[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Change default PATH for Jessie / wheezy+1


On Wed, 8 Aug 2012 12:40:42 +0100
Roger Leigh <rleigh@codelibre.net> wrote:

> As a distribution we have to decide on a default, and that is ip.
> We took the effort to remove all use of ifconfig from ifupdown and
> other related tools for wheezy precisely to make it removable and
> optional, so that it can eventually be removed.

It's perfectly fine to make it optional so the system doesn't require
it, but complete removal seriously affects usability. Ifconfig is much
more human-oriented, and it's not Linux-only, as some people mentioned

> While it's fine for an end user to continue to use ifconfig, we
> should continue to remove its use by ourselves and in programs in
> Debian.  Warning the user that they are using an obsolete tool is
> IMO entirely justified, particularly when there is a much better
> and more capable replacement.

I don't think any warning is justified. I use ifconfig quite often, and
seeing any warnings is very annoying. It's the same situation as with
idn(1), which used to display information about it's LGPL license every
time you run it, which for me as a user is just on-screen rubbish which
prevents me from receiving the information effectively.

Also, when it becomes an optional package, it's completely user's
choice to install it, so we shall respect it and not to warn anyone.
A kind of warning may be put as the last paragraph of the package
description, however, so users know what they're doing when they
install it.

Ifconfig is perfectly fine for many tasks and I can't seen why it
should be wiped or anything. Same applies to route(8), which produces
more readable output for IPv4 (but not for IPv6). As for netstat(8), I
don't know a better tool.

P.S. There are complaints about net-tools that they use old APIs. Okay,
complainers are free to port them to newer ones, probably adding
support for multiple IPv4 addresses or anything, but please keep the
interface as close as possible to what we have now.

WBR, Andrew

Attachment: signature.asc
Description: PGP signature

Reply to: