[tor-dev] Code review request for bug #9701

[michael at schloh.com]

Hello Ken,

On Thurs., May 15, 2014, Ken Keys wrote:
>On 5/15/2014 2:10 PM, michael at schloh.com wrote:
>> On Thurs., May 15, 2014, Ken Keys wrote:
>>>>On 5/15/2014 9:38 AM, michael at schloh.com wrote:
>>>>>> If anybody's feeling particularly friendly, then it would be nice of
>>>>>> you to do the favour of reviewing the proposed solution to bug #9701:
>>>>>>
>>>>>> https://trac.torproject.org/projects/tor/ticket/9701
>>>>>>
>>>>>> It relates to clipboard turds of the Tor Browser in violation of
>>>>>> the premise of disk avoidance by design. The code in question is
>>>>>> attached to the bug report.
>>>>>>
>>>>1. Some versions of the SunOS--maybe Solaris too, I'm not sure--had
>>>>a provision where you could copy text without any explicit clip board
>>>>action. You could just select the text and middle click in the
>>>>destination. Will this change break that behavior or are these
>>>>systems still supported?
>>>>
>> I couldn't find any official requirements doc stating that all POSIX
>> or certain versions of certain POSIX/LINUX distros are supported.
>>
>> Regardless of this, I think the highlight and middle click thing
>> is a X11 feature. Good idea to bring this up, but since my Linux
>> tests included mouse as well as middle click action this probably
>> extends to SunOS and other POSIX OSs using X11. Is that your gut
>> feeling too? You think we need more testing using real (VM) SunOS
>> and other platforms?
>>
>I had presumed that the feature was implemented by the window manager
>rather than the underlying X11 system.
>
Hmm, that could be (although didn't CDE have middle pasting
as well?) The problem is that tests combining the set of
(Gnome|Enlightenment|KDE|CDE|Unity|Xfe|LXDE|?) with
(Distro|32/64Arch|Version) is prohibitive.

That's why I was trying to make the change as surgical as possible.
The proposed solution changes nothing for selections <1Mo and
reverts to pre-file cache logic (year 1999) for selections >1Mo.

>I need to get my Linux box up and running again. I don't think it
>would be necessary to test on the actual OS as long as the feature
>can be exercised on the test rig. If the feature is not covered by
>a test case it wouldn't help to run it on the actual OS.
>
I don't know of a test rig that middle clicks, but I'm kind of new
here. So you don't think it would help to do one or two tests on a
less popular or exotic platform with a non Gnome/KDE/Unity? The
problem with Solaris is that all the release since 5 years have
used Gnome.

Regards,
Michael
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3379 bytes
Desc: not available
URL: <http://lists.torproject.org/pipermail/tor-dev/attachments/20140516/38787f6c/attachment.bin>