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

Re: [tor-bugs] #3015 [Obfuscation/BridgeDB]: Enhance bucket functionality



#3015: Enhance bucket functionality
----------------------------------+--------------------------
 Reporter:  kaner                 |          Owner:  isis
     Type:  enhancement           |         Status:  closed
 Priority:  Medium                |      Milestone:
Component:  Obfuscation/BridgeDB  |        Version:
 Severity:  Blocker               |     Resolution:  fixed
 Keywords:  bridgedb-dist         |  Actual Points:
Parent ID:                        |         Points:  1
 Reviewer:                        |        Sponsor:  SponsorM
----------------------------------+--------------------------
Changes (by isis):

 * status:  needs_information => closed
 * points:   => 1
 * resolution:   => fixed
 * sponsor:   => SponsorM
 * severity:   => Blocker


Comment:

 Replying to [comment:9 isis]:
 > I should also add that, because the design of `Bucket.py` is in contrast
 to the current development direction for #12505, #12506, #12029, #12030,
 and #12031, and because:
 >
 >  * the code in `Bucket.py` is no longer used, for reasons mentioned in
 [https://trac.torproject.org/projects/tor/ticket/3015#comment:7 my
 previous comment],
 >  * the code in `Bucket.py` never had unit/integration tests, and
 >  * the SQL schema governing how the bridges assigned to `Bucket`s are
 stored in the database is clunky, highly inefficient (one shouldn't have
 to do a
 [https://gitweb.torproject.org/bridgedb.git/tree/lib/bridgedb/Bridges.py?id=9ec5a96a437edd203ea0979a9a60c0297619f426#n453
 for loop over all the bridges] which BridgeDB has seen in the past 9 years
 to find the ones currently assigned to `Bucket`s), and gaining in
 inefficiency linearly every time a new bridge somewhere comes into
 existence,
 >
 > I am tempted to remove `Bucket.py` nearly entirely until it can be
 redesigned with some particular use-case(s) in mind. In removing the code,
 I'd like to retain:
 >
 >   * functionality for extracting data about bridges still assigned (or
 historically assigned) to buckets from the databases (unfortunately, this
 will still retain the nasty `for` loop linked to above, unless we want to
 update the SQL schema once more before migrating to the new databases in
 #12030)
 >
 > and remove:
 >
 >   * `Bucket.py`,
 >   * the ability to dump buckets to `.brdg` files,
 >   * the ability to create new buckets via the `BUCKET_FILES` config
 setting, and
 >   * the `SIGUSR1` handler for dumping buckets to `.brdg` files


 Fixed in my `fix/3015-remove-buckets`
 [https://gitweb.torproject.org/user/isis/bridgedb.git/log/?h=fix/3015
 -remove-buckets branch].

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/3015#comment:10>
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