a11a1acda3
Pin ctrlc version
...
The 3.1.4 update bumped the version of `nix` being used and can't
compile on 1.34.0
2020-03-07 19:58:04 -05:00
3d2ba4a161
Minor build cleanup
2020-03-07 19:38:28 -05:00
bors[bot]
c0de7f580e
Merge #24
...
24: Update to Aeron 1.25.1 r=bspeice a=bspeice
Co-authored-by: Bradlee Speice <bradlee@speice.io>
2020-01-22 04:19:55 +00:00
99a6c2d551
Update to Aeron 1.25.1
2020-01-21 22:58:35 -05:00
bors[bot]
871556edd1
Merge #23
...
23: Rename to aeron-driver-sys, update to 1.25.0 release r=bspeice a=bspeice
If this goes well, going to go ahead and release the aeron-driver-sys crate, no reason to hold that back.
Co-authored-by: Bradlee Speice <bradlee@speice.io>
2020-01-13 02:20:18 +00:00
b9dbfdd29f
Cargo fmt
2020-01-12 21:19:26 -05:00
b86a876fdf
Main API needs 1.34
2020-01-12 20:58:31 -05:00
07d336a90c
Pin minimum support at Rust 1.33
...
No promises about maintaining that, just want to know if/when I do something that breaks it
2020-01-12 20:26:08 -05:00
27f9a03617
Attempt to use the job matrix instead of re-installing in Windows
2020-01-12 20:10:56 -05:00
b23d08a084
Don't install clippy/rustfmt until after we mess with the Windows target
2020-01-12 19:36:42 -05:00
8b8acd5f4b
Set Rust runtime to MSVC on Windows
2020-01-12 19:33:04 -05:00
23f2a52c6e
Show full rustup info
...
Don't think Aeron will compile with Gnu ABI on Windows
2020-01-12 19:05:43 -05:00
e27a9a317c
Driver error message now consistent on Windows
2020-01-12 19:00:55 -05:00
80c1aed824
Try just removing sh.exe outright
...
Hilarious things happen in Travis when /bin goes away
2020-01-12 18:44:49 -05:00
437e3f9d05
Attempt to remove sh.exe from PATH
2020-01-12 18:37:15 -05:00
56fc28dc47
Attempt to find sh.exe on Windows path
2020-01-12 18:15:54 -05:00
53e7a3b370
Move the clippy/fmt checks to start; we want to fail fast
2020-01-12 18:15:28 -05:00
18ddd92948
Switch back to MinGW makefiles
2020-01-12 17:56:12 -05:00
05f0ad5067
Echo path for Windows debugging
2020-01-12 17:49:54 -05:00
39cc12340c
Cargo fmt
2020-01-12 17:08:45 -05:00
64fccd6a7e
Force Windows to use the MSYS Makefiles generator
2020-01-12 17:03:43 -05:00
b9b5b74b7c
Rename to aeron-driver-sys, update to 1.25.0 release
2020-01-12 16:47:29 -05:00
bors[bot]
531bc0d64c
Merge #21
...
21: Driver proxy r=bspeice a=bspeice
Start implementing the proxy, makes interactions significantly easier.
Co-authored-by: Bradlee Speice <bradlee@speice.io>
2019-11-08 04:07:26 +00:00
a5c8bbb05b
Minor docs note about C++ behavior
2019-11-07 22:53:42 -05:00
3240282354
Re-enable deny(docs)
...
It's better than leaving it out, but I don't think I've yet done a good job at explaining what each element's purpose is.
2019-11-07 22:52:26 -05:00
4884108e04
Fix driver termination
...
Turns out when you attempt to write 0 bytes... you write 0 bytes.
2019-11-07 21:06:11 -05:00
88985418e0
Nicer flyweight API
2019-11-07 20:50:08 -05:00
bors[bot]
1061f302c9
Merge #20
...
20: Update to Aeron 1.23.1 r=bspeice a=bspeice
bors r+
Co-authored-by: Bradlee Speice <bradlee.speice@gmail.com>
2019-11-07 03:19:36 +00:00
Bradlee Speice
b33748ef7d
Update to Aeron 1.23.1
2019-11-06 22:18:38 -05:00
5725c32c72
Start work on flyweights and driver proxy
...
Implemented flyweights four different ways, can't say I like any of them. This is at least a starting point.
2019-11-03 21:19:32 -05:00
505b9a4bd6
Refactor - don't use a client module
2019-11-03 16:19:03 -05:00
bors[bot]
15969581b6
Merge #19
...
19: Broadcast receiver r=bspeice a=bspeice
Doesn't include some tests from Aeron/Agrona because of the complexity of mocking `AtomicBuffer`. Will eventually need to do more work to test this, but it's sufficient for now.
bors r+
Co-authored-by: Bradlee Speice <bradlee@speice.io>
2019-11-03 04:52:33 +00:00
1b49cd0326
Add in the tests
2019-11-03 00:50:42 -04:00
6df284f73c
Add the copy broadcast receiver
2019-11-02 23:12:36 -04:00
4ad3dbcecd
Start work on the broadcast receiver
2019-11-02 22:34:20 -04:00
bors[bot]
98002c00be
Merge #18
...
18: Don't give the full buffer to handler functions r=bspeice a=bspeice
Much simpler from an API perspective, unlikely to be a real performance issue.
bors r+
Co-authored-by: Bradlee Speice <bradlee@speice.io>
2019-11-03 00:50:36 +00:00
79e8c26a23
Don't give the full buffer to handler functions
...
Much simpler from an API perspective, unlikely to be a real performance issue.
2019-11-02 20:50:06 -04:00
bors[bot]
f6dedbe268
Merge #17
...
17: Aeron tests r=bspeice a=bspeice
Port over tests for `ManyToOneRingBuffer` from Aeron C++
Co-authored-by: Bradlee Speice <bradlee@speice.io>
2019-11-03 00:22:05 +00:00
47a102b063
Finish the buffer read messages
2019-11-02 20:21:15 -04:00
bf3a39a46f
Cargo fmt
2019-11-02 19:12:42 -04:00
daecbf2e2f
Finish write tests from Aeron
2019-11-02 19:12:27 -04:00
fff850f6b7
Code formatting
2019-11-02 18:05:15 -04:00
b567336bb0
Start porting over existing tests from Aeron
2019-11-02 18:03:40 -04:00
bors[bot]
b902f580c9
Merge #16
...
16: Atomic buffer trait r=bspeice a=bspeice
Implements #12 . Turns out I had been making it too hard on myself previously.
bors r+
Co-authored-by: Bradlee Speice <bradlee@speice.io>
2019-11-02 20:01:57 +00:00
9139bf7234
CnC termination test working again
2019-11-02 16:00:57 -04:00
b235655f71
Formatting
2019-11-02 15:50:33 -04:00
8fac817ba3
read now working again
2019-11-02 15:50:23 -04:00
ed766ce86b
write now working again as well
2019-11-02 15:40:52 -04:00
02638d20c0
claim_capacity now working again
2019-11-02 15:18:52 -04:00
fd23f2891a
AtomicBuffer trait part two
...
I have a much better grasp of what I need to do now.
2019-11-02 14:59:52 -04:00
