[Info-vax] Using VMS for a web server

[Bill Gunshannon]

In article <4f422b94-1faa-40ce-83a5-8a0c7d10f44a at googlegroups.com>,
	johnwallace4 at yahoo.co.uk writes:
> On Sunday, 7 June 2015 04:28:10 UTC+1, David Froble  wrote:
>> Stephen Hoffman wrote:
>> > On 2015-06-06 16:02:11 +0000, David Froble said:
>> > 
>> >> So, what are we saying?  That there can be such a thing as "too good"?
>> > 
>> > Yes, there is such a thing.  Over-designing some product is 
>> > comparatively easy.  If you miss your target market, you're in deep 
>> > trouble.
>> 
>> But I don't think that that was the issue or question.  The question 
>> was, can you hit your target market so well that there are little to no 
>> problems, and your support people don't get much experience in solving 
>> the non-existent problems.
>> 
>> Such a solution is not over designed, it is very well designed.
>> 
>> For such a solution to be considered "bad" because it is well designed 
>> is distasteful to me, and things go much further downhill from there.
>> 
>> Should we design out nuclear power stations so that they periodically 
>> throw out some radioactive gases, so we can keep the hazmet people on 
>> their toes?  That's in my opinion rather similar to tossing rocks at 
>> Jan-Erik's VMS based solution(s) that don't have problems.
>> 
>> > If you're building a barn, then you'll find using trees are expensive 
>> > and heavy, you need a number of large trees, and constructing a classic 
>> > post-and-beam requires effort, skill and (lately) a ginormous CNC mill. 
>> > Using an engineered solution such as truss involves more complex pieces, 
>> > and trusses are prone to sudden failures in extreme conditions, but 
>> > trusses are cheaper, and you can use more of what might have been a 
>> > brace or just scrap wood within a post-and-beam structure.  There are 
>> > yet more expensive and better solutions than using a post-and-beam 
>> > design, too.  There's a market for post-and-beam barns, but it's not 
>> > nearly as big as that of the pole barn, or of the ordinary commercial 
>> > prefab construction that you can see getting delivered to building sites 
>> > by the truckload.  But I digress.
>> 
>> Yes, you do, and I'll digress some more.
>> 
>> I happen to have one of those post and beam barns.  I think replicating 
>> it would be very expensive, and for anything it could be used for, 
>> rather stupid.  It has those 12-14 inch square oak beams.  I'd really 
>> hate to pay for one of those today, let along the dozens in the barn.
>> 
>> As for failures, one of the joints is pulling apart.  But that barn will 
>> most likely out-live me, even if the problem is not addressed.  The 
>> building is unbelievable.  I do plan on pulling it back together one of 
>> these days.  That's going to be fun.
>> 
>> If I was smart, I'd disassemble the barn and find buyers for all the 
>> lumber.  With the proceeds I could probably put up several truss based 
>> buildings.  But the old barn does have "character".
>> 
>> > Getting back to software, a well-run, existing production application 
>> > installation usually tries to get to at least a local minima of whatever 
>> > they're optimizing for (usually cost and effort, increasingly based on 
>> > data collection and analysis), and a major upgrade or a replacement 
>> > installation often looks at getting into whatever the current global 
>> > minima might be.
>> > 
>> > Yes, there are vendors which target better-grade products -- not usually 
>> > "too good" -- and there can be profits here for the best of those 
>> > vendors.   But it's a whole lot of work and a whole lot of investment to 
>> > ensure that you're meeting and variously exceeding the expectations of 
>> > your customers.  Spend too much getting to "perfect", and you'll likely 
>> > lose your customers, too.  Tradeoffs.
>> 
>> Sometimes it's not cost that is involved, it is good design.  Many times 
>> it doesn't cost more to do the job "right", and sometimes it cost less.
> 
> Beautifully put.
> 
> Over here, the bit about "doing it right" vs "doing it shiny" is
> exemplified by London's Millenium Bridge. Very shiny on paper (or
> modern equivalent), functional disaster in the real world, slightly
> embarrassing for a country that used to have a world leading bridge
> design industry, cost a small fortune to fix. But shiny. 
> 
> If people regularly designed bridges the way that people regularly
> design complex computer systems...

They would fall down.

bill

-- 
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg999 at cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>