Oops, posted empty early heh - sorry, BUT!
Go to
http://blogs.computerworld.com/14576/who_writes_linux_big_business?page=1and
scroll down to a Comment submitted by Michael Geiser
<snip>
- It looks like many developers are also reviewers. I'd be VERY concerned if
the developers are allowed to review and approve their own work products
- Page 13 indicates that authors are allowed to approve their own work
product ("An interesting (if approximate) view of kernel development can be
had by looking at signoff lines, and, in particular, at signoff lines added
by developers who are not the original authors of the patches in question.")
It looks like 25% of the kernel code is signed off by the developer who
checks it in.
Remember back in Nov 2003 when a BitKeeper to CVS gateway was in an attempt
to add a root exploit back door to the Linux kernel? A patch in the file
kernel/exit.c was attributed to David Miller (who sign offs on about 10% of
the kernel changes, BTW). That just seems like a bad idea.
- The velocity of change and the number of contributors in just the kernel
is mind-numbing. The report authors even note this on page 15 ("The huge
rate of change and number of individual contributors"). Looking at 2.6.30; I
see 81 days between release dates and 11,989 "patches" in that time.
I have to think regression testing that time frame can't have been too
comprehensive.
</snip>
Back on topic though, I am saddened by the way managers cut their cloths to
their employers but it's not exactly surprising in today's world. The BBC
used to be able to resist these people by having jobs for life but not no
more.
Regards,
Nico Morrison
__________________________
[email protected]
Skype: nicomorrison
http://nicomorrison.com
http://theflowerraj.com
__________________________
2009/8/20 Nico Morrison <[email protected]>
>
> Regards,
> Nico Morrison
> __________________________
> [email protected]
> Mob: 07875 596 316
> Skype: nicomorrison
> http://nicomorrison.com
> __________________________
>
>
>
> 2009/8/20 Tim Dobson <[email protected]>
>
> Nico Morrison wrote:
>>
>>> I LIKE the idea of people writing the Linux kernel code outside of
>>> company hours.
>>>
>>
>> Does anyone do this?
>>
>> From What I can see here:
>> www.cs.tut.fi/~tta/demography.pdf<http://www.cs.tut.fi/%7Etta/demography.pdf>(Specifically
>> sections 5 and 6) a few must, but a considerable number appear
>> to be *large* companies.
>>
>> I am appalled that a coder can approve his or her own patch. Interesting
>>> stuff.
>>>
>>
>> Er well in kernel development they can't. They can run it on their own
>> system etc and pass it around, but to get it into the mainline kernel there
>> are processes and QA proceedures the kernel project has put in place.
>>
>> Tim
>>
>>
>> -
>> Sent via the backstage.bbc.co.uk discussion group. To unsubscribe,
>> please visit
>> http://backstage.bbc.co.uk/archives/2005/01/mailing_list.html.
>> Unofficial list archive:
>> http://www.mail-archive.com/[email protected]/
>>
>
>
