#3 iscii2utf8.pl-new doesn't work properly on remote connection

[Sunil Mohan Adapa]

If I run it directly on my machine it works fine and no
perl warning etc.
But if I connect to some other machine, connect back to
my machine for there and run the script, it does not work.
Here is the behavior:
The ISCII input file does not get converted much
although diff shows that there is difference of few
bytes in each line.
The perl warning popup when the script is run.
something like this:
Malformed UTF-8 character (overflow at 0x82000000, byte
0x39, after start byte 0xbf) in pattern match (m//) at
./iscii2utf8.pl-new line 132.
Malformed UTF-8 character (overflow at 0x82000000, byte
0x39, after start byte 0xbf) in pattern match (m//) at
./iscii2utf8.pl-new line 132.
Malformed UTF-8 character (overflow at 0x82000000, byte
0x39, after start byte 0xbf) in pattern match (m//) at
./iscii2utf8.pl-new line 132.
Malformed UTF-8 character (overflow at 0x82000000, byte
0x39, after start byte 0xbf) in pattern match (m//) at
./iscii2utf8.pl-new line 132.

All this work perfectly fine on local machine. I tried
setting the same locale as on the local machine in the
remote connection, but same problem occured.
I tried setting the default locale on both, same
problem persisted.

[Indian Linux]

Logged In: YES
user_id=415115

error comes at the foll char

READ:
Malformed UTF-8 character (overflow at 0x82000000, byte
0x11, after start byte 0xbf) in
pattern match (m//) at ./iscii2utf8.pl line 134.

sample iscii
Ϻ ڡ

itrans from which it was generated
kajaraare cha.nchal naino.n me.n suuraj chaa.Nd kaa Deraa

so corresponding to 'D' in itrans input

These errors not encountered when script run on RH 7.2
also local (7.3 ) -> telnet RH 8.0 -> telnet local (7.3)
doesnt give error
but local (7.3) -> telnet RH 8.0 gives abv error
seems there is some RH8 specific bug

[Pramod R]

Logged In: YES
user_id=639622

It looks like telnet/ssh is eating some utf-8 characters. Can you try
telnet -8 or the equivalent for ssh ?

-Arun sharma

I had the some problem on mandrkae 9.0, but it was run locally.
On mandrake 9.1, the problem was not there. The problem may have to do with
the locale or charset the locale uses. Mdk 9.1 defaults to unicode.

[Sunil Mohan Adapa]

Logged In: YES
user_id=755045

#!/usr/bin/perl -w
while ((my $c = getc)) {
if ($c =~ /[A]/) {
}
}

the above script has the same problem with our input.