Re: RFC4941bis implementations

[Fernando Gont <[e-mail]>]

On 2/4/20 17:20, Alexandre Petrescu wrote:
> Le 02/04/2020 à 21:21, Fernando Gont a écrit :
>> On 2/4/20 16:09, Alexandre Petrescu wrote:
>>> Le 02/04/2020 à 19:26, Fernando Gont a écrit :
>>>> Folks,
>>>>
>>>> FWIW, I produce a patch for Linux kernel (net-next) that implements 
>>>> rfc4941bis: 
>>>> https://www.gont.com.ar/code/fgont-patch-linux-net-next-rfc4941bis.txt 
>>>> (up & running here).
>>>
>>> Does it work with a plen 65 in RA?
>>
>> No. Why would that matter in this context?
> 
> That is another way to tell me to shut up in this context, which I will.
> 
> Not before I tell that this patch is another way to write new software 
> propagating further in the future that limit.

1) Typically, you address one issue in each patch. This patch didn't 
remove support for non-64 prefixes. The support wasn't there. Since the 
patch needs to be reviewed, it is better for the patch to do one thing, 
that's clearly defined somewhere (rfc4941bis). One this gets solved, it 
is possible, if deemed appropriate, to improve on this patch and relax 
the requirement on the prefix length.

2) I'm not sure how winning solves problems. You haven't wrote a patch 
to implement rfc4941bis, and you haven't implemented a patch to relax 
the requirement on the prefix length, either. Still, you even have the 
idea of obstructing progress of rfc4941 out of frustration that 
rfc4941bis does not boil the ocean you'd like to boil.

Thanks,
-- 
Fernando Gont
SI6 Networks
e-mail: [e-mail]
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492