Donald J. Ankney wrote:
If you're working for an employer where cost (both initial and TCO) are
not part of the solution criteria, are they hiring?
Well, in all fairness to this statement, I have an unfair advantage. I
own both business I operate, so I make the choice and live with the
consequences of the choices I make. So, you bet that I pick what I fell
is the best for the job, I try anyway and if that cost more money, so be
it! I value my time in sleeping and peace of mind!
But as far as the subject of hiring is concern, when someone good cross
my path, I always react to it. Rare that it happen, but when it does I do.
However, I am for sure looking to find a person(s) that will enjoy
building in an OpenBSD way ONLY under the BSD license a complete hosted
PBX solutions to replace that platform I am using now. I thought I pick
the best one, but it's all the same in the end. You get stuck in lock in
and screw over by the companies anyway. This person can be either full
time, part time, work from home, as an ahoc of their own job, I really
don't care about that. I am very logical and practical men. I care about
the end solutions and the quality of it. How we get there is totally
irrelevant to me, but I will get there! If interested, or anyone
interested, this can be taken off list. I never maid it a secret here,
but never really posted a job requirements if you like because I think
it wasn't appropriate may be! But as you asked, well here is the answer
to that question.
So, Yes, I am looking for long term on that, start from the ground and
stay with it and expand it after the fact and enjoy the freedom it may
provide in the future to continue contributions to the OpenBSD project
in anyway possible.
I think we're approaching things from very different positions. To me,
an operating system doesn't provide solutions. It's the platform on
which solutions are implemented. Judging from your examples, your job is
focused far more on switching and routing than mine is. OpenBSD does
ship with a fairly complete toolkit for those tasks.
One business I have is an ISP, so yes that a fair statement, the other
is a web design firm with heavy traffic and database as well. But it may
not be as different as you think however. I agree with you as far as the
OS is concern. To me, it needs to be rock solid to run what you may want
to run on it. Example, you saw me talking about Cisco for example. Well
their call manager solutions a few years ago when I was looking at
various solutions was running on NT4 and required you to run NT4 for
their solutions. I went to a demo, but as soon as I saw the engineer
turning on his monitor and logging in his call manager management
system, I asked a simple question and only one to him. He was from
Cisco. The question was simple. Is your system required Microsoft NT 4
to run your call manager PBX systems and the answer was yes. I walk out
of the room and that was it for me. Later on I found that that it
doesn't support virtual hosting PBX anyway, so it wouldn't have worked
never the less, but the bottom line here is that I need something stable
and Microsoft wasn't it period!
So, the platform OS is the start, pick a good one, then you are half way
there. Then there is the more challenging one that you may not be able
to run what you may want on it. Not that it doesn't run I grant you
that. But does it run well however, that's important.
Just like the MySQL ProgreSQL discussions going on here. MySQL use
treads, ProgreSQL doesn't, so on OpenBSD, until the rtreads is complete,
it's more likely that ProgreSQL run better then MySQL, does it mean you
can't use MySQL, no, but it depends on your requirements. I use MySQL
and I am very happy. I had to do tuning to make it work properly
however, but it sure fit my needs. However, I am considering seriously
giving a try to ProgreSQL. Is it because I have problem with MySQL, no,
just that it progress so well in the last 7 years, that may be it's time
I give it an other run in all fairness. It's not what it used to be when
I was running MySQL 3.22.x many years ago.
I don't think we are that far apart.
The main difference might be that you are force to run some applications
because the users wants that, oppose to me where I look at the choice of
applications that does about the same things and I pick witch one I
think after testing works best for the task at hand and then tell the
users, that's what they will have to use and get use to it! I value
their input, but in the end, we will not run three different version of
similar things, but one. Can we switch in the future, sure if all the
justifications are there and it improve the security and stability.
I have to give you a win however in the case where yours will switch to
something that look better may be. They want it because it's cool. I am
sure you have to deal with that. I don't! That's not a valid
requirements for me. The look doesn't make your job easier or safer, or
doesn't keep the system more stable. Usually, and I generalize here a
bit, cooler it looks, more bugs and holes you have in it and more often
the designer pay attentions to the look for the users instead of how it
is done inside.
I'm a systems administrator, so my outlook is toward data
access/storage/security and end-user experience. An OS shouldn't ship
with those sorts of tools -- if I wanted a that sort of mess, I'd use
RedHat.
Agree. Some tools however, like PF, essential to security makes sense to
be part of the kernel however, so shipping it with it makes it better
and faster and it does make sense here! But yes, an OS should have
everything under the sun with it!
I'm not looking of OS or hardware-level support. When I implement a
solution, it either needs to be simple enough to debug myself if I find
problems or I need to have a mechanism to report bugs to the developers
with a reasonable expectation that they will be fixed in the source. The
latter is especially critical when the only solution I can find is a
closed-source solution. I'd rather not use closed-source ever, but
sometimes that's just how the cards come up.
No argument from me on that.
What you call features, I call end-user requirements.
Well.... On that I don't agree, but that's OK.
Not all tasks have the same criteria. It's not changing the playing
field, it's evaluating each task/job as it comes and setting the
solution criteria appropriately. My bottom line can't be quantitatively
measured by network efficiency. I have 5 subnets full of end-users
sitting at Windows and Mac workstations trying to do whatever it is they
do. My bottom line is how well I can build/manage/design services that
meet their needs. Again, I think we have very different jobs.
Not that different, but different approaches for sure. I am not trying
to let users use anything under the sun that they may like if that's
what you mean. So, if that's the case, then yes it's different. But I
have to ware many hats and trust me, it's not only routing by a very
long shoot!
Regards,
Daniel
PS: I only reply to this one because of the asking for the job part. (:>
Just in case someone point out to me I said I wouldn't reply anymore.
Sorry and yes I shut up now.
