xtics bug?

xtics bug?

Post by Rufus T. F » Sat, 04 Feb 2006 04:11:54

Hi Ethan and others,
here a minimal script that reproduces the possible xtics bug:

gnuplot> reset
gnuplot> unset xtics # <-- must be unset first
gnuplot> set border 1+2+8 # remove top border
gnuplot> set xtics nomirror # xtics at bottom, but *not* at top
gnuplot> p sin(x) # voil?- tics at top!

I would expect to see no top border with no mirrored xtics, in fact
nothing at all at the top.

The problem seems to be related with the nomirror option to {x|y}tics.
And it is introduced only when having unset them first and then
reenabled them. It does not occur when xtics had not been unset before.

The border unsetting is not really necessary.

Please correct me if I'm wrong!
Cheers, RTF

xtics bug?

Post by Hans-Bernh » Sat, 04 Feb 2006 04:56:22

Hmm... 4.0 release binary on Windows does exactly what you expect.
What are you using?

Hans-Bernhard Broeker ( XXXX@XXXXX.COM )
Even if all the snow were burnt, ashes would remain.


xtics bug?

Post by Rufus T. F » Sat, 04 Feb 2006 05:11:44

$ gnuplot --version
gnuplot 4.1 patchlevel 0
$ uname -srom
Linux 2.6.11-kanotix-7 i686 GNU/Linux

Did you $reset; unset xtics; set xtics nomirror; replot ??
Try to set size 1,.5

xtics bug?

Post by merrit » Sat, 04 Feb 2006 06:26:50

In article < XXXX@XXXXX.COM >,

How very strange.
Even stranger is that repeating the command makes the
unwanted tics go away.

unset xtics
set xtics nomirror
set xtics nomirror
plot <foo>

Nope. You found a bug.

Ethan A Merritt

xtics bug?

Post by merrit » Sat, 04 Feb 2006 06:40:37

In article <drtteq$d0o$ XXXX@XXXXX.COM >,

Apparently you found an intentional bug (does that make
it a "feature"?) Someone put in a section of code that
explicitly turns on the tic mirroring in this case.
I have absolutely no idea why.

(cc'ed to the person who added this code)
Ethan A Merritt

xtics bug?

Post by Petr Mikul » Sun, 05 Feb 2006 08:15:55

This message is in MIME format. The first part should be readable text,
while the remaining parts are likely unreadable without MIME-aware tools.

It has been fixed in cvs. Please update.