Status of click-to-play plugins in Firefox 24/26

Lawrence Mandel lmandel at mozilla.com
Fri Nov 1 20:50:57 UTC 2013


----- Original Message -----
> We've been testing a myriad of sites (and plugins) ever since the first
> implementation of Click-to-Play landed on mozilla-central (I forget which Fx
> version that was). I can't give you a concrete list right now of all the
> sites we've tested going back to when we started testing CTP but here is a
> list of what was tested most recently (in Fx 26b1):
> https://etherpad.mozilla.org/firefox26b1-plugins

Thanks Anthony. I like that this is well documented. What I am really after is have we identified the different patterns of Java usage (with code examples) on the Web and do we know how Firefox behaves once click-to-play is enabled in each of these cases? For ex, do GMail and Yahoo Mail instantiate and interact with Java in the same way? How do we respond on both sites?

Lawrence

> 
> Anthony Hughes
> QA Engineer, Desktop Firefox
> Mozilla Corporation
> 
> 
> ----- Original Message -----
> > From: "Lawrence Mandel" <lmandel at mozilla.com>
> > To: "Benjamin Smedberg" <benjamin at smedbergs.us>
> > Cc: "release-drivers" <release-drivers at mozilla.org>, "Firefox Dev"
> > <firefox-dev at mozilla.org>
> > Sent: Thursday, October 31, 2013 8:13:27 AM
> > Subject: Re: Status of click-to-play plugins in Firefox 24/26
> > 
> > ----- Original Message -----
> > > On 10/30/2013 11:52 AM, Lawrence Mandel wrote:
> > > > From the short list of sites that you have gathered have you been able
> > > > to extract patterns of Java usage (i.e. how Java is programmatically
> > > > used within a page)?
> > > There were two bugs in the Firefox UI that prevented sites from working,
> > > which may be common patterns across more sites. But the sample of real
> > > sites I have is too small to actually draw many conclusions.
> > 
> > How much testing have we done of real sites? If you need to create a list
> > of
> > sites to test, jjensen has used simple scripts in the past to sift through
> > large amounts of Web content in order to pull out useful information, like
> > the use of specific CSS properties. His scripts may be able to be
> > re-purposed for identifying sites that make use of Java content.
> > 
> > Lawrence
> > 
> > > 
> > > The patches in my queue make all of the known sites work except for
> > > http://www.lidl-druckservice.de/flyer-din-a5.html which doesn't work in
> > > Chrome and appears to just be broken.
> > > 
> > > > If so, are the usage patterns and our solution for them documented
> > > > somewhere? Lawrence
> > > https://developer.mozilla.org/en-US/docs/Site_Author_Guide_for_Click-To-Activate_Plugins
> > > is still the recommended solutions for site authors. Nothing much has
> > > really changed there:
> > > 
> > > * recommend using callbacks for plugin activation
> > > * recommend makin plugins visible at least until they are activated
> > > > Of course we need to keep in mind that users just want these sites to
> > > > work
> > > > so that they can do whatever their task happens to be. While we want to
> > > > secure our users they do have an option if a site doesn't work in
> > > > Firefox
> > > > - try another browser.
> > > We want to absolutely be no worse than Chrome on any site that has
> > > significant market share. However, we also made the decision that
> > > letting a few people walk was probably the correct decision in this
> > > case, because the improved stability and performance for everyone else
> > > is noticeable.
> > > > In terms of uplifting, is there a good reason to keep this change in
> > > > 26?
> > > > If
> > > > UI and string changes are required, why not defer to a later release?
> > > At this point, it does not look like any string changes will be required.
> > > 
> > > The hard part about this is that we aren't going to get any meaningful
> > > testing of CtP except on the beta channel. So in order for this not to
> > > sit in several long 6-week cycles, we need to iterate on it rapidly on
> > > the beta channel. At this point I'm not committed to actually shipping
> > > this in 26, but I do think we should try to make the necessary
> > > improvements and uplift them as quickly as safety allows, in order to
> > > validate our opinion on whether this is safe enough to ship.
> > > 
> > > --BDS
> > > 
> > > 
> > _______________________________________________
> > firefox-dev mailing list
> > firefox-dev at mozilla.org
> > https://mail.mozilla.org/listinfo/firefox-dev
> > 
> 



More information about the firefox-dev mailing list