MIME type application/opensearchdescription+xml

[xyzzy]

Hi, is MIME type application/opensearchdescription+xml as you'd expect it to be?  Obviously nobody is going to create an RFC, and the MIME types registry requires a "recognized standard" (or similar) for subtypes in the standard tree.  OpenSearch is no SDO like the W3C or ECMA, but could be considered as a "single purpose SDO" (= maintain the OpenSearch specification + extensions).  I plan to test this theory, check out RFC 4288 for the details.

-Frank

http://www.opensearch.org/User:Xyzzy 

[DeWitt Clinton]

Interesting, I hadn't thought to inline the registration template right into the spec itself.   I'm curious also to see if it works.  

Minor nits:

1) How about, instead of:

 If specific extensions are desired ".a9.xml" (or similar) might do the trick.

We go with:

   File extension: .osdx (by convention)

Microsoft uses that, and it doesn't seem to collide with anything (after a cursory web search). 

2) Instead of the acronym 'OSDD', how about we use 'OSD'.  There's only a small amount of formal prior art ('osd' appears once in the spec text).   In the wild, 'OSD' seems to be more prevalent by about a ratio of 10:1.

3) Under Author, do you think it appropriate to list someone other than me?  You, perhaps, especially considering you've done all of the work?  Or even just "The OpenSearch Community" or somesuch if that's viable and you prefer?  I don't feel strongly, other than it just feels weird for me to be credited authorship for other people's work.

4) Not sure how I feel about the large block of pre-formatted text in the middle of the spec.  I'd probably just wiki-format it for consistencies sake, even if it deviates with RFC conventions (considering the rest of the spec does anyway).  I'm happy to make that change myself once you're comfortable with the content, if that's easiest.

All mostly minor suggestions.  Thanks again for thinking of this!

-DeWitt

--
You received this message because you are subscribed to the Google Groups "OpenSearch" group.
To view this discussion on the web visit https://groups.google.com/d/msg/opensearch/-/r93szKewrH0J.
To post to this group, send email to opens...@googlegroups.com.
To unsubscribe from this group, send email to opensearch+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/opensearch?hl=en.

[xyzzy]

Hi, I posted the "preliminary review request" before I saw your reply here:  I forgot to tell Google Groups that I want e-mail copies of direct replies....  No problem, I can tweak the "preliminary review request" as long as needed.  At some point in time the request goes to the IESG, and then the template in the spec. and the template forwarded to the IESG *SHOULD* be identical ;-)

* s/OSDD/OSD/, yes I like that much better, I'll update the Wiki.

* Author: No, that should be the same as in Author.  Same idea as in the expired I-D. 

* File extension .osdx: I'll add that to the text, but I've never heard of this convention before.  My own "invention" .a9.xml has the nice feature that anything working for .xml "automagically" includes .a9.xml.

* Format: The spec. must contain the complete template according to RFC 4288 section 4.10 second paragraph. Please delay any "un-pre-ttifying" until it is registered.  But you can hide the template deep in some sub-sub-sub-section of an appendix or similar, admittedly the <pre> stuff is ugly.  For starters I'll try "font-size: smaller". 

For others:  There is zero guarantee that this registration will work.  The IESG can simply reject it claiming that "opensearch.org" is no SDO (standards development org) they've ever heard of.  After that the IAB could overrule the IESG in an appeal process, but I don't volunteer for an appeal.  OTOH *iff* the IESG accepts it there will be an ordinary IETF Last Call, and the IETF community could reject the registration as the worst case of "media type squatting" ever daring to meet reality.  Which is precisely what it is -- in parts I consider this registration as a "happy IANA" experiment:  The mailing list happiana@ietf discusses issues with IANA registries, among others URI schemes and media types.

-Frank

[xyzzy]

Update, now I've found two *.osdx  below  C:\Program Files\Microsoft SDKs\Windows\v7.1\Samples\winui\shell\shellextensibility\OpenSearch  on my windows 7 SP1 box.  I think I'll put the revised template in a revived draft and post it as "not for publication as RfC" (or similar) only for info, if that's still possible (used to work six years ago).

-Frank

[xyzzy]

On Thursday, October 20, 2011 3:18:05 PM UTC+2, xyzzy wrote:

I'll put the revised template in a revived draft and post it as "not for publication as RfC" (or similar)

done 

[xyzzy]

Hi, 

after some trouble with the Internet Draft submit tool and the "not for RFC" boilerplate I've fixed that in version 03 and sent a registration request to the  IESG.  If you happen to be at IETF 82 here's your chance to explore some dark corners of happIANA and I-D.freed-media-type-regs.

Regards,

 Frank