12 Apr
2016
12 Apr
'16
11:28 p.m.
Hi Bhargava,
On Tue, Apr 12, 2016 at 02:25:31PM +0000, Bhargava Abhyankar wrote:
Currently we validate the source code patches in unit
test environment
and the source code behavior is verified by GDB tool. Going forward we
have planned to test the same source in the integration test setup.
So in other words, this code has not yet been used in practise, by real
MS interfacing with OsmoBTS and OsmoPCU. I thik this is bad news, and
we will have to go for the 'temporary branch' solution that I proposed.
Patches that look fine in terms of coding style and that have been
reviewed positively could be pushed to something like a
'radisys/reviewed' branch. They would remain there until they have been
tested in some way beyond the basic unit test.
I hope you can understand that we cannot merge patches where even the
original author / submitter has not tried to use them in a real setup.
I'm not talking about a comprehensive end-to-end regression test, but
just something very basic and simple, like a single type of MS
attaching, establishing a PDP context and doing a file transfer over
GPRS.
Feel free to share what kind of "integration test setup" you have in
mind for further testing of the code.
--
- Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)