| Title: | The Digital way of working |
| Moderator: | QUARK::LIONEL ON |
| Created: | Fri Feb 14 1986 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 5321 |
| Total number of notes: | 139771 |
Ordinarily, I am continually impressed with the scope of vision in
Digital's product planning. We all know it comes at significant cost
in development effort and time-to-market. Thus it comes as a surprise
when announcements don't seem consistent with our strategic goals.
The decision to include Ingress in ULTRIX appears to undercut
Digital's commitment to Rdb and the entire distributed, OLTP effort.
Many customers are likely to choose Ingress as their database platform
to reduce the number of different database managers and software
development environments they have to manage, even if they don't read
this as signaling eventual abandonment of Rdb.
Understand that any customer's I.S. organization needs to minimize the
number of different tools they try to support. Moreover, those in
heterogeneous environments find 3rd party arguments for
transportability very appealing, since they've had to give up the
simplicity of a single-vendor solution anyway.
The issue really isn't one of compatibility of SQL (as assumed by most
writers in ULTRIX_SQL note 19.*). The customer now faces two
completely different 4GL environments if they follow DEC's
recommendation--Ingress on RISC, Rdb on VMS. On the other hand, Ingress
can offer a fine implementation on VMS. This is a powerful incentive
for a customer to try to standardize their internal development on
Ingress. Possibly the ISVs will make the same choice.
The observer would have to wonder if Digital had decided that Rdb was a
point product and the future belongs to Ingress.
The potential loss of software revenue for database products is the
least of our worries. A whole major business strategy depends on our
selling customer I.S. organizations on our DECtp program. Since Rdb is
a cornerstone in this plan, the impact of incorporating Ingress in
ULTRIX could be ..., well bad.
I understand that there wasn't funding for an Rdb port in time to meet
the time-to-market needs. Still, customers could continue to buy
Oracle or Ingress while we spent the two years to port Rdb. This would
have slowed our penetration of the workstation market. That's the
obvious tradeoff.
This may not be the best place for this discussion, but a quick look
in the other Notes files led me to believe that most of the
knowledgeable people are looking at their more narrow product areas.
And, more to the point, we all have a vested interest in protecting and
promoting the high-margin segments of our business.
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1096.1 | Sounds like a marketing question to me | CVG::THOMPSON | My friends call me Alfred | Mon Apr 30 1990 16:53 | 4 |
I'd bring it up in NODEMO::MARKETING. A lot of people interested in such stuff read that conference. Alfred | |||||
| 1096.2 | The goose is not in jeopardy | BANZAI::COUGHLAN | DBS Product Management | Tue May 01 1990 16:50 | 38 |
If the strategy says `provide a Digital-supplied database management system
for ULTRIX (VAX & RISC) systems', and the "roll your own" approach says
it will take n years (where n>1), what is the correct tactic to
implement the strategy? Better get one from outside and private-label
it. If we're going to put our name on it, it shouldn't be a schlocky
one, it should be a technolgy leader, right? And since we have lots of
"A Digital RDBMS shuld do xxxx" requirements, but we can't satisfy all
of them without rolling our own in n years, and we can't wait n years,
let's pick one that meets as many of the xxxx requirements as possible.
A short list of potential choices comes up real quick, and a detail
evaluation of technology, marketing, and business considerations gets
to the final choice. Surprise, it's Ingres.
Reality meets strategy. "Roll your own" tactics make us all (including
me) feel better, but time marches on, and we don't always have a lot of
time.
I don't see that providing a Digital RDBMS on Ultrix that's not a pure
port of Rdb/VMS undercuts the importance of Rdb/VMS, our DB strategy, our
OLTP strategy, or anything else. It provides the opportunity to
enhance account control instead of abdicating it to third parties, if
we have the will to sell it instead of snipe at it.
The expectation is that there are UNIX shops, and VMS shops, and a
small intersection of mixed shops, so the "single 4GL and system
management paradigm across all platforms" is not the driving factor in
a decision. Where it is, the Ingres Tools provide a common
4GL, and we must punt on system management.
In the long term, there will be a single Digital relational database
offering across all Digital platforms. Funding for a port is not the
issue: time to implement the correct development tactics is the issue.
We've bought the time, we have a good solution today, and a better one
is in the pipeline. The only thing that could have prevented this
would have been better prescience on the part of the management of
Rdb/VMS 4-5 years ago.
| |||||
| 1096.3 | mixed is the majority | NYEM1::MILBERG | I was a DCC - 3 jobs ago! | Tue May 01 1990 23:51 | 22 |
re .2
the comment about "VMS shops and UNIX shops and a small intersection of
mixed shops"
In my account's engineering organization, scattered all across the
country (many divisions), EVERYPLACE is a mixed shop. The "old" VMS
large, timesharing machines are coexisting (for the time being) with
the "new" UNIX (ours and others) workstations.
The issues of the future are servers - both compute and database.
Enough has been said about the 'cost of mips' on RISC vs. CISC so you
can guess where the compute servers are heading. The issue of
'database compatibility' is being raised TODAY and integration is the
'must have' if mixed environments are to continue and VMS is going to
survive tomorrow.
Please don't remind me how much better VMS is in data management,
adequate and usable are what is bought.
-Barry-
| |||||
| 1096.4 | Today's environment is HETEROGENEOUS! | JAWJA::GRESH | Subtle as a Brick | Wed May 02 1990 09:42 | 4 |
I agree completely with Barry when he says "mixed is the majority".
If I didn't know better I'd swear that he's been calling on my account.
-Don
| |||||