Re: Password quoting

2013-09-27 Thread guns
On Fri 27 Sep 2013 at 10:52:44AM +0200, Oswald Buddenhagen wrote: > i think a good rule of thumb is "when you use weird characters, quote > the whole string". that also works with the conf filetype. I agree, this is a minor point, and I am happy to drop it now that double quoting works reliably.

Re: Password quoting

2013-09-27 Thread Oswald Buddenhagen
On Thu, Sep 26, 2013 at 09:19:04AM -0500, guns wrote: > The only reservation I have is that all configuration arguments must now > backslash-escape literal " and \ regardless of whether or not they are > enclosed in double quotes. This is reminiscent of a shell's input parser > correct. > and is u

Re: Password quoting

2013-09-26 Thread guns
On Wed 25 Sep 2013 at 09:14:54PM +0200, Oswald Buddenhagen wrote: > i took a somewhat more aggressive and "creative" approach to the > problem, after all. please review and test the attached patches. After testing the patches, I can confirm that my concerns have been addressed: passwords that con

Re: Password quoting

2013-09-26 Thread Oswald Buddenhagen
On Wed, Sep 25, 2013 at 10:48:04PM -0500, guns wrote: > I am unable to do any tests because the patches do not apply to my > copy of the repository. > whoops, sorry, i entirely forgot the "preparatory" patch. >From e4ace86e387a9496d31c8f7fb628689eb3ecdcd9 Mon Sep 17 00:00:00 2001 From: Oswald Budd

Re: Password quoting

2013-09-25 Thread guns
On Wed 25 Sep 2013 at 09:14:54PM +0200, Oswald Buddenhagen wrote: > On Tue, Sep 03, 2013 at 11:34:31PM -0500, guns wrote: > > The changes are done conservatively, and the style is in line with the > > rest of the code. I am happy to make any and all amendments that you > > propose and will be awai

Re: Password quoting

2013-09-25 Thread Oswald Buddenhagen
hi, On Tue, Sep 03, 2013 at 11:34:31PM -0500, guns wrote: > The changes are done conservatively, and the style is in line with the > rest of the code. I am happy to make any and all amendments that you > propose and will be awaiting your response. > i took a somewhat more aggressive and "creative

Re: Password quoting

2013-09-04 Thread Oswald Buddenhagen
On Tue, Sep 03, 2013 at 11:34:31PM -0500, guns wrote: > The changes are done conservatively, and the style is in line with the > rest of the code. I am happy to make any and all amendments that you > propose and will be awaiting your response. > thanks, i'll definitely have a look at these patches

Re: Password quoting

2013-09-03 Thread guns
On Mon 2 Sep 2013 at 12:01:35PM +0200, Oswald Buddenhagen wrote: > On Mon, Sep 02, 2013 at 03:25:07AM -0500, guns wrote: > > On Mon 2 Sep 2013 at 09:17:29AM +0200, Oswald Buddenhagen wrote: > > > patches welcome. > > > > The patch is trivial, but agreement is harder won. > > > it's not entirely t

Re: Password quoting

2013-09-02 Thread Oswald Buddenhagen
On Mon, Sep 02, 2013 at 03:25:07AM -0500, guns wrote: > On Mon 2 Sep 2013 at 09:17:29AM +0200, Oswald Buddenhagen wrote: > > lastly, the PassCmd output is *not* treated in any way. the problem is > > at the IMAP level: the spec does not define a means to escape quotes in > > a quoted string, so ev

Re: Password quoting

2013-09-02 Thread guns
Implementation Recommendations", so I suppose this is not mandatory for compliance, but the IMAP servers I interact with accept \\ and \" within in double-quoted strings.¹ > patches welcome. The patch is trivial, but ag

Re: Password quoting

2013-09-02 Thread Oswald Buddenhagen
hi, On Mon, Sep 02, 2013 at 12:49:51AM -0500, guns wrote: > mbsync currently assumes that all password inputs are double quote > escaped¹. This is true for "Pass" config entries, the output of > "PassCmd", and also for passwords entered at the terminal. [...] > ¹ The following lines in src/compat/

Password quoting

2013-09-01 Thread guns
Hello, mbsync currently assumes that all password inputs are double quote escaped¹. This is true for "Pass" config entries, the output of "PassCmd", and also for passwords entered at the terminal. Thus if one's IMAP password is the three character sequence: a\b it must entered in one of the follo