RE: [cddlm] early draft of deployment API
Team, Plese plan on reading the document and providing the feedback either directly to Steve or prior to the meeting on next Wednesday. Thanks, Dejan.
-----Original Message----- From: owner-cddlm-wg@ggf.org [mailto:owner-cddlm-wg@ggf.org] On Behalf Of Steve Loughran Sent: Thursday, May 19, 2005 8:45 AM To: 'cddlm-wg@ggf.org' Subject: [cddlm] early draft of deployment API
This is all in CVS, but for those without access, here is the latest copy
1. moved over to MUWS (consistently) 2. purged MOWS 3. highlighted everywhere in the docs the namespace needs to be fixed.
I've just noticed that my processing-of-options section names two faults to be returned in specific cases, but does not define the faults. I have them defined in a (non normative) doc, constants.xml, and will knock up a quick XSL to make an HTML doc. but it cannot be normative, as we will think up new faults as we go along. Or to be precise, they will find us.
ignore errors in the visio-derived images; visio is playing up on that vmware instance, and my laptop is on loan to someone else. I'll fix them from home, or even ssh down that vmimage and use it at work.
-steve
Milojicic, Dejan S wrote:
Team,
Plese plan on reading the document and providing the feedback either directly to Steve or prior to the meeting on next Wednesday.
Thanks,
Dejan.
There will be one more revision out tonight, with fixed namespaces and pictures, plus integration with the latest CMP revision, if needed. If anyone can give me feedback on yesterdays draft before then great, otherwise wait until later to work with the next copy.
This is my draft for review/submission I have spent the entire day dealing with 'issues' with MS word, the primary one being that it would crash whenever the doc was opened. This edition is a rebuilt document which I hope will be more stable. And I have provided it in PDF for everyone's pleasure. I'm not going to be overcritical of MSWord, as it has the edge of both usability and stability over OpenOffice, and any other current competitor. But it does seem capable of creating corrupt documents, documents it cannot even recover from. You can tell this is the case, because every version of the app comes out with more crash recovery tools. So when it crashes it posts a dumprep to microsoft, offers to open up the doc when it restarts, and even warns you that it crashed on load last time. I think the effort may have been better spent, however, in writing apps that dont crash on load. Just a thought. Returning to the doc, here are the changes since yesterday -updated namespaces -updated pictures -correct case of ResourceId everywhere -clear separation of System and Portal EPRs from the endpoints themselves. -moved namespace list to beginning of the doc (eliminates forward references) Overall the spec is probably making the correct use of the MUWS stuff, though that spec is so complex its hard to be sure. I havent had a chance to read stuarts doc (sorry!), but since stuart himself patched the deploy API schemas last night, I'll trust him that they are in sync. Please have a look if you can, and tell me where it is all horribly wrong. I will be at the phone conf on wednesday (on vacation the following week), and want to generate the non-normative fault and option list docs by then, in between actually doing some of the parser.. Presumably we also need to submit the XSD and WSDL and miscellaneous XML files to the editor too? -steve
participants (2)
-
Milojicic, Dejan S -
Steve Loughran