Re: [Nolug] Radius & Tacacs+

From: Dennis J Harrison Jr <>
Date: Tue, 20 Jan 2009 14:17:54 -0600
Message-ID: <>

On Tue, Jan 20, 2009 at 1:23 PM, Jeremy (mailing list box)
<> wrote:
> Dennis J Harrison Jr wrote:
>> 'The' Developer?
> Yeah, the Samba team. *rolls eyes* Just like Microsoft is referred to as
> the developer of Windows, Valve as the developer of Steam, ERSI as the
> developer or ArcGIS, the kernel team as the developer of the Linux kernel.
There are over 30 people who have write access to the git repos right
now. Many more write patches and concentrate on different system
within the samba project. One outdated sentence on a wiki is not the
collective opinion of all those people.

>> Anyway, you shouldn't be afraid to self certify the software you
>> use... since at the end of the day, the user is going to look to you,
>> or whoever you train. So if you have the responsibility to repair and
>> progress. Shouldn't you also have the knowledge? I am not saying use
>> every bit of new fangled 'gee wiz' code out there. Just that if you
>> wait for 'The' Developer (lol) to certify code before you even touch
>> it... that you will miss out on a lot of what open source (and
>> hopefully free) can bring to the table.
> Actually, when I have the time to mess with software that interests me, I
> play with it. But I keep software that isn't deemed release quality,
> especially when the developer of the software says it isn't, away from my
> production networks and servers.
But you should have a reason other then some blanket statement.

All Americans are fat and loud. All British have bad teeth. All
Canadians look like Terrence and Philip.

All Alpha software is unfit for use. All production servers have to
run 6 year old code that still requires constant patching.

All blanket statements are true. Free thought is horrible. Let us
keep far far away from it!

> Uptime, stability, disaster recovery and performance are the things I focus
> on in my production environment. I'm not going to put anything I am not
> 100% sure of into the mix that will jeopardize any of those aims.
So what are you 100% sure of?

>> Don't we want to progress and solve new problems instead of rehashing
>> the same old problems over and over? And how can we as developers be
>> expected to do that when the majority of users are just hangers on
>> waiting for that special moment in time when the sun and moon and
>> stars all align and I make it easy for them to use my code to make
>> money without any thought / input of their own beyond following the
>> configuration templates?
> Are you serious? Are you complaining because software users out there are
> not eager to jump up and down, run everything in debug mode and submit
> patches or documentation? Seriously, do you not know the world in which we
> work?
No, I am complaining because your ivory tower mentality should at
least be coupled with the ability to bring forward progress.

> The fact is that it has always been that way and is not like to change now
> that computer technology is available to the masses. If we were in the
> sixties or seventies, when computer users were likely to be admins or
> developers, your expectation might have traction. However, most users
> don't have a clue how any of this computer or internet "stuff" works.
> That's the reason I have the job I do and get paid the money I do - people
> don't care how it works, it just has to work.
Which is why you have a responsibility to make it work 'better'. If
you can do something to improve a situation, you should.

Why would using Samba4 over windows server be considered 'better' to
me? Because ownership of data and portability of that data are
concerns we should be paying attention to. The philosophical
directive of the samba team (and the community which also gets a voice
in direction via personal input to developers) is much more consumer
rights focused then Microsoft can currently afford to be.

>> I also take it personally that you think I would abuse a relationship
>> with anyone (customer or otherwise) to the point of absolute
>> dependence. That would be irresponsible, and something I stay very
>> far away from.
> Having worked with companies that foisted incomplete, beta and, in some
> cases, alpha quality software upon us to "use" or "resell" to the clients (a
> product called Stellent comes to mind from my past), I am highly suspicious
> of anyone who uses beta or alpha quality products in production
> environments, especially when they charged by the hour or incident.
Unless you have ANY SHRED of proof towards me foisting anything to
anyone, then keep your libel to yourself.

What determines the difference between alpha / beta / and stable?

Do you have a clear cut definition that everyone prescribes to?

Does test driven development help us draw better lines or does it
obfuscate the process?

The only Samba4 test that is still failing on a typical debuntu
buildout is python-ldb support.

You need to get some facts to back up your theories. Such as:

A list of failures related to Samba4

or a list of features that are not implemented currently that you need

Something in between, or along those lines. Otherwise, what are you
really arguing here? That the word 'Alpha' means 'buy windows' ?

Bring some meat into this, bring in some facts... Something other then
conjecture. I don't know how to fix a problem you won't define.

If you are saying that Samba4 doesn't work as described, you're wrong.

> J
> ___________________
> Nolug mailing list
Nolug mailing list
Received on 01/20/09

This archive was generated by hypermail 2.2.0 : 02/17/09 EST