So who's running VMS anyway?
Richard Maher
Over the years I've had occasion to become a tad concerned about the level
of ineptitude-driven insecurity being experienced by your average HP/VMS
management apparatchik. Now, they certainly would not be alone in the quest
for a yogi, a maharishi, or a memphis-meltdown life coach, to abrogate their
responsibilities to, but the extent to which several of them have come under
the influence of various industry Svengalis has given me cause for much
consternation. Don't get me wrong, I'm more than happy to hear what any
"industry expert" has to say, it's just that I'd much rather hear it from
the organ-grinder himself rather than after being regurgitated through the
prism of any monkeys, lackeys, or minions, that may be scattered throughout
HP/VMS. So it was with some enthusiasm, nay relief, that I greeted the news
of Eric Newcomer's forthcoming keynote address to the VMS Bootcamp.
So I looked up the Bootcamp agenda hoping to see if his keynote address
would be broaching such issues as "Web Services - Ten years of a failed
technology; where next?" but I couldn't find anything. Then I came across
Eric's blogg under IONA's *"Thought Leadership"* section
at: - http://blogs.iona.com/newcomer/archives/000564.html
but was a bit surprised not to see any proud mention of the "OpenVMS
Bootcamp" in the list of upcoming conferences at which he will be speaking.
Anyway, I'm sure that was just a simple oversight. What's important is the
substance of what he has to say and what influence, if any, it may have on
ex-colleagues, friends, and compatriots on the starship VMS.
And just when you think you've seen all this IT industry has to offer (us if
not so much the customers), you get slapped in the face by something truly
remarkable - "Middleware for Middleware"! Not only do they have to buy
middleware so that their disparate and/or heterogeneous systems can talk to
each other, they have to purchase further software (or perhaps
smack-bang-ware) 'cos the first lot of interoperability claims were all
bullshit and never gonna work anyway.
Look, I must have misunderstood this, so if someone attends Eric's talk
could they please come back and advise later. But hold on, maybe the idea
has merit. This "universal adapter" (or perhaps "Enterprise Service Bus"? -
Now where have I heard that before? Spooky! Presumably any name other than a
"hub" as that would be all a bit Tibco(ish) with its bollocks ODBC adapters)
may soon have real application in the VMS arena. With VMS Middle Management
now actively supporting the port of gSOAP to VMS, just as "the first
production release" of WSIT hits the shelves, their customers are going to
need some middleware-for-middleware to rosetta-stone the ideosyncracies.
Then there's the GlassFish port, that'll be coming on stream, as soon as
gSOAP approaches a deliverable, and this universal-adapter stuff is becoming
critical! My God, how did we ever live without it? We have the solution; the
IMM team have just been a bit slow in shipping us the problem. (And then in
another couple of years we can come up with a "kernel-disambiguator"
codename "Tannoy" or "middleware-for-middleware-for-middleware"?)
Yep, this ideas got legs alright! If I was counselling Ann McQuaid I'd be
sorely tempted to advise her to provide a few million in seed money for a
VMS port of Artix. But, I'm not picking on Eric or IONA, let's not leave
Tibco out, their market capitalization and customer-base appear to be a lot
bigger but I don't see why we still shouldn't slip'em a couple of mill loose
change to bring ActiveMatrix onboard to Rendezvous HUBlike with our
Universal Adapters. But why stop there? I'd also recommend setting up
several HP/VMS Web Services - "Centres of excellence" in say several
geographical regions. Nowhere in particular, but for the sake of argument
let's say, um, plucking something random, um, Dublin, Munich and
Christchurch?
Honestly, you all make me bloody sick! If you didn't laugh, you'd cry :-(
Can not one of you pricks pause long enough to stop and ask the VMS
client-base what *they* want? Or does your arrogance, predilection for
"Thought Leadership", and unwillingness to cut your losses on a failed
architecture, prevent you from even countenancing an alternate point of
view? Here's a news flash for Ann McQuaid, if she ever makes it out of the
HP/VMS ashram, *"It's called RIA!"* and you can find out all about it at one
or two places including
http://en.wikipedia.org/wiki/Rich_Internet_application
the "I" in RIA stands for Internet (something that one or two people are
using these days to conduct business ma'am)
Please point me at the HP/VMS websites exhibiting RIA and their accompanying
tutorials! Please tell me who any long suffering VMS customer goes to, to
get his applications on the web! Please tell me in what year the last Java
Client GUI was developed (in the world) as I haven't see nor heard of one in
years! Please tell me why, when Mark Daniel has the Web Server performance
stats, and JFP has the Python, MySQL, django ports, and Tier3 Software has
provided one of the few publicly accessible RIA/VMS examples, what the fuck
it is exactly that your employees do!
Ann McQuaid said: -
[We will also continue to enhance OpenVMS for the HP Integrity and
AlphaServer platforms with new HP Integrity systems; storage performance and
connectivity; performance and scalability; the integration of industry
standards around security, integration software, Web Services, Java, and
UNIX®/Linux® interoperability; and virtualization capabilities.]
So, you bet the farm on Web Services; so did many others. But coming back,
Gartneresque, year after year saying *this* is the year they'll take off is
like someone still holding dotBOMB shares and expecting them to come good,
or when you keep sending money to that Nigerian Prince so that when he turns
up in the US he can repay you ten-fold. It just ain't gonna happen! Learn to
let go! Even Tibco have come up with something Ajax like the General
Interface so they can cash in - What are you waiting for???
Regards Richard Maher
PS. Speaking of shares, it's amazing what one can glean from the NASDAQ
website! When I was trying to track down Eric and IONA I stumbled across
ISE.IE and NASDAQ.COM and had a look at the share-price to see what sort of
industry-leader gets invited to give keynote speaches at a VMS BootCamp. To
be honest I was a bit under-awed with the stock chart for the last 12 months
(couldn't find any cause for the Jan08 rally; any ideas?) and the chart for
the last 10 years was a real eye-opener! You can also find out who's selling
their share-options when they mature (even if they're in itty-bitty
bundles). You can compare IONA to TIBCO to IBM to the NASDAQ overall index
and even HP. The good (or bad) news is that HP shares look to be
over-performing the market like a company posessed. No need to change
anything there; the market loves them. Let me add here that I'm a novice at
stock charts and for all I know IONA have split their stock every six months
for the last 10 years and their market capitalization has actually increased
exponentially - who knows? The charts are there if anyone's interested.
PPS. Here's a tip for the customers out there. I've got another
universal-adapter; it's called a "Socket" and they come free with pretty
much every computer or network appliance that you buy. Could be worth
looking into?
Jose Baars
not
special to OpenVMS.
I agree that middleware in general is solving a problem that never has
been identified properly.
Attempts at defining 'the problem' have been :
- Heterogeous networks, TCP/IP and....( DECNET? SNA? IPX? Really
Important! )
- Asynchronous delivery to avoid dependency on network outages ( when
was your network really down? how about saving it to a file first?)
- send once receive many ( see the socket manual under MULTICAST),
- guaranteed delivery once and once only ( this was called handshaking
and checksumming since before I was born)
- shielding pogrammers from network intricacies
( Why? Hire two programmers: one for the hetwork stuff, one for the
business logic, in stead of having a middleware department for years
after that
How? By having them use a product that comes with a 30000 page
manual set?
All these 'problems' are either outdated, do not deserve
more than a footnote in the design of an interface, or shift
any possible problems from the one-time design and programming
effort to the daily troubles of system maintenance.
So, the bulk of middleware has as main features the following :
- false promises
- failing or no support
- numerous bugs
- ridiculous license fees
- huge daily maintenance costs
- unreliable vendors
- security issues
> PPS. Here's a tip for the customers out there. I've got another
> universal-adapter; it's called a "Socket" and they come free with pretty
> much every computer or network appliance that you buy. Could be worth
> looking into?
What's more, the manual of this adapter is a fraction of the length of
the
average middleware.