lkml.org 
[lkml]   [1997]   [Jul]   [18]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: Take A deep Breath
Date
In article <linux.kernel.Pine.LNX.3.95.970717095932.13756A-100000@pc7537.hil.siemens.at>,
Ingo Molnar <mingo@pc7537.hil.siemens.at> wrote:

>to make this more clear: for example David S. Miller implemented the core
>of the new networking stuff (have you seen those K6 lmbench numbers posted
>recently to linux-kernel?) in 1 day (after lots of heavy thinking). I say
>such an effort is plain impossible and unaffordable for any commercial
>entity, with strict source control, style and documentation rules.

I'm a commercial programmer, working for a fairly large company
(McAfee associates); I've been forced to do quick turn-around major
patches to various things, ranging from rewriting firewall proxies to
going through the 2.0.28 kernel and taking the braindamage out of
all the isa net modules. In, yes, one day; the documentation came
later.

>Same applies to the new dentry stuff. Commercial OS vendors invest
>billions of dollars a year into their systems, and Linus implements this
>new revolutionary technique (not present in any OS i'm aware of) which
>also impacts 60% of the core kernel in ... 2 weeks?

And how long will it be before a stable production kernel is made
with the dentry code in it? There is *very little* that is different
between the Linux development cycle and a good development cycle at a
commercial entity, except that the development codebase is published
and everyone and their sister can see the daily reports of kernels
exploding.

(Oh, and the small detail that almost nobody wants to get paid for
doing this. That's probably unique in the OS development world.)

____
david parsons \bi/ Mumble mumble rant mumble.
\/

\
 
 \ /
  Last update: 2005-03-22 13:39    [W:0.047 / U:0.040 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site