First, let me thank everyone for their responses to my query. Most of them have been both educational and helpful.


I want to touch on a few of the responses with comments of my own.

Offline, Alex Brodsky wrote:
AB > found an interesting web page for you - http://www.tachyonsoft.com/cpindex.htm
with lots of code pages for translation.

Thank You Alex! And thanks to you as well, Dave Bond. This link is a thorough description of a very large number of code pages, completely cross referenced my individual characters. It is a wonderful resource, and I want to thank Dave Bond for putting it together and Alex for telling me about it.






Also offline, Paul Gilmartin wrote:
DBC >> Should I be asking, what are your favorite translations?

PG > OEMVS311

A Google search quickly revealed that this "OEMVS311" is a translate table that converts back and forth between ISO 8859-1 (ASCII Latin-1, aka code page 819) from 1047 (EBCDIC Latin-1). A review at Tachyon of these code pages showed that they both have every single character I mentioned in my original post save one (The TM symbol doesn't seem to occur in ANY character set). I guess that's about as good as it can get for me, so thank you, Gil.






Also offline, Chaye Wala agreed with the 1047 choice for EBCDIC, but preferred 437 for ASCII. I looked at it, but 437 does not have a definitions for the broken vbar, the copyright symbol and the registered trademark symbol, so for my needs, 1047 remains better.





Richard Tsujimoto wrote:
RT > [snip] I think the problems were with square brackets. Using 037 for MVS didn't work, but I noticed that FTP didn't have a problem. So, I switched to CCSID 924, which is what FTP used on MVS.

Richard Tsujimoto later reported:
RT > I checked my notes and 1047 did work, but I opened an ETR with IBM last September, asking what CCSID I should use. They responded with 924.

Yes, 924 does handle square brackets fine, but it doesn't have broken vbar, and its definition for cent sign (X'B0') is different from what has been documented on the green card (X'4A') going back several decades, so for my needs, 1047 remains better.






A couple of people took issue with my indifference to reversibility:

Tony Harminc wrote:
TH > Not having reversible tables is almost always a mistake, in my experience. It seems like a good idea to start with (who needs that unpronounceable scripty-looking capital B thing anyway), but it lets you down when you need it.

And Shmuel Metz wrote:
DBC >> Complete reversibility doesn't matter.

SM > To you. But if you're asking for what others prefer, then you have to accept what matters to them as part of their answer.

Perhaps. Nevertheless, there are three reasons why reversibility is not a priority to me.

- First, most of the files for which I want to do this translation are on a one-way trip to a display device or printer.

- Second, by reversibility, I am referring to the entire set of 256 byte-wide values. Perhaps I am wrong, but once you get past the set of commonly defined characters, I know of no consensus about how to translate the remaining bytes.

- Third, most character sets are not congruent (although [as Tony Harminc points out] 1047 and 8859-1 might be, I haven't yet had time to examine them thoroughly). Generally, each set will have characters defined that are not defined in the other.






Finally, Mike Flint suggested:
MF > I use translations:
        EBCDIC CP 1146 to/from ASCII CP 850
- dealing with translations purely on mainframe.

On inbound from off-mainframe servers, we use ISO-8859-01 and 5348, and convert to 1146.

Also use UTF-16BE (CP 1200) on outbound to off-mainframe servers, so this has been converted from 1146.

But here we use dollar signs, not pounds sterling, so these would not work so well for me.






So thank you all, I appreciate the input. It has been most helpful.

[:)]

Dave Cole              REPLY TO: [EMAIL PROTECTED]
Cole Software          WEB PAGE: http://www.colesoft.com
736 Fox Hollow Road    VOICE:    540-456-8536
Afton, VA 22920        FAX:      540-456-6658

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to