use of ISO-8859-1 encoding in ACL handling code

[Michael Schwartz]

Hi,

Is there a reason boto.s3.bucket.set_xml_acl uses ISO-8859-1 for encoding ACL content? It can't represent all possible languages (e.g., I just tried with Arabic). Would it cause problems if I switch this to use UTF-8 instead?

Thanks,

Mike

[Mitchell Garnaat]

Hmm, I think you need to ask yourself that question 8^)

https://github.com/boto/boto/commit/2463033353c6d8e879d9702cbadebf1bca0f3418

It seems like utf-8 would be a better choice and I can't think of any reason why 8859 would be better.  Any recollection of what motivated the origin commit?  Perhaps it was just a desire to have some explicit encoding rather than nothing at all, which was the case prior to this commit.

Mitch

[Michael Schwartz]

Thanks for digging into the history Mitch - you're right, I was the one who originally set that encoding :-)

If I recall correctly I used ISO-8859-1 in that commit because I saw that encoding being used elsewhere in boto. But looking now I don't see that encoding used anywhere else, so I'll change it to UTF-8 now.

Thanks,

Mike

[Mitchell Garnaat]

Yeah, I think that should be fine.  I don't anticipate this causing any problems for anyone.

Mitch