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

Re: New busybox breaks everything

On Thu Oct 21, 1999 at 09:58:23AM -0400, Adam Di Carlo wrote:
> David Huggins-Daines <dhd@eradicator.org> writes:
> > On Wed, Oct 20, 1999 at 11:45:05PM -0400, David Huggins-Daines wrote:
> > > 1) Merge back in these functions from the old revision of busybox (this
> > > probably isn't very hard, it just consists of adding the code back in)
> > 
> > Er, on closer inspection this turns out not to be true.  This new busybox is
> > quite different from the old one.
> > 
> > Perhaps we should just revert to the old one?
> Why?  What is breaking?  Can someone describe the problems introduced
> by the new busybox?  We can kick them over to Erik for fixing...
> As for whether this is official or an unofficial update, I think it's
> semi-official.
> We can certainly roll back if we need to, but they built fine for me
> at least... (not tested).

I got approval from Bruce Perens (the "official" though otherwise
occupied maintainer of busybox) to make an "interim release". This 
is as official as it is going to get.

Please do let me know of _any_ busybox problems. Some problems will be
simple integration issues (i.e. dbootstrap should now be run from the
/etc/init.d/rcS init script, the tar implementation is different and
now acts like tar is supposed to (i.e. tar -xvf <file>), etc). Other
problems are going to be the result of bugs in busybox. While I'll
be happy to help on with the integration effort, I am _very_ _very_
interested in squashing any busybox bugs. If you see a busybox bug, no
matter how big, small, stupid, or ugly, I want to know about it ASAP and
I'll have a fix for it checked into CVS within hours of your report.

I am being _paid_ by Lineo to work on busybox, and IMNSHO the new
busybox (mod bugs) is much better -- particularly in the context of the
rescue disk. It is smaller (for the same feature set), most all tools
now act very much like their full GNU counterparts, option parsing has
been made much better, and IMHO the code is generally much cleaner.

Please let me know what is broken.


Erik B. Andersen   Web:    http://www.xmission.com/~andersen/ 
                   email:  andersee@debian.org
--This message was written using 73% post-consumer electrons--

Reply to: