Fix spelling; add a comment.


svn:r10943

svn:r10941

svn:r10940

 Fix router_choose_by_bandwidth to no longer be biases by floating-point roundoff issues.  This runs through the list of routers yet another time, and uses an additional bitfield, but this should be okay: the function did not appear in profiles before, and shouldnt start appearing now.


svn:r10939

 Add a bit-array type with reasonably fast inline functions.


svn:r10938

 We can have multiple authority certificates for an authority at a time: make the code reflect that.


svn:r10937

 Code to add signatures to a pending consensus directory.


svn:r10936

 whitespace fixes


svn:r10935

 Patch from Robert Hogan: set conn->dns_server_port correctly so that we can close dns server ports when they change, thus avoiding crashes and dangling references and other sources of unhappiness.


svn:r10933

 Patch from tup: translate LFs to CRLFs in NS replies and
 ensures the data section is properly terminated with "\r\n.\r\n


svn:r10932

 Patch from tup: correct grammar for asynchonous reply lines in control-spec.txt


svn:r10931

 ADDRMAP events should never have specified local time.  Extend them to have the proper gmt time, and make GETINFO address-mappings always do the right thing (since it is new in this version).


svn:r10930

 Fix whitespaces and compilation in and around dirvote code.


svn:r10929

 Make buffer large enough to hold ip6.arpa addresses.  This code was not actually used yet, so no user-visible bug existed. (Based on patch by croup.)


svn:r10928

 Some dirvote code to handle generating votes and slinging them around.  More code is still needed.


svn:r10927

 Make all code to generate v1 directories into conditional code.


svn:r10926

 Better fix for GCC 4.2 warning on routerlist.c


svn:r10925

back in after they've been removed.


svn:r10913

 Use tor malloc replacements in eventdns.c for now: there might be goblins there...goblins that only dmalloc can find!


svn:r10912

svn:r10911

svn:r10910

 Once more with feeling.


svn:r10907

 esc_router_info static field should not be cleared postfork, to avoid races.


svn:r10906

 Fix/note some relatively trivial mem usage issues


svn:r10905

svn:r10903

svn:r10902

(we always picked one router past the one we wanted!) and could
conceivably read past the end of the smartlist too.

Nick, I don't know what gcc 4.2 was worried about, but we need
to find a better fix than this. :)


svn:r10901

the median'


svn:r10899

svn:r10898

at least 100KB/s, and consider their bandwidth adequate to be a
Guard if it is at least 250KB/s. This fix complements proposal
107. [Bugfix on 0.1.2.x]


svn:r10897

instead. same functionality, saves a bit of bandwidth. and might even
work.


svn:r10896

 slightly smarter heuristic about when to use obsolete ::a.b.c.d format.


svn:r10893

 Patches to proposal 117 from coderman (from or-dev, 18 Jun)


svn:r10892

 Fix some bugs in ntop/pton.


svn:r10891

 Use our own version of inet_ntop and inet_pton everywhere, to avoid partitioning attacks.


svn:r10888

 Actually, OSX (and some BSDs) accept [f:f:f:f:f:f:f:f::] as a perfectly fine ipv6 address.


svn:r10887

 Same as the last patch, only more so.


svn:r10886

 OSX writes ::3 as ::0.0.0.3; choose a less ambiguous example.


svn:r10885

 Apparently, this problem can occur on the whole BSD family. "Yay."


svn:r10884

 Apparently, OSX does not define s6_addr32 or s6_addr16. How silly.


svn:r10883