On 2015-02-27 22:12- Phil Rosenberg wrote:
Hi Alan
Yes I believe this is doable in wxWidgets. I'm sure I remember
coding it up when I added newline support so I think the code is
already in the driver if you look. I guess I never realised this
needed plumbing into the core code.
Should
On 2015-02-27 14:21-0500 Jim Dishaw wrote:
> I have identified an issue while testing on a relatively clean
machine. If the user is using XQuartz (which is likely) and has some
of X11 installed via MacPorts (possible because xorg-libX11 might be a
dependency) then plplot won't configure correctly
Hi Alan
Yes I believe this is doable in wxWidgets. I'm sure I remember coding it up
when I added newline support so I think the code is already in the driver if
you look. I guess I never realised this needed plumbing into the core code.
Should be a very easy fix if you know how to feed the sizes
Hi Phil:
With regard to the x26.02 issue with a much-too-wide legend box. That
issue does not happen for any device driver (i.e., cairo and qt) that
provide get_string_length capability as indicated by the
has_string_length flag. Here are the complete current locations in
the code that have some
I have identified an issue while testing on a relatively clean machine. If the
user is using XQuartz (which is likely) and has some of X11 installed via
MacPorts (possible because xorg-libX11 might be a dependency) then plplot won't
configure correctly and the build fails.
I have sorted throug
Phil,
I don't think this would necessarily need propagating to all bindings for
the release. Presumably the default would be to use wxPLViewer which
would mean you could use the wxwidgets driver stand-alone from any
language? This is probably most usage. If you want to embed it then
you would n
That would be my preference too in an ideal world. However that change
would then need propagation to all the other bindings, which is
unlikely to be feasible on the timescales of this release cycle.
On 27 February 2015 at 14:06, Jim Dishaw wrote:
> I vote for option 1 because it mirror the philo
I vote for option 1 because it mirror the philosophy of plsfnam() and I think
it is useful for the windowed drivers. When I wrote a new Windows driver, I
thought something like this would be handy if someone wanted to embed a plot in
a window that was part of an application.
I think the impact
Okay, I am really struggling with this. Jim is pretty much correct.
I had thought I would be able to use pl_cmd() before calling plinit to
store the data I needed. However this causes a segfault because the
dispatch table is not initialised.
I know this is really late in the day for this stuff, b
Sorry
Ignore my last email - the comment didn't mean what I thought it did -
it was reference to NULLing the pointer to avoid it getting freed for
a second time.
On 27 February 2015 at 10:08, Phil Rosenberg wrote:
> Hi
> Does anyone know where PLStream::dev is freed?
>
> In the wxWidgets driver p
Hi
Does anyone know where PLStream::dev is freed?
In the wxWidgets driver plD_tidy_wxwidgets function there is a comment
saying it is freed in plcore.c, but I can't find it. Other drivers
also seem to not bother tidying up their memory allocations for this
pointer.
The reason I ask is that if thi
Hi Jim
There is already a way to do it via pl_cmd. It's just a question of when it
needs to be done. I will look to see if it is possible to do before plinit and
store the wxDC buffer until plinit is called.
Phil
-Original Message-
From: "Jim Dishaw"
Sent: 27/02/2015 01:27
To: "Phil
The current status is that the comprehensive test script is failing on
Linux for the Qt5 case. So it is not yet worth anyone's time to do the
epa_build of PLplot (which epa_builds Qt5 (and much else) and plplot
and does the Qt5 variant of comprehensive testing) or Qt5 variant of
comprehensive testi
13 matches
Mail list logo
