How to encode help-files (utf8 or ansi)?

[Florian Chavez Juarez]

Hi everybody,

I have a rather technical question regarding the help-files for user-written commands. I am developing a few user-written commands and up to now I encoded my help-files in ANSI. Up to Stata 13 the special characters (e.g. á) were correctly displayed. As of Stata 14 they are no longer recognized and weird symbols are displayed. Of course, I can simply convert/encode them in UTF8 and the problem is solved for Stata 14, but what's about the older versions of Stata? It's important that the help-files can also be read by Stata9-13. How would you proceed, what type of files (UTF8,ANSI) would you submit to SSC?

Is there any trick to provide help-files that can be correctly read by both the old and the new Stata?

Thanks a lot for hints!

Kind regards,
Florian

[Nick Cox]

Short answer is not to think about encoding at all, but to use SMCL directives for anything for which this question might arise.

With me it's a small point of principle not to be bang up-to-date on SMCL and the idea there is that my help files should remain readable a few versions back of the current version.

For some reason, Florian singled out SSC, but that's my advice concerning the Stata Journal too.

[Florian Chavez Juarez]

Thanks a lot, I think this is the best solution indeed! I was not aware of the fact that SMCL also provides codes for special characters such as á. Thanks a lot.