Re: Firefox 2.0.0.1 trashes Java Console


"John Ersatznom" <j.ersatz@xxxxxxxxxxxxxxx> wrote in message
news:emeec0$bj$1@xxxxxxxxxxx
Oliver Wong wrote:
Actually, I think what you experienced was the intended behaviour.
You mentioned elsewhere in the thread that you were using "1.5.7".
"2.0.0.1" is not considered to be an automatic upgrade for "1.5.7", but
"1.5.9" *is*. It's like how Apache 2.0 and Apache 1.3 are separate
products, and both of them are still actively developed.

Let me get this straight. You're claiming that the intended behavior is to
nag the user about every single minor bugfix or tweak update, but not to
bother informing them about a major upgrade's availability?
[...]

No. The intended behaviour is to notify the user of updates available to
software that is installed on his/her computer, and not to notify them of
updates to software which isn't installed on his/her computer. Given that
the FireFox 2.0 and Firefox 1.5 series are considered different products by
Mozilla, a FireFox 1.5 user should not be notified about a FireFox 2.0
update, just as much as a Opera user shouldn't be notified about a FireFox
2.0 update.


That would be equivalent to Windows XP Update constantly nagging you about
bugfixes (which it does on a weekly basis) while Microsoft quietly stocked
store shelves with shrinkwrapped boxes of Windows Vista without any
fanfare and simply hoped people would accidentally stumble onto them while
shopping some day.

Mind you, having Windows Update nag you to upgrade to Vista would be
annoying, because Vista costs money and is as much a downgrade
(performance, DRM) as an upgrade,

Actually, yeah, your comparison is apt. Some people feel that 1.5 is
better than 2.X. Rather than piss those people off, Mozilla felt that 2.0
should be considered a completely seperate product, and 1.5 will continue to
be actively developed in parallel with 2.0

but Firefox 1.x to 2.0 is a straight upgrade, and free, rather than a
lateral move that costs money.

People disagree with this assertion.


It should be "first version NOT supported", like String's
"substring(begin, end+1)" interval idiom, so Sun would have said "need a
new version for 2.1" rather than "OK up to 2.0" and everything would be
fine. :)

Except that Sun (and other plugin developers) probably can't predict
the future. How will they know whether or not a new version of the plugin
will be needed for 2.1, unless they have a copy of Firefox 2.1 to test
with? Contrast this with the fact that they DO have a copy of Firefox
2.0, so they can state "It works OK with Firefox 2.0. Anything higher,
and we're not making any promises."

The claim was that Sun should have said it would work with 2.0.0.* or
2.0.* or similarly. The latter definitely amounts to "it may not work with
2.1" and the former to "it may not work with 2.0.1". It's a matter of
whether you specify the end point or one past the end point, and in this
case it looks like requiring specifying the end point was the more
confusing choice.

Right. I guess I was focusing more on the "need a new version" instead
of "*maybe* need a new version".


More generally, plugin developers would have the problem of guessing
exactly what the first future version would be where it no longer worked
(or the last where it did). Maybe Mozilla should take a page from Sun's
book here, and just publish an interface that plugins will always work if
they adhere to, and keep a separate version (like serialVersionUID) for
just the interface, which changes only when an incompatible change to the
interface is made that won't work with old plugins.

Better still, they could have done what most software has been doing since
the 80s or earlier, and been backward compatible with older versions. Any
plugin that worked in Firefox x would then work in Firefox x+1, but
changes to the interface, when such were made, would mean some plugins
designed for x+1 might not work in x.

Well, dem's the apples...

- Oliver


.

Relevant Pages

  • Re: Which Anti Virus to use?
    ... I have Firefox set as default. ... The plugin thing is quite easy and powerful, ... >>goes up in one of the Lightnings at Thunder City! ... Saw a lot of old stuff flying over the house after Fairford ...
    (microsoft.public.windowsxp.general)
  • flash9 checklist (was: flash-9, 10 on FreeBSD)
    ... Preliminary checklist for getting flash9 to work in native firefox: ... for RELENG_6 too (merging the former commit), ... If the plugin doesnt show up in firefox' about:plugins, ... it so the appropriate dependencies can be added to the relevant ports. ...
    (freebsd-questions)
  • [SLE] Segmentation Faults - Solved
    ... in Firefox, Mozilla & Opera, making them impossible to run. ... started after having tried installation of an update to the Vlc plugin. ... repository' that all SuSE browsers under SuSE would use. ...
    (SuSE)
  • Re: Firefox 2.0.0.1 trashes Java Console
    ... It's like how Apache 2.0 and Apache 1.3 are separate products, and both of them are still actively developed. ... Except that Sun (and other plugin developers) probably can't predict the future. ... How will they know whether or not a new version of the plugin will be needed for 2.1, unless they have a copy of Firefox 2.1 to test with? ... Maybe Mozilla should take a page from Sun's book here, and just publish an interface that plugins will always work if they adhere to, and keep a separate version for just the interface, which changes only when an incompatible change to the interface is made that won't work with old plugins. ...
    (comp.lang.java.programmer)
  • Re: Issues with Windows Media Player Plugin for Firefox in Firefox 3
    ... Media Player Plugin for Firefox. ... the event API does not appear to work ...
    (microsoft.public.windowsmedia.player)