Re: [Plplot-devel] Problem pushing changes to the repository

2015-01-29 Thread Alan W. Irwin
On 2015-01-29 21:35- Phil Rosenberg wrote: > This is only a guess, but how are you accessing the repository? If you cloned it using the git:// address then I think this might be read only. You need to access with ssh to push changes I think. @Arjen: if you log in to SF and look at

Re: [Plplot-devel] Problem pushing changes to the repository

2015-01-29 Thread Phil Rosenberg
This is only a guess, but how are you accessing the repository? If you cloned it using the git:// address then I think this might be read only. You need to access with ssh to push changes I think. Phil -Original Message- From: "Arjen Markus" Sent: ‎29/‎01/‎2015 20:35 To: "PLplot develop

[Plplot-devel] Problem pushing changes to the repository

2015-01-29 Thread Arjen Markus
Hi everyone, I made some changes to the Fortran bindings to support the new plmap* routines. This works fine - a clean comparison with the C x19 example. However, when I try to push the changes (after diligently following the receipe for our git workflow ;)), I get the message: fatal: Coul

Re: [Plplot-devel] segfault bugfix for plbuf.c

2015-01-29 Thread Alan W. Irwin
On 2015-01-29 08:06- Phil Rosenberg wrote: > As Jim said resizing xwin or the current wxWidgets driver uses the buffer, but that's not great for automated testing. The new wxWidgets driver uses the buffer for all plotting when called from a console application (i.e. for all the examples) inclu

Re: [Plplot-devel] segfault bugfix for plbuf.c

2015-01-29 Thread Jim Dishaw
I agree the plexit call is bad, I put it in primarily for debugging and I've been thinking of alternatives. I'm going to implement your idea. I'm going to add a data type suffix to each command. Arrays will also have a size passed (which is what plbuf currently does). The data type prefix wil

Re: [Plplot-devel] segfault bugfix for plbuf.c

2015-01-29 Thread Phil Rosenberg
Hi Jim In general plexit calls are bad. In fact it is on my to do list to go through and eradicate them. However, I'm not sure what alternative there is as the rendering cannot continue, because once an unknown command is found there is no way to know how many bytes to skipto find the next command

Re: [Plplot-devel] plGetCursor

2015-01-29 Thread Phil Rosenberg
That's great, thanks for all the answers. Andrew I agree it's not the way you'd do it if you were writing a wxWidgets application. But I guess if, like the examples, you are writing a console application, then it gives you a mechanism to interact with the plot. Phil On 29 January 2015 at 02:15,

Re: [Plplot-devel] segfault bugfix for plbuf.c

2015-01-29 Thread Phil Rosenberg
As Jim said resizing xwin or the current wxWidgets driver uses the buffer, but that's not great for automated testing. The new wxWidgets driver uses the buffer for all plotting when called from a console application (i.e. for all the examples) including the initial plot, which makes it a useful