IT - the perfect solution

[NEOWatcher]

I'm splitting off another discussion started in the CT section.

Somehow between me and Jay, and some others, we touched a nerve and the thread went OT somewhere around here
(http://www.bautforum.com/showpost.php?p=974303&postcount=95)

Anyway, we are discussing how a technological solution always turns out to be a better choice, whether the decision maker knows how that can be or not.

I'm sure everyone has experienced "A solution looking for a problem".

[Fazor]

I don't know WHO could have started that thread down that path :whistles innocently:

I'm in insurance now, so I don't deal too much with technical mumbo-jumbo, just the equally complex legal variety. Although from the user end of things, it's painfully obvious that the company likes to do the whole "we need this function in this program, so instead lets write a whole new progam". When said new program comes, the new feature works lousilly at best, and the old features that we depend on also become unreliable. And they wonder why we have a hard time getting excited when a new system is announced...

[mugaliens]

Anyway, we are discussing how a technological solution always turns out to be a better choice, whether the decision maker knows how that can be or not.

I'm sure everyone has experienced "A solution looking for a problem".

Is that why it takes me an hour to build travel plans and about forty mintutes to file travel claims when I return?

Back in the 1990s, it used to take HR 15 minutes to build the travel plans and me about 10 minutes to file a claim.

Then again, that's when it was done on paper.

What computers are good for:

1. Excellent replacements for typewriters

2. Terrific for building models of systems whereby one can change one or more variables and compare the results

3. Graphic information

4. Conducting initial statistics

5. Rapid communication and sharing information

6. Storing large amounts of information and being able to rapidly search that information

7. Controlling the environment. At least the one on the inside. I have sensors tied into my light switches which send information about temperature, humidity, and whether or not anyone happens to be in a room. I can manually adjust the temperature, but the computer "learns" my preferences, and how they vary with varying humidity, time of day, day of week, and even time of year, then adjusts the temperature in the house to match what it thinks my preferences will be based upon historic data. This not only means I'm almost always comfortable, but it also means I save a lot of money when I'm not here, as the computer knows that, too (since I programmed it in). Computers control large building environments in much the same way for similar savings.

What computers are NOT good for:
1. Making decisions. They're a good aide for decision-making, but the moment the system being modeled changes, the computer model is flawed. Because most systems are in a constant state of flux, we must have humans in the loop.

2. Organizing information. Give me a sharp secretary with a good cabinet and paper filing system any day! If he/she's computer savvy, all the better, as that person will impose the same level of organization on electronic files that they do with paper files. But if you leave it up to the average person to organize their own files, about 80% of the time you wind up with a mess that their replacement will pull their hair out trying to figure out what's what. This is why nearly all large organizations have file plans which force all users to store information in a predetermined way.

3. Visual recognition. Let's face it - birds have far better image processing power and capability than the best computer in the world.

Well, that's a start - I'm sure others can add to this list!

[NEOWatcher]

1. Excellent eplacements for typewriters

There's something amusingly ironic about that typo.


Pretty good synopsis, but there is one thing that seems to waiver back and forth between the two and corresponds with your #7.

Controling process flow. When something needs to progress from person to person, a computer is a good way for it to happen consistantly.

Unfortunately, in application, it turns it into a technical nightmare, and is being magnified with Sarbanes.

[mugaliens]

There's something amusingly ironic about that typo.

Wow! That was quick!

But thanks for helping me prove my point, as the problem is fixed, and I didn't have to retype the entire thing like I used to have to do with a typewriter.

[NEOWatcher]

The timing on this one couldn't have been better.

Lets add "a replacement for management skills".

I just got an email saying I will get an email requiring me to complete an online course on "mutual respect". In the months that follow, more tests to guide our behavior.

You see, the company is trying to take pride in thier reputation. :rolleyes:

[Mister Earl]

Let's not forget the plight of us techs.

:(

[Pinemarten]

I wired a sawmill in BC, Canada that was capable of control from head office in Portland, Ore.

They could watch the lumber prices and orders from there, and adjust the mill to compensate.
The mill could go from 1x1" at 10' to 6x6" at 20' just by entering it in the software. It would change the whole line at once.

Industrial computers and controllers have advanced immensely.
Machines can be monitored and controlled in a windows GUI.
Product can be adjusted and quality control monitored.
Motor controllers monitor the loads etc, and even indicate bearings needing grease.
There are things called CNC machines, mostly lathes, where they make a drawing in AutoCad, put a crushed Volkswagen in and it spits out 40 machine guns. Well, not quite that sophisticated yet....

You are right, it is a long list.

[Ronald Brak]

Personally there are expert systems I trust more than human experts.

Here is a Micheal Chriton quote. I found it on the internet, so it must be accurate:

...A similar circumstance applies to psychologists, who are most accurate in making diagnoses when they are young, and tend to rely on checklists. Later, as experienced practicioners, they rely on clinical judgement and misdiagnose. This means that psychologists become demonstrably less skilled as they become more experienced. A sort of inversion of expertise.

[Maksutov]

I've dealt with some really big products over the years. These include submarine power units which were over two stories high and weighed 80 tons, earth moving equipment that made the operators look like Lilliputians, and various components for 747s.

When the design and then the processing got to sub-assemblies and final assemblies, the little computer screens displaying CAD/CADD/CAM etc., renditions of what we were building just didn't cut it.

When one is working with really big, complicated assemblies, engineering drawings on paper provide information that computer monitors do not. One aspect is being able to see relationships that would be lost between the pixels on the biggest computer screen.

This is especially true when one is having a design finalization meeting or discussing what went wrong in an assembly that's 30 feet long, weighs five tons, and is loaded with all kinds of sensors, moving parts, hydraulics, structurals, etc. That's why the table at such a meeting must be large. We need room to roll out the drawing and decide if it's OK, and later, how to fix what went wrong, both without missing any details that might impact our decision making process.

[mugaliens]

There are things called CNC machines, mostly lathes, where they make a drawing in AutoCad, put a crushed Volkswagen in and it spits out 40 machine guns.

And here I was thinking it was old thimble collections - silly me!

Still, we're probably not far off from that one.

[springa]

Personally there are expert systems I trust more than human experts.

Here is a Micheal Chriton quote. I found it on the internet, so it must be accurate:

...A similar circumstance applies to psychologists, who are most accurate in making diagnoses when they are young, and tend to rely on checklists. Later, as experienced practicioners, they rely on clinical judgement and misdiagnose. This means that psychologists become demonstrably less skilled as they become more experienced. A sort of inversion of expertise.


Hmm, I find it difficult to believe that a checklist is better than experience for making psychological diagnoses. Psychology seems like one of those fields where there is no substitute for experience, although I suppose experience could be detrimental if it gives someone too many prejudices or pre-conceived notions.


Regarding computers and technology in general, so much depends on how they are implemented and whether an organization has the resources and expertise to get them functioning at full capacity. I used to work for a financial services company where we kept upgrading to new digital imaging and workflow systems that were technically better, but in reality each upgrade slowed things down because apparently both the memory and bandwidth were not adequate to have the systems work at full speed.

[Moose]

Here is a Micheal Chriton quote. I found it on the internet, so it must be accurate:

Michael Crichton is demonstrably a poor source for reasonably objective assessments of professional expertise, particularly that far outside of his field (writing mass-market fiction.) It doesn't seem to stop him from trying, though.

[mugaliens]

Michael Crichton is demonstrably a poor source for reasonably objective assessments of professional expertise, particularly that far outside of his field (writing mass-market fiction.) It doesn't seem to stop him from trying, though.

No, it doesn't, and while some people consider him to be poor with respect to objectivity, I've discovered, over time, that he's remarkably dead on target with his somewhat controversial subjects.

It's the controversy itself which I suspect raises the question of objectivity.

[Moose]

No, it doesn't, and while some people consider him to be poor with respect to objectivity, I've discovered, over time, that he's remarkably dead on target with his somewhat controversial subjects.

Please don't do that.

[Pinemarten]

Hmm, I find it difficult to believe that a checklist is better than experience for making psychological diagnoses. Psychology seems like one of those fields where there is no substitute for experience, although I suppose experience could be detrimental if it gives someone too many prejudices or pre-conceived notions.


I don't know about psychology, but I have seen it in veterinarians whose profession is far more crucial to humanity.

I used to work on a farm with 600 brood sows. We started losing 2-3 a week due to 'acute peritonitis' (perforated gastric ulcers).

Our vet couldn't figure it out.
This was a concern to me because I was the poor schmuck that had to drag the carcasses out the door.

This went on for 3-4 months, samples were sent to labs, we paid more vets etc.

After a month or two I decided to look at our copy of The Merck Veterinary Manual. A bible to most vets.

Section on ulcers- One of the main causes is high salt content in the diet. When high salt supplements such as whey are added to the diet.....

We had just started feeding them whey (7500 litres/600 sows/1.5 days), before the problem started.
I mentioned this. We reduced the salt in their grain based 'chop' to zero.
No more heavy lifting.

[Moose]

It is an interesting line of thinking. My mother occasionally tells me about the six months or so I was sick when I was but a year old. It was a very difficult six months for my folks. I was constantly in distress/pain. I'd seen doctor after doctor, specialist after specialist. I'd been tested for cystic fibrosis (I think that's what Mom said) and a bunch of other stuff.

Finally a med student asked the right questions. It turned out to be my tonsils. Nobody thought of it because tonsils don't usually inflame before five or six.

[Ronald Brak]

Hmm, I find it difficult to believe that a checklist is better than experience for making psychological diagnoses. Psychology seems like one of those fields where there is no substitute for experience, although I suppose experience could be detrimental if it gives someone too many prejudices or pre-conceived notions.

There is the possiblility that older psychologists don't use checklists becauses checklists were considered important when they were trained.

And all I know is the quote had the name Micheal Chriton at the end of it. It may not have been by Micheal Chriton the writer, it could have been by someone who knows what he's talking about.

[yuzuha]

I don't know WHO could have started that thread down that path :whistles innocently:

I'm in insurance now, so I don't deal too much with technical mumbo-jumbo, just the equally complex legal variety. Although from the user end of things, it's painfully obvious that the company likes to do the whole "we need this function in this program, so instead lets write a whole new progam". When said new program comes, the new feature works lousilly at best, and the old features that we depend on also become unreliable. And they wonder why we have a hard time getting excited when a new system is announced...

Actually, it is more complicated than that. The users want something, but are not sure exactly what they want or exactly how it should work and don't quite know how to ask for it. Their manager then concieves his own sand-castle based on what he thinks his people need, rather than what they tried to ask for. He then puts in an ill-formed request to a sales team who missinterprets the request and churns out an incorrect and sketchy product requrements doc. Then the analysts dream up something that will sort of fit what they interpret the customer wants based on the requirnments doc, which usually involves creating a whole new architecture that is incompatable with the old architecture... but, programming isn't allowed to change the interface because the customer doesn't want to retrain their staff, even though the new architecture is not compatible with the old interface. So, we dig out the bubble-gum and bailing wire and duct tape a horse and buggy to the front of the new Corvette we wrote, when what the actual users really wanted was a pickup truck. That pretty much sums up my 30 years of experience in DP... that, and that computers are quite adept at producing mountains of reports and retaining terabytes of data that nobody looks at, or needs, other than to make some manager somewhere feel secure in the belief that they are measuring something useful, rather than just wasting paper and resources.

Maybe I'm a bit cynical, but I still think that if you want good software you need to teach the users to be programmers, or eliminate all the layers of bureaucracy between them and the programmers.

[LurchGS]

Yuzuha hit it on the head for me. At best, I'm a weekend programmer.. I'd ask for a tweak or feature on our home-grown central app, and MAYBE two years later I'd get it. Wrong, doing everything but specifically what I asked for.

I finally got tired of it and started writing my own apps. Nothing fance, but Python/Mysql/TCLTK pretty much do everything I need done. SInce it's for *me*, I don't care what it LOOKS like, so long as the fields are where I want them, etc.

SInce I document the heck out of my code (mostly so I have some idea what the heck I was thinking if I have to look at it 6 months down the road), it's trivial for others to modify for their own uses. And I overload on the HELP tools.

Anybody can come up to my app, fire it up, and step-by-step through the process of [whatever].

Sure, a 'professional' programmer could do what I need done in an afternoon and it takes me a week. But I get more than just a working tool out of the deal. I learn how to do new things, and I have immense satisfaction when I'm done creating my tool.

OF course, this can't be done by everybody. My company is small (12 employees - 25% are sales dweebs) - try this in Conglomco and you'll need lederhosen to prevent road rash under your wallet. Further, I'm one of the primary owners, so I can pretty much do what I please.

It's good to be the king

[Maksutov]

[edit]So, we dig out the bubble-gum and bailing wire and duct tape a horse and buggy to the front of the new Corvette we wrote, when what the actual users really wanted was a pickup truck....I imagine that if enough duct tape is used, that horse and buggy winds up pretty gooey.

Re records retention, most of what's saved is, as you said, useless and a waste. However, re quality records, there's typically a legal reason for doing so, i.e., to be able to demonstrate that the product was properly designed for its intended application and manufactured correctly.

During the first wave of product liability cases in the 1970s, those companies that had no records to demonstrate manufacturing competence found themselves paying out loads of money.

Poorly designed and/or manufactured products can lead to irate customers, as illustrated here (http://www.dilbert.com/comics/dilbert/archive/images/dilbert2007043054828.gif).

[mugaliens]

Please don't do that.

And do you also object to my comment, "It's the controversy itself which I suspect raises the question of objectivity."

Hmmm?

[mugaliens]

LurchGS - I'm a systems analyst, which means I document and codify business rules which I then hand to the programmers so they can model what we do by hand via computer, hopefully, to some gained advantage.

Well, a little off-topic, but just wanted to clarify my position...

[HenrikOlsen]

However, re quality records, there's typically a legal reason for doing so, i.e., to be able to demonstrate that the product was properly designed for its intended application and manufactured correctly.
Used correctly, you can be nasty to deserving people:)

Production wants records so they can show that things where produced in compliance with specs, so production has their backsides covered and can pass the buck on to design.
Design wants records so they can show the design is in compliance with what management asked for, so they can pass the buck on to management.
Sales want records so they can point at the shiny ISO9001 certificate when trying to sell something.
Management wants it because they've learned it's a GoodThing(tm), without realizing that used correctly by everyone else it means they end up responsible for everything that goes wrong, with noone to blame because all the usual targets have the paper to prove they did it as ordered.

[LurchGS]

Mug - cool

My sister got her doctorate in Systems, though she's not in the field - she spent time as an actuarian (we do love our numbers), then as creator/keeper of the NTFS' railroad crossing database and as an investigator. (She was on the board that investigated that train/truck collision near Chicago a few years ago - where it was determined that the truck driver had been driving much longer than was legal). Now, though, she's with the FRA, and I'm not sure what's she's doing.. not talked to her in a couple years.

[Maksutov]

Mug - cool

My sister got her doctorate in Systems, though she's not in the field - she spent time as an actuarian (we do love our numbers), then as creator/keeper of the NTFS' railroad crossing database and as an investigator. (She was on the board that investigated that train/truck collision near Chicago a few years ago - where it was determined that the truck driver had been driving much longer than was legal). Now, though, she's with the FRA, and I'm not sure what's she's doing.. not talked to her in a couple years.Would that be NTSB?

I thought NTFS was what Microsoft replaced FAT with when they intro'd NT.

[LurchGS]

LOL - yeah, NTSB.. you can tell what I'm doing off-camera, can't you? Putzing with XP and quickly going less sane than I was this morning.

[SpectateSwamp]

Yuzuha hit it on the head for me. At best, I'm a weekend programmer.. I'd ask for a tweak or feature on our home-grown central app, and MAYBE two years later I'd get it. Wrong, doing everything but specifically what I asked for.

I finally got tired of it and started writing my own apps. Nothing fance, but Python/Mysql/TCLTK pretty much do everything I need done. SInce it's for *me*, I don't care what it LOOKS like, so long as the fields are where I want them, etc.

SInce I document the heck out of my code (mostly so I have some idea what the heck I was thinking if I have to look at it 6 months down the road), it's trivial for others to modify for their own uses. And I overload on the HELP tools.

Anybody can come up to my app, fire it up, and step-by-step through the process of [whatever].

Sure, a 'professional' programmer could do what I need done in an afternoon and it takes me a week. But I get more than just a working tool out of the deal. I learn how to do new things, and I have immense satisfaction when I'm done creating my tool.

OF course, this can't be done by everybody. My company is small (12 employees - 25% are sales dweebs) - try this in Conglomco and you'll need lederhosen to prevent road rash under your wallet. Further, I'm one of the primary owners, so I can pretty much do what I please.

It's good to be the king

Been programming since 1970.
I do it pretty well the same as you do.

I do the odd flowchart when things get sticky.
When there is no deadline, a day or two helps
bring out the best solutions.

I do lots of comments as well. Leaving behind hints as to what
should be changed here or there. Or what other possibilities
exist at certain points in the code.

I comment out testing lines - rather than use debuggers
There might be a group of prints that test a main function.

That's how I developed my Desktop Search -
MultiMedia player. The interface is ruff. Just 3 prompts.
A skin; should not be a problem.

Been hounding the MicroSoft crowd at:
http://channel9.msdn.com/showuserthreads.aspx?userid=31672

Looking everywhere for a Desktop Search ShowDown - but no takers.

I have more control over my media: Video, Music, Pictures and Text
than anybody. It's the only program I need. Everything is portable.

[mugaliens]

Please don't do that.

uh...

(LOL)

...done!

Seriously - a lot of people disagree with him, but he's perhaps one of the most avid researchers of any modern author. I've read stories of how Heinlein and Clancy would keep dozens of files of articles to support their content, but some of Crichton's novels involve literally hundreds of background reports, publications, essays, etc., by leaders in their fields.

Is he popular?

Not with some people - but only because he simply tells things like they are and refuses to be politically correct.

And good for him!