On Fri, 2007-01-12 at 22:47 +0100, Stefan Trethan wrote:
> > - do people here consider it lame to use autorouting?
> YES! YES! and YES!
> It's almost as lame as thinking the component libraries that come with a
> software will be of much use.
...
> my opinion is that
> if you use an autorouter you shouldn't even be making boards.
I got into programming in the mid 1970s. Back then, there were those who
screamed that anyone who uses a C compiler shouldn't be writing
software.
There were even those who considered assemblers to be lame, and that
anyone who couldn't/wouldn't memorise all the hex/binary opcodes,
addressing modes and argument fields, and calculate jump displacements
mentally, weren't good enough to program.
There'd be those of us making PCB boards for whom a depth of electronic
knowledge is just a means to an end (that of getting a piece of custom
hardware working to a level good enough for small volume sales, without
shelling out $thousands to a hardware dev), and there'd be those for
whom deep electronic knowledge is the end in itself. I belong to the
former group, and I by no means expect this position to command any
respect in this list.
It's all a matter of relativity. You could call it 'lame' to purchase
blank copper board from electronics stores instead of making one's own.
You could even call it lame to use bought copper, instead of
mining/smelting your own.
Back to the point - if I can get Eagle's autorouter to reach 100% on a
single-sided board without many convoluted trace routes, I'm happy. Time
saved. If/when my products go larger scale commercial, I can outsource
the one-time task of designing a better PCB, because at that stage I'll
be able to afford it. A specialised h/w dev will do it all in SMD, and
oversee the manufacturing.
Cheers
David