Content-Type: multipart/alternative; boundary="-748019302-551550357-1382026716=:69198" ---748019302-551550357-1382026716=:69198 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi All=0AI've attached a patch which should give support for using wxGCDC. = It is not finished or watertight, but it does work on my windows machine so= i thought now was a good point to send it round to check for any platform = dependant differences, so if someone could try it on a linux machine or a M= ac that would be good. i think I might have broken the AGG backend with thi= s patch but I'll fix that later if needed.=0AThe only real hitch i came acc= ross was that setting clipping regions are not consistent between wxGCDC an= d other DCs. I think this might have been cleared up in more recent release= s and in fact the clipping method used in Plplot has been depreciated. Beca= use of this some flag is needed to indicate whether a wxDC or a wxGCDC is b= eing used. We therefore have the following overarching changes:=0A=A0=0ASet= ting backend=3D2 now uses the wxDC backend, but uses a wxGCDC.=0AI've added= a backend=3D3 option which is an indicator that the user is going to pass = in a wxGCDC using plcmd(PLESC_DEVINIT, gcdc)=0AFrom the command line settin= g backend=3D3 should give the same results as backend=3D2, but I must confe= ss I haven't tested it.=0AWe now have subpixel accuracy on points, for back= ends 2 and 3=A0which you should be able to see if you draw a horizontal lin= e of 1 pixel width - if it lies on a pixel boundary it will have grey edges= .=0A=A0=0APlease let me know how any tests you do go.=0A=A0=0APhil ---748019302-551550357-1382026716=:69198 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable
Hi All
I've attached a patch which should give= support for using wxGCDC. It is not finished or watertight, but it does wo= rk on my windows machine so i thought now was a good point to send it round= to check for any platform dependant differences, so if someone could try i= t on a linux machine or a Mac that would be good. i think I might have brok= en the AGG backend with this patch but I'll fix that later if needed.
=
The only real hitch i came accross was that setting clipping regions a= re not consistent between wxGCDC and other DCs. I think this might have bee= n cleared up in more recent releases and in fact the clipping method used i= n Plplot has been depreciated. Because of this some flag is needed to indic= ate whether a wxDC or a wxGCDC is being used. We therefore have the following overarching changes:
 
Setting backend= =3D2 now uses the wxDC backend, but uses a wxGCDC.
I've added a b= ackend=3D3 option which is an indicator that the user is going to pass in a= wxGCDC using plcmd(PLESC_DEVINIT, gcdc)
From the command line se= tting backend=3D3 should give the same results as backend=3D2, but I must c= onfess I haven't tested it.
We = now have subpixel accuracy on points, for backends 2 and 3 which you s= hould be able to see if you draw a horizontal line of 1 pixel width - if it= lies on a pixel boundary it will have grey edges.
=  
Please let me know how any tests you do go.<= /font>
 
Phil
---748019302-551550357-1382026716=:69198--