On 03/22/2013 10:11:52 AM, Georgi Chorbadzhiyski wrote:
Around 03/22/2013 04:47 PM, Ilya Kuzmich scribbled:
> Implements -t -p ans -s. Same functionality as busybox logger sans
> numeric facility. Do we really need this?
>
> There is one sublte corner case though:
> logger "", logger "" "", or even logger </dev/null
> gnu silently ignores all of them
> busybox logs logger "" and logger "" "", but ignores logger
</dev/null
> this toy logs them all
I think toybox already have that.
http://lists.landley.net/pipermail/toybox-landley.net/2012-April/000422.html
Ok, re-triaging this. (Sorry if I'm a bit woozy right now, been sick
all week.)
The strings and mdev stuff are moot, mke2fs needs a restart. (I should
also move toysh.c and login.c into pending, they need work.)
Your logger is cleaner, but I mentioned the issues with both last
message.
What were the license terms on the test.c you derived yours from? It
looks like a lot of people touched it over the years.
The two portability ones... Hmmm, I'm torn. I gave a talk at CELF last
week about my plans going forward:
http://www.youtube.com/watch?v=SGmtP5Lg_t0
http://landley.net/talks/celf-2013.txt
And it boils down to "I want to support android via musl, not sure how
much I care about OSX at all at this point". I have a test environment
for glibc, uClibc, and musl. I can get a test environment for bionic
(and should prioritize that).
I can't get a mac test environment easily because it's proprietary.
Even if I went into a store and bought a copy (which I haven't),
running it under qemu is fiddly due to hardware anti-circumvention crap:
http://lists.gnu.org/archive/html/qemu-devel/2008-02/msg00444.html
(It's basically the way everybody knows decss, everybody knows the blue
ray key leaked, but there's still enough looming legal threat to
prevent upstream packages from cleanly integrating this knowledge.)
Rob
_______________________________________________
Toybox mailing list
Toybox@lists.landley.net
http://lists.landley.net/listinfo.cgi/toybox-landley.net
