John Doty wrote: > On May 10, 2009, at 3:10 PM, Joerg wrote: > >> John Doty wrote: >>> On May 9, 2009, at 2:51 PM, Joerg wrote: >>> >>>> Stefan Salewski wrote: >>>>> On Fri, 2009-05-08 at 17:13 -0700, Joerg wrote: >>>>> >>>>>> IMHO that is fundamentally wrong. How many successful race car >>>>>> drivers >>>>>> these days do you think can disassemble and re-assemble a Ferrari >>>>>> engine >>>>>> _and_ tune it properly? >>>>>> >>>>> What I heard about Michael Schumacher was that his strength was >>>>> more his >>>>> technical understanding about the car, which makes it possible to >>>>> discuss with the tech team to improve the cat, than his driving >>>>> skills. >>>>> >>>> Sure they know the technology, just like a pilot must know the inner >>>> workings of a jet engine or like I know how C and assembler is >>>> written. >>>> But that does not mean those people can perform the work of an >>>> expert >>>> mechanic or engineer in those fields. >>> The best can. When Yeager was test flying the Mig-15, he wired the >>> pyros on the ejection seat himself. >>> >> Wiring pyros is not designing an ejection seat. > > Understanding a system as hazardous as a Mig-15 well enough to know > how to prepare it, without any documentation, so that he could > survive a test flight that probed the limits of its performance > required more engineering savvy than most aerospace engineers posess. > Design is easy: you only have to master your own approach. Getting > an alien approach to work is much harder. Similarly, good design > review is much harder than good design (and therefore unfortunately > rare). >
It can be harder or not. Most of my work is re-design other folks' designs. Sometimes it is a bear to do but when the original job was done (almost) well enough to pass everything then it's fairly easy. From what I've heard the Russians did a pretty good job with the MIG series of aircraft. >> >>>> In fact most can't, and they don't >>>> have to. >>> Just because most people are content with mediocrity does not mean we >>> should cater to their laziness. And note that this kind of laziness >>> makes their job harder: master the tools, and the computer becomes an >>> enormously more powerful device in your hands. >>> >> Every sector of a trade has their strengths and weaknesses. Or >> should I >> consider engineers who can't design high voltage ICs, RF amps or >> fix EMC >> problems in their sleep "mediocre" just because I can do those things? > > That's not what matters. It's the ability, and the willingness, to do > *whatever* the job requires that sets the first rate apart. What > first rate engineers can "do in their sleep" is pick up whatever > specialized skills are needed for the job at hand. And don't tell me > such people don't exist: I've worked with several. > Sure, I do that all the time. Learning about x-ray technology right now (not medical this time). But a top notch engineer must know when it is better for the overall NRE budget and product performance to call in an expert on a certain matter. Unfortunately there are way too many groups or individuals who try it on their own not matter what. For reasons that completely elude me. >> I >> would never dare to say that, because it's not true. They just have >> other specialties. >> >> >>>>>> I know several fine electronics engineers who are not at all >>>>>> versed in >>>>>> fixing a PC, let alone install an OS. In fact, this is the >>>>>> majority of >>>>>> top notch engineers that I know. >>>>>> >>>>>> >>>>> It's hard for me to imagine an engineer who can not install an OS, >>>>> when >>>>> so many 12 years old school boys can do it. I can imagine other >>>>> "top >>>>> notch" people, like (financial) managers, artists, maybe >>>>> mathematicians >>>>> -- but that is not out target group. >>>>> >>>> Maybe hard to imagine for you but that how life is :-) >>> Nope. Life for the first rate is studying a new thing every day, >>> stretching yourself, learning how to exploit different methodologies >>> and points of view. >>> >>> Von Neummann once recommended a specific vacuum tube to the engineers >>> working on the early computers. He understood the issues and knew >>> that this specific new tube had worth characteristics. That's first >>> rate. >>> >>>>> Of course gEDA for Windows would mean more users. But would those >>>>> additional user contribute something to the project? >>>>> >>>> Oh yes. Without feedback from lots of folks who do engineering >>>> and CAD >>>> for decades you cannot create a good CAD tool. >>> Unless you walk in the developers' shoes, you cannot give truly >>> effective feedback. And developers who never walk in users' shoes >>> will never really understand what they need. Fundamentally, gEDA is >>> better because we don't have that inefficient kind of division of >>> labor. >>> >> Not so. For example, in medical electronics we receive the most >> valuable >> feedback from the best cardiologists in the country, guys I'd trust >> 100% >> if my time on the table came. Yet most don't have the foggiest idea >> how electronics work. And that's perfectly ok. > > I have seen this illusion created many times in the space program. > The engineers and astronomers believe they have had a profitable > exchange when, in fact, they have talked past each other without > understanding. > > There's a reason that Hubble observations are planned by software > written by an astronomer with space operations experience who taught > himself AI. The software specialists who wrote the original version > produced a system that formally met every requirement but didn't > work. They didn't do their homework, and therefore didn't understand > the problems they were supposed to be solving. > It is not an illusion at all if this feedback process is managed well. Rest assured, I sure hope not but should you ever develop a cardiac event that requires stent placement with local ultrasound guidance there is a system that has been optimized very well, by exactly the process I outline above. It has proven itself time and again, since the 90's. Competitors have tried, even HP threw in the towel. Number of competitors with electronic IVUS to date: Zero. >>>>> KiCAD was available from the beginning for Windows. Based on your >>>>> logic >>>>> the development of KiCad should be very fast, because of all these >>>>> "top >>>>> notch engineers" who can use it and who can contribute. ... >>>> It has improved trmendously over the last three years. >>>> >>>> >>>>> ... I do not >>>>> know >>>>> much about KiCAD, but it seems to be not too bad, and I know some >>>>> people >>>>> who used it on Windows. ... >>>> IMHO it's at a more useful stage right now than gEDA. No flames >>>> please, >>>> that's just my personal opinion, as someone who's done CAD quite >>>> extensively for over 20 years. Kicad is a very good CAD program, >>>> but has >>>> some quirks left. >>> They won't get fixed by complaining. >>> >> Mentioning and explaining details of a bug is not complaining. >> Papering >> over stuff like that is what keeps SW in the nerd corner. > > You have not identified any bugs. All you've done is complain that > the software doesn't fit your prejudices. > I have (politely) outlined things that many engineers in industry cannot live with, and have told me so. If a refdes shake-up doesn't qualify as something that needs improvement, then I don't know what does. I seriously doubt that this fine group sees it that way. If they did, yeah, then I'd simply rest my case ;-) >>>>> .. But most development seems to be still done by >>>>> the original author. >>>>> >>>> Yes, and therefore even more amazing. But I never understood why the >>>> Charras team and the gEDA team don't join forces. Very good things >>>> could >>>> come out of that. >>> We have completely different and fundamentally incompatible visions. >>> That doesn't mean we can't respect each other, ... >> >> Fully agree. >> >> >>> ... but I think joining forces is crazy. >>> >> Don't agree :-) > > You agree we have fundamentally incompatible visions, yet you think > we can join forces? Crazy... > That is a necessity with many, many products. If there isn't input from a wide group of people you end up with a not so good product or with a niche product. That is not the case with one-off designs and such. But CAD? -- Regards, Joerg http://www.analogconsultants.com/ _______________________________________________ geda-user mailing list [email protected] http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
