GLMPresentation >> window looks strange

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

GLMPresentation >> window looks strange

Guillermo Polito
Hi!

We were browsing with Steph users of #announce: and we arrived to the
following method:

GLMPresentation >> window
     self announce: (GLMWindowRequest new
         action: [ :aWindow | ^ aWindow ]).
     ^ nil

And we could not stop noticing that, while the announcement, probably
originally meant to decouple the Presentation and the window, is now
coupling them in a non-clear way... I mean, the coupling is there, but
is now hidden in an announcement.

Also, the usage of the non-local return makes the assumption that the
announcement will be executed on the same process... And that is at most
strange, but also looks buggy.

Why is this method not a normal accessor?

Guille

Reply | Threaded
Open this post in threaded view
|

Re: GLMPresentation >> window looks strange

Denis Kudriashov
And this is not only place. I saw it is common approach to get view from  GL presentation. And it makes code analysis so hard.

2016-06-07 10:54 GMT+02:00 Guille Polito <[hidden email]>:
Hi!

We were browsing with Steph users of #announce: and we arrived to the following method:

GLMPresentation >> window
    self announce: (GLMWindowRequest new
        action: [ :aWindow | ^ aWindow ]).
    ^ nil

And we could not stop noticing that, while the announcement, probably originally meant to decouple the Presentation and the window, is now coupling them in a non-clear way... I mean, the coupling is there, but is now hidden in an announcement.

Also, the usage of the non-local return makes the assumption that the announcement will be executed on the same process... And that is at most strange, but also looks buggy.

Why is this method not a normal accessor?

Guille