On 23 December 2012 19:33, tim Rowledge <tim at rowledge.org> wrote:
>
> On 23-12-2012, at 11:15 AM, Frank Shearar <frank.shearar at gmail.com> wrote:
>
>> On 22 December 2012 23:36, Bert Freudenberg <bert at freudenbergs.de> wrote:
>>>
>>> We're talking about _where_ the window opens.
>>
>> OK. I suspect the reason I've not noticed anything like that is
>> because I'm running 1024x768 on a Lucid box. The window opens rather
>> close to the top left, pretty much where I'd expect it to. I just
>> opened a TrunkImage.image on a Mac and it looks like it opens it right
>> in the centre of the screen. Is that what you see?
>
>
> Last I recall there was no info in the image header to specify where the main squeak window would open and it was a platform decision at start time. Now, with the Ffenestri stuff in proper use one could revise that by allowing the image code to specify where the first window opens, with appropriate policy objects to discover the screen size and look at OS preferences etc.
>
> Are non-RISC OS vm *still* opening a main window by default? I hope not. It's a stupid thing to do; no window should even created until (and unless!) some displaying to Display (or suitable other object) is done.

As far as I understand it, only the Windows VM has issues running
headless (because of Windows' bonkers split between console
applications and windowed applications). Certainly when testing the
build scripts (thus headless, on a Linux machine), the builds behave
just as you'd expect a command line application to.

Which has its own problems, since there's a great deal of code in the
image that assumes that popping up a Debugger is a sensible thing to
do. I know Camillo Bruni and the Pharo guys have done a great deal of
work on command line issues: there might well be some tasty
easy-to-pluck fruit hanging off our neighbour's tree...

frank

> tim
> --
> tim Rowledge; tim at rowledge.org; http://www.rowledge.org/tim
> Throughout history, every mystery solved has turned out to be...
> *not magic*
>
>