1460180 - Remove the RDF atoms

Status: RESOLVED FIXED

(Core :: XPCOM, enhancement)

Description

[Nicholas Nethercote [inactive]]

They aren't used now that rdf/ is gone.

Comment 1

[Nicholas Nethercote [inactive]]

Attachment: Bug 1460180 - Remove the RDF atoms.

They aren't used now that rdf/ is gone.

Review commit: https://reviewboard.mozilla.org/r/242566/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/242566/

Comment 2

[Nathan Froyd [:froydnj]]

Comment on attachment 8974249 [details]
Bug 1460180 - Remove the RDF atoms.

https://reviewboard.mozilla.org/r/242566/#review248580

Comment 3

[Kris Maglione [:kmag]]

This will make it a bit harder for Thunderbird to move the RDF service into comm-central, but I suppose they'll find a workaround.

Comment 4

[Nathan Froyd [:froydnj]]

Oooh, that's a good point; I hadn't thought about that.  Thunderbird has no good way of injecting static atoms.

I guess we could make those atoms conditional on some Thunderbird-specific #define?

Comment 5

[Kris Maglione [:kmag]]

(In reply to Nathan Froyd [:froydnj] from comment #4)
> Oooh, that's a good point; I hadn't thought about that.  Thunderbird has no
> good way of injecting static atoms.
> 
> I guess we could make those atoms conditional on some Thunderbird-specific
> #define?

I think that would probably be reasonable, yeah. Maybe just `#ifdef MOZ_RDF_SERVICE` and let them define it when they're building with RDF.

Comment 6

[Kris Maglione [:kmag]]

Comment on attachment 8974249 [details]
Bug 1460180 - Remove the RDF atoms.

https://reviewboard.mozilla.org/r/242566/#review248614

::: xpcom/ds/nsGkAtomList.h:2488
(Diff revision 1)
>  GK_ATOM(nsuri_xhtml, "http://www.w3.org/1999/xhtml")
>  GK_ATOM(nsuri_xlink, "http://www.w3.org/1999/xlink")
>  GK_ATOM(nsuri_xslt, "http://www.w3.org/1999/XSL/Transform")
>  GK_ATOM(nsuri_xbl, "http://www.mozilla.org/xbl")
>  GK_ATOM(nsuri_mathml, "http://www.w3.org/1998/Math/MathML")
>  GK_ATOM(nsuri_rdf, "http://www.w3.org/1999/02/22-rdf-syntax-ns#")

I believe this can go, too, along with the reference in nsNameSpaceManager.

Comment 7

[Jorg K (CEST = GMT+2)]

Or how about:
// Thunderbird and Seamonkey still use the RDF service.
#if defined(MOZ_THUNDERBIRD) || defined(MOZ_SUITE)

Comment 8

[Kris Maglione [:kmag]]

(In reply to Jorg K (GMT+1) from comment #7)
> Or how about:
> // Thunderbird and Seamonkey still use the RDF service.
> #if defined(MOZ_THUNDERBIRD) || defined(MOZ_SUITE)

I'd rather avoid those kind of app-specific ifdefs. You know at build time whether you're building the RDF service. You should be able to define a specific feature macro easily enough.

It also makes it easier to find code that needs to be excised when the RDF service is finally removed from comm-central.

Comment 9

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/4f27d46e7abe
Add a comment about RDF atoms. r=me

Comment 10

[Nicholas Nethercote [inactive]]

All static atoms are now in a single place (nsGkAtoms) and generated from scripts. This makes conditional definition of atoms pretty much impossible. So I'm going to leave the RDF atoms in the list. I have landed a patch that adds a comment explaining that they are used in comm-central.

Comment 11

[Tiberius Oros[:tiberius_oros]]

https://hg.mozilla.org/mozilla-central/rev/4f27d46e7abe