61408 - [FIXr]please document mimeTypes.rdf helper app database fields

Status: RESOLVED FIXED

(Core Graveyard :: File Handling, defect, P1)

Description

[James Salsman]

From bug 46135:

> Scott and/or Ben, could you please provide ...
> documentation for each of these helper app fields:
> 
> Description -- ?
> Value -- ?
> FileExtensions -- what format(s) are allowed?
> Path -- ?
> SaveToDisk -- ?
> AlwaysAsk -- ?
> HandleInternal -- ?
> PrettyName -- ?

> are there any general docs on working with nsIRDFDataSource? 

> In particular, how do I properly handle the likely case that
> an old RDF database might be used when looking for new fields?

Thank you!

Comment 1

[James Salsman]

(I asked these questions a while ago, and I'm pulling them 
out to a new bug because the answer will help solve both 
the blocked bugs marked here.)

Comment 2

[James Salsman]

Resetting the QA Contact to original -- I have no idea 
who it should be, but the RDF QA Contact might be the 
most likely to know where the information is supposed 
to be documented once it has been.  Sairuh and Shrir 
can watch via cc.

Comment 3

[James Salsman]

See also:

http://lxr.mozilla.org/seamonkey/source/uriloader/exthandler/nsExternalHelperApp
Service.cpp

Comment 4

[James Salsman]

I'd like to take this opportunity to point out these two confusing 
adjacent lines in mimeTypes.rdf, which Ben G. and I discussed briefly 
without any real resolution:

 22         <NC:handleInternal>true</NC:handleInternal> 
 23         <NC:handleInternal>false</NC:handleInternal> 

Clearly the former is what you would expect in the definition, as it 
is for text/html, so can line 23 be deleted?

Thank you, Scott, in advance.

Comment 5

[Ben Bucksch (:BenB)]

> <NC:handleInternal>true</NC:handleInternal>
> <NC:handleInternal>false</NC:handleInternal>

Bug 56224

Comment 6

[James Salsman]

Is there any chance of this ten-minute task being done any time soon?

Comment 7

[James Salsman]

Michael:  Would you please ask that this be made an immediate priority?

Comment 8

[Karl Johan Kleist]

If this still is a blocker for bug 58811, please re-consider priorities.
Mozilla gets badwill from the fact that the demos here
< http://java.sun.com/products/javawebstart/demos.html > don't work.

Comment 9

[Boris Zbarsky [:bzbarsky]]

1) This is not a 10-minute task.  I know how the fields are used, and I estimate
   it would take me an hour and a half or so to document them.  I happen not to
   have that hour and a half at the moment....
2) It's assigned to the wrong person, imo (ccing law, who would be more likely to
   be able to document this)

Comment 10

[James Salsman]

Even ten minutes of documentation on this database would help with the block on 
a lot of the content-type dispatch issues.  Please help now.

Comment 11

[James Salsman]

PLEASE PLEASE PLEASE!

Comment 12

[Boris Zbarsky [:bzbarsky]]

ugh, taking.  This has gotten ridiculous; here's hoping drivers take a
comment-only patch during freeze.  ;)

Comment 13

[Boris Zbarsky [:bzbarsky]]

wrong component, too.

Where do people think the documentation should go?  I suppose I could stick it
in the default mimetypes.rdf as simple XML comments; it would get clobbered in
the in-profile generated versions, I presume, but...

The other alternatives would be to put this documentation in
nsExternalHelperAppHandler, in a document on mozilla.org, etc.  Please let me
know (I'd rather have on location and have other places pointing to it than
having multiple copies of the same text; this makes me lean toward putting it on
mozilla.org, somewhere in http://mozilla.org/catalog/libraries/uriloader/ or
something).

Comment 14

[Boris Zbarsky [:bzbarsky]]

Attachment: something like this..

Comment 15

[Boris Zbarsky [:bzbarsky]]

Note that I plan to create an nsIRDFMIMEService which would be able to return
MIMEInfo objects or persist such objects out to mimeTypes.rdf.  At that point I
will move the documentation to that service implementation, since there will be
no incentive for anyone to munge the datasource directly.

Comment 16

[timeless]

Comment on attachment 112432 [details] [diff] [review]
something like this..

>+   NC:useSystemDefault - "true" if we should handle per defauly OS setting,
defauly => default

>+ Each ""urn:mimetype:externalApplication:major/minor" resource can have the
""urn: => "urn:

Comment 17

[Boris Zbarsky [:bzbarsky]]

fixed.

Comment 18

[Boris Zbarsky [:bzbarsky]]

fixed, even.

Comment 19

[Chris Petersen]

Marking verified per last comments.

Comment 20

[James Salsman]

Boris,

I hope this is a quick question:

Does NC:externalApplication and/or NC:path use input filename %s substitution 
and/or stdin?

Can you answer for at least the big three platforms please?

Thank you.

Comment 21

[Boris Zbarsky [:bzbarsky]]

It does not, on any platform.  See bug 57420.  The reason that I did not put any
more comments than I did in there is that that field is a serialization of a
file object which is more than just a file path (at least on MacOS).

In any case, please do not reopen bugs just to ask a question....