It's not clear to me that the licensing question raised during the review of pdf.js was fully answered so we should make sure it does before releasing this. From bug 714712 comment 81: ::: browser/app/profile/extensions/uriloader@pdf.js/LICENSE @@ +28,5 @@ > + FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL > + THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER > + LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING > + FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER > + DEALINGS IN THE SOFTWARE. I'm no licensing genius but a single license file and no license headers in the source code is not the norm for us. Has somebody that knows better than me (like gerv) confirmed that this is ok?

My response from further down that bug was: The current Mozilla license policy: http://www.mozilla.org/MPL/license-policy.html states that code that originated with Mozilla that becomes part of the core products needs to be MPL 2.0. I plan to have a discussion about whether the Apache License 2.0, as a non-copyleft license, should be added to that list, but I have not yet managed to get sufficient bandwidth from Mitchell to start it. Several of the Mozilla legal team feel that, for a number of reasons, we should not be shipping code using licenses without patent protection clauses, such as MIT or BSD. (This particularly applies to B2G and Gaia, and there is a bug open on sorting that out too.) So, as I see it, we have the following options: 1) Switch to MPL 2.0 2) If a license with no copyleft at all is desirable for some reason, either: 2a) get the Apache discussion started and concluded and, if the result is to allow Apache, switch to Apache; or 2b) Try and convince the legal team that MIT is OK in this case. 2a) is not looking likely in the next few weeks; even getting started requires me getting Mitchell and Harvey in the same meeting, and that's difficult. Gerv

At what point does this need to be resolved? Before release? Before beta? before aurora? Feels to me like as long as we figure this out before release we're probably ok, giving us a little over 12 weeks. But correct me if I'm wrong.

(In reply to Dave Townsend (:Mossop) from comment #2) > At what point does this need to be resolved? Before release? Before beta? > before aurora? The longer it takes to resolve, the more branches we have to update, and the more of a pain it is. I'd have to consult on whether we need contributor permissions for the change; if we do, the more non-employee contributors we have, the more of a pain it is. Gerv

(In reply to Gervase Markham [:gerv] from comment #3) > (In reply to Dave Townsend (:Mossop) from comment #2) > > At what point does this need to be resolved? Before release? Before beta? > > before aurora? > > The longer it takes to resolve, the more branches we have to update, and the > more of a pain it is. > > I'd have to consult on whether we need contributor permissions for the > change; if we do, the more non-employee contributors we have, the more of a > pain it is. > > Gerv Any updates here? PDF.js is on Aurora/Beta now, and we plan to put it on the release channel for FF16.

pdf.js should be switched to either Apache 2 or MPL 2. MPL 2 unless the dev team have a notable preference for Apache 2. mossop: how do we make that happen? Gerv

(In reply to Gervase Markham [:gerv] from comment #5) > pdf.js should be switched to either Apache 2 or MPL 2. MPL 2 unless the dev > team have a notable preference for Apache 2. mossop: how do we make that > happen? Brendan?

The project was created under the MIT license on github and still lives on github. Is it normal to relicense projects imported in moz central? Andreas Gal and Chris Jones were the creators so would it be best to ask their initial intentions?

Where it lives isn't the issue. The question is whether it was significantly started on Mozilla time or on someone's personal time. If it was started on Mozilla time, it needs to follow our licensing policies. If it was started on someone's personal time, then of course they get to choose the licence. Although even in that case, I would strongly recommend licences with patent clauses, such as Apache 2, over ones without, such as MIT (both are non-copyleft). Gerv

(In reply to Gervase Markham [:gerv] from comment #8) > Where it lives isn't the issue. The question is whether it was significantly > started on Mozilla time or on someone's personal time. If it was started on > Mozilla time, it needs to follow our licensing policies. If it was started > on someone's personal time, then of course they get to choose the licence. > Although even in that case, I would strongly recommend licences with patent > clauses, such as Apache 2, over ones without, such as MIT (both are > non-copyleft). > > Gerv Andreas and Chris could you respond to the above?

...so it seems like we need to hear from Andreas/Chris, both on the history of the code and, if it was started on their own time, whether they'd be willing to go with Apache anyway. :-) Gerv

It was started on personal time. I thought we'd relicensed to apache already ... if not, I'm 100% fine with doing so.

I've emailed Andreas asking him to make a statement in this bug regarding the licensing and am assigning this to Brendan for now so that we have someone on owning getting this patch made & landed in time for 16.

I spoke with Andreas and he also is fine with the Apache license. We'll update on github and then the update will then make it into mozilla central before FF17 is uplifted.

Brendan: that's great. Let me know if you need anything from me. Please do resolve this bug once the m-c copy is updated. Gerv

Fixed by bug 789615