[MERGE][bzr-email] Implement mail sending using smtplib

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Marius Gedminas wrote:
...

> That's a good reason to say "We want".  "And we want it to fail if
> something tries to write a Unicode string into it."

I changed it to:
    # We can use a cStringIO because show_diff_trees should only write
    # 8-bit strings. It is an error to write a Unicode string here.

> 
>>>> +class InstrumentedSMTPConnection(SMTPConnection):
>>>> +    """Instrument SMTPConnection.
>>>> +
>>>> +    We don't want to actually connect or send messages, so this just
>>>> +    fakes it.
>>>> +    """
>>>> +    
>>> What does bzr's coding style say about trailing whitespace?
>>>
>> I'm not sure what you are calling 'trailing whitespace'.
> 
> The last lines I quoted above that remark has four spaces after the '+'.
> In other words, its repr is '+    \n'.
> 
>> Are you saying
>> "adding a fully blank line at the end of docstrings"?
> 
> No.
> 
>> Which we don't do.
>> Or "adding whitespace to the end of a single line"? which we ask not to
>> do, but we are avoiding going through the rest of the codebase to clean
>> it up, as it can cause conflicts in what would otherwise be a trivial merge.
> 
> Yes.
> 
> Was this also a move of existing code from another place?
> 
> I did not want to raise an alarm.  I just thought I would point it out
> because my text editor conveniently highlights the trailing whitespace.
> 

No. It was just a mistake. I have my editor set to show trailing
whitespace as well, but apparently:
let python_highlight_space_errors = 1

Doesn't show whitespace only lines. It only shows when you have trailing
whitespace at the end of a real line (and I've seen that it doesn't show
trailing whitespace in strings either.

Thanks for the catch, I'll clean it up. I'll sometimes go back with:

grep -rnI "\s+$"

Unfortunately you've now pointed out that my vim setup is incomplete.
Vim itself seems a little schizo as to whether a blank line gets
whitespace indentation or not.

>>>> === modified file README
>>>> --- README
>>>> +++ README
>>> ...
>>>> +Also, if using 'smtplib', the messages will be sent as a utf-8 text message,
>>> Some pedantic people insist that UTF-8 be spelled "UTF-8", with all the
>>> capital letters, as it is an abbreviation for UCS (or Unicode)
>>> Transformation Format.
>> I'm pretty sure we have never been pedantic about UTF-8, especially
>> since capital letters imply stress that isn't meant to be there. I can
>> change it in public documentation, though.
> 
> Sometimes I wonder whether I am too pedantic, but I usually manage to
> convince myself that that's a good trait in a programmer, like paranoia
> in a sysadmin.  :-)
> 
> I've never perceived stress when I see an acronym/abbreviation in
> capital letters.  That's an interesting point of view.  (I did notice
> that you spelt "RFC" with capital letters.)
> 
> Cheers!
> Marius Gedminas

That's just personal inconsistency. :)

Being precise is a good trait. Being pragmatic is another one. Sometimes
they conflict.

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFujkTJdeBCYSNAAMRAgofAJ4+WOhMxNldcj74S2dv8LtaButTJgCg2foq
iGNG0lnPNh/4dLUzzOfJ4Po=
=W93k
-----END PGP SIGNATURE-----