[BusyBox 0000980]: patch to avoid "broadcast +" syntax
Roberto A. Foglietta
roberto.foglietta at gmail.com
Sat Nov 25 17:55:02 UTC 2006
Denis Vlasenko ha scritto:
>
> I just tried 1.1.3 and it works for me too, without any patching.
>
The only thing you have demonstrated is bb-1.1.3 with or without my
patch works well on YOUR system while works different if is used on
others systems for example mines.
This should convinced yourself that my patch AT LEAST do not introduce
any bugs or any different changes in "just working system" and fix only
the problem in which systems it happens.
You embraced the opposite conclusion, instead: it is not usefull!
So why did I felt the need to have a patch which doesn't do anything?
> I do want to fix busybox bugs, but it does not mean that I will apply
> any random patch.
>
I know I am pretty a newbie in delivering patches for community but
"random" sound a little bit offensive!
;-)
Rob has a great experience about a lot of systems but he did not used
to listen the others just because in the 99% of the cases he was right
or better... so he did not care patches because one day he would
rewritten the entire applet (and he did sometimes).
Under my point of view things are not changed at all: now patches are
under feedback but they did not still applied because "on your desktop
system everything seem works fine".
Good for me too but why desktop users have to use bb when they
probably are dreaming a good 3D layer for kde or gnome, instead?
Anyway I do not want waste your time in discussion about this patch.
I think just few people could happily/easily put their hands in
ifconfig without destroying or rewriting it as whole. So ifconfig is not
going to be changed in the near future. So my patch will apply cleanly
in THAT systems in which it makes the difference. If the people who are
going to build THAT systems have to apply my patch THEY will appreciate
my work much more than if it would anonymously accepted in bb.
;-)
Finally I think not all NIC are the same: a NIC could be smart enough
to set broadcast by itself as a useful feature in the today's desktop or
server systems: why does NIC-BIOS should pollute other subnets risking
the dummy user thinks the NIC is not performing so good? If a NIC-BIOS
is able to boot an entire system why it should not be able to set the
broadcast itself? But a NIC could be projected also to blindly obey or
to save FPGA-gates as in some embedded/telecommunications systems. In
these different systems broadcast has to be setup correctly by software.
Does it make sense for you?
Now you could ask: who cares about some strange embedded systems? May
be the busybox maintainers if I am right about desktop users whose are
all dreaming the 3D desktop instead a smaller footprint core system.
;-)
Cheers,
--
Roberto A. Foglietta
http://roberto.foglietta.name
http://linux.genova.it
More information about the busybox
mailing list