Damien Miller wrote:
> On Thu, 22 Jun 2006, Brad Davis wrote:
>
>
>>For those of us who use OpenBSD to create "semi-embedded" networked systems,
>>adding awk or sed or grep or cut or tr to an image that must include ifconfig
>>(or a stripped down ifconfig) isn't always acceptable. "Disk" space isn't
>>always free.
>
>
> please.
>
> [djm@fuyu djm]$ ls -l `which sed`
> -r-xr-xr-x 1 root bin 26596 Jun 7 22:47 /usr/bin/sed
>
> If your embedded image is so impoverished that you can't fit sed, then
> it is unlikely that you will be able to write meaningful shellscripts
> anyway.
>
And if you can't write and add to your image a simple C program to solve your issue then how in the world can you create a bootable image?

OpenBSD is a standard kit, I don't expect it to suit my needs, but it's standard, and I can make it fit my needs.
If the management goes ballistic, which happens now and then, my latest stable copy is still stable.
I'll base all my local releases on the latest release _I've_ decided as 'stable'.

I don't trust Theo for one cent, _I_ myself decide on what _I_ call stable.
So far, in the end, Theo's been right, but it's still down to _my_ call.

So basicly, fark off.
The team delivers a system that is secure while being standard, if you want special functions then write 'em, publish 'em and use 'em, and if others call you stupid then the most likely reason is that you're stupid, so eat shit and die.

There's nothing wrong in publishing changes, but if you publish silly changes you should be prepared for 'silly' answers.
--
// hdw