although of course it is longer, it captures all the semantics and rules about the XDI graph model that I could think of (including liternal context nodes and multiplicity), plus some that we hadn't even called out explicitly yet (like the fact that a literal statement can only exist on an attribute -- you can't have one on an entity or a root.) I also incorporated the IRI character ABNF that Joseph developed (thanks Joseph).

Please do take the time to review it when you can:


As always, please do comment or contributed on the Discussion page, and also post a note here on the list. I'd like to move review on this forward as far as possible this week, so we can move this into Last Call as soon as we can.

Thanks,

=Drummond 












[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]