Software Prophet

This page was prompted by a comment by TomStambaugh in XpIsGeniusFriendly. He wrote:

"Some of my friends have suggested that I, and many like me, have a worldview that the ancients called "prophet" - as in, for example, the "prophet crying in the wilderness". Names like Moses, Nehemiah, and Ezra come to mind. Sometimes I, rightly or wrongly, "see" things that others don't (yet)."

Have other WikiZens experienced similar situations?

Personally, I have found that its taken me a long time to learn to act on my "prophecies". I.e. I tended to "see" things that others didn't, but to be hesitant about coming forward with my ideas. Starting out as a junior programmer, I found myself seeing opportunities and risks that others didn't mention. However, they were the "experienced ones" and I wasn't, so therefore I assumed that I must have been wrong :-) Often, as time passed, it turned out that I was right. So I slowly grew in confidence and started to speak up more, like a "prophet crying in the wilderness".

However, that's not the end of the story. Even now, with greater confidence in my convictions, it can still be hard to get "buy in" from co-workers. I spent about 12 months advocating a particular improvement to our software, only to be greeted with a total lack of enthusiassm. This was probably because considerable effort had been invested in the old way of doing things, so people didn't believe that we could achieve a new, radically better and faster solution. Finally, we signed a contract that committed us to making the kind of improvements I had been advocating. (Yes, we signed up to do something that most of my colleagues thought was impossible, but that's another story. Does the phrase DontConfuseSalesWithDelivery? ring a bell? :-) So, I decided to clarify exactly what we were getting ourselves into. That weekend, in my own time, I produced a working prototype of the newer, faster solution. It went on to form an essential part of the solution that we delivered, and it worked as I had predicted.

This strikes me as a very inefficient approach, having to build it before anyone would believe it was possible!

Has anyone else had similar difficulty in being a prophet, and if so, how have you solved it?

-- Anonymous Prophet

Story of my life, build it before they will believe. Sometimes not even then. It gets very frustrating.

In my experience, people don't want to try something new until they have no other option.

I hope others have a more positive story.

'Fraid not. :-/ I HaveThisPattern all the time, regarding at least half a dozen things at any given time.

When I was in college (mid 1980s) I spent some time arguing with Mike Godwin (of GodwinsLaw fame) about the Macintosh OS. I was appalled that it didn't have pre-emptive multi-tasking. I argued that Apple should have licensed a Unix kernel and built their GUI on top of that. He argued that people didn't really need multi-tasking on a personal computer.

Some things seem like obvious winners to me. Sometimes I'm wrong, though. I like to focus on the times I'm right. That makes me feel like a prophet.

-- EricHodges

A truly frustrating bit of my industry experience is that between 1983 and 1984, I led a team at Cadmus that designed and built just such a package on top of Unix. It was called "CadMac?", and provided a source-code compatible interface to a reimplementation of the Macintosh toolbox. It included the LayerManager?, which I invented as a pragmatic way to allow multiple parallel processes to share the same screen real estate. LayerManager? provided a "waitNextEvent()" routine (analogous to the toolbox's "getNextEvent()") that suspended the calling process until an event was available for it. Apple bought Cadmac in 1985 or 86, mostly to bury it. The LayerManager? did find its way into MultiFinder and its descendents. Many of us fought valiantly and ultimately unsuccessfully to try to persuade Apple to release its paradigm to competing hardware vendors. Cadmac would have been straightforward to port to the PC's of its time. I'm convinced that it would have stopped Microsoft in its tracks, and would have led to Apple's dominance of the software world. SteveJobs and Mr Pepsi disagreed. The rest is history. -- TomStambaugh

You've heard, perhaps, about this new automotive technology where they use solenoids to actuate the valves, and do away with the camshaft? A few manufacturers are playing with the idea at the present time, including BMW, IIRC.

http://www.halfbakery.com/idea/Cam-less_20Engine

I had that idea somewhere around 1989. <smug grin, followed by forehead-slap for never telling anyone about it> -- MikeSmith

I think a lot of people have had this idea. I had it as soon as I heard about camshafts. The asked the guy who was telling me about camshafts why they didn't use linear actuators and he said they aren't robust or reliable enough.

Well then, shall we throw down some half-baked ideas that we'll never develop and see if we can refine them and give them away freely?

My first contribution is TetrahedralComputing?. Turn all computer components into tetrahedrons and put PCI-style system interfaces on all sides. Use the asymmetric multiprocessing and hotplug features of Linux to add new CPUs, memory, drives, etc whenever you need a bit more.

In short, Beowulf clusters that tile in 3D

Prophecy? Fallacy? -- ShaeErisson

: Cooling issues and the pain of getting to a component in the middle of the pile come to mind. Also, regular tetrahedra don't tile 3D space.

Hmm... WhatIf? the tetrahedron contained plumbing for 'pass-through' ventilation; then you could cool the whole thing using a couple well-placed fan tetrons. Given some structural 'bus' tetrons, you could make much accessible space for components for which that would be useful, while still allowing other components to cluster tightly. You could pack anything not relating to persistence and inherently parallel as tight as you want: if any component is approximately as good to you as any other, why would you want to get anything other than the top one out? There might be something to this. :) -- WilliamUnderwood

Throwing around half-baked ideas is more like brain-storming, especially if they are ideas "that we'll never develop". In my view, the more interesting case is compelling ideas - compelling in that they beg to be built - that have been or can be implemented and that are currently unpopular. -- TomStambaugh

The issue isn't so much having an insight. There are insights aplenty among the big brains and deep thinkers. Perhaps within the population the havers of insight are scattered and few. But, as a profession and large, shared conversation we need have any given insight only once - provided it gets shared and used. There's the rub. The issue isn't insight, it's adoption. How do you bring your insights to others who do not have them? Part of the answer is in the question, actually. "How do you . . . " posed as a problem to answer with techniques and a series of tasks is answered in part by "How do you . . . " to describe how one approaches the problem. Stridency, for example, turns people off from good ideas, no matter how good the ideas.

One approach that works is by simple demonstration free of proclamation and self-aggrandizement. Consider as an example having an insight that in electronic collaboration that less is more, then just inventing a wiki engine, vs. proclaiming the rightness of wiki.

-- JamesBullock

Don't worry about people stealing your ideas. If your ideas are any good, you'll have to ram them down peoples throats. - Howard Aiken

There are two kinds of people: those who do the work and those who take the credit. Try to be in the first group, there is less competition there.." - Indira Gandhi

The very best way to predict the future is to create it. - Michael Kami

-- Brian

Michael Kami, eh? I guess AlanKay belongs in the second group then.

The attribution is as I first saw it. As for its accuracy I’ve always been a skeptic first and foremost.

CategoryWikiUser