On Wed, 17 Mar 2004, Roozbeh Pournader wrote:Done that.
On Wed, 2004-03-17 at 19:39, Nadim Shaikli wrote:
by using or even writing our own Bidi/me is very shocked to hear that!
/me is not shocked at all.
Why in hell did we spend those many hours working on FriBidi's 100%
Unicode compatiblity if one will still need to write a new bidi engine
for many free software applications?
There are 10 kinds of application developers:
* Those who start their own bidi implementation
* Those who use FriBidi of ICUSwitched to that. Am using ICU at the moment, and hating every second.
The Putty case proves nothing. I can understand someone not wishing for runtime dependancy on another library. Not wanting any compile time dependancy on an external library means that you, by definition, have to write your own. I don't think that's a case for fribidi's future in any way.That's something to think about. I seriously believe we should change the license to a more liberal license, and the Putty case is proof enough.
For whatever it's worth, it's a me too on this one. Notice that there is nothing to prevent any open source software of any license from using fribidi in any reasonable way, unless they say "they don't want compile time dependancy".No, I seriously believe the opposite. I like the Copyleft, GPL, LGPL, and FSF. And I want to walk in the road. Why? Because that's the way many great pieces of software are walking in: Linux, GNOME, blah, blah...
BTW, in the case of PuTTY, there are a 10 things to consider:Well, technically there is a Linux version of putty compiled using winelib. That was, however, done mostly to test winelib, and not to bring putty to Linux.
* The weakest point in bidi support in PuTTY is not the bidi alg, but the semantics of the terminal.
* PuTTY is a no-op in Linux IMO
shachar
roozbeh
--behdad behdad.org
-- Shachar Shemesh Lingnu Open Systems Consulting http://www.lingnu.com/