[racket-dev] bug reports: include prefs file?
Right now, there are two big text fields: "Description", and "Steps to
Reproduce".
How about changing it to be 2-3 big text fields, but one of them
defaults to having all this synthesized info in it in plain-text form,
and then the user is free to edit it, or even expected to?
This removes more GUI than it adds.
Perhaps 3 fields, with the new one being "Additional Information", and
something like the format below.
If someone wants to do this, I will write the procedure to sort out the
preferences and generate this text.
INTERACTION HISTORY:
(display "Hello, world!")
(+ 1 2 3)
POSSIBLY MORE-SENSITIVE PREFERENCES:
...
COLLECTIONS:
...
LINKS:
...
OTHER PREFERENCES:
...
HUMAN LANGUAGE: english
VERSION: 5.1.3.10--2011-09-24(09b0a46/a)
ENVIRONMENT: unix "Linux matthias 2.999-686 #1 SMP Fri Sep 2
20:66:05 UTC 2025 i686 GNU/Linux" (i386-linux/3m) (get-display-depth) = 32
MEMORY USE: 94206368
Neil Van Dyke wrote at 09/27/2011 11:09 PM:
> The prefs seem potentially more sensitive than the info traditionally
> hidden behind "Show Synthesized Info".
>
> I'd like to see the "Show Synthesized Info" button go away, if you're
> going to include sensitive prefs in the info. Either the information
> should be exposed while user is writing bug description, or there
> should be a confirmation step after submitting, that pops up a window
> that presents this info that will be added to the bug report and gives
> them a chance to edit or opt-out of it. An advantage of exposing
> during writing bug description is that the user then knows what info
> is provided automatically, so they don't waste time on it.
> Implementing the confirmation dialog seems easiest right now, because
> you can mostly just take the code for "Show Synthesized Info", and you
> don't have a UI design&implementation problem for how to expose the
> info while writing description.
>
> I'm not only being a privacy hippie here. I know of Racket projects
> in which the collects info alone (which Dr* has long included in bug
> reports) could threaten business opportunities of the owner of the
> code, would raise concerns about security that people would then be
> obligated to examine, and could also constitute the bug submitter
> violating an NDA or other restrictions on how they handle certain info.
>
> Robby Findler wrote at 09/27/2011 10:37 PM:
>> Would you think it unwise if it was another field behind the
>> "synthesized info" button? (Perhaps with some other name, but in
>> roughly the same manner.)
--
http://www.neilvandyke.org/