Alternative symbols for easy rotation

On 18.11.2015 15:41, Rachael wrote:

When placing a common library part such as a resistor or capacitor it's

often necessary to rotate it to fit in with how you are drawing a certain

part of your schematic. For any one device, library symbols can currently

only be drawn in a single rotational position so when you then rotate the

part your text is now no longer how you would like it and you are required

to smash the device and manually replace all the text in the right

positions and orientations. This is okay for a handful of components but

not good for a large design. It's fine for a hobbyist just throwing

together small projects for their own use but as a professional design

engineer, when I present clients with schematics I have drawn they need to

look professional and this is causing additional work to achieve this.

 

It would be nice if several versions of a symbol (up to 4) could be

provided which would automatically be cycled through upon rotate if they

were available. This would mean we could create nice sets of symbols in our

libraries and rotate away to our hearts content safe in the knowledge that

we wouldn't be building up a mountain of smash and reposition work down the

line.

 

A quick fix I tried for this was to have -H and -V variants of resistors,

capacitors etc but this just became a pain in the BOM! (pun intended!)

Is vertical reading really that difficult? All my schematics are a mix

of H and V and nobody I met ever had problems interpreting that.

Morten Leikvoll wrote on Thu, 19 November 2015 06:00

Is vertical reading really that difficult?

It may not be stictly speaking all that "difficult", but that's not the

point.  It is certainly annoying and very unprofessional.  Every time you

decrease clarity and annoy those who have to read the schematic, you

increase the chance of someone misinterpreting it.  Even one such case out

in the field can be more expensive than to have drawn the schematic

properly in the first place.

Quote:

All my schematics are a mix of H and V and nobody I met ever had

problems interpreting that.

Being correct is only a necessary but not sufficient requirement for a good

schematic.  It's not about whether in the end someone was able to correctly

interpret your schematic, but how easily, with little chance of error, and

without the feeling the designer just threw slop at them.  Don't

underestimate the latter.

Neatness counts, A LOT.  Your production tech, field tech, or customer,

might not complain, but what attitude did they have when dealing with your

schematics?  Maybe they weren't even aware of it, but the "eh, who cares"

attitude exhibited by your schematic probably caused a little of the same

in them.  Eh, that solder joint is probably good, I'm not going to bother

to check with the jewler's loupe like I usually do. Eh, that cap fell on

the floor where I might have stepped on it, it's probably not cracked, good

enough, not worth my getting out of this chair to get a new one.  And so on

it goes.

When you give someone sloppy work, you set the tone and give them

permission to do the same.  Or, if you're working for them, they may decide

they like the other engineer's work better, even if they aren't even

concious of why.

I can tell you for sure that if you worked for me, you'd be making all your

text horizontal.  And if I actually had to say something to you to make

that happen, I probably wouldn't tell you this, but your lack of

professional pride in your work would be something in the back of my mind

when deciding who to assign to what task.

--

Web access to CadSoft support forums at www.eaglecentral.ca.  Where the CadSoft EAGLE community meets.

On 05.12.2015 15:17, Olin Lathrop wrote:

Morten Leikvoll wrote on Thu, 19 November 2015 06:00

Is vertical reading really that difficult?

 

It may not be stictly speaking all that "difficult", but that's not the

point.  It is certainly annoying and very unprofessional.  Every time you

decrease clarity and annoy those who have to read the schematic, you

increase the chance of someone misinterpreting it.  Even one such case out

in the field can be more expensive than to have drawn the schematic

properly in the first place.

I didn't even want to reply to this imho irrelevant topic, but since its

early morning, and my fingers need to warm up, here I go.

I think this topic belongs in the nitpicking category. It is very hard

to make all the different people happy, especially the few that has some

unique preferences or minor reading disabilities. For this topic, the

smash function covers the vertial reading disabilities and I see no

reason to change this in Eagle. I do however see the potential for

scewing up readability by placing the refdes/value totally wrong, or

even delete them. Its a function I try to stay away from, although in

some rare cases I may choose to use it.

Quote:

All my schematics are a mix of H and V and nobody I met ever had

problems interpreting that.

 

Being correct is only a necessary but not sufficient requirement for a good

schematic.  It's not about whether in the end someone was able to correctly

interpret your schematic, but how easily, with little chance of error, and

without the feeling the designer just threw slop at them.  Don't

underestimate the latter.

We are all different pepople with different skills and habits and some

of us work better together than others. I accept that, but if a customer

ever threw this request at me, I would first assume he/she was either:

-a documentation geek, demonstrating his powers

-a slave of some company rules invented by someone not even able to read

schematics.

-someone unskilled at reading schematics who is comparing to something

else he/she saw.

In any case, Im sure I could easily get them to accept vertical text.

And Im sure that's an easier job than to get this apparently

inconsistant feature into Eagle.

Neatness counts, A LOT.  Your production tech, field tech, or customer,

might not complain, but what attitude did they have when dealing with your

schematics?  Maybe they weren't even aware of it, but the "eh, who cares"

attitude exhibited by your schematic probably caused a little of the same

in them.  Eh, that solder joint is probably good, I'm not going to bother

to check with the jewler's loupe like I usually do. Eh, that cap fell on

the floor where I might have stepped on it, it's probably not cracked, good

enough, not worth my getting out of this chair to get a new one.  And so on

it goes.

Yes neatness counts, but when we have different views of what's neat,

you dont get far. I feel bad when I read schematics where the

refdes/value placement is unpredictable. I prefer drawing the sch to

make space for labels where they are, relative to the symbol.

The production attitude you mention is slightly off topic, but is

important in production line, not that much during design. Personally I

dont give a damn on details like that for my prototypes, at least in my

current work environment where there is no safety risks. Actually I dont

mind some sloppyness on prototypes, it may help me identlfy weaknesses

in both the design, and production process. I never expect production to

be 100% correct all the time, there are lot of humans involved. But I do

add selftests wherever I can.

When you give someone sloppy work, you set the tone and give them

permission to do the same.  Or, if you're working for them, they may decide

they like the other engineer's work better, even if they aren't even

concious of why.

 

I can tell you for sure that if you worked for me, you'd be making all your

text horizontal.  And if I actually had to say something to you to make

that happen, I probably wouldn't tell you this, but your lack of

professional pride in your work would be something in the back of my mind

when deciding who to assign to what task.

You make me smile and think all the thoughts I listed above. If I worked

for you and you asked me to do this on Eagle, we'd be having a hot

argument about time spending and resource utilization, and maybe end up

having dark clouds hanging over both our heads. If eagle had this as an

option already, I would still think the thoughts and not care.

If you were working for me, I think you'd do good production quality

control, at least if you behave toward people less skilled than you

From reading your posts over the years, I know we are different people,

but at least we have something in common, since we are both die hard

Eagle users.

On 12/8/2015 6:16 AM, Morten Leikvoll wrote:

Ok, this is growing out of proportions, so Im gonna end this short.

 

Yes, you are right on many of your statements (also regarding my attitude)

 

The reason why I have this attitude regarding vertical reading is kinda

related to the tool we are actually discussing. We are discussing Ealge.

I doubt any high requirement space/safety would even allow Eagle as a

design tool for their gear. I may be wrong, but I think they will have

their standards to work along, and alignment of text would be very low

down on that list of need-to-have featurs. Eagle would fail early here.

Eagle has too many issues that would not make me feel good to use in

those high requirements products.

Hi Morten,

You would be very surprised at the areas where EAGLE is allowed and the

high requirements it's used to meet. I can't name drop, but the point is

just to say that I see value in this being added to EAGLE.

I have sent an enhancement request to the devs on this issue. Now it's

up to them.

Best Regards,

Jorge Garcia

Morten Leikvoll wrote on Mon, 07 December 2015 03:28

-a slave of some company rules invented by someone not even able to

read

schematics.

That may be true, but it would be reality nonetheless.  Sometime

requirements are requirements.  All-horizontal text is sometimes a

requirement, whether you think it's silly or not.  I happen to make my

schematics that way all the time, but that doesn't matter when a large

corporate customer insists the consultant does it that way, or they'll get

a different consultant.

This really should be something Eagle could allow us to do more easily.  I

know I'd use it heavily, whether required to do so or not.

Quote:

In any case, Im sure I could easily get them to accept vertical text.

Again, sometimes it's a requirement that you don't get to negotiate if you

want the job.

Quote:

And Im sure that's an easier job than to get this apparently

inconsistant feature into Eagle.

You seem to be misunderstanding something.  I recognize you don't think the

feature is important, but inconsistant makes no sense.  All we're asking

for is the ability to have alternate symbols within a single device.  These

could then be optionally tied to specific rotation angles to make selecting

them particularly simple.

Right now you have to either rotate and smash parts, or make whole

alternate devices just to get symbols in different orientations.  Both

those make more work and give greater chance of inconsistancy.

Quote:

I feel bad when I read schematics where the

refdes/value placement is unpredictable.

Predictability makes little sense in this context.  There is nothing to

predict when the schematic is in front of you.  If you really meant

consistancy, then that makes little sense too.  In your system, the

direction towards the refdes and value are inconsistant because they depend

on how the part was rotated.  You're already not getting all refdes to the

right, for example.  You might get that for a vertical resistor, but then

the refdes would be above (and tilted 90 deg in addition) for a horizontal

resistor, or maybe below and tilted the other way, depending on whether the

stock part was rotated 90 or 27 deg.

With the proposed system, you could have refdes and values to the right for

vertical resistors and refdes above and value below for horizontal

resistors, for example.  That would be more consistant since each would be

pre-defined as separate symbols for their orientation, so less chance of

differences across the schematic.

Quote:

Actually I dont

mind some sloppyness on prototypes, it may help me identlfy weaknesses

 

in both the design, and production process.

Wait, you're actually saying manufacturing errors are good for your

prototyping process!?  Sorry, but that's just silly.  Designing a circuit

so that any resistor can be replaced by a capacitor or any other component

that comes in the same package makes absolutely no sense, and is totally

impractical anyway.  This is not designing to part tolerance ranges, but

basically random replacement of parts.

Playing along for a little, let's consider how this is supposed to work.

Your tech installed a SOT-23 zener where you specified a SOT-23 NPN

transistor.  Your circuit doesn't work and you trace it down to this

manufacturing error.  What are you going to do about it?  Redesign the

circuit to be tolerant of either a zener or transistor in that spot?

Surely you must see how silly that is.  And next time it could be a PNP

instead of a NPN, or a MOSFET.

Quote:

If I worked

for you and you asked me to do this on Eagle, we'd be having a hot

argument about time spending and resource utilization,

No, we wouldn't.  Certainly not for very long.

--

Web access to CadSoft support forums at www.eaglecentral.ca.  Where the CadSoft EAGLE community meets.

Morten Leikvoll wrote on Mon, 07 December 2015 03:28

-a slave of some company rules invented by someone not even able to

read

schematics.

That may be true, but it would be reality nonetheless.  Sometime

requirements are requirements.  All-horizontal text is sometimes a

requirement, whether you think it's silly or not.  I happen to make my

schematics that way all the time, but that doesn't matter when a large

corporate customer insists the consultant does it that way, or they'll get

a different consultant.

This really should be something Eagle could allow us to do more easily.  I

know I'd use it heavily, whether required to do so or not.

Quote:

In any case, Im sure I could easily get them to accept vertical text.

Again, sometimes it's a requirement that you don't get to negotiate if you

want the job.

Quote:

And Im sure that's an easier job than to get this apparently

inconsistant feature into Eagle.

You seem to be misunderstanding something.  I recognize you don't think the

feature is important, but inconsistant makes no sense.  All we're asking

for is the ability to have alternate symbols within a single device.  These

could then be optionally tied to specific rotation angles to make selecting

them particularly simple.

Right now you have to either rotate and smash parts, or make whole

alternate devices just to get symbols in different orientations.  Both

those make more work and give greater chance of inconsistancy.

Quote:

I feel bad when I read schematics where the

refdes/value placement is unpredictable.

Predictability makes little sense in this context.  There is nothing to

predict when the schematic is in front of you.  If you really meant

consistancy, then that makes little sense too.  In your system, the

direction towards the refdes and value are inconsistant because they depend

on how the part was rotated.  You're already not getting all refdes to the

right, for example.  You might get that for a vertical resistor, but then

the refdes would be above (and tilted 90 deg in addition) for a horizontal

resistor, or maybe below and tilted the other way, depending on whether the

stock part was rotated 90 or 27 deg.

With the proposed system, you could have refdes and values to the right for

vertical resistors and refdes above and value below for horizontal

resistors, for example.  That would be more consistant since each would be

pre-defined as separate symbols for their orientation, so less chance of

differences across the schematic.

Quote:

Actually I dont

mind some sloppyness on prototypes, it may help me identlfy weaknesses

 

in both the design, and production process.

Wait, you're actually saying manufacturing errors are good for your

prototyping process!?  Sorry, but that's just silly.  Designing a circuit

so that any resistor can be replaced by a capacitor or any other component

that comes in the same package makes absolutely no sense, and is totally

impractical anyway.  This is not designing to part tolerance ranges, but

basically random replacement of parts.

Playing along for a little, let's consider how this is supposed to work.

Your tech installed a SOT-23 zener where you specified a SOT-23 NPN

transistor.  Your circuit doesn't work and you trace it down to this

manufacturing error.  What are you going to do about it?  Redesign the

circuit to be tolerant of either a zener or transistor in that spot?

Surely you must see how silly that is.  And next time it could be a PNP

instead of a NPN, or a MOSFET.

Quote:

If I worked

for you and you asked me to do this on Eagle, we'd be having a hot

argument about time spending and resource utilization,

No, we wouldn't.  Certainly not for very long.

--

Web access to CadSoft support forums at www.eaglecentral.ca.  Where the CadSoft EAGLE community meets.