Text 218, 90 rader
Skriven 2005-04-28 10:49:00 av Michiel van der Vlist (2:280/5555)
    Kommentar till text 209 av Carol Shenkenberger (6:757/1)
Ärende: Changing FTS-0001.
==========================
Hello Carol,

 > Lets be frank.  I can not afford at this time to destabilize
 > my system as I'm now Z6C.

Even without the ZC hat, I fully understand you don't want to mess with a
system as stablke as yours when you will be away for extened times. I still
don';t know how you do it; runnin a system that keeps on running in the
autopilot for 8 month.

 > I plan (among the moelhills, and a more topmost one) to setup a
 > secondary system with IP capability and if it proves *stable* to
 > eventually swap to it.

It will be difficult to get the same stability as you have now. I had the same
decision to make when I added IP capability. My POTS only system has been
running on a DOS machine until Sept 2003. There was no way that I could make it
take care of IP too, DOS is ony stable in a single task environment, I never
could get a /stable/ multitasker for DOS.

S I switched to Win98 for the IP part and later swapped to POTS part to the
Win98 machne as well. It works but it isn't by far as stable as the good old
DOS system. It needs a reboot now and then.

Maybe you should go Linux.

MvdV>> flaws. I either have to learn to live with them or look further. But t
MvdV>> problem is that whenever you find an alternative that seems to solve o
MvdV>> problem, you always run into a new incompatibility problem with existi
MvdV>> software. So one has to settle for something less than perfect.

 > Havent even checked if Chinese script might have some problem.
 >  While we have a few nodes listed there, they do not travel
 > the mainstream echos often enough to have generated any
 > 'complaints'.

Chinese script in FidoNet?

 > I do not know if a future China will grow in fidonet, but
 > awareness should be there that their codepage may conflict
 > with some standard as well.

I have never heard of a Chinese codepage. Unlike the Japanese Kanyi set, the
Chinese alphabet has thousands of characters. How can they ever get that into
an 8 bit character set?

You know Carol, that is one of the reasons I am sceptical when claims are made
about Fidonet nodes in China. FidoNet is based on 8 (actually just 7) bit
character text only.

There is no way that one can decently fit Chinese in such a scheme. You need a
graphic environment or at least a 16 bit character set. So if there is Fidonet
in China, they will have to make do with an ugly transliteration or a VERY
limited representation of their character set. Or they will have to use another
language. I don't think that will be satisfactory for the average user.

A medium that does not allow decent comunication in one's own laguage
won'tattrackt many users and those users certainly won't stay if an alternative
becomes available.

 > I am in favor of allowing any codepage to be used as long as the
 > effect is cosmetic on those who do not share the same one. I am
 > adverse to mandating any one if it costs us nodes.

So am I. In fact I am adverse to ANY change that is not downward compatible.
That is why I have problems with numbers in the phone numbers field that aren't
phone numbers. That is why I have problems with ION hosts.

Force people to upgrade their software and many of them will leave. We can't
afford that.


Other han that, these codepages are a kludge as well. Many languages need more
than the 96 characters that the ASCII set can provide. Codepages can solve most
of the problems as most languages (except for Chinese) can do with 2 times 96
characters.

But... there is still no way that I can use German and Cyrillic in the same
text. If I use codepage 866, I have to forgo of the umlauts. And vice versa.

The only way around this is to allow 16 bit characters. Or preferebly a mixed
set, such as UTF8. UTF8 is downward compatible...

But I doubt if I will see the day that all of FidoNet goes UTF8...


Cheers, Michiel

---
 * Origin: http://www.vlist.fidosoft.de (2:280/5555)