I have a possible opportunity to start a software project from
scratch, so I have been again considering which DBMS, which
language(s), which toolsets, etc. in order to get ballpark figures for
two different scenarios. Without giving you any of the requirements
except a need for 24/7 web access and reliability and excellent
scalability, which two of these would you be most inclined to choose
and why?

UniData
Cache' (I tried Tony's suggestion of Alt-130 and it doesn't work in
the google groups editor as far as I can tell)
OpenQM
SQL Server
Oracle
DB2

If and only if you cannot chose any of those, let me know which other
you would choose. Thanks. --dawn

dawn wrote:
UniData.

If for no other reason than WYKIWYL.
And it's easier to spell than Caché.

Any reason you left UniVerse off the list?

--
frosty

Dawn,

If I was in your shoes, I would probably go with Oracle or SQL
Server. The main reason why I go this route is because there are a
helluva' lot more job opportunities with Oracle or SQL Server. (Go to
sites like Monster.com or CareerBuilder and compare the number of
opportunities between Oracle, SQL Server and UniData.) If you need
to hire programmers, analyst or administrators, you will definitely
get more "hits" in your hiring search.

<The following is one man's opinion. I'm stepping up on the
soapbox.>

I have been programming in the PICK arena for over 15 years and I'm
seeing the job pool getting smaller and smaller every year. At this
time, there are more opportunities in the Oracle and Microsoft
worlds. Plus, I have very disappointed with IBM's lack of marketing
of the U2 family. They have a good product but (again, my opinion)
have done little or nothing to promote it.

<I am now stepping down from the soapbox.>

I cannot offer an opinion DB2, OpenQM or Cache' since I have never
used them.

That's my $0.00000002 worth. Good Luck in your search!

(Good Luck...Good Hunting!)




dawn wrote:

On Feb 8, 12:22 pm, "frosty" <fros... (AT) bogus (DOT) tld> wrote:
I'm guessing -- What You Know Is Why You Like?

Yes, but Jim's recent tip helps.

Yes, I know I will not chose UniVerse because I've worked with UniData
in one capacity or another since about '90, so there is no reason I
can come up with for going with UniVerse in said. Not only WYKIWYL,
but TDYKIBTTDYD

cheers! --dawn

On Feb 8, 12:25 pm, "Boicee" <gbo... (AT) matheson-trigas (DOT) com> wrote:
Since I don't know you, I'm not sure if you are aware that my shoe
closet includes such things as
http://www.tincat-group.com/mewsings...codd-dead.html

I don't sweat the job opportunity thing and I already know (quite a
bit of) SQL. Think of this as me using my own dollars to write and
maintain the software. If it takes fewer dollars each year on the
whole (hardware, software, peopleware) to get a highly usable,
realiable, maintainable, scalable, secure, fast solution with Oracle
or SQL Server, then that would be a reason for me to go with such a
database.

Oh, I'm well aware. There is a wave a-comin' however, that will start
moving us away from SQL (as these vendors already know).

Yes, but if (a big "if") I wanted "angel developers" (like angel
investors except they give hours instead of dollars in order to
participate in profit-sharing), then I might find more takers where
people already know me, right?

Good, that's what I'm lookin' for.

I have to agree with that. I think there are some steps in the right
direction, but it is a massive ship to move, rather than the seemingly
more agile Intersystems company.

That has never stopped m..., I mean, I understand. ;-)

Thanks for your input, much appreciated. --dawn

Obviously, QM and DesignBais

"dawn" <dawnwolthuis (AT) gmail (DOT) com> wrote

On Feb 8, 9:35 am, "dawn" <dawnwolth... (AT) gmail (DOT) com> wrote:

Too obscure, even in the MV market. It's still the industry's best
kept secret. I admire IBM for continued enhancements and keeping the
software alive but it seems like U2 is always in maintenance mode and
the company doesn't really know what it's going to do with it. IBM
has had a long time to properly market U2 outside the confines of the
MV space, and they haven't done so. IBM is trying to get people to
migrate to U2 just because they're IBM. For many companies that line
works but for others it's just not good enough. Companies that buy
from IBM because it's a big name aren't getting the bang for their
buck because IBM still hasn't recognized U2 as a first class citizen
in their DB2 family. Until that happens U2 is just another technology
that's subject to getting deprecated if the economy takes a downturn.
All that said, it's a safe bet if you are intent on staying in the MV
space.

Cach Alt (don't let go) numeric pad 130, let go Alt é ... works for
me using Google Groups interface.
This platform and InterSystems, company behind it, are quickly coming
up on my list. Definitely the business choice for many reasons.
Technically the non-MV side of the core is still a bit cryptic and,
like we have our P word heritage, they have a small M word stygma.
But all of the add-ons like Java, SQL, XML, etc hit all of the
standards, and it's all much easier to use than the bolt-on solutions
offered by the current MV vendors.

Good software but general perception is that it's still "Martin Inc".
That's not a criticism, it's just the way it is. I have the same
problem with the perception of Nebula R&D but we're slowly growing
too. "OPENqm" cannot be used for commercial use, so QM is the product
being discussed. That highlights one of the basic problems - product
identity is a core issue for this product and company. People simply
don't understand where it fits. I keep giving Martin the opportunity
to explain to us in business terms why QM isn't just another MV
platform but he keeps missing the marketing opportunities and pitches
OpenQM simply as a migration option. And (more on-topic) that dove-
tails with a lack of larger testimonial sites that will convince the
rest of us that QM is indeed a production quality environment that can
be used in larger installations. When the perception of the business
in the public eye matches the perception of the software by those who
tinker with the open source side, then QM might be a contender. That
said, at this point it's just another MV DBMS, so for a business that
is selling software to a mainstream audience, neither QM nor any of
the other MV choices will be ideal.


There are lots of people out there who know the software, so it's
probably the safest bet of the bunch. Here's another M word that
people need to consider. The solution requested was defined as
"scalable" and "reliable". If someone wants this then they should be
prepared to pay for it and with SQL Server they will certainly pay but
they will also certainly have a platform which _can_ match the
requirements - notice I don't say it _will_ match requirements, that
depends on the implementation.

While the cost is coming down as Oracle attacks the mid-market SMBs
with a vengeance (I wrote a paper a couple years ago saying this was
going to happen...) the cost of DBAs and administration is still in the
pricey area. There's no doubt the software can satisfy your
requirements and that you can get the people to support it. This is
another safe bet but the added security will cost you. If you want to
go cheap or MV you won't get the same business security. The decision
for Oracle is probably more political than technical.

No clue. It's interesting however, that you, I, and anyone I know
will mention SQL Server and Oracle because we know something about
them, but we mention DB2 because ... well, because that's IBM and there
has to be something there even though we don't know what it is. I get
the same sort of impression when I read business magazines and they
give a final obligatory nod to DB2 amongst other databases. People
don't write demos to show how software integrates with DB2, but they
do for MySQL, SQL Server, and Oracle. For all of the mentions of DB2
and related initiatives, I don't know anything about it, and yet as
some segment of their target audience, I should know a little
something through the same sort of osmosis that we get with other
products. I think a lot of developers in the world share my shoes on
this one. So anyway, it seems like if you adopt DB2 you get a whole
world full of stuff, it's just not the same world everyone else lives
in - and you should be prepared to pay the price for that obscurity.
But hey, you can't go wrong when you buy IBM, right?


Perhaps surprisingly, I'd consider MySQL. Just check out the
mysql.com website (and of course thousands of others) to see what they
say about scalability, reliability, etc. When it comes to worldwide
acceptance, low cost of ownership, and all of the other things that
are important for business and technical people, plus the advantage of
being as Open Source as it gets, I think this should be on your list.

Nuf outta me.

TG@ does.this.guy.ever.stop.yackinNebula-RnD.com

On Feb 8, 1:25 pm, "Jeff Caspari" <FDFDF... (AT) sneakernet (DOT) com.invalid>
wrote:
I suspect that solution makes more sense for existing Pick software,
but it is on my list with a note that it might be more difficult than
I would like to do AJAX front-ends that work well in FireFox, for
example. Thanks. --dawn

Lots to respond to, Tony. Thanks for your angle on this. For a quick
response, I will say that I agree that MySQL should be on the list and
don't recall why I left it off (except that Oracle bought out InnoDB
and I haven't looked at it since that time), so it is on my list now.
My only interest in DB2 is not as an IBM DBMS, but because they are
early in the XQuery business, thereby permitting both the relational
model and a "XML data model" but I'm not overly fond of being an early
adopter on that front at this point.

My short list is UniData, Oracle, MySQL (thanks for the reminder) and
Cache (dang, it still doesn't work, pressing Alt, holding down while
typing 130, then no more Alt and type e, no go, ah well). I do not
want to use a Windows server for the DBMS, but only for reliability,
scalability, and security reasons ;-) I have done some work on each
of these platforms other than Cache' during my career, but agree that
IBM is not doing what I would want them to do to keep me interested,
while so far Cache' seems to know how to market to me (and I make my
technology choices baesd only on that which markets to me, Al
Franken ;-)

Thanks. --dawn

On Feb 8, 1:59 pm, "NebulaTony" <26q7zm... (AT) sneakemail (DOT) com> wrote:

"dawn" wrote:

Dawn, DesignBais was AJAX before AJAX had a name. It's AJAX through
and through. You and I still haven't done a demo and you still really
don't know anything about it as is evident from a couple points in
your brief comment. One of these days I really would like to show you
the environment, not as a sales pitch, but to share with you what
we're so excited about.

Regards,
T