[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: [tor-bugs] #21236 [Metrics/Metrics website]: Put a visualization of Tor Browser downloads and updates on the Metrics website



#21236: Put a visualization of Tor Browser downloads and updates on the Metrics
website
-------------------------------------+------------------------------
 Reporter:  karsten                  |          Owner:  karsten
     Type:  enhancement              |         Status:  needs_review
 Priority:  High                     |      Milestone:
Component:  Metrics/Metrics website  |        Version:
 Severity:  Normal                   |     Resolution:
 Keywords:                           |  Actual Points:
Parent ID:                           |         Points:
 Reviewer:                           |        Sponsor:
-------------------------------------+------------------------------

Comment (by karsten):

 Let's have the coding discussion in Berlin.  I'm not opposed to the
 refactoring you propose.  I'm mostly concerned about the timing and about
 not knowing in which direction we'll be heading with metrics-web.  And if
 we do a first step into some direction, that might limit or at least
 influence our future choices, which I'd want to avoid.  But let's talk
 more about this in Berlin when we all still memorize this code. :)

 What we should talk about though is the log-line matcher issue you
 describe above.  I believe there's no bug!  Our `LOG_LINE_PATTERN` should
 not match a log line with fewer or more capturing groups.  Note that there
 are ''four'' capturing groups in the pattern of which none are optional
 but of which we only use ''three''.  I believe that
 `logLineMatcher.groupCount() != 3` always returns `true`, because there
 are four capturing groups, regardless of the pattern, so your code change
 doesn't actually change behavior there.  What we ''can'' do is turn the
 fourth, unused capturing group into a non-capturing group (see
 [https://gitweb.torproject.org/karsten/metrics-
 web.git/commit/?h=task-21236-4&id=57edde652bbdf2fd1ff511921d184f66000b696a
 57edde6 in my task-21236-4 branch]).  But that's just a tweak, not a
 bugfix.  If I overlooked a bug here, please let me know!

 And yes, removing the disagreement data and format makes sense:
 [https://gitweb.torproject.org/karsten/metrics-
 web.git/commit/?h=task-21236-4&id=6e40b957cb165f9bfe9a806192f3fbd9adaefbd4
 6e40b95 in my task-21236-4 branch].

 Alright, I'll merge and deploy later today (assuming that these trivial
 two changes don't require much review).  And I'll write a short
 announcement to tor-project@ that this graph now exists.  But I'll hold
 back the blog post until we have a more thorough analysis of this graph,
 per suggestion from yesterday's Vegas meeting.

 Thanks again for all the input!

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/21236#comment:23>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
_______________________________________________
tor-bugs mailing list
tor-bugs@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs