Hi,
I just add you bug report by Arnaud Quette.
I am responsible of this regression bug.
Source of this bug was a bug with Pulsar ESV models.
Previously, mge-utalk was sending too many "Ax 1" command. In consequence,
Pulsar ESV UPS were over flooded. Then UPS restart and all switching plug
were cut OFF than ON (who cut power to plugged PC).
To avoid it, I modify enable_ups_comm() to only send init command ("Ax 1")
if link was not established with UPS. As "Si" command is answered only when
link is established, I used it as a test. I ignored by that time that echo
could broke it :(
By removing this test, Pulsar ESV UPS will bug again.
What I suggest is to make a double test (with and without echo) to avoid
both bugs.
Regards,
Martin Loyer
