Re: [bug-inetutils] Re: GNU Inetutils, #gnu & AMS

[Debarshi Ray]

>       The maintainers decides what patches get commited, if the
>       maintainer does not like a patch as it is, he or she is free to
>       rewrite or reorganise it.
>
>   That is true.
>
>   However, it is sad to see this sort of dispute between two GNU
>   developers.  Could you please both make an effort to agree on what
>   should be done with this code, so that you can have a good
>   relationship again?
>
> I have a clear stance, and I won't budge from it; having lots of small
> tiny files is a mess.  I want to include it in inetutils, it is lovley
> code, as usual from rishi.  I am only disapointed that such a silly
> issue is the culprit to the whole mess.

I keep saying this everywhere, so I will repeat again and quote myself
from the list.

"Ah, and don't forget that the code that you plan to copy-paste into
one file supports only IPv4, bare minimum sanity checks and a single
kernel. Add multi-protocol support, robust sanity checks and multiple
kernels and you will get something much worse.

If you don't believe me please have a look at existing route
implementations. I suggest the Net Tools implementation used on
GNU/Linux systems, NetBSD implementation and FreeBSD implementation.
The NetBSD and FreeBSD implementations do differ."

If you want an example of another GNU program doing this, have a look
at GNU Parted. I never started out with the current layout of code. I
went back and forth and reached the conclusion that this was the most
optimum layout. The BSD support is in the works. Once it arrives (will
take around 15 days) and we support more protocols you will get the
picture.

Happy hacking,
Debarshi