Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
12/03/10 07:07
Read: times


 
#179838 - You are still just the machine shop with ready blue prints
Responding to: ???'s previous message
Richard said:
The people for whom I've done work seldom fail to know what they want


That's just one type of customer.

But I don't think you are every capable of seeing anything with someone elses eyes. That's just not possible for you.

There are more than one type of customers.

There are more than one type of products.

A lamp timer can be specified in measurable metrics. You can specify button press sequences. Size of display. Current consumption. Color. Shape. Materials.

But there are other products. How do you document the "feeling" that makes a customer select MP3 player A instead of MP3 player B? What makes a user select one car instead of another. Or one house instead of another?

A customer can't put metrics to such things, so in a number of situations, you need to help with simulations and prototypes (sometimes just mockups) to let them touch and feel. And you need to bring in the customer's customers. You really can't write in a requirements specification exactly what "feels comfortable", or what "feels wow".

The military, or the industry, doesn't care about "comfortable" or "wow". So they can write technical requirements specifications. But do try to realize that there are other products than that. How do someone order the development of a new tooth brush? By giving 50 pages of full specificatins? Hardly. If they had 50 pages of specifications, they wouldn't order any development. They would just ask the factory to produce. Somewhere they must involve industrial designers etc. And they need to try different strength and thickness of the fibres. And possibly different length. And possibly vary the parameters for the tip compared to other parts of the brush.

I don't think you have every really thought about the development process. You have only every thought about the narrow "engineering" situation where most of the development have already been done by the customer. You are just the hired hand that clues together a couple of components and slaps in a couple of thousand lines of code. No "invention" part in the work.

Have you never, ever, had a customer wanting you to try to produce multiple - different - solutions, for them to evaluate?

Whatever your customers may have supplied to you in requirements documentation, they are not representative of all situations. Some products can't just be fully specified before the development work starts - the specifications may actually be the majority work of the development process. That really should be obvious, if you spend some time and think about it.

But the thing is that you have, year in and year out, continued to only comment about the specific case where the customer can supply a full specificatin, similar to a machine shop being given complete blueprints that they just have to produce from. Why, just why, can't you see that that view can not be projected onto all development processes and that you obviously can not again, and again, tell people what failures they are if they do a development project without full specifications?

List of 80 messages in thread
TopicAuthorDate
Discussing ARMs on 8052.com...?            01/01/70 00:00      
   Same same...            01/01/70 00:00      
   positive            01/01/70 00:00      
   SevensAndNines...            01/01/70 00:00      
      Sevens and Nines            01/01/70 00:00      
         Try the chat here then            01/01/70 00:00      
   I really want to discuss it here            01/01/70 00:00      
      for me is ok and interesting :-)            01/01/70 00:00      
         nevermind at least            01/01/70 00:00      
      OK            01/01/70 00:00      
   Scope of the forum            01/01/70 00:00      
      on other hand            01/01/70 00:00      
         I have to agree            01/01/70 00:00      
            Flaw in the argument            01/01/70 00:00      
               Fresh blood is needed            01/01/70 00:00      
   the closest thing to ...            01/01/70 00:00      
      But related to Keil tools            01/01/70 00:00      
         not really policed - yet...            01/01/70 00:00      
            Actually well policed and few to answer            01/01/70 00:00      
               "Policed"? [ed]            01/01/70 00:00      
                  Not by Keil            01/01/70 00:00      
                     Agreed.            01/01/70 00:00      
                        what I said            01/01/70 00:00      
                        That's what should happen here            01/01/70 00:00      
               Not a suitable platform            01/01/70 00:00      
   i vote aye            01/01/70 00:00      
      second the motion            01/01/70 00:00      
      a nice typo... or intention?            01/01/70 00:00      
         re: typo -- or intention?            01/01/70 00:00      
   But there's ARMs and ARMs...            01/01/70 00:00      
      Definitely too large span of ARM chips            01/01/70 00:00      
         many of the concepts in use with 8051 chips are applicable            01/01/70 00:00      
         the problem with 'embedded'            01/01/70 00:00      
            It's the small side that is i need of a forum            01/01/70 00:00      
               Architecture Agnostic            01/01/70 00:00      
      I suppose you're right ...            01/01/70 00:00      
         Metric to use depends on past experiences            01/01/70 00:00      
            apply the right parametres            01/01/70 00:00      
               once again, I have to agree            01/01/70 00:00      
                  There's lot of analysis involved - about money            01/01/70 00:00      
                     Yes, that's all true for the PC world ... but what about us?            01/01/70 00:00      
            Feature Creep            01/01/70 00:00      
               Often easier to shrink after you have a product            01/01/70 00:00      
                  Absolutely!            01/01/70 00:00      
                     Isn't that a product of poor analysis?            01/01/70 00:00      
                        If requirements are properly analysed...            01/01/70 00:00      
                        no, it is a result of 'poor' customers            01/01/70 00:00      
                           Don't (necessarily) blame the customers!            01/01/70 00:00      
                              no blame of the customer            01/01/70 00:00      
                                 It's often like pulling teeth ...            01/01/70 00:00      
                                    Consultant?            01/01/70 00:00      
                                       Managing expectations            01/01/70 00:00      
                                       You can have "the talk" ...            01/01/70 00:00      
                                          You are still just the machine shop with ready blue prints            01/01/70 00:00      
                                             You can't engineer an enigma            01/01/70 00:00      
                                                Still not admitting different development requirements            01/01/70 00:00      
                                                   holydays time            01/01/70 00:00      
                                                      See? Now here's a guy who knows how to solve the problem            01/01/70 00:00      
                                                         Still not seeing the picture            01/01/70 00:00      
                                                            Not exactly ...            01/01/70 00:00      
                                                   I think you're confusing marketing with engineering            01/01/70 00:00      
                                                      and neither have you            01/01/70 00:00      
                                                         I just haven't worked for FOOLS            01/01/70 00:00      
                                                      RE: I haven't addressed your examples because            01/01/70 00:00      
                                                         Well, the world's going mad ... but I'm not on board            01/01/70 00:00      
                                                            Summing it up            01/01/70 00:00      
                                                               free advice            01/01/70 00:00      
                                       too much ARM            01/01/70 00:00      
                                          Pardon??            01/01/70 00:00      
                                    Contract            01/01/70 00:00      
                              Yes, it's true, but educating them should be their cost.            01/01/70 00:00      
                           I have to disagree ... provisionally            01/01/70 00:00      
               I don't see that as a problem ...            01/01/70 00:00      
         No, they can't            01/01/70 00:00      
            That's true, but aren't most problems quite small?            01/01/70 00:00      
   Discussing Discussing ARMs on 8052.com ...            01/01/70 00:00      
      Yes, among others            01/01/70 00:00      
      Then that's a completely different question!            01/01/70 00:00      
   An observation            01/01/70 00:00      
      correction            01/01/70 00:00      

Back to Subject List