summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Tweak commentbug23857Simon MacMullen2011-06-011-2/+3
|
* Make sure the global name server is definitely up and synced by the time we ↵Simon MacMullen2011-06-011-0/+3
| | | | treat the cluster as "up".
* Merging bug24137 to defaultMatthew Sackman2011-05-313-11/+3
|\
| * These "--novalid"s are less meaningful in the context of bug 24137 since you ↵bug24137Simon MacMullen2011-05-311-2/+2
| | | | | | | | need to have docbook installed to get this far and so can't trigger the bug. But it can't harm to put them in, and it's possible that there's some other bug I haven't thought of that this fixes.
| * Remove more pointless namespaces.Simon MacMullen2011-05-311-4/+0
| |
| * Don't validate against the docbook DTD when building the usage erl - we ↵Simon MacMullen2011-05-261-1/+1
| | | | | | | | don't need docbook installed to build the server, but without this --novalid we instead go to the internet if docbook is not installed, which can make building flaky.
| * What on earth were those namespaces imported for? They're not used for ↵Simon MacMullen2011-05-261-4/+0
| | | | | | | | anything, and they help the build to fail if we can't contact docbook.org.
* | Merging bug24079 to defaultMatthew Sackman2011-05-272-9/+13
|\ \ | |/ |/|
| * In light of bug21782, solve the asymmetry problem the other way.bug24079Matthew Sackman2011-05-272-4/+4
| |
| * Support multiple source dirsSimon MacMullen2011-05-122-2/+3
| |
| * If we see a list of modules there already, trust it.Simon MacMullen2011-05-111-2/+5
| |
| * Use erl files rather than beam files to generate the app module list.Simon MacMullen2011-05-112-5/+5
| |
* | Merge bug 24136.Simon MacMullen2011-05-251-1/+1
|\ \
| * \ Merge bug 24133Simon MacMullen2011-05-250-0/+0
| |\ \
| * | | Assert equivalence of x-message-ttlbug24133Matthew Sackman2011-05-251-1/+1
| |/ /
* | | R14B03 moved hostname and ip_port out of inet.hrl and into inet.erl where ↵bug24136Matthew Sackman2011-05-252-2/+5
|/ / | | | | | | they export_type'd them. This is a backwards incompatible change. Thus there's little alternative to bumping the erts requirement for using specs
* | Merged bug24107 into defaultEmile Joubert2011-05-248-125/+174
|\ \
| * \ Merged bug24129 into defaultEmile Joubert2011-05-240-0/+0
| |\ \
| * | | Add a "tracing" column to the list_vhosts command. A bigger change than ↵bug24129Simon MacMullen2011-05-234-8/+53
| | | | | | | | | | | | | | | | expected, as vhosts didn't have the standard info item infrastructure.
| * | | Inline exchange/1, multihead publish/7Simon MacMullen2011-05-231-17/+9
| | | |
| * | | Rename and move things a bitSimon MacMullen2011-05-231-15/+14
| | | |
| * | | Abstract the typesSimon MacMullen2011-05-231-6/+7
| | | |
| * | | Don't display it as binary.Simon MacMullen2011-05-231-4/+4
| | | |
| * | | Update docs.Simon MacMullen2011-05-231-6/+2
| | | |
| * | | Actually use the exchange we have. Not very scientifically measured, but ↵Simon MacMullen2011-05-231-4/+4
| | | | | | | | | | | | | | | | this seems to knock about 1s off the time to to "time MulticastMain -x 2 -y 0 -z 30" with tracing on (from ~82s to ~81s for me).
| * | | Fix dumb mistake.Simon MacMullen2011-05-231-1/+1
| | | |
| * | | --Simon MacMullen2011-05-201-2/+2
| | | |
| * | | Switch to hard coding the exchange name.Simon MacMullen2011-05-205-29/+50
| | | |
| * | | Fold publish and republish together.Simon MacMullen2011-05-202-45/+40
| | | |
| * | | Optimise looking up trace exchange and message construction.Simon MacMullen2011-05-202-19/+43
| | | |
| * | | Rewrite the rabbitmqctl interface. Define trace_exchanges in rabbit.app.Simon MacMullen2011-05-205-70/+46
| |/ /
* | | Since we should not normally see warnings now, let's zhuzh them up.bug24107Simon MacMullen2011-05-231-1/+2
| | |
* | | Rename variableSimon MacMullen2011-05-231-2/+2
| | |
* | | Introduce variableSimon MacMullen2011-05-231-2/+4
| | |
* | | Exclude mochiweb from the xref check.Simon MacMullen2011-05-231-1/+3
|/ /
* | merge bug24118 into defaultMatthias Radestock2011-05-205-43/+47
|\ \
| * \ merge bug24117 into defaultMatthias Radestock2011-05-195-32/+34
| |\ \
| | * | BQ:idle_timeout => BQ:timeoutbug24117Matthew Sackman2011-05-195-12/+12
| | | |
| | * | whitespaceMatthew Sackman2011-05-181-1/+0
| | | |
| | * | Well it's no shorter, but it closer matches the originalMatthew Sackman2011-05-181-14/+14
| | | |
| | * | BQ:needs_idle_timeout :: State -> Bool ==> BQ:needs_timeout :: State -> ↵Matthew Sackman2011-05-185-26/+29
| | | | | | | | | | | | | | | | (false | idle | timed); which better reflects the different needs
| * | | simplifyMatthias Radestock2011-05-191-21/+14
| | | |
| * | | Merge bug24116 into defaultRob Harrop2011-05-191-8/+17
| |\ \ \ | | |/ / | |/| |
| | * | And now, after testing it, actually make it work by correcting one of the ↵bug24116Matthew Sackman2011-05-181-10/+15
| | | | | | | | | | | | | | | | counters...
| | * | Do not read the msg surviving dropwhile1Matthew Sackman2011-05-181-8/+12
| |/ /
* | | some more inliningbug24118Matthias Radestock2011-05-191-21/+18
| | |
* | | cosmeticMatthias Radestock2011-05-191-17/+17
| | |
* | | fix error handling and construct xname on init rather than every messageMatthias Radestock2011-05-191-14/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | I have no idea why that 'catch' was there. There is nothing to go wrong during publish except the exchange may not be found. Which the 'catch' wasn't catching since that is just an ordinary {error, not_found} return. Cue explosion. So now we handle that and print a nice error message. Also, there is no point constructing the exchange name for every message; doing that once, on init, is enough. That has the further advantage that now the implementation matches the spec (rabbit_exchange:name() is a qualified name, not a binary).
* | | inliningMatthias Radestock2011-05-191-32/+25
| | |
* | | sometimes clarity trumps performanceMatthias Radestock2011-05-191-9/+4
| | | | | | | | | | | | | | | I can't actually detect a difference in performance here, though I'm sure the new code is slower than the old one.
View Lorry Controller Status