Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > sci.electronics.design > #451531 > unrolled thread
| Started by | o pere o <me@somewhere.net> |
|---|---|
| First post | 2017-03-12 22:51 +0100 |
| Last post | 2017-03-17 08:56 +0100 |
| Articles | 7 — 4 participants |
Back to article view | Back to sci.electronics.design
OT: Example needed o pere o <me@somewhere.net> - 2017-03-12 22:51 +0100
Re: OT: Example needed Don Y <blockedofcourse@foo.invalid> - 2017-03-15 15:11 -0700
Re: OT: Example needed tabbypurr@gmail.com - 2017-03-15 21:06 -0700
Re: OT: Example needed o pere o <me@somewhere.net> - 2017-03-16 09:41 +0100
Re: OT: Example needed Martin Brown <'''newspam'''@nezumi.demon.co.uk> - 2017-03-16 10:25 +0000
Re: OT: Example needed Don Y <blockedofcourse@foo.invalid> - 2017-03-16 11:16 -0700
Re: OT: Example needed o pere o <me@somewhere.net> - 2017-03-17 08:56 +0100
| From | o pere o <me@somewhere.net> |
|---|---|
| Date | 2017-03-12 22:51 +0100 |
| Subject | OT: Example needed |
| Message-ID | <oa4fqa$the$1@dont-email.me> |
I need an example of someone specifying properties X, Y and Z, thinking of an object A, and getting a completely different object B which has the specified properties but is completely unsuitable for the task. The funniest, the best! I have to argue against the obsession to evaluate performance of humans based on allegedly "objective" figures. Pere
[toc] | [next] | [standalone]
| From | Don Y <blockedofcourse@foo.invalid> |
|---|---|
| Date | 2017-03-15 15:11 -0700 |
| Message-ID | <oace1f$ps9$1@dont-email.me> |
| In reply to | #451531 |
On 3/12/2017 2:51 PM, o pere o wrote: > I need an example of someone specifying properties X, Y and Z, thinking of an > object A, and getting a completely different object B which has the specified > properties but is completely unsuitable for the task. > > The funniest, the best! I have to argue against the obsession to evaluate > performance of humans based on allegedly "objective" figures. If it was even *mildly* commonplace, you'd think YOU would have a great example to offer! :>
[toc] | [prev] | [next] | [standalone]
| From | tabbypurr@gmail.com |
|---|---|
| Date | 2017-03-15 21:06 -0700 |
| Message-ID | <958d4a72-beb3-4642-936b-cafe1df76223@googlegroups.com> |
| In reply to | #451531 |
On Sunday, 12 March 2017 21:52:55 UTC, o pere o wrote: > I need an example of someone specifying properties X, Y and Z, thinking > of an object A, and getting a completely different object B which has > the specified properties but is completely unsuitable for the task. > > The funniest, the best! I have to argue against the obsession to > evaluate performance of humans based on allegedly "objective" figures. > > Pere read amazon reviews, plenty there - though 'completely different' is stretching it rather.
[toc] | [prev] | [next] | [standalone]
| From | o pere o <me@somewhere.net> |
|---|---|
| Date | 2017-03-16 09:41 +0100 |
| Message-ID | <oadivs$f2t$1@dont-email.me> |
| In reply to | #452082 |
On 16/03/17 05:06, tabbypurr@gmail.com wrote: > On Sunday, 12 March 2017 21:52:55 UTC, o pere o wrote: >> I need an example of someone specifying properties X, Y and Z, thinking >> of an object A, and getting a completely different object B which has >> the specified properties but is completely unsuitable for the task. >> >> The funniest, the best! I have to argue against the obsession to >> evaluate performance of humans based on allegedly "objective" figures. >> >> Pere > > read amazon reviews, plenty there - though 'completely different' is stretching it rather. > I don't have a *good* example, that's why I am asking. To get an idea, I was thinking on something like this: consider (UK) football and a billionaire who wants to build *his* football team. (B)illionaire asks: How to you find good players? (A)nswer: You have to look how well they play... This takes a long time. B: I want them now! Do good players have something in common? A: Well... Their thigh perimeter is usually big... B: Ok, go find me the guys with thicker thighs in the country! A: ... After some time: B: Oh, they don't score goals. Why? Pere
[toc] | [prev] | [next] | [standalone]
| From | Martin Brown <'''newspam'''@nezumi.demon.co.uk> |
|---|---|
| Date | 2017-03-16 10:25 +0000 |
| Message-ID | <oadp7d$1be9$1@gioia.aioe.org> |
| In reply to | #451531 |
On 12/03/2017 21:51, o pere o wrote: > I need an example of someone specifying properties X, Y and Z, thinking > of an object A, and getting a completely different object B which has > the specified properties but is completely unsuitable for the task. How about thick survivalists on Amazon grumbling that a geologists hammer does not make a particularly good ice climbing pick. https://www.amazon.co.uk/product-reviews/B0002OVCMO/ref=cm_cr_dp_hist_one?ie=UTF8&filterByStar=one_star&reviewerType=all_reviews&showViewpoints=0 They are both after all a pointy bit of metal with a handle. > The funniest, the best! I have to argue against the obsession to > evaluate performance of humans based on allegedly "objective" figures. > > Pere The main problem with objectives that haven't been properly thought through is that individuals meeting their targets may actually work against the wider interests of the company to perform well. Two that spring to mind from companies I have worked with are: Snr accountant on quadratic bonus based on how quickly he gets debtor invoices paid and how long he can avoid paying creditors. Short term outcome rotating position on stop list of three equivalent suppliers (and hence 120 days free credit - more if they were not on the ball). Medium term bailiffs turning up at factories to take away photocopier. Ultimate outcome tramp soaked in skunk oil sat in posh reception at HQ. Factory manager on quadratic bonus based on stock levels at end of year below some target. Outcome complete manufacturing stall in January. The ship it and be damned model of modern software development is also a victim of executive bonuses trumping sound engineering decisions. You can always ship a patch later but the hardware needs to be sold *NOW*. -- Regards, Martin Brown
[toc] | [prev] | [next] | [standalone]
| From | Don Y <blockedofcourse@foo.invalid> |
|---|---|
| Date | 2017-03-16 11:16 -0700 |
| Message-ID | <oaekks$5bg$1@dont-email.me> |
| In reply to | #452102 |
On 3/16/2017 3:25 AM, Martin Brown wrote:
> The ship it and be damned model of modern software development is also a victim
> of executive bonuses trumping sound engineering decisions. You can always ship
> a patch later but the hardware needs to be sold *NOW*.
Actually, I think that "bias" in thinking -- despite the often OBVIOUS
impracticality of it for specific products/markets -- is a bigger problem.
Yeah, you *might* design the hardware to support a post-production patch.
But, how practical is it to patch the code in a refrigerator, after the sale?
And, how do you factor in the costs of those "patches" ("fixes")? If the
owner has to bring the device *to* you for YOU (or your agent) to apply
the patch, what is the cost of that inconvenience (to the customer -- is
he likely to AVOID the update and just walk away thinking your product sucks)?
What is the cost to *you* (presumably, "your agent" must be compensated for
his time in some way)?
[Would you ship a hardware design with the expectation of replacing R27
at some future date -- with the *right* value of R27?]
Even for network-connected devices, there is a cost: providing and maintaining
the server to support the upgrades AND the hassle incurred by users (customers)
when/if problems with the updates are incurred.
Finally, it encourages a sense that the product is never "completed" but always
in a state of "work in progress" -- there's always a ready excuse for anything
that doesn't work quite right, etc. (and the user/customer is left with this
lasting impression of your product and you as a supplier)
[toc] | [prev] | [next] | [standalone]
| From | o pere o <me@somewhere.net> |
|---|---|
| Date | 2017-03-17 08:56 +0100 |
| Message-ID | <oag4nq$d1v$1@dont-email.me> |
| In reply to | #452102 |
On 16/03/17 11:25, Martin Brown wrote: > On 12/03/2017 21:51, o pere o wrote: >> I need an example of someone specifying properties X, Y and Z, thinking >> of an object A, and getting a completely different object B which has >> the specified properties but is completely unsuitable for the task. > > How about thick survivalists on Amazon grumbling that a geologists > hammer does not make a particularly good ice climbing pick. > > https://www.amazon.co.uk/product-reviews/B0002OVCMO/ref=cm_cr_dp_hist_one?ie=UTF8&filterByStar=one_star&reviewerType=all_reviews&showViewpoints=0 > > > They are both after all a pointy bit of metal with a handle. > >> The funniest, the best! I have to argue against the obsession to >> evaluate performance of humans based on allegedly "objective" figures. >> >> Pere > > The main problem with objectives that haven't been properly thought > through is that individuals meeting their targets may actually work > against the wider interests of the company to perform well. +1! This is exactly the point I wanted to support with examples. I think I saw something funny long time ago, but can't remember where... > Two that spring to mind from companies I have worked with are: > > Snr accountant on quadratic bonus based on how quickly he gets debtor > invoices paid and how long he can avoid paying creditors. Short term > outcome rotating position on stop list of three equivalent suppliers > (and hence 120 days free credit - more if they were not on the ball). > Medium term bailiffs turning up at factories to take away photocopier. > Ultimate outcome tramp soaked in skunk oil sat in posh reception at HQ. > > Factory manager on quadratic bonus based on stock levels at end of year > below some target. Outcome complete manufacturing stall in January. > > The ship it and be damned model of modern software development is also a > victim of executive bonuses trumping sound engineering decisions. You > can always ship a patch later but the hardware needs to be sold *NOW*. Full agreement here also! While it is true that software can (and should) be evolved, this is no excuse for shipping crap in the first place. Pere
[toc] | [prev] | [standalone]
Back to top | Article view | sci.electronics.design
csiph-web