[Charlug] posscon

Dennis Clark boomfish at gmail.com
Wed Apr 22 13:54:02 EDT 2009 

Software has become more like a commodity than an asset, and businesses
ought to learn to treat it as such. The "complete" solutions offered by
vendors usually require you to pay for much more functionality than you
actually need, which then often requires you to buy bigger hardware to
support the software bloat.

Businesses are usually better off using expertise (either in-house or
external experts) to integrate best-of-breed (open-source) technologies to
create solutions that actually fit the business needs. The external experts
in this case are mostly for evaluating technologies that are either still
emerging or have not yet been integrated into the organization. I don't see
much of a need for paying expensive consultants for things like web servers
and application servers, but it may be worth doing so for technologies like
ActiveMQ and ServiceMix. It can be good to have expert on-call support
available for such technologies at least until the technology matures or you
build enough in-house expertise to support it yourself.

One sticking point I see business managers may have with keeping expertise
in-house is knowledge retention. Businesses like to keep their IT staff
slim, but that makes them vulnerable to knowledge loss if a key employee
leaves. External experts can help in this regard by providing training
services for newer staff. On the other hand, my experience is that
businesses tend to underestimate the value of IT-related business process
knowledge: if such knowledge leaves before it is captured, it has to be
relearned (often by trial-and-painful-error).

I see SaaS as an emerging technology to enable further outsourcing of IT
from an organization, and like any form of outsourcing its success depends
on how tied the business processes being outsourced are tied to the
bussiness's core competencies.

I'm sure we can all agree that the days of "money is not an issue" are over.
My hope that this does not result in a blind adoption of open-source simply
as a cost-cutting measure without looking into the actual business needs;
unfortunately I'm already seeing signs that this might happen in my own
workplace.

Cheers,

Dennis


On Mon, Apr 20, 2009 at 3:09 PM, Bob Evans <bobevans19 at gmail.com> wrote:

> Interesting points, Dennis.
>
> In the old model, I would agree that there was a need to rely on an
> "expert". I had many years of dealing with IBM, for example, on what is n=
ow
> referred to as their 'enterprise server' or zOS machines. (We still call
> them mainframes when talking amongst ourselves...)
>
> The IBM mainframe model is a perfect example (I think) of the 20th century
> strategy: proprietary hardware running proprietary software. Got a proble=
m?
> Turn around. Your IBM CE (Customer Engineer) is standing right behind you,
> ready to sell you a software or hardware solution to your dilemma. I put =
up
> with this all through the eighties and most of the nineties.
>
> At some point in there, Microsoft got into the act. They aped IBM, even
> though they touted themselves as the "new guys" in corporate America. But
> essentially they played the same game: upgrade to our new release of
> Windows, or Office or Microsoft Bob, and all will be well.
>
> The joke for years, even though it wasn't that funny, was that no manager
> in the mainframe world ever got fired for recommending an IBM solution. S=
ame
> thing held true later, when it was often noted that no manager ever got
> fired for recommending a Microsoft solution. After all, money was never an
> issue.
>
> Things changed (I think) for the better around 1995.
>
> That's when free software, or Open Source, if you must, began to make
> inroads. First on the net: Linux servers, running Apache & MySQL, Perl, e=
tc,
> made sense on many many levels. Cost was a factor, but software superiori=
ty
> was a much bigger factor. (My opinion, but I've heard it repeated enough =
to
> think it might have merit...)
>
> Today? Who knows. I hear a lot about software as a service. Gotta believe
> that has potential. People really don't need it, yet, so it might not take
> off. But if it does... Bye bye to a lot of current players.
>
> As to the point about relying on a vendor, I find that the quality of the
> employees in house usually out weighs the merits of buying that expertise.
> But many managers won't go along with me on that, because today they
> actually believe that the Red Hats & Oracles and Novells of this world wi=
ll
> answer the phone when they call. After all, Microsoft & IBM would answer =
if
> you had any reason to call them...
>
>
>
> We shall see...
>
> Stay tuned.
>
>
>
>
>
>
>
>
> On Mon, Apr 20, 2009 at 2:31 PM, Dennis Clark <boomfish at gmail.com> wrote:
>
>> (I've trimmed the Ccs because I'm sure the RedHat folks get enough email
>> as it is without me adding to it).
>>
>> I agree that the OS is becoming less important, but my reasons differ fr=
om
>> yours. Many key applications these days are online applications, the Web
>> browser is the one desktop technology that most of those applications re=
ly
>> on. With the standard Web browser technology (HTML and Javascript) being=
 so
>> limited, we're in the midst of a new browser war ... except now there ar=
e 5
>> big players: Microsoft, Mozilla, Apple, Google, and Opera. (Some of you =
may
>> question my addition of Opera, but it's a big player in most of the embe=
dded
>> device market). With most of these browsers able to run on multple
>> platforms, browser support becomes a much bigger and more important issue
>> than desktop OS support for online applications.
>>
>> As an Adobe ColdFusion developer I hear a lot about Adobe's Flash Player
>> and AIR client. Although they are not open-source, they have been ported=
 to
>> all the modern desktop OSs and browsers and provide a much richer front-=
end
>> experience than what is possible with standard browser technology. The
>> technologies for server back-end connectivity for Flash Player and AIR a=
re
>> open source(*) so you can connect such clients to just about anything. T=
his
>> means that you can build server applications for Flash/AIR using open-so=
urce
>> and provide rich front-end interfaces for them without having to worry a=
bout
>> the desktop OS or even the browser type and version (beyond the Flash Pl=
ayer
>> requirement, that is).
>>
>> As for 1-800 support, I see the role of "open-source support" as changing
>> from general help lines to a fast line to technical expertise. In fact I
>> have a conference call scheduled this week with Progressive Software (aka
>> FUSE), a company that provides technical support for enterprise Apache
>> technologies such as ActiveMQ. ActiveMQ is free and I have been able to
>> install it myself, but we may tap the FUSE team for architecture and des=
ign
>> advice.
>>
>> You may mock the concept of paid support, but from an IT management
>> viewpoint it's very risky to rely completely on an unpaid open-source
>> community for supporting your mission-critical applications that result =
in
>> lost business and revenue for every minute of downtime. With paid
>> open-source support you not paying for the rights to use a software prod=
uct
>> or even for the ability to call someone who can read a help desk script.
>> Rather, you are paying for a guarantee that you will get the prompt
>> attention of someone who knows much more about the technology than you do
>> during those times you really need it.
>>
>> (*) There are a couple of data management features in Adobe LiveCycle th=
at
>> are not available in its open-source versions, but there are already
>> projects underway to build replacement open-source components that imple=
ment
>> those missing features.
>>
>> -- Dennis
>>
>>
>> On Mon, Apr 20, 2009 at 11:26 AM, Bob Evans <bobevans19 at gmail.com> wrote:
>>
>>> No argument from me. I just don't think it is at the OS level anymore. =
OS
>>> is becoming a non issue. Just as Microsoft is falling out of the pictur=
e,
>>> like IBM before them, I don't see Red Hat as more than a footnote. The =
model
>>> of "corporate run software company" is less and less viable.
>>>
>>> There are fewer & fewer reasons to call 1-800-redhat (or whatever). And
>>> having that phone number is their only real reason to be, in the corpor=
ate
>>> venue. If I know as much (or more) about not only Linux, but Red Hat Li=
nux
>>> specifically, why do I need them?
>>>
>>> As we move forward, now with a generation of younger people who grew up
>>> with a somewhat mature open source world, the concept of relying on a
>>> "vendor" of software that I can retrieve online just as easily as they =
can
>>> becomes somewhat absurd...
>>>
>>> Stay tuned. Things could change. They often do.
>>>
>>> -be
>>>
>>> On Mon, Apr 20, 2009 at 11:09 AM, Tim Jowers <timjowers at gmail.com>wrote:
>>>
>>>> Sorry I could not make the event. I think whoever organized the event
>>>> gets it. http://www.posscon.org/ I'll try to find the presentations
>>>> archives.
>>>>
>>>> Many Linux/Open Source people do not get it. Businesses who use Open
>>>> Source have a competitive edge. Tech businesses who use open source
>>>> win because they "build on the work of others". My point is not "will
>>>> open source win", but "open source has won". E.g. RedHat may be seen
>>>> as boring now because they already have won. The G1 phone runs Linux.
>>>> Linux is the OS of choice going forward. The battle for the
>>>> programming language of choice rages. Maybe it will be Java, Ruby, or
>>>> Mono. Which is open source?
>>>>
>>>> TimJowers
>>>>
>>>> On Mon, Apr 20, 2009 at 8:42 AM, Leam Hall <leam at reuel.net> wrote:
>>>> > Yeah, but Nalley crashed the party so he doesn't really count. :)
>>>> >
>>>> > If you're talking about business in Linux, Red Hat is it. No one else
>>>> really
>>>> > competes well in the datacenter market. I'm pretty solid in my
>>>> conviction
>>>> > that the OS will be a major player for 3-5 years, however, I do agree
>>>> that
>>>> > past that point something else will have the option of causing big
>>>> change.
>>>> > My current guess is the growth of application blades  or 1U boxes
>>>> where the
>>>> > OS is re-built to support a finite set of functions. Cisco moving in
>>>> to the
>>>> > server space is a good possibility.
>>>> >
>>>> > In that space I think Linux will win because of brain-share. *BSD has
>>>> great
>>>> > value but the business support isn't behind it as much. Whether or n=
ot
>>>> RH is
>>>> > the vendor of choice in 2015 remains to be seen. I think it's possib=
le
>>>> but
>>>> > there's too many smart people working on the problem for one place to
>>>> > guarantee being the de facto winner. I'd give it 40% for RH, 40% for=
 a
>>>> > technology that exists today but is still in infancy, and maybe 20%
>>>> for
>>>> > something not yet developed past the skunkworks stage.
>>>> >
>>>> > On the zLinux issue, the problem is companies have these huge boxes
>>>> they
>>>> > can't really write off but don't have the skills to fully utilize.
>>>> I've done
>>>> > some work with zLinux that required me understand a little about the
>>>> > mainframe. I was blown away! There are some really cool things about
>>>> the
>>>> >  mainframes that make them totally interesting.
>>>> >
>>>> > The issue with zLinux is that you're using a VM and lose so many par=
ts
>>>> of
>>>> > the frame that are useful. The mindsets are highly divergent and
>>>> unless
>>>> > there's a real and over-whelming business case I feel zLinux is a
>>>>  solution
>>>> > looking for a problem.
>>>> >
>>>> > That's my thinking, anyway, and what I base my career choices on.  :)
>>>> >
>>>> > Leam
>>>> >
>>>> >
>>>> > Bob Evans wrote:
>>>> >>
>>>> >> Well, certainly the Fedora/Red Hat advocates(3)  outnumbered any
>>>> other
>>>> >> Linux
>>>> >> advocates(0) in the speaker category. The closest in speaker count
>>>> numbers
>>>> >> were employees of Blue Cross in Columbia, touting their zSeries
>>>> >> implementation of Linux.
>>>> >>
>>>> >> (zSeries =3D IBM mainframe)
>>>> >>
>>>> >> Some of you might remember IBM's demo of that little trick several
>>>> years
>>>> >> ago
>>>> >> at a LUG meeting at CompUSA on South Blvd. We tried it at BOFA,
>>>> too...
>>>> >>
>>>> >> So the nineties have officially arrived at Blue Cross. Good to
>>>> know...
>>>> >>
>>>> >> Regarding your other comment, about Red Hat in 3-5...
>>>> >>
>>>> >> The likelihood of there being a discrete OS, other than in firmware,
>>>> in
>>>> >> 3-5
>>>> >> years seems highly unlikely.
>>>> >>
>>>> >> But we'll see.
>>>> >>
>>>> >>
>>>> >> On Mon, Apr 20, 2009 at 7:51 AM, Leam Hall <leam at reuel.net> wrote:
>>>> >>
>>>> >>> Not sure how you see it a Red Hat event. Greg DK pretty much did
>>>> what he
>>>> >>> did at BarCamp Charlotte, and the one Keynote by the RH CIO wasn't
>>>> enough
>>>> >>> to
>>>> >>> get you going.
>>>> >>>
>>>> >>> It was a Collegiate event for people who are new to or maybe even =
on
>>>> the
>>>> >>> fence about Open Source. Red Hat was just a bit player.
>>>> >>>
>>>> >>> I think RH will be a major enterprise player for at least 3-5 years
>>>> and
>>>> >>> eliminate most of the commercial Unices.
>>>> >>>
>>>> >>> More later, breakfast calls...
>>>> >>>
>>>> >>> Leam
>>>> >>>
>>>> >>> Bob Evans wrote:
>>>> >>>
>>>> >>>> I actually talked myself out of going, opting for family time. I
>>>> was IN
>>>> >>>> THE
>>>> >>>> PARKING LOT when I realized that this was not something I really
>>>> care
>>>> >>>> about
>>>> >>>> any more.....
>>>> >>>>
>>>> >>>> Looking over the sponsors,it was essentially a Red Hat event.
>>>> >>>>
>>>> >>>> I'm not anti-Red Hat. I just don't see them being the future of
>>>> >>>> software,
>>>> >>>> open or otherwise. They've certainly had a degree of success
>>>> courtesy of
>>>> >>>> the
>>>> >>>> open source world, so I guess they feel obligated to hold these
>>>> >>>> things...
>>>> >>>>
>>>> >>>> It WAS a nice drive, though. And a beautiful day...
>>>> >>>>
>>>> >>>> -be
>>>> >>>>
>>>> >>>>
>>>> >>>>
>>>> >>>>
>>>> >>>>
>>>> >>>> On Mon, Apr 20, 2009 at 6:57 AM, Leam Hall <leam at reuel.net> wrote:
>>>> >>>>
>>>> >>>>  Well...
>>>> >>>>>
>>>> >>>>> After being slightly numbed by all the "thank yous" up front I
>>>> started
>>>> >>>>> to
>>>> >>>>> nod during the first keynote. The business track I attended in t=
he
>>>> AM
>>>> >>>>> was
>>>> >>>>> aimed at "why is Open Source okay". True material but not what I
>>>> had
>>>> >>>>> been
>>>> >>>>> hoping for.
>>>> >>>>>
>>>> >>>>> Spent a little time handing out swag for Fedora, answering
>>>> questions
>>>> >>>>> and
>>>> >>>>> doing a little bit of encouraging. Joined in a conversation
>>>> helping a
>>>> >>>>> newbie
>>>> >>>>> over lunch, and that was about it. Hopefully the developers track
>>>> was
>>>> >>>>> more
>>>> >>>>> useful to the attendees. After musing, munching, and going over
>>>> the
>>>> >>>>> remaining schedule I put time with family over time listening to
>>>> stuff
>>>> >>>>> I
>>>> >>>>> already knew.
>>>> >>>>>
>>>> >>>>> I'd love to hear from businesses that are significantly open
>>>> source and
>>>> >>>>> how
>>>> >>>>> they integrate everything together. After so many years in the
>>>> >>>>> enterprise
>>>> >>>>> level game and seeing so many millions of dollars wasted on tools
>>>> that
>>>> >>>>> don't
>>>> >>>>> interoperate I'd like to see someone speak about real business u=
se
>>>> of
>>>> >>>>> open
>>>> >>>>> source and real cost savings. One speaker seemed happy that the
>>>> use of
>>>> >>>>> a
>>>> >>>>> dozen or so open source tools saved $100,000. Great! I've doubled
>>>> that
>>>> >>>>> with
>>>> >>>>> a few lines of Bourne shell in just the right spot.
>>>> >>>>>
>>>> >>>>> Hopefully the June conference is better.
>>>> >>>>>
>>>> >>>>> Leam
>>>> >>>>>
>>>> >>>>> Bob Evans wrote:
>>>> >>>>>
>>>> >>>>>  So... did anyone go to this thing?
>>>> >>>>>>
>>>> >>>>>>  _______________________________________________
>>>> >>>>>>
>>>> >>>>> CharLUG mailing list
>>>> >>>>> CharLUG at charlug.org
>>>> >>>>> http://charlug.org/cgi-bin/mailman/listinfo/charlug
>>>> >>>>>
>>>> >>
>>>> >>
>>>> >>
>>>> ----------------------------------------------------------------------=
--
>>>> >>
>>>> >> _______________________________________________
>>>> >> CharLUG mailing list
>>>> >> CharLUG at charlug.org
>>>> >> http://charlug.org/cgi-bin/mailman/listinfo/charlug
>>>> >
>>>> > _______________________________________________
>>>> > CharLUG mailing list
>>>> > CharLUG at charlug.org
>>>> > http://charlug.org/cgi-bin/mailman/listinfo/charlug
>>>> >
>>>> _______________________________________________
>>>> CharLUG mailing list
>>>> CharLUG at charlug.org
>>>> http://charlug.org/cgi-bin/mailman/listinfo/charlug
>>>>
>>>
>>>
>>> _______________________________________________
>>> CharLUG mailing list
>>> CharLUG at charlug.org
>>> http://charlug.org/cgi-bin/mailman/listinfo/charlug
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://charlug.org/pipermail/charlug/attachments/20090422/b2287101/att=
achment-0001.htm

More information about the CharLUG mailing list