On 25 Jun 2015 at 10:12, Beman Dawes wrote:
It looks to me like it is high time to change the Boost guideline for C++ source file encoding to 7-bit ASCII without BOM or UTF-8 with BOM, and to change the inspect program accordingly.
Comments?
+10. I'd suggest UTF-8 with or without BOM. Just assume it is, if it's 7-bit clean now it's automatically UTF-8 clean. Don't forget to adjust all the .gitattributes in all boost git repos to say UTF-8 encoding for source files, some git tools won't render diffs right without that. As an aside, I recently had merry hell getting AFIO to get sent per commit to the wandbox online compiler and wasted a full day trying. Louis' python script which worked finally, while Krzysztof shell script did not. I had been trying to get the latter working. I figured out the problem: I suspect AFIO has a unicode char somewhere in its source code which only appears when you try to pipe it through scripting in order to upload it to a JSON REST API. Python errored out on the bad string, whilst the shell script just magically failed silently. That alerted me to the need to tell Python the source code has a codec, and then it all worked. Bear these sorts of problem in mind if you switch on UTF-8, as debugging that stuff was not obvious. BTW wandbox scripting instructions are now at https://svn.boost.org/trac/boost/wiki/BestPracticeHandbook#a14.USERFRI ENDLINESS:Considerlettingpotentialuserstryyourlibrarywithasinglemousec lick for anyone interested. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/