[ge-talk] [haiku-development] Re: Notification Server?

[Jonas Sundström]

"Ryan Leavengood" <leavengood at gmail.com> wrote:
 ... 
> I suppose I should have known better, but frankly I had no
> idea what this would blow up into. It sure seemed like a 
> simple idea. 

Part classic "bikeshed color" discussion.

> But certainly not as important as our few other deviations
> from R5, like the layout library.

Lots of people have suggested simple ideas and been turned down. I 
would guess that all the current deviations from the "Just R5" doctrine 
have been done by core developers (for lack of a better term) and none 
by anyone not having commit rights. *grumble* I suppose it makes me 
seem a little ridiculous, but sometimes I do think it's unfair what 
gets in and what doesn't.

Discussion often fails, but it seems like the project rewards those who 
simply commit. Of course, it may be that some discussion takes place 
that I have missed, and that people don't just commit without having 
asked their peers if its a good idea.
 
> But I really think this shows how important it will be to get
> organized after R1 is out. There is a huge chance for things to 
> really
> go awry once the simple "make it like R5" mantra is no longer
> applicable. Even though it seems like the Glass Elevator stuff has
> played out, this goes to show how even simple things can become
> complicated.

Yeah, post-R1 will be interesting. Like the Chinese sign (or was it 
Japanese?) for disaster -- part threat, part opportunity. IIRC.

> > 2) Having said that, what Chris said blends in nicely to the 
> > "design"
> > (ideas, really) that I was talking about earlier: BNotification 
> > wraps
> > BMessage and BMessanger to make a "Kit" (really this should go in 
> > the
> > Interface Kit, at some point). Send() sends the message to some 
> > predefined
> > mime type. By default, there is a Snarl-esque little popup view 
> > thing.
> > Maybe an icon in the deskbar and allows you to turn them on or off 
> > or go to
> > preferences.
> 
> What is funny after all this discussion and "controversy" what Chris
> and you just described is exactly what I originally envisioned when I
> started this thread. Though I agree the conversation was still 
> useful.

Just because you, Michael, Chris and just about everyone else
agreed (and me too) on -how- it could easily be implemented 
this does not necessarily mean that it's a good idea to do actually do 
it. 

We could just as well agree on how to easily exploit the security
shortcomings of Haiku, or how to make a fun version of Space
Invaders using the files on one's desktop.. ;P

That something seems fun, interesting or easy is not always the
cue to start doing it. I'm sure someone was intrigued by the idea
of land mines or atomic bombs.

Not that I think popups are anywhere near as harmful. :]

/Jonas Sundström.