[BusyBox] The 1.0.1 commit list.

[Mike Frysinger]

On Thursday 14 July 2005 04:03 am, Rob Landley wrote:
> 9539: Insufficient description.  (Is it a bug fix?)

if you look at the udhcp module it'd have a good message ;)

really these are multiple changes, some bugfixes, some not ... i'd look at the 
changes in the udhcp module to pick out individual changes/reasons

> 9620: What's with "$ident"?

cvs feature, can probably be punted now that we use svn ...

> 9545, 9633, 9641: loop.c is cursed.  What does the #ifdef accomplish?

no idea :/

> 9650: Is this a new feature, or a bugfix?

depends on who you ask :)  i say it's a bugfix so please move to 1.0.1

> 10109: This workaround for a clearly broken closed-source compiler doesn't
>        even say _which_ compiler...

iirc this was also posted to the busybox list with more details ... i dont 
recall it being an obscure compiler, just not a common one ...

> 10116: Does anything actually use this?

this was added to support the new copyfd code ... but now netcat uses it and i 
think we could probably change more things to use it

> 10155: Don't comment warnings, _fix_ warnings.

if you have a fix for it feel free ... i didnt add anything because the code 
works correctly and i want to break it ... i dont use rpm myself :p

> 10159: how the heck do I move this one over?

you'll have to do it by hand i think:
svn pe svn:ignore <dir>
pe being short for propedit

> *** The following touch the web pages.  We should just grab the new ones or
> something.
> 9447, 9472, 9573, 9697, 9700, 9704, 9787, 9821, 9822, 9823, 9834, 9922,
> 10064, 10122, 10124

i dont think there's much point in maintaining web pages outside of trunk, 
maybe just punt the dir ?  would make for a smaller release tarball :)

> Note: we should also zap the patches/ directory from 1.0.0

agreed

> *** These are new features, which belong in 1.1:
>
> 9766,9871, 10529: the new to-do list.
> 9418: makefile O= and PREFIX= support.
> 9542: readlink -f.
> 9767,9768: Sort rewrite.
> 9837,9838: ash read -t (timeout) support.
> 9868: nice (and renice rewrite).

agreed

> 9969,9970,9972: Support for some random closed source libc.  (Why include
> this?)

it's for libgloss/newlib actually ... they arent closed source

> 10096: test (add [[ and == support)
> 10097: cp -H, -L

i'd like to see these added to 1.0.1

> 10123, 10133, 10154, 10575: fakeidentd
> 10147, 10152, 10182: printenv (also known as "set with no arguments"...)
> 10148, 10149, 10343: sum
> 10160, 10174, 10175: ed (does anybody _use_ ed???)
> 10163, 10169, 10170, 10622: stat
> 10171, 10172, 10173, 10177, 10178, 10179, 10180, 10259, 10260, 10261,
> 10262, 10275, 10276, 10277, 10278, 10279, 10280, 10281, 10299, 10342,
> 10508, 10509, 10510, 10511, 10512, 10513, 10514, 10515, 10516, 10517,
> 10520, 10521, 10522, 10523, 10524, 10525, 10526, 10527, 10528, 10530,
> 10531, 10532, 10533, 10545, 10546, 10608: e2fsprogs
> 10202, 10203: ln -S and -b flags.
> 10206, 10234: zcip
> 10243, 10244, 10248: fancy lsmod for 2.6.
> 10253: nc -w option.
> 10264: Fix ash "standalone shell".
> 10288, 10325, 10333: eject
> 10294: etherwake
> 10298, 10311, 10313, 10314, 10315, 10316, 10317, 10321, 10322, 10323: comm
> 10347: sed -T
> 10485: makedevs
> 10556: ipcs, ipcrm

agreed

> *** This is just wrong:
> 10137, 10165: A space between # and define is not an improvement.

err see my other post about spec requiring # be the first character ?
-mike