887463 - [webvtt] Remove webvtt library

Status: RESOLVED FIXED

(Core :: Audio/Video, defect)

Description

[Caitlin Potter (:caitp)]

As the legally binding IP of this [github.com/mozilla/webvtt] repository belongs to me, and I no longer have commit access, I am revoking Mozilla's license to use this software.

In order to avoid violating intellectual property and copyright law, the media/webvtt directory must be removed from mozilla-central. This includes all works derived from this initial repository.

See: https://github.com/mozilla/webvtt/issues/419

Comment 1

[Caitlin Potter (:caitp)]

Legal team will be involved if this is not accomplished shortly

Comment 2

[Caitlin Potter (:caitp)]

Further, the github.com/mozilla/webvtt repository and all backups are to be deleted.

Comment 3

[Ralph Giles (:rillian)]

I've been talking to people about how to address this request. More when I know how to respond.

Comment 4

[Caitlin Potter (:caitp)]

There are two ways to address this request, and only one of them is sensible.

1) The pertinent code is deleted from mozilla-central [the media/webvtt directory], code making use of this directory is commented out for the time being, and the mozilla github repository and all backups thereof are deleted. This is not a particularly sensible solution, as it requires more time and energy to be used, both to delete code, and to write a workable replacement.

2) Much more sensible; either the upstream mozilla/webvtt repository is deleted, or ownership of it is transferred to me. It is obvious that nobody will be maintaining this code if it is not me, people have other things to do -- and I feel that not having rights to this repository is preventing me from maintaining it. Yes, I could say that the canonical repository has moved, but that does not really reflect reality. I'm not opposed to the review process, but clearly no review is actually happening until it's time to check into mozilla-central, as this is the only time there are actually eyes on the code.


Option 2) takes considerably less time and energy to enact than option 1), and has the benefit of not requiring any working code to be thrown away.

- I am the SOLE MAINTAINER of this code
- I am also the SOLE INTELLECTUAL PROPERTY AND COPYRIGHT OWNER of this code, despite the
  LICENSE, until I am handed a diploma.
- On these grounds, if I am NOT IN CONTROL of the upstream repository, then I will NOT
  MAINTAINING the code any longer, and will NOT GRANT LICENSE to use the code.

The code within the repository, legally, is my own. If ownership of the repository is not transferred to me, then Mozilla will no longer have the right to use it, and a bunch of time and energy will be wasted for no good reason.

I'm not bending these conditions, I will be speaking with legal on Tuesday if this isn't resolved. I do not wish to have my work stolen from me.

Comment 5

[Caitlin Potter (:caitp)]

Time is running out, all I need to see is a merged patch removing the library from the tree and deleting the GitHub repository, or else I need to see either the Mozilla repository deleted, or ownership transferred to me (contact GitHub support). The third option really sucks, and nobody wants to see that happen.

I am not patient with thieves.

Comment 6

[Andreas Gal :gal]

Please review the bugzilla code of conduct. bugzilla is not a place for legal discussions. I will assign the bug to me and facilitate an email reply from our legal team directly to the reporter. Please do not add any further comments to this bug that are outside the scope of bugzilla's purpose.

Comment 7

[Caitlin Potter (:caitp)]

This is not a legal discussion, this is a bug for the delivery of requirements established in legal discussions, which includes the removal of code for which Mozilla is not granted license, if the alternative, smarter decision is not chosen.

This is a bug, and it can be resolved by a patch. There is no code of conduct violation here.

Comment 8

[Caitlin Potter (:caitp)]

Sorry Ralph, but you really leave me no choice in this matter. If you didn't think this would have happened, you should have thought harder about what you've done.

Comment 9

[Ralph Giles (:rillian)]

Attachment: remove webvtt parser

Patch removing webvtt parser. Call sites are either removed or stubbed out, and corresponding tests are disabled.

I'm still seeing a timeout on test_seek.html, so there will be one more small change to fix that.

Comment 10

[Ralph Giles (:rillian)]

https://tbpl.mozilla.org/?tree=Try&rev=6078c74b73be

Comment 11

[Ralph Giles (:rillian)]

Attachment: remove webvtt parser v3

Fixes a clang uninitialized warning:

--- a/content/media/TextTrackCue.cpp
+++ b/content/media/TextTrackCue.cpp
@@ -185,7 +185,7 @@ TextTrackCue::ConvertNodeTreeToDOMTree(nsIContent* aParentCo
 already_AddRefed<nsIContent>
 TextTrackCue::ConvertInternalNodeToContent(const webvtt_node* aWebVTTNode)
 {
-  nsIAtom* atom;
+  nsIAtom* atom nsGkAtoms::span;
 
   nsCOMPtr<nsIContent> cueTextContent;

Comment 12

[Ralph Giles (:rillian)]

https://tbpl.mozilla.org/?tree=Try&rev=2e5df2288f08

Comment 13

[Ralph Giles (:rillian)]

Attachment: remove webvtt parser v4

my paper bag is brown.

Comment 14

[Ralph Giles (:rillian)]

https://tbpl.mozilla.org/?tree=Try&rev=dfef9f35e57a

Comment 15

[Ralph Giles (:rillian)]

Attachment: remove webvtt parser v5

Fixed another clang warning:

--- a/content/media/TextTrackCue.cpp
+++ b/content/media/TextTrackCue.cpp
@@ -214,7 +214,6 @@ TextTrackCue::ConvertLeafNodeToContent(const webvtt_node* aW
   if(!document) {
     return nullptr;
   }
-  nsNodeInfoManager* nimgr = document->NodeInfoManager();
   return cueTextContent.forget();
 }

This one builds on the try macs: https://tbpl.mozilla.org/?tree=Try&rev=00bfec82599a

Carrying forward earlier r+.

Comment 16

[Ralph Giles (:rillian)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/0219548b1ef2

Comment 17

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/0219548b1ef2

Comment 18

[Gervase Markham [:gerv]]

For the avoidance of doubt: the change made here was for technical reasons, as part of replacing the C implementation provided by Seneca College with one written in Javascript provided by Mozilla. See bug 891661 for the details of the replacement code.

Gerv