[nSLUG] Article on why monoculture gets picked

D G Teed donald.teed at gmail.com
Wed Apr 25 16:11:21 ADT 2007


On 4/25/07, Stephen Gregory <nslug at kernelpanic.ca> wrote:

> That is utter foolishness. There are plenty of reasons to use
> something other then Apache. Lighttpd is lighter weight and easier to
> configure as it doesn't have all of the features of Apache. The ROOT
> DNS servers use a mix of software to prevent a single vulnerability
> from taking down the entire Internet. I use dnsmasq because it is
> considerably smaller then bind and fits on a embedded system.


I'm curious, why would an embedded system use a DNS server?

There are some times when an alternative emerges that is worthwhile
or fits a niche requirement.  But in general, I feel there are too many
options and too many resources spread thin.  Ian Campbell mentioned
issues with stability and developers of lighttpd - this would be
an example of the pitfalls of using software from small nucleus
projects.

There are niche requirements which have spawned the development
of software such as busybox, but I question whether we need
everyone's shot at making lightweight software for no other reason
than to try and make something simple or leaner.  In many cases,
the feature requests drive the thing back to the bloat they wanted to
avoid.  Projects like Firefox are a demonstration of that Honda Accord
kind of transformation.

Really, I'll never evaluate all 800+ Linux distros, nor try every type of
shell, etc.
so it is pointless to have the extremes of variety.  However, I might look
at a few imap server options or content management options.

What this article is about, is asking people to stem off the
desire to create something you named after yourself or your favorite
author.  That was in vogue 10 years ago, but it isn't so cool anymore.
Developers, whenever possible, should try to work with an
existing project, if possible, to enhance it as required.

Is that so outlandish?

If the reason for working on this stuff is to have fun, then
isn't it more interesting if it is really useful as well?  That's
what I'd like to see happening - and it should be possible
with less of the "I'll do it my way" and more community.

The quote from the article is wrong. Forking over a disagreement in a
> license is why we have Xorg and  OpenSSH. In the case of Xorg the
> fork was absolutely necessary as XFree86 was no longer free.
>

There are some times when license changes do create a real need, but
at other times the quabble is about apache license versus BSD versus
LGPL, etc.

Some people feel strongly about the shades of grey between various open
source style licenses.  I don't, and while there are real reasons to be
concerned
with the legal questions and their political use and abuse, I'd rather
focus on the energy and time put into making installers with recent kernels
and stuff like that.

--Donald


!DSPAM:462fa7dd72505005216706!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nslug.ns.ca/mailman/private/nslug/attachments/20070425/5b5f7eda/attachment-0001.html>


More information about the nSLUG mailing list