On Thursday, 12/15/2005 at 09:54 EST, "Horlick, Michael" <[EMAIL PROTECTED]> wrote: > I did the 'convxlat 00371252' and yes, the TCPXLBIN file matches the one on > TCPMAINT's 592 disk.
Good. We can quit worrying about that. > Let's take the case of 1 character , a circumflex (â) > > In the 00371252 TCPXLATE file , at the bottom at x'42' the ASCII is x'E2' which > is correct , no ? In Western Windows using native Windows services, â is indeed 0xE2. (You can prove this by using the Accessories -> System Tools -> Character Map.) > In the CANADIAN TCPXLATE file , at the same position the ASCII is x'83' This is correct for OS/2 (code pages 850 and 858). > Doing the PPS FRENCH FILE A (TCPXLBIN CANADIAN it prints as â > Doing the PPS FRENCH FILE A (TCPXLBIN 00371252 it prints as an upper gamma ? > (according to my Greek colleague) This tells me that your print server is actually operating using code page 437, the old IBM PC code page, meaning it isn't using Windows APIs, but DOS. See http://en.wikipedia.org/wiki/Code_page_437. Perhaps the CHCP command could help. You can confirm this by printing capital circumflex a. It very likely appears as superscript underscore a when you use CANADIAN. OS/2 and DOS both treat 0x83 as â, which was the source of my confusion. Are you using the built-in LPD capability of Windows? Or are you running some other program? If necessary I can create a 00370437 translation table, but it won't be until after the start of the new year. Alan Altmark z/VM Development IBM Endicott
