kicad-developers team mailing list archive

[Dick Hollenbeck <dick@xxxxxxxxxxx>]

On 11/22/2013 05:00 AM, Simon Turner wrote:
> Dear all
> 
> Sadly having to report more problems (following on from the KiCAD 
> autorouter fiasco). I thought I'd try the auto module placement tool - 
> What a waste of time.
> 
> Now I drew my outline, I placed and locked critical parts and then set 
> to autoplace. Where did my first footprint get put ? straight on top of 
> another one !
> 
> I had previously tried drawing a 3" square box around the "pile" that 
> resulted from the netlist import and all it did was start to create 
> another "pile" of parts in the top left corner of my outline. It's like 
> the PCB footprints are being ignored and the algorithm is just looking 
> at getting the pads as close together as possible.
> 
> I also noted that the so called placement process is pathetically slow 
> and used only 5% of one of my 4/8 processors, for this size of board it 
> is literally faster to place the parts myself (even if the auto place 
> worked) - This can't be a dissimilar process from autorouting and the 
> java router would route this board in seconds yet part placement could 
> take hours.
> 
> I know you guys put a lot of effort and I note that significant progress 
> has been made lately but surely these sorts of things can be tested 
> first ? It just makes the project look silly.
> 
> I'd also like to suggest that a bit more interaction goes on with the 
> rest of the world and that new releases of the code are more 
> forthcoming. for example if i try to install KiCAD from the software 
> centre of Ubuntu (which is undoubtedly one of the most popular versions 
> of linux for some unknown reason and if I understand correctly the same 
> code source is used for similar variants of linux) I get a version that 
> I'd say is about a year out of date as functions that I'm used to 
> finding are not even there. Yes I tried a manual install and was pleased 
> to find that it was more up to date than my windows version but that was 
> very tedious and updating failed while the packages I was apparently 
> missing from ubuntu could not be found in the software centre or 
> installed for some reason. Some of us are merely budding electronics 
> engineers, not linux and software experts and if the most accessible 
> version is 1 year out of date then I'm afraid it makes the whole thing 
> look laughable. As if it's not enough that I have to worry about my 
> design working and that I've got the right footprint with correct pin 
> associations (because some of the libraries are well dodgy) and that 
> routing has all gone to plan I have to cross check that the software is 
> behaving properly (like the other day I tried the auto router and had 
> tracks ploughing through each other - good job I checked or I'd look 
> like a right idiot), that will not have people coming back for more.
> 
> Kind regards and with thanks for all your hard work so far.
> 
> Simon


Simon,

Here's my theory of happiness.

Think of your expectations, think of reality, and think of happiness as all being
measurable and quantifiable with a number.  Three things, all measurable and quantifiable.

Think of your expectations as a high bar, like what an athletic high jumper has to clear.
  This is set by choice.

Reality is not set by choice, it is what we perceive as real.

Happiness is the difference between reality and expectations, and it can be negative,
positive, or zero.

So imagine *you choose* to set your expectations at 2 meters off the ground.  If reality
does not exceed 2 meters, you will be unhappy.  If reality exceeds two meters off the
ground you will be happy.

Having said that, it is now possible to be happy anywhere, even in prison.  Simply lower
your expectations, which are set by choice.


Practical applications of the theory:

When we purchase something, it is sensible to set our expectations off the ground, to set
the "expectations bar" at a height in anticipation of reality of value equal to or above
what that amount of money can typically purchase.


Now, the important question to ask yourself is this:

"Why are my expectations so high for something that I have done nothing to receive?"

If your neighbour gives you his used car for free, would it make sense for you to complain
about it to the point of making him unhappy for giving it?


So I have no ability to understand why you are so unhappy, other than that you have
established your expectations too high.



Now consider my perspective, for a moment, which is entirely different than yours.  I have
sunk $360,000.00 of my personal funds into KiCad.  Clearly this provides safety margin for
the establishment of some level of expectation of received value (reality of function),
above ground level.

Just this week I said KiCad was barely useable.  This is because there is a lot of shit
code in KiCad, none of it mine.

Do you think I am happy?  Do you think I am happy, that as a major sponsor, no one felt it
appropriate to list SoftPLC Corporation as a major sponsor of the work on any website?

Do a commit count some time.  Check out version 1 of the code and take a look at it.

=========================

Dude, as to testing, you are doing the testing.  That is not too much to ask in my
opinion, that users do the testing.  If you disagree, then don't use the software.

As to releases, don't expect them, compile your own from source.  As to fixes, submit your
own or hire somebody to fix them.

Nobody here owes you anything.