Re: [arcmedia] [apps-discuss] Proposed charter for arcmedia

[Stian Soiland-Reyes <[e-mail]>]

OK, I didn't know the draft was only to be discussed on the arcmedia list.


I'll give some feedback as I'm trying to implement it :)

https://github.com/ResearchObject/specifications/blob/gh-pages/bundle/draft/draft-arcmedia-archive-robundle-00.txt#L142

I added a "Based on media type:   application/zip"  as my format (as
many others) is a +zip - I think this could be the case for many
specialized formats.


The draft for archive/ should perhaps mention something about when to
use +zip and friends?  And also that a specialization of an archive
need not be an archive itself - e.g. application/jar is not meant to
generally be unpacked as an archive, and hence does not claim to be
+zip or (if it was made today) archive/.


Under Encoding considerations I detailed the encoding of the filepaths
within the archives - I am not sure if this is the right place or if
we should ask about this specifically (just to remind people about
it). There could also be rules about what are legal file paths that
perhaps should be expressed.

https://github.com/ResearchObject/specifications/blob/gh-pages/bundle/draft/draft-arcmedia-archive-robundle-00.txt#L175


It might be useful to ask if the archive has a mechanism to look up
media types of the archived files - e.g. I have them in
/manifest.json, Adobe UCF/epub has them in META-INF/manifest.xml.

.

I dropped the ancient "Macintosh file type code" - I have no idea how
to make one or find out what a ZIP one is called on a Mac OS 9. This
might be useful for older formats though - e.g. for StuffIt for Mac OS
9? Are we required to have this due to older RFCs for media types?




For Fragment identifier I added that I would prefer not to have one as
I already have two other measures:

https://github.com/ResearchObject/specifications/blob/gh-pages/bundle/draft/draft-arcmedia-archive-robundle-00.txt#L229

... but perhaps for the general formats (.zip, tar.gz, etc) it would
be very nice if we can agree on a common default for fragment
identifiers. I suggest using URI-style paths, with escaping done by
the IRI.


http://example.com/archive.tar.gz#/folder%20with%20spaces/within/archive.txt


Some formats, e.g. .dmg - might not fit this pattern as it could in
theory contain multiple partitions - but generally a /-based system
should work.

Fight over if paths in tar files start with / or not begin now!




On 30 December 2014 at 12:13, Murray S. Kucherawy <[e-mail]> wrote:
> On Mon, Dec 29, 2014 at 9:58 PM, Stian Soiland-Reyes
> <[e-mail]> wrote:
>>
>> Hi - what's the status of the chartering of the arcmedia WG? I don't
>> see that it has been added to http://datatracker.ietf.org/wg/ and
>> there is no further emails on the arcmedia mailing list.
>
>
> A mailing list was created shortly after the November meeting, and the
> proposed charter was posted there and here.  There has been no response of
> any kind.
>
> -MSK



-- 
Stian Soiland-Reyes, myGrid team
School of Computer Science
The University of Manchester
http://soiland-reyes.com/stian/work/ http://orcid.org/0000-0001-9842-9718