RE: Help with saveas and fltk

[Rick M. Cox]

-----Original Message-----
From: Ben Abbott [mailto:address@hidden 
Sent: Friday, June 14, 2013 9:13 PM
To: Rick M. Cox
Cc: 'Nicholas Jankowski'; 'help-octave Octave'; 'Tatsuro MATSUOKA'
Subject: Re: Help with saveas and fltk

On Jun 15, 2013, at 9:52 AM, Rick M. Cox wrote:

>  From: address@hidden [mailto:address@hidden On Behalf Of
Nicholas Jankowski
> Sent: Friday, June 14, 2013 6:48 PM
> To: Ben Abbott
> Cc: Rick M. Cox; help-octave Octave; Tatsuro MATSUOKA
> Subject: Re: Help with saveas and fltk
>  
> i have an xp machine i could throw 3.64 on... will try it when i get a
chance.
>  
> On Fri, Jun 14, 2013 at 9:32 PM, Ben Abbott <address@hidden> wrote:
>> On Jun 15, 2013, at 3:44 AM, Rick M. Cox wrote:
>> 
>> >  From: Benjamin Abbott [mailto:address@hidden
>> > Sent: Friday, June 14, 2013 3:36 AM
>> > To: Nicholas Jankowski
>> > Cc: Rick M. Cox; help-octave Octave
>> > Subject: Re: Help with saveas and fltk
>> >
>> >> On Jun 14, 2013, at 6:25 PM, Nicholas Jankowski <address@hidden>
wrote:
>> >>
>> >>> On Thu, Jun 13, 2013 at 10:55 PM, Ben Abbott <address@hidden>
wrote:
>> >>>
>> >>>> On Jun 13, 2013, at 10:25 PM, "Rick M. Cox" <address@hidden>
wrote:
>> >>>>
>> >>>>  From: address@hidden [mailto:address@hidden On Behalf Of
Nicholas Jankowski
>> >>>> Sent: Thursday, June 13, 2013 6:57 PM
>> >>>> To: Rick M. Cox
>> >>>> Cc: Ben Abbott; help-octave Octave
>> >>>> Subject: Re: Help with saveas and fltk
>> >>>>
>> >>>>> On Thu, Jun 13, 2013 at 9:44 PM, Rick M. Cox <address@hidden>
wrote:
>> >>>>>
>> >>>>>> From: Ben Abbott [mailto:address@hidden
>> >>>>>> Sent: Thursday, June 13, 2013 5:58 PM
>> >>>>>> To: Rick M. Cox
>> >>>>>> Cc: 'Nicholas Jankowski'; 'help-octave Octave'
>> >>>>>>
>> >>>>>> Subject: Re: Help with saveas and fltk
>> >>>>>>
>> >>>>>>> On Jun 13, 2013, at 03:12 PM, "Rick M. Cox"
<address@hidden> wrote:
>> >>>>>>>
>> >>>>>>>> Nick and Ben:
>> >>>>>>>>
>> >>>>>>>> I hope some of this helps:
>> >>>>>>>>
>> >>>>>>>> 1.      I haven't seen the "D:\Program" error in a long time.
>> >>>>>>>>
>> >>>>>>>> 2.      I installed Octave 3.7.2+ and Ghostscript v9.07 in paths
with no white spaces.  Even though I don't use it anymore, I never deleted
Octave 3.6.4.  It still resides in a path with white spaces.  I certainly
could delete it if you thought it was a problem.
>> >>>>>>>>
>> >>>>>>>> 3.      I thought I correctly added Ghostscript to the system
path/environment variables per the setup instructions.
>> >>>>>>>>
>> >>>>>>>> 4.      When I type system ("gswin32c.exe") I get a Ghostscript
9.07 header.
>> >>>>>>>>
>> >>>>>>>>
>> >>>>>>>> I'm cheered that my routines work on your systems.  I'd like it
to work on mine.  J
>> >>>>>>>>
>> >>>>>>>> Thanks again,
>> >>>>>>>>
>> >>>>>>>> Rick
>> >>>>>>>
>> >>>>>>> I'm not sure it will help resolve the problem, but please type
EXEC_PATH at Octave's prompt and copy-n-paste the text it returns into you
reply.
>> >>>>>>>
>> >>>>>>> Also tell us if you've installed more than one version of
Ghostscript and where they are.
>> >>>>>>>
>> >>>>>>>  Ben
>> >>>>>>
>> >>>>>>
>> >>>>>> Ben:
>> >>>>>>
>> >>>>>> I am sorry but I don't know how to copy something from the Octave
3.7.2+ Command Window.  I can see the results of the EXEC_PATH command but I
can't copy them.  I have looked for help on the web but haven't found the
right thing yet.
>> >>>>>>
>> >>>>>> As to your second question there is only one copy of Ghostscript
installed on my machine.  When I tried to install it to a path with no white
spaces, I had to uninstall the previous version first.  The path to
Ghostscript on my machine is:
>> >>>>>>
>> >>>>>> D:\Ghostscript\gs9.07
>> >>>>>>
>> >>>>>> I added this path to the Environment Variables on my system.
>> >>>>>>
>> >>>>>> Rick
>> >>>>>
>> >>>>>
>> >>>>> Rick, also check to see if there's one in the Octave folder. The
windows builds might have them included assuming most windows users wouldn't
know what ghostscript is. It would be located here:
>> >>>>>
>> >>>>> for my copy of version 3.6.4, its located at:
>> >>>>>
>> >>>>> C:\Octave\Octave3.6.4_gcc4.6.2\gs
>> >>>>>
>> >>>>> with the main executable as
C:\Octave\Octave3.6.4_gcc4.6.2\gs\bin\gswin32.exe
>> >>>>
>> >>>> Nick,
>> >>>>
>> >>>> Since Rick is running the mingw version of Octave, is there a
problem using gswin32.exe instead of the mingw gs.exe?
>> >>>
>> >>> Ben
>> >>>
>> >>> I believe on mine the executable included in octave is gswin32.exe,
there is no gs.exe
>> >>>
>> >>> and threadhopping to your other question, the MinGW version includes
a shortcut that starts Octave in what I believe is the msys shell.  It
doesn't add things to the system path such that runnign commands from the
dos shell will automatically do anything.
>> >>
>> >> On win7, I have an msys octave-3.6.2 that is able to print using
gswin32.exe.  Which version of windows are you running?
>> >>
>> >> Ben
>> >
>> > Ben:
>> >
>> > I am running Windows XP Pro SP3.  I use Octave 3.7.2+ with the GUI.  I
had the same problem though with Octave 3.6.4.
>> >
>> > Rick
>> 
>> Except that the problem appears to be related to WinXP, I have no other
insight at the moment.  Is anyone else following the discussion running XP?
>> 
>> Ben
> 
> Nick and Ben:
>  
> I appreciate Nick's offer and look forward to what he finds out.
>  
> To add to the mystery every other plot that I have tried so far has saved
just fine.  There seems to be something about these semicircles and arcs.
>  
> Rick

Perhaps you can experiment with removing different curves to see if you can
isolate the problem.

(1) No constant resistance circles
(2) No constant reactance arcs
(3) No text
(4) etc ...

Also, your arcs don't aren't constructed  using equal length lines.  Try the
attached version.

Ben

Ben:

Here is one last message for tonight.  I tried plotting a simple cubic
equation.

x = linspace (-10, 10, 201);
y = x.^3;
plot (x,y)
print "testcubic.eps"

The result is attached.  In all of these plots it looks like some origin is
selected and a first subset of the whole data set is plotted.  The second,
third and so forth subsets are plotted.  In the on-screen versions it always
looks just fine.  In the saved file versions it looks like the first point
in each subset resets to the origin which results in the 'rays' going out to
each subset.  Your version of scdraw does the same thing as mine but with a
different origin.  The cubic plot sets an origin in the lower left corner.

I don't know which program is responsible for what but it looks like the
data generated when fltk is the graphics tool includes some character or
command that the on-screen renderer has no problem with but the program that
generates the saved file gets confused by.

I'm revealing my age here but in the old days of mechanical plotters one had
to remember "pen up" and "pen down" commands to keep from getting output
that looks like what happens on my system.

Rick

testcubic.eps
Description: PostScript document