Opened 9 years ago
Closed 9 years ago
#238 closed defect (wontfix)
synchronous IP autodetect
| Reported by: | dawuud | Owned by: | |
|---|---|---|---|
| Priority: | major | Milestone: | undecided |
| Component: | network | Version: | 0.7.0 |
| Keywords: | Cc: |
Description
i'm not a fan of autodetect but if we have to have autodetect then it looks like it can be implemented synchronously like this: https://stackoverflow.com/questions/166506/finding-local-ip-addresses-using-pythons-stdlib/166591#166591
using this python module which makes this operation way more portable than os.subprocess... https://pypi.python.org/pypi/netifaces
note that this task is part of tasks: http://foolscap.lothar.com/trac/ticket/236 https://tahoe-lafs.org/trac/tahoe-lafs/ticket/2491 https://tahoe-lafs.org/trac/tahoe-lafs/ticket/517#comment:31
Change History (5)
comment:1 Changed 9 years ago by
comment:2 Changed 9 years ago by
BTW I think this code is going to live in Tahoe, not Foolscap. Do we have a Tahoe ticket for it?
comment:3 Changed 9 years ago by
there's this... we could rename the ticket...
"synchronous node startup" https://tahoe-lafs.org/trac/tahoe-lafs/ticket/2491
comment:4 Changed 9 years ago by
| Component: | unknown → network |
|---|
comment:5 Changed 9 years ago by
| Resolution: | → wontfix |
|---|---|
| Status: | new → closed |
We've decided to stop doing autodetection: applications will either have to do address-detection / port-allocation themselves, or (better) be told these values through user-provided configuration data.
So this ticket can go away.
This or something like it should work and be very portable... in my dev branch: https://github.com/david415/foolscap/tree/238.synchronous-ip-autodetect.0
Am I expected to write unit tests for this? At the very least the test should fail if an empty list is returned. Note that this netifaces module also supports IPv6...