On Friday 24 March 2017 17:56:47 Pau wrote:
Would be a nice feature for lime-config to check if
the proper
lime-proto-* is installed. In case not do not write it to /etc/config/lime
list protocols 'batadv:%N1'
list protocols 'bmx6:13'
list protocols 'olsr:14'
list protocols 'olsr6:15'
list protocols 'olsr2:16'
So in my case olsr* is not installed and it's kind of confusing to have
it in config/lime (it should be in lime-defaults instead).
Also, as far as I understand this new feature, there should be not
special macros on /etc/config/lime. For instance:
option hostname 'LiMe-%M4%M5%M6'
list protocols 'batadv:%N1'
So /etc/config/lime should reflect what's the current real state on the
node, while /etc/config/lime-defaults has the "rules" which are/have
been used to generate this current state.
Then hostname should be "LiMe-WhatEver" and batadv should be something
like batadv:123
Those things depends on the specific configuration so have to be solved in a
different manner, more over we should be careful of the side implications, for
example if we substitute %Nx in the config with the actual value but then the
user change the SSID the batadv vlan will be not changed, this can be a
desirable side effect in some use cases, but is not other so we should talk
and think a bit about those kind of things...
Cheers!
Cheers.
On 24/03/17 17:46, Gio wrote:
> On Friday 24 March 2017 17:31:50 Pau wrote:
> [CUT]
>
>> IMO we should have all existing options on lime-defaults. This way it is
>> easier for the user to know which is the current state of the node
>> without the need of looking at cryptic /etc/config files.
>
> I agree in principle because of that i put the warning and stacktrace
> printing>
>> I realized that many existing options are not even present on
>> lime-defaults, for instance 'country' or 'txpower' for WiFi
config
>> section.
>
> those are someway special, because the script accept any option we put
> there except some, but definitely we should document that someway ;)
>
>> We will test your commit here.
>
> Thanks you!
>
> Cheers!
>
>> Thanks!
>>
>> On 24/03/17 13:41, Gio wrote:
>>> I have been thinking about the unexpected behavior made evident by
>>> modifications in the branch reproducible_config
>>>
>>> Look at line 5 of this paste for an example of misbehavior
>>>
http://pastebin.com/deGbPvmB
>>>
>>> I have been investigating for what can be causing it and discovered we
>>> are
>>> accessing configurations that are not declared in our default file, so I
>>> have added a couple of them that were obviously missing and added some
>>> code to avoid insurgence of this again in the future
>>>
>>> So please test 568ffd9b80a8aec053ae3ff16c6f5ff1647d17bd and feedback ;)
>>>
>>> Cheers!
>>>
>>> _______________________________________________
>>> lime-dev mailing list
>>> lime-dev(a)lists.libremesh.org
>>>
https://lists.libremesh.org/mailman/listinfo/lime-dev
>
> _______________________________________________
> lime-dev mailing list
> lime-dev(a)lists.libremesh.org
>
https://lists.libremesh.org/mailman/listinfo/lime-dev