[GRASS-dev] some impressions on the wx interface

[Glynn Clements]

Michael Barton wrote:

> > And as I said before - I am missing the d.* commands  sooo much!
> 
> The old d.* commands will not work in GRASS 7...at all because of the  
> change in the underlying display architecture.

The commands work fine; they just behave differently in terms of where
the resulting image ends up.

> Remember, they dumped  displays to a generic X window.

They generated output via the current monitor, which may or may not
have been an X monitor.

> To get an image into a canvas (TclTk  
> or wxPython) is considerably more complicated, though you can do a lot  
> more with it once it's there.
> 
> The reason for the message is that Jachim Czpicky originally, and  
> Martin Landa subsequently have thought about the possibility of having  
> the command parser recognize a d.* command and display the result in a  
> wxpython canvas.

Trying to recognise d.* commands prior to execution is unwise; e.g. 
this probably won't work if the command is a user script which invokes
d.* commands.

It would make more sense to assume that any command may generate
output, and to allow for this by setting all relevant environment
variables.

> This sort of works in the Mac and Unix from the  
> wxpython command parser, but I don't think that the python scripts to  
> do this from a terminal are functional (Martin can correct me on this  
> if I'm wrong).
> 
> However, it seems to me better to wait until the display architecture  
> is more finalized in GRASS 7 to do this.

I'm unaware of any planned changes to the way that the display library
works. I'm not anticipating any significant changes, although specific
features can be added if necessary.

-- 
Glynn Clements <glynn at gclements.plus.com>