#176 red border GOP

[marius schebella]

hey,
when I have a subpatch that contains no element in it, but is a GOP, then I don't see the red GOP border.
that does not seem to be a problem. BUT...
when I use this empty subpatch and create an object in it via message command, then that red border suddenly is visible in the parent patch.
see attached file.
marius.

[marius schebella]

[stffn]

Logged In: YES
user_id=1658640
Originator: NO

This seam very similar to the report #1351274 called "pd 0.39-1 GOP bounding box doesn't appear if no objects".

[Hans-Christoph Steiner]

Logged In: YES
user_id=27104
Originator: NO

This is the same as bug #1351274

http://sourceforge.net/tracker/index.php?func=browse&group_id=55736&atid=478070&set=custom&_assigned_to=0&_status=1&_category=100&_group=100&order=artifact_id&sort=DESC&offset=100

[Hans-Christoph Steiner]

Logged In: YES
user_id=27104
Originator: NO

This is the same as bug #1351274

[marius schebella]

Logged In: YES
user_id=1856112
Originator: YES

it's the same, but #1351274 does not show up when I browse the bugs. wrong settings? strange.

[Hans-Christoph Steiner]

Logged In: YES
user_id=27104
Originator: NO

Try this link:

http://sourceforge.net/tracker/index.php?func=detail&aid=1351274&group_id=55736&atid=478070

[Jakob Leben]

Logged In: YES
user_id=2064182
Originator: NO

The reason, I think, why red border doesn't show up when no elements present is: The same GOP technique is used to show the contents of a subpatch that contains data structures - for example an audio array. In this case you really don't want the red border to be drawn over the graphical representation of data structures, do you? So, red border gets drawn only when there are elements which are at the same time not data structures. This I wouldn't consider a bug then. What's your opinion?

Now, why it gets drawn when an object is created via message command: the red border gets drawn in the garray_vis function (g_array.c). This function apparently gets called every time an object is created, exactly because that's the point when the red border might need to be drawn. I think the reason why when you have the subpatch open the GOP on the parent patch is covered over with a grey square is just to avoid the problem of having to visualize GOP on the parent and the subpatch window at the same time but differently (for example: on the parent patch without the red border).
But the code obviously doesn't handle the case when a GOP's window is closed and when you create an object it calls garray_vis which draws the red square. Therefore we should add some checking inside garray_vis.

This is my idea. Maybe I'll try to solve the problem or I hope this text will be of help to whoever else.

[Jakob Leben]

Logged In: YES
user_id=2064182
Originator: NO

umm, sorry, that was graph_vis (g_graph.c) instead of garray_vis (g_array.c).

[Jakob Leben]

Logged In: YES
user_id=2064182
Originator: NO

allright, the problem with red rectangle being drawn in a wrong situation was in glist_add (g_graph.c).
I posted a patch: https://sourceforge.net/tracker/index.php?func=detail&aid=1953772&group_id=55736&atid=478072

For reasons I explained below I don't consider a bug the fact that red rectangle is not drawn immediately after a subpatch is created and set to GOP. I'm saying that because the bug no. 1351274 seems to still be open.

[Hans-Christoph Steiner]

Logged In: YES
user_id=27104
Originator: NO

this is a duplicate, please do not post here, instead use this one:

http://sourceforge.net/tracker/index.php?func=detail&aid=1351274&group_id=55736&atid=478070