Quicktime and innerHTML
in [JavaScript]
|
Prev: Yellowfin 5.0 release
Next: Buy Ambien with cod. Cheap Ambien. Ambien cheap. Cheap Ambien fedEx.
From: VK on 25 Apr 2010 19:36 On Apr 26, 3:01 am, Stefan Weiss <krewech...(a)gmail.com> wrote: > Conspiracy theories aside, you have a very unusual definition of the > word "bug". As long as Microsoft is the responsible party, you prefer to > use euphemisms like "differences" or "intentional functionality > break[s]". I see a great career in politics or marketing :) Do you consider "intentional functionality break" to be softer form of "bug"? Uhmm... :-) If I wanted it to do A and it appeared to do B - it is a bug. If I thought it should do A and so programmed it but by proper specs reading it supposed to do B - it is a bug as well. If I knew that it must do B but by my own considerations programmed it to do A instead - it is not a bug, it is an "intentional functionality break". Feel the difference. Say Mozilla Bugzilla has many confirmed bugs. At the same time Flash plugin will have scriptable interface blocked if you use the automatic installation on prompt. In order to have the scriptable interface enabled, you have to download plugin installation file first and then run it manually (<1% of average users). It is not a bug, it is an "intentional functionality break" because ActionScript security model differs from Gecko JavaScript security model and they don't like it.
From: VK on 25 Apr 2010 21:12 > On Apr 26, 3:01 am, Stefan Weiss <krewech...(a)gmail.com> wrote: > > Conspiracy theories aside, the trick is in not using <object>, use <embed> instead for IE as well and it works (though the .src change is still blocked on IE). The working demo is at http://javascript.myplus.org/qt/qt.html All files listed for download at http://javascript.myplus.org/qt/
From: Andrew Poulos on 26 Apr 2010 02:47 On 26/04/2010 11:12 AM, VK wrote: >> On Apr 26, 3:01 am, Stefan Weiss<krewech...(a)gmail.com> wrote: >> >> Conspiracy theories aside, > > the trick is in not using<object>, use<embed> instead for IE as well > and it works (though the .src change is still blocked on IE). The > working demo is at > http://javascript.myplus.org/qt/qt.html > All files listed for download at > http://javascript.myplus.org/qt/ Hmm, I'm surprised it works. Does it work for all flavours of IE? Is there a browser that only plays QT with an OBJECT element? Andrew Poulos
From: VK on 26 Apr 2010 06:32 On Apr 26, 10:47 am, Andrew Poulos <ap_p...(a)hotmail.com> wrote: > On 26/04/2010 11:12 AM, VK wrote: > > >> On Apr 26, 3:01 am, Stefan Weiss<krewech...(a)gmail.com> wrote: > > >> Conspiracy theories aside, > > > the trick is in not using<object>, use<embed> instead for IE as well > > and it works (though the .src change is still blocked on IE). The > > working demo is at > > http://javascript.myplus.org/qt/qt.html > > All files listed for download at > > http://javascript.myplus.org/qt/ > > Hmm, I'm surprised it works. > > Does it work for all flavours of IE? > Is there a browser that only plays QT with an OBJECT element? IE 4.x-5.x are not in the test list for a long time so cannot speak for them. The rock bottom MS fallback testing now is IE6 / Win XP SP2. In this environment it runs just fine with the same results as IE8/ Vista SP2 Taking into account the history of standardization of NPAPI ( http://en.wikipedia.org/wiki/NPAPI ) I would speculate that <object><embed> combo was ever needed for IE 3.x - IE 4.x versions only. Since IE 5.x so since 1999 this combo usage is a useless and mostly harmful cargo cult ( http://en.wikipedia.org/wiki/Cargo_cult_programming ) Please note that while NPAPI is standardized, older IE DOM access to it is not with documents.embeds being missing or broken. So <embed> is used for either IE or al. but both ID and NAME has to be set to the same value and then the access goes either al.: document.embeds['PluginName'] or IE: document.getElementById('PluginName')
From: VK on 26 Apr 2010 12:43
Updated and extended test suit is at http://javascript.myplus.org/qt/qt2.zip Just for fun I checked Apple's claim that QuickTime supports all current and legacy interfaces, including LiveConnect, so I adapted the test case for NN4 as well and tried it on Netscape Navigator 4.79 (a.k.a. "Netscape Communicator") / Win XP SP2 Indeed it blinked "Java loading..." in status bar and then played everything nicely and smoothly, two or three movies at once on the same page. [ NN4 has its own small foot-print optimized JVM included into installation, so to repeat the test one doesn't need to install Java. Just get the NN 4.79 itself at http://browser.netscape.com/releases ] Firefox 3.6.3 / Win Vista SP2 Cannot play more than one movie at once, others are on hold. Also acts weird with two movies both inserted over document.write: The last written is the first played, the first written doesn't autoplay. One needs to mouse hover it or even keep mouse pressed to make it play - the working variant seems random. That shows that Mozilla products' media capabilities have been noticeable deteriorated over the past 10 years - which is not an anyhow huge secret though. Internet Explorer 6 / Win XP SP2 No complains, nicely and smoothly - one, two or three movies at once, either over document.write or innerHTML Internet Explorer 8 / Win Vista SP2 ditto Safari 4.0.5 / Win Vista SP2 ditto Chrome 4.1.249 / Win Vista SP2 ditto Opera 10.51 / Win Vista SP2 ditto Conclusions: 1) not so great for Gecko. It is realized by the team as well, so it is announced to make a new high productivity engine by the end of this year with 2D graphics handling using hardware resources directly (some German name for the engine, has a mental block right now) - that caused a panic in QA for such short terms but in fact the time limits are very narrow and this year is exactly all what Mozilla has to steady stay on the mass market. 2) yes, object-embed combo is a useless and often harmful cargo cult left from the Browser Wars times. Use <embed> only with id/name. |