On Saturday, October 8, 2011 at 3:09 PM, Richard Johnson wrote:
> On Sat, Oct 08, 2011 at 12:47:49PM -0700, Peter Hosey wrote:
> > On Oct 8, 2011, at 12:17:58, Richard Johnson wrote:
> > > 1.3 was prematurely released, with major changes that almost certainly 
> > > break the notifications you may rely upon. Use Growl 1.2.2 or earlier 
> > > until fixes are in for the problems with the Growl 1.3 API changes.
> > 
> > The change that broke applications that use old frameworks wasn't an API 
> > change (which would be *in the framework*).
> > 
> > The change that broke old frameworks was the move from a .prefPane to a 
> > .app.
> 
> That was part of the API, in actual practice.
> 
> Your arguing differently does not change the fact that the programming
> interface used by many applications (the API) to find Growl and to send
> notifications for display by Growl was changed. Further, that deliberate
> and ill-thought-out, ill-planned-for change broke apps that users were
> relying upon.
> 
> <sigh>
> 


*sigh* indeed. The problem here is two fold:


1) Applications that used the isInstalled method that is part of the framework, 
but NOT recommended whatsoever in any of our documentation. 
http://growl.info/documentation/developer/implementing-growl.php is our 
documentation for implementing Growl, unchanged from the 1.2.2 documentation 
other than some simple formatting. I have not yet updated it, but it's going to 
change soon since the 1.3 framework DOES have changes which will require some 
updates for those applications which want to use Growl in a Sandbox.

2) Applications which use a really old version of the framework.

You can argue about this until you are blue in the face, and we can argue back. 
But in reality if an application is not working, out of the box, with Growl 1.3 
then they did something which needs to be corrected. It can be as simple as 
just dropping in place an updated framework. Or as complicated as fixing the 
way that the framework is used.

This is all assuming that this is the only thing which is done. Some developers 
went so far as to do a file system check to ensure that Growl.prefpane was on 
the file system. While I can see why they did it, they just didn't have to. 
Bottom line is that was never a recommendation.

OS X does not define an API as one thing or another. In fact our switching from 
GrowlHelperApp.app to Growl.app is quite supported by OS X. The problem becomes 
that other applications did understandable but improper things. 

Another thing to note is that if they were not to stop working now, they would 
have as soon as they started implementing Sandboxing support. So it's either we 
find all of the culprits now, provide a way forward and this whole problem goes 
away in a few months, or Growl becomes useless altogether by the end of the 
year for about half of the population if we had not planned ahead.
> > No APIs were removed in any hitherto released version of the framework.
> 
> You broke Growl for the majority of users with your changes. This was
> deliberate on the part of the dev team, though perhaps without a full
> understanding of the consequences.
> 

We understood the consequences entirely.
 
> You need to own that and provide clear communication about how you're
> fixing it.
> 

Well, here we go. I agree that for 1.3 any application which implements Growl 
support in a fashion which breaks it is going to be broken until that is 
resolved on their end. Our plan for that getting it fixed is that application 
developers will need to fix their end. Bottom line is, Growl 1.3 works well for 
some people, and for others it doesn't, but only temporarily until these issues 
are addressed.

We do have a workaround for the isInstalled applications almost done getting 
tested. The 1.3 framework responds to isInstalled with a YES response, instead 
of a NO, every time. So the workaround to drop the framework into place will 
work for those applications afterwards. We also want to update the Growl 
Version Detective to address making dropping the framework in easier, i.e. 
doing it for the end user. However, that's going to take some time to do that. 
We have other more pressing issues, like the menu item that everyone wants 
disabled, fixing the source code repo, and fixing the crasher for the queue'ing 
displays. Also making sure that we're responding to end user emails within a 
timely response period is important, so keep in mind that we're doing that as 
well.
 
> The continual insane nitpicking denial we're getting instead doesn't play
> well, especially when it's now a for-pay application.
> 


Keep in mind that Peter is trying to do his best here, but he was not working 
on Growl 1.3. He said as much. He's answering your technical questions though 
with the knowledge he has, but he doesn't have say in the direction of The 
Growl Project, so he's limited in what he can say. I think you're being a tad 
bit harsh on him here, since he is answering your questions.

Nobody is denying anything here. We knew Growl 1.3 was going to cause some 
applications to have problems, and when we ran into those applications we let 
the developers know about it. I had a ticket open with Sparrow about the issue 
well before the 1.3 release came out for instance. Dropbox's forum build with 
the fix is because we made sure that they were aware there was a problem after 
someone on our beta list reported it to us. However, it is unreasonable to 
expect that we can test every single application out there, or that we are able 
to enforce our own release schedule upon other applications.
> Am I going to have to change my advice to "Find an alternative to using
> Growl, as the developer community is off the rails, and Growl thus will be
> an utter train wreck for the foreseeable future."
> 

I think once the 1.3 framework is out, most of the issues you are seeing will 
be mitigated here. I think you're being unfair about this since it's only been 
one week since we made a change after 8 years of not doing so. Also keep in 
mind that if we hadn't done this, the project likely would have stopped 
existing entirely.

You can still use 1.2.2 currently if the applications you are using will not 
work with 1.3, but please contact the developers of those applications to let 
them know that they need to update. I'll have documentation up next week, but 
I've been unable to this week due to responding to people emailing us and using 
this list.

Chris
 
> Best of luck!
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Growl Discuss" group.
> To post to this group, send email to [email protected] 
> (mailto:[email protected]).
> To unsubscribe from this group, send email to 
> [email protected] 
> (mailto:[email protected]).
> For more options, visit this group at 
> http://groups.google.com/group/growldiscuss?hl=en.

-- 
You received this message because you are subscribed to the Google Groups 
"Growl Discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/growldiscuss?hl=en.

Reply via email to