Dot allowed as characters allowed in netlist?

On Sunday, March 11, 2018 at 2:50:26 PM UTC-7, John Larkin wrote:
> On Sun, 11 Mar 2018 16:31:33 -0500, "Tim Williams"
> <tiwill@seventransistorlabs.com> wrote:
> 
> >Who actually has problems parsing strings anymore?

> Some things don't work right if file names have embedded spaces. And...

Yeah, string handling problems are completely endemic.   I'm young enough to
have known better than to substitute lowercase L for a digit 1 at the
early typewriters-to-computers transition,  and old enough to remember editing and
recompiling SPICE to allow models using lowercase to be incorporated.

And even today, my database on OS #1 spits out right-single-quote when I ask for
an apostrophe (0xD5 character instead of 0x27), and an EPUB converter on OS #2 offers
to either 'UnSmarten punctuation' or 'transliterate to ASCII', but doesn't say what order
those operations are.   If you don't do either, though, it morphs to a unicode character
that I cannot identify... and if you do both, the apostrophe disappears
entirely.   It also mangles whitespace, generating spurious chapter headers.

I want the document on my smartphone, which is on OS #3.   So, there's three operating
systems and three programs, plus a couple of transfers (and a cloud), all of which
is easy to do, as long as you want it done wrong.

Decades ago, I'd  just make a FORTRAN prettyprint program.   Now, not so easy.
But, if you expand from three programs to six, there's ways to edit the whitespace,
and identify/substitute the D5 character, and do the load-via-the-cloud step.   

Six programs and three operating systems for moving a paperless
printout to a pocketable gizmo.   And, getting it wrong (in more ways
than I've yet mentioned).   Because, text is a minefield.

It's not just me.   Everyone is putting up with poorly documented
text limitations (and assistances: by Leaver's law, everything a smart
program does for you, it also does to you).

<https://xkcd.com/327/>

On Mon, 12 Mar 2018 00:30:32 GMT, Steve Wilson <no@spam.com> wrote:

>John Larkin <jjlarkin@highlandtechnology.com> wrote:
>
>> SI format clearly distinguishes between exponents and dimensional
>> units. The 4k7 thing is just ugly and sloppy.
> 
>> Modern decimal points don't fall off drawings. Properly drawn decimal
>> points never did.
> 
>There is a readability problem with decimal values. For example, .47uF on a 
>large drawing can be misinterpreted to mean 47uF.
>
>One solution is to add a zero in front of the decimal so it reads 0.47uF
>
>I know you already know that. Just mentioning it for others.

Some Zener diodes use the form 6V2  in their  part numbers

Takes less space also.

As for 0.47uF, I always add the leading zero.

Also, on schematics, I have learned to not  use 4-way tie points
because of the sometimes ambiguity of the wire crossing or connecting
whether or not the dot is huge and visible.

On Mon, 12 Mar 2018 00:30:32 GMT, Steve Wilson <no@spam.com> wrote:

>John Larkin <jjlarkin@highlandtechnology.com> wrote:
>
>> SI format clearly distinguishes between exponents and dimensional
>> units. The 4k7 thing is just ugly and sloppy.
> 
>> Modern decimal points don't fall off drawings. Properly drawn decimal
>> points never did.
> 
>There is a readability problem with decimal values. For example, .47uF on a 
>large drawing can be misinterpreted to mean 47uF.

The best fix for that is to type 470 nF. 

>
>One solution is to add a zero in front of the decimal so it reads 0.47uF

Yes. In engineering school, we were taught to always use a leading
zero before a decimal point :  0.47 uF. 

We were also taught to draw big, bold decimal points.

>
>I know you already know that. Just mentioning it for others.


-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

"boB" <boB@K7IQ.com> wrote in message 
news:qsrbad1tgvvdct9rnc8n5ub1gqi09mc7e6@4ax.com...
> As for 0.47uF, I always add the leading zero.

Same here.  Curiously(?), I always speak it as "point four seven", but write 
it "zero point four seven".  The zero is implied in written form, and 
omitted in spoken form.

All about minimizing ambiguity.  The spoken word should be concise, so that 
the listener is not burdened and fatigued by irrelevance; while the written 
word should be visually unambiguous, and can take some liberties with 
concision(?).

As most readers verbalize written text, one cannot take too many liberties 
on that aspect, or the reader will be as lost as the listener.


> Also, on schematics, I have learned to not  use 4-way tie points
> because of the sometimes ambiguity of the wire crossing or connecting
> whether or not the dot is huge and visible.

It often looks better too, though sometimes it doesn't.  There's always a 
solution that is both appealing and sensible.  Lazy drafters tend not to 
look for them (or think about appeal or readability at all, for that 
matter).

Tim

-- 
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Contract Design
Website: https://www.seventransistorlabs.com/ 

On Sun, 11 Mar 2018 20:14:07 -0700, boB <boB@K7IQ.com> wrote:

>On Mon, 12 Mar 2018 00:30:32 GMT, Steve Wilson <no@spam.com> wrote:
>
>>John Larkin <jjlarkin@highlandtechnology.com> wrote:
>>
>>> SI format clearly distinguishes between exponents and dimensional
>>> units. The 4k7 thing is just ugly and sloppy.
>> 
>>> Modern decimal points don't fall off drawings. Properly drawn decimal
>>> points never did.
>> 
>>There is a readability problem with decimal values. For example, .47uF on a 
>>large drawing can be misinterpreted to mean 47uF.
>>
>>One solution is to add a zero in front of the decimal so it reads 0.47uF
>>
>>I know you already know that. Just mentioning it for others.
>
>Some Zener diodes use the form 6V2  in their  part numbers
>
>Takes less space also.
>
>As for 0.47uF, I always add the leading zero.
>
>Also, on schematics, I have learned to not  use 4-way tie points
>because of the sometimes ambiguity of the wire crossing or connecting
>whether or not the dot is huge and visible.
>
>
>

I don't understand all this dot/decimal paranoia. Make them clean and
visible. Deliberate offsetting of natural 4-way connections makes ugly
schematics, as does the 4K7 notation.






-- 

John Larkin         Highland Technology, Inc

lunatic fringe electronics 

On Sun, 11 Mar 2018 16:31:33 -0500, "Tim Williams"
<tiwill@seventransistorlabs.com> wrote:

>Who actually has problems parsing strings anymore?
>
>Don't write your PCB tools in God damned PHP.
>
>On a related rant:
>
>Stop naming fucking files in ALL_CAPS_WITH_UNDERSCORES.  We have long file 
>names, case and spaces now.  Like, for two or three decades.

Spaces are awful when using command line tools. 

You either have to use quotes or some wild card or file name
completion, if available.

>
>We can write appealing, pronouncable file names now.  Like "Power Supply Rev 
>1".
>
>Tim

On 2018-03-11, Joerg <news@analogconsultants.com> wrote:
> Sorry for the occasional non-political post ...
>
> In the past I have seen special character issues with netlists across 
> foreign country boundaries. For example, in some countries they use a 
> comma as the decimal "point" and NXP also has those dreaded "...,215" in 
> some part numbers. That can blow a netlist out of the water.
>
> How about a real decimal point (a dot) inside a footprint?
>
> Main reason I ask is for mounting hole designators. A client wants all 
> this in metric dimensions and that will require fractions of a 
> millimeter. For clarity I'd like to list that as "4.5MM_DRILL" footprint 
> or similar. If dots are not allowed I could only use 4500UM_DRILL" or 
> such and that's less understandable.

  "4m5" ?

-- 
This email has not been checked by half-arsed antivirus software 

John Larkin <jjlarkin@highlandtechnology.com> wrote:

> On Mon, 12 Mar 2018 00:30:32 GMT, Steve Wilson <no@spam.com> wrote:

>>There is a readability problem with decimal values. For example, .47uF
>>on a large drawing can be misinterpreted to mean 47uF. 
 
> The best fix for that is to type 470 nF. 

Depends on your POV. I'm from the old school of mmf and cps. Switching to Hz 
was easy, but I could never get my mind around nF. Eventually I figured out 
1nF was 1,000mmf, but I get completely lost at 10 nf.

So I had to check. Sure enough, 470nF = 0.47mfd / uF

  http://www.justradios.com/MFMMFD.html

So you are right, but I'll stick with 0.47uF

Otherwise, I'll probably make huge mistakes:)

On 12/03/2018 08:12, Steve Wilson wrote:
> Depends on your POV. I'm from the old school of mmf and cps. Switching to Hz
> was easy, but I could never get my mind around nF. Eventually I figured out
> 1nF was 1,000mmf, but I get completely lost at 10 nf.
> 

mmF can be mega confusing - I worked with one old-timer who used it for 
*milli-micro* (aka nano) but books of that era also used *micro-micro* 
(aka pico). Usually the context help avert diaster.

piglet

piglet <erichpwagner@hotmail.com> wrote:

> On 12/03/2018 08:12, Steve Wilson wrote:
>> Depends on your POV. I'm from the old school of mmf and cps. Switching
>> to Hz was easy, but I could never get my mind around nF. Eventually I
>> figured out 1nF was 1,000mmf, but I get completely lost at 10 nf.
 
> mmF can be mega confusing - I worked with one old-timer who used it for 
> *milli-micro* (aka nano) but books of that era also used *micro-micro* 
> (aka pico). Usually the context help avert diaster.
 
> piglet

You are right. The Radiotron Designers Handbook, 1954 Edition, states 

"Note:  The abbreviations mF or mmF should not be used under any 
circumstances  to indicate microfarads or micro-microfarads, because mF is 
the symbol for milli-farads  (1 x 10-3 farad). Some reasonable latitude is 
allowable with most symbols, but here  there is danger of serious error and 
misunderstanding."

p176, http://preview.tinyurl.com/hmnpj2r

I guess there are few threads here that actually talk about capacitor 
values. But in LTspice, I often use 1e-12, 1e-9, 1e-6, 1e-3, etc, when 
there can be confusion about the value using standard nomenclature.

For completeness, here are some other books from the era, as well as two 
bonus links:

  Electrons And Holes In Semiconductors
  by William Shockley 1950
  http://preview.tinyurl.com/y77jxe6q

  Radiation Lab Series
  http://www.febo.com/pages/docs/RadLab/

  Reference Data For Engineers
  http://preview.tinyurl.com/ycjcp6cp

  Terman, 1943
  https://preview.tinyurl.com/ya5ho5dt

  Jim Williams Lab Notes
  http://www.introni.it/riviste_jim_williams.html

  Bob Pease Lab Notes
  http://www.introni.it/riviste_bob_pease.html