[Battlemesh] WBMv8 -- a personal perspective
nemesis at ninux.org
Mon Aug 10 17:50:17 CEST 2015
On Mon, 10 Aug 2015 17:04:41 +0200, Sven Eckelmann <sven at narfation.org>
> On Monday 10 August 2015 14:13:24 nemesis wrote:
>> I would also encourage more participation of ALL the routing
>> developers, because in my opinion the batman-adv team did not
>> follow the
>> process as much as the other routing protocol developers.
> Can you please tell us more about that? We've debugged different
> problems for
> the wibed team (it seems to be that some people more or less
> seriously like
> blame batman-adv for everything), provided workarounds or suggestion
> what we
> think should be be changed in OpenWrt to fix crashes/problems in
> components. At some point the wibed team seemed to have a basic
> firmware with
> batman-adv already running. Maybe I remember this wrong but at least
> I was
> always under the impression that batman-adv was already there and
> and changes should only be provided by the routing daemon developers.
> that point I was only asked by Pedro about the current version of
> in the Chaos Calmer openwrt-routing feed. This was followed by a
> discussion of
> Marek and Igor about whether the firmware team should switch to Chaos
> for WiBed or stay at the current version (which I assumed was
I am sorry about the misunderstanding.
We should have found the time to send an email explaining the situation
or even better we should have had a collective discussion at some point.
The problem is that all this stuff happened between wednesday and
friday, wednesday we left for the social event at 16:00, so we got back
working on thursday with nothing encouraging, at that point we started
to worry, but we informed everyone of the situation verbally, we thought
it was enough, our bad.
> The only thing I heard after that point were some rumors about some
> adv people" which were allegedly started by Juliusz , the
> existence of a
> second firmware team (which already seemed to have stuff running -
> which meant
> for me that I also already have batman-adv working)
The only problem was that we tried to get all the protocols to use the
but we had to make a separate one for batman-adv, see here:
That's the thing we'd like to have avoided.
I wrote what I wrote, now that I'm seeing your point of view I think I
should have communicated better the difficulties we were encountering
and we should have had a collective discussion to decide how to proceed.
> and then some people
> running around making fun about batman-adv because it allegedly
mmm ... whoever did that is either malignant and/or stupid.
We noticed duplicated packets while testing the batman-adv specific
config, then realized there was a misconfiguration.
Then fixed the misconfigurations and still got the duplicated packets
randomly until they disappeared... only to know that they reappeared
randomly during tests.
Then it turned out it was a wifi driver bug. That's what the battlemesh
is for, finding bugs, isn't it?
> I've tried to debug the latter one in a smaller setup because the
> other things which I overheard didn't seem relevant for the actual
> tests. The
> 7 WDR4300 I brought with me were already used in a different testbed.
> Krishna was kind enough to help out with his nodes and tested some
> things with me and at least partially succeeded. The problem with the
> driver was reported to OpenWrt, a working version was of the wifi
> driver was
> identified and everything was explained to persons who I've
> recognized as
> relevant for the ongoing tests (Amadeus, Henning, ...). Of course,
> this was
> already quite late and resulted only in smaller discussions about
> what we
> should do differently about the firmware stuff in the next battlemesh
> and no
> change in the "new test firmware".
> During that time or after that I was not approached anymore by anyone
> configuration stuff or heard anyone requiring help with batman-adv
> for the
> battlemesh test. Maybe you've talked about it with Antonio (at least
> I was
> near you while you've talked very intensely with him in italian) but
> I have a
> big flaw - I don't speak italian at all.
I want to thank you and everybody for your work.
I think the communication problems can be solved by having regular
meetings to explain the status of the tests.
> Sorry for not being the most social person. I don't talk much and try
> not to
> disturb person when they are already occupied. But this doesn't mean
> that I am
> not listening or that I don't try to help. This was only my second
> and also the first one after a two year pause from batman-adv and
> So I didn't want to join any team before the battlemesh started
> because I
> considered myself more a burden to any team which already knew what
> was done
> the last years. I was still trying to help ad hoc when required and
> hindering someone.
> Even when still thought and talked positive about the event with
> txt.file and
> Amadeus - no, I don't think so anymore after reading these mails. If
> it is
> required to create a second team which doesn't want to cooperate with
> us (or
> at least not with me) - ok. But then reading this is more than
> I said that I would attend the next battlemesh and most likely a
> meeting in
> Sarantaporo. But I've now decided not to do it anymore (at least not
> such an event). I met many very kind and intelligent people. But this
> help when at the end I am under the impression of being expelled and
> bad about having attended the battlemesh in the first place.
That would be a shame Sven. Don't take it personal for what I wrote, it
was a partial and wrong view of how the things unfolded.
I had fun, we worked really hard to try to get something. I even
dreamed the frigging UCI configs at night. That was pretty hilarious.
More information about the Battlemesh