On Sun, May 20, 2007 at 05:12:30PM +0200, Radovan Garabik wrote:
> I started to implement the idea from resize.c, but then I stumbled upon
> a simpler way of finding terminal width, and implemented that (I hope it
> is sufficiently generic), with resize(1) as a fallback.
Aha, reading TIOCGWINSZ from termios, makes sense. I didn't pay much
attention to which paramater is being fetched when I was extracting
getcols.c, I see now that it does the whole shebang with termio,
TCGETA + TCSETAW. Seems excessive even for the original program...
> > The problems remain. Firstly, the end is at the 82nd character, which is
> > twelve characters short of the actual number of columns. Secondly,
> > The first column in the pydf output expanded the most; the illustration
> > (sixth column) expanded too, but way too little compared to the amount of
> > empty space remaining.
>
> Yes, I am aware of the issue, this is due to somewhat inefficient width
> calculation (and default configuration suited to my previous small
> monitor :-)), see 'format' in /etc/pydfrc.
> I have plans to re-design the formatting routines, therefore I have not
> closed the bug yet.
Regarding format in pydfrc, that data structure is a completely
non-intuitive thing to use in a configuration file. But now that I see that,
it's obvious how to improve it - make the default for 'bar' be something
other than 0.1% of screen size?! :)
With a standard 80 column screen, I see good results with 20.0 (percent).
Now if there only was a way to set this in an environment variable, so that
I can propagate it to all my machines through my .zshrc rather than a
separate .pydfrc... :)
--
2. That which causes joy or happiness.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
