Re: r2llib-0.08 and biditext-0.04 "sniffing baboons" release (fwd)

[Tzafrir Cohen]

I accidentally sent this to mulix in private mail instead of to the
list...

---------- Forwarded message ----------
Date: Thu, 26 Jul 2001 11:01:40 +0300 (IDT)
From: Tzafrir Cohen <[EMAIL PROTECTED]>
To: mulix <[EMAIL PROTECTED]>
Subject: Re: r2llib-0.08 and biditext-0.04 "sniffing baboons" release

On Thu, 26 Jul 2001, mulix wrote:

> On Wed, 25 Jul 2001, Tzafrir Cohen wrote:
>
> > On Wed, 25 Jul 2001, mulix wrote:
> >
> > > On Tue, 24 Jul 2001, Tzafrir Cohen wrote:
> > >
> >
> > > are you sure using the version string 0.9.2 on biditext is smart, at
> > > this point in time?
> >
> > The switch to 0.9 by Ilya was because of an important feature was added,
> > that has changed the behaviour (identifying charset). Although it is
> > probably more fitting of 0.1 than to 0.9 .
>
> actually, i was referring to the fact that your last "old style
> biditext" version was also 0.9.2 and i'ts a little confusing, since this
> is "new style biditext". i'll maintain my numbering (current version
> mulix-0.04) until the time we decide that the new style is to be the
> "official" biditext.
>
> > I used 'rxvt -fn heb8x13'
> >
> > behaves quite differently from 'rxvt -fn fixed'
> >
> > But now I see that 'rxvt -fn heb6x13' is not as bad as 'rxvt -fn heb8x13'
> >
> > I'm running this from a remote X terminal, BTW.
>
> sorry, both are exactly the same here.
> i'll try running thetests via remote display on the local lan tomorrow.
>
> please try again, and if this persists for you, i'll add profiling code
> to biditext, so that we'll have some hard numbers to play with.

This is not so much a matter of network latency. I try running it from a
computer with XFree/cygwin near-by I I seem to get the sam results asa at
home.

I tried setting the color of the xterm, and the number of colors in the X
server of XFree/cygwin to 8 (like the poor X terminal), but it still shows
well.

But anyway, it is within the X code, and not within the filesystem code,
probably, because it is X server dependent. I still don't have a testcase
where the hebrew reversing (w/cheking the file) has a noticable effect on
my computer (PII350).

-- 
Tzafrir Cohen
mailto:[EMAIL PROTECTED]
http://www.technion.ac.il/~tzafrir