Messages in this thread | | | Date | Thu, 31 Dec 1998 01:33:53 +0100 | From | Marcin Dalecki <> | Subject | Re: small addition to keyboard driver propose with patch |
| |
"Stanislav V. Voronyi" wrote: > > In message <368A5066.4557CA9B@cs.net.pl> dalecki writes: > > >Yes that's what I'm doing. I preffer to work under X11 if there is need > > If you can work only under X its good, but I work as sysadmin > in ISP, and almost all machines which I have its a diferent kind of > servers, and almost on all of this machines X even not instaled.
There UNICODE doesn't matter for console.
> > >for > >fluffy features. However one can't disable the unicode desease in the > >current kernel! > > Unicode is not desease. In Poland you have only 3 or four letters > diferent from ASCII. And diferences between english l and polish is minimal,
This is just plain wrong (by a magnitude):
±¡êʳ£ñѼ¬¿¯óÓ¶¦æÆ (ISO8859-2)
And by the way Polish has more similiarities with Ukrainian then english obviously. Even more then with Russian :-). Razgawaritje po ruski? (Ehm. I know there are hardly Ukrainians talking ukrajiñski jêzyk.)
> but I live in Ukraine & need 37 cyrrilic letters absolutely different from
Before Stalin told them different the Ukrainians used the same scripting system as in Poland.
> any latin. And as for me Unicode is very useful. Even on textual console.
For You ISP job?! BTW. the implementation I'm suggesting isn't going to make Your work any more difficult. It's the CLIENT side wich is of relevance.
> And if this Unicode stuff annoyed for you it can be implemented > as a configurable option.
Not easy. and it isn't currently this way...
> >Read twice. With the design propsal properly implementid. (This still > >needs to be done.) > >You should be able to emulate anything you where talking about entierly > >in > >user space without polluting the console implementation in kernel. > > I prefer in-kernel console driver that look like virtual terminal. > Yes, we can leave in kernel minimum (vt52) function for single screen console > and all other do in user space. But this "user space" program even for user > nobody must be running with superuser level. Imho 10kb of code in kernel > is murch better that potential 10k security holes in "user space multiconsole > implementation" that must be worked with uid 0 even for user nobody.
Yes it's easier to mess around in the kernel, then to write proper user level applications, since you don't need to know what even consoles and virtual terminals are really about. And you don't need to know all the other stuff about the OS too. (Please just prove me wrong by providing the corresponding apps.) But this doesn't imply that complicated stuff (I just don't beleve it can be done in 10k. Even the simplest proper unicode mapping will requie at least 64k.) shouldn't be done in user land. There You cand read files with different glyphs. You don't need to worry about size, compatibility, and more of nice gooddies like this... ;-).
And please show me where are all those unicode aware console applications you need that badly to support? Unicode editor for example? I could actually use one myself and I would be already just plain happy if the border drawing characters worked properly anywhere.
Please just do man screen(1).
Marcin Dalecki / free programmer
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |