[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [tor-bugs] #21059 [Core Tor/Tor]: shared-rand-current-value violates spec
#21059: shared-rand-current-value violates spec
--------------------------+------------------------------------
Reporter: atagar | Owner: asn
Type: defect | Status: needs_review
Priority: High | Milestone: Tor: 0.3.0.x-final
Component: Core Tor/Tor | Version:
Severity: Major | Resolution:
Keywords: | Actual Points:
Parent ID: | Points: 0.1
Reviewer: | Sponsor:
--------------------------+------------------------------------
Comment (by asn):
Replying to [comment:24 teor]:
> Replying to [comment:23 atagar]:
> > Hi teor, are you sure you want to go this route? The problem in this
ticket isn't that clients are unable to accept misordered fields. Stem can
still read the consensus just fine. Rather, the trouble is that its
**descriptor validator** was telling us that tor's actual behavior doesn't
match the spec.
> >
> > I see two options...
> >
> > * Fix the ordering in the spec, and be careful in the future that we
don't keep making this mistake.
> > * Loosen the requirement from MUST to SHOULD so it's no big whoop if
tor's order matches the spec or not.
> >
> > Considering that I found three separate instances of these fields
being misordered I don't think this is a detail we want to keep contending
with.
>
> I think we agree here.
>
> I also think my patch does what you want, with the extra constraint that
the arbitrary order of consensus lines may vary between consensus methods,
but must be consistent for a particular consensus method:
>
Hey, why do you think this consensus method consistency rule is important?
I feel it makes the text confusing.
Personally, given the inconsistencies that atagar noted, I think the
patches in comment:6 and comment:20 look good to me.
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/21059#comment:26>
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