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

Re: [tor-bugs] #2411 [Tor Client]: expand getinfo circuit-status to show build_state flags



#2411: expand getinfo circuit-status to show build_state flags
-------------------------+--------------------------------------------------
 Reporter:  arma         |          Owner:  rransom           
     Type:  enhancement  |         Status:  needs_review      
 Priority:  major        |      Milestone:  Tor: 0.2.3.x-final
Component:  Tor Client   |        Version:                    
 Keywords:               |         Parent:  #2114             
   Points:               |   Actualpoints:                    
-------------------------+--------------------------------------------------

Comment(by nickm):

 Replying to [comment:12 rransom]:
 >
 >  * Rearranging the code to put event-specific flags for `FAILED` and
 `CLOSED` events in the middle of the keywords describing a circuit's
 current state would significantly uglify Tor's code.

 Not greatly IMO.  I'd just have the function that generates a state
 description take an optional reason_flags argument, and make an XXX note
 that those events should get moved to the end later.

 >  * The `PURPOSE` event extension was added after the `REASON` and
 `REMOTE_REASON` fields were, so the spec has previously been interpreted
 as allowing new event extensions to be placed anywhere after the non-
 extended event.  PyTorCtl is thus quite broken.

 Like I said on IRC, I don't believe that the spec was "interpreted" on
 that point when I added PURPOSE back in 2008: I think I just ignored the
 possibility that stuff would break based on a tighter/looser reading of
 the spec, and nobody caught it in time.  In fact, I think that stuff might
 have broken when I added it, which wasted people's time.

 Regardless of whether we think that that pytorctl's breakage is conformant
 or nonconformant, the fact remains that this change will break pytorctl
 as-is, which means that we need one of the two options suggested above.
 I'd rather take the "fix in parallel" approach so we don't need to delay
 this feature, but if you're okay with holding off on merging it until
 pytorctl and its users won't break, I can live with that.

-- 
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/2411#comment:14>
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