Some questions to the roadmap for next time

392 views
Skip to first unread message

White Fox

unread,
Feb 26, 2024, 9:35:02 AMFeb 26
to dev...@kicad.org
Hi guys

I joined this mailing list nearly a half year ago, just for interest,
but this time I have a question to further development, so I come with
it. Because this is my first mail here, let me introduce myself:

At first I have to say, that I have no practical experience with KiCAD.
I'm working with Altium (and, sorry, love that) with a lot experience
for this tool also as beta test group member. But as I'm always
interested at more green gras, I often joined some tool discussions to
keep an oversight what other tools can do. So even if I never designed a
board with KiCAD, I was always inquisitive about KiCAD and its
capabilities. Last years, Altium develops its product into a direction
which comes with a bad smell, and so I discuss often with my boss about
to move to KiCAD. I know that KiCAD is also used in some companies with
success.

As far I know, KiCAD has no support for multiboard projects. I looked
out for something like a road map and found something on gitlab, but
maybe outdated. Last edit was 2020.
So I would like to know, what KiCAD devs think about multiboard
projects. Are they on track for KiCAD development, was this already
discussed? Or will this overload development capacities?

I have, unfortunately, absolutly no experience in C++ (but have in C,
even if I hate C, and in Java, OOP is known to me). And I did some
multiboard designs, so I think I could at least contribute some
experience. (Unfortunately, Altium supports multiboard projects, but in
a poorly way, it could be much more powerfull.)

The same I wrote for multiboard projects I could repeat for pcb
variation support, but I better do not and just mentioned it.

I'm looking forward for your answers.
Greetings from germany

Oliver

--
Das Nötige ist einfach und das Komplizierte unnötig
-Michail Kalaschnikow


As a german, please let me tell you: Germans are not rude.
If you invite a german to have lunch with you and he says simply 'No', be assured that he means:
«I'm so glad that you ask me to spend lunch time with you, I'm sure it would be a great time,
but unfortunately, I'm deeply sorry for that, I have a meeting with my boss and a few minutes later,
I promised our most important customer a call.
And moreover, I'm a little embarassed for that, I feel a bit sick today, maybe I'm not a good lunch companion today.
But I really hope so that we can share this great time another time.»

The reason for this kind of highly efficient conversation, which is typical for germans, is quite simple:
He knows that you are probably very busy with highly important business,
and it is a special kind of German courtesy not to want to waste your precious time.
So please, don't misunderstand short and precise communication as rowdyness.

Jon Evans

unread,
Feb 26, 2024, 9:47:56 AMFeb 26
to dev...@kicad.org
Hello Oliver,

On Mon, Feb 26, 2024 at 9:34 AM 'White Fox' via KiCad Developers <dev...@kicad.org> wrote:

As far I know, KiCAD has no support for multiboard projects. I looked
out for something like a road map and found something on gitlab, but
maybe outdated. Last edit was 2020.

Personally I have come to the conclusion that long-term roadmaps are of limited use, especially for KiCad.
At the moment we focus more on planning for the next upcoming release.  Anything that is not being actively worked on for that next release is considered "unplanned".
 
So I would like to know, what KiCAD devs think about multiboard
projects. Are they on track for KiCAD development, was this already
discussed? Or will this overload development capacities?

This falls into the category of things that have been talked about, but there are no immediate plans and nobody is working on it.
 
 I think I could at least contribute some
experience. (Unfortunately, Altium supports multiboard projects, but in
a poorly way, it could be much more powerfull.)

The first step would be to have some discussion of what problems we are trying to solve, and then we can brainstorm possible ways of solving them.
Often, people who are experienced with other tools like to talk about features in terms of the other tools: "KiCad should do it like X", or "KiCad should not do it like Y".
What I find more useful is to make sure to discuss what people actually need to do, independent of what tool they use, and agree on that first.
It is then easier to discuss the advantages and disadvantages of various possible ways of doing it.

For "multiboard projects" I think if you ask 5 people, you will possibly get 5 different answers about what this feature means, and what they want to do with it.
So, figuring this out (and getting it documented) is a good first step.
 

The same I wrote for multiboard projects I could repeat for pcb
variation support, but I better do not and just mentioned it.

Variants are planned and being more actively worked on, although there are a number of things that have to happen first, and I think schematic variant support will come before PCB variant support.

Best,
Jon

Rafał Pietrak

unread,
Feb 26, 2024, 1:14:20 PMFeb 26
to dev...@kicad.org
Dear everybody,

W dniu 26.02.2024 o 15:47, Jon Evans pisze:
[----------------]
>
> For "multiboard projects" I think if you ask 5 people, you will possibly
> get 5 different answers about what this feature means, and what they
> want to do with it.
> So, figuring this out (and getting it documented) is a good first step.

If that could be of any help, and should anybody have already ideas
and/or experience regarding multiboard design in form of loose comments
or discussion, I can gather those into an initial documentation and put
it on a github somewhere.


-R

White Fox

unread,
Feb 26, 2024, 1:22:27 PMFeb 26
to dev...@kicad.org
Hi Rafal

Well, this seems to me exactly like what I would start with, thank you.
Where is this snippet from? I can't remember such an email thread here.

greetings
Oliver

Rafał Pietrak

unread,
Feb 26, 2024, 2:09:52 PMFeb 26
to dev...@kicad.org
Hi Oliver,

You've asked for snippet source - it came from this email by Jon.

-R

W dniu 26.02.2024 o 15:47, Jon Evans pisze:
> --
> You received this message because you are subscribed to the Google
> Groups "KiCad Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to devlist+u...@kicad.org
> <mailto:devlist+u...@kicad.org>.
> To view this discussion on the web visit
> https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCCF2av3Efih-fP_ZVcum3Q45tBgpMAQ6nrN5GZ6mOMabg%40mail.gmail.com <https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCCF2av3Efih-fP_ZVcum3Q45tBgpMAQ6nrN5GZ6mOMabg%40mail.gmail.com?utm_medium=email&utm_source=footer>.

White Fox

unread,
Feb 26, 2024, 3:07:16 PMFeb 26
to dev...@kicad.org
Oh, sorry. The fault was on my side.
My email provider moved the mail from Jon into the junk folder. I missed
it, because I did not thought about and wondered about Rafals mail.
Thanks for clarifying, sorry for this confusion.

Back to topic:
Even if I worked with another tool and have no experience with KiCAD, I
think I know that KiCAD is not Altium (Eagle, Cadence, PADS, ...). And
even I'm inspired by, I thik I have some more ideas and wishes which are
nowhere implemented yet, so don't worry about my working background. ;)

Where would you normally discuss topics like this? The mailing list
would be fine for me, but I'm afraid of waisting and confusing by
discussing multiple topics in that way here. Maybe you have a better
place already?

greetings
Oliver

Jon Evans

unread,
Feb 26, 2024, 4:39:07 PMFeb 26
to dev...@kicad.org
Normally it is easiest to have an initial discussion on a GitLab issue as comments.

At some point if we agree on a direction, the next step is to make a specification document (typically with Google docs but not required).

To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.
To view this discussion on the web visit https://groups.google.com/a/kicad.org/d/msgid/devlist/5449b0b8-8ed5-408b-8736-b44e0ebefd65%40gmx.de.

Rafał Pietrak

unread,
Feb 29, 2024, 2:57:26 AMFeb 29
to dev...@kicad.org
Hi Oliver,

The subject somehow died out, but I'd like to followup:
1. I've looked up current kicad github for guidance, and it looks pretty
"overpopulated" - lots of small repositories without any sort of
"oversight". Many marked as "legacy", but still visible ... and in
consequence creating "information noise".
2. so i might have missed a place where such discussion (documentation
gathering) currently takes place - pls advice.
3. if there is none and I create one, how do I invite kicad (and other
tools) users to join the discussion?
4. is there a naming convention to use for a project, or "multiboard"
would be just fine?

-R

W dniu 26.02.2024 o 19:22, 'White Fox' via KiCad Developers pisze:

Kliment (Future Bits)

unread,
Feb 29, 2024, 3:06:23 AMFeb 29
to dev...@kicad.org
Hello Rafał,

KiCad development was never on GitHub - it was originally on Launchpad
and later moved to GitLab. The issue (and therefore feature request)
tracker is at https://gitlab.com/kicad/code/kicad/-/issues/ - you can
search for related issues there, or start a new one.

Hope this helps

-Kliment

White Fox

unread,
Feb 29, 2024, 3:18:07 AMFeb 29
to dev...@kicad.org
Hi everyone

@Kliment: Thanks for that information

@Rafal:
Well, I thought about how to describe, and I would had choosen another
term then multiboard, I would had prefered something like module(s) project.
The reason for that is, that multiboard is the naming for that domain in
Altium, and to start a planning from scratch, it may be a good idea to
get a little distance to what I know and use before. But this is just my
personal preference, and the term module may be already used in other
tools. So, you may name it in a way you think it's best.
Nevertheless, I would start planning from scratch with a definition of
all important and used terms to clearify what we discussing about, so
the exact naming may be less important.

I have a gitlab and a github account, just tell me where to join the
discussion.

regards
Oliver

Rafał Pietrak

unread,
Feb 29, 2024, 7:08:40 AMFeb 29
to dev...@kicad.org
Hmmm... I just tried to setup an account there (on gitlab), and it asked
me for a phone number "for confirmation" - that's unacceptable invasion
of privacy. So, (regrettably) I'm not be able to join you there.

BR.

Rafał

W dniu 29.02.2024 o 09:06, Kliment (Future Bits) pisze:

Seth Hillbrand

unread,
Feb 29, 2024, 12:25:19 PMFeb 29
to dev...@kicad.org
GitLab is where all development happens.  If their requirements are too onerous then, while we are sorry to lose you, there are extremely limited ways that you'll be able to be involved in KiCad at all.

You can still offer translations at https://hosted.weblate.org/projects/kicad/ but that is all.  Everything else happens on GitLab.  We do not accept patches through the mailing list.

It is interesting to me that you would consider phone numbers to be private information.  We used to print huge books with everyone's name, address and phone number in them.  Then everyone got a copy of the book.  Of course, I put my phone number in my e-mail signature, so maybe I'm missing something.  Regardless, GitLab isn't using your phone number for nefarious purposes, they offer a free service with free CI time on their shared machines.  Before they asked for verification information, cryptobros would abuse this for bitcoin mining and everyone suffered.  Now, projects like KiCad are able to use the CI machines for testing builds and finding bugs.

Seth

KiCad Services Corporation Logo
Seth Hillbrand
Lead Developer
+1-530-302-5483
Long Beach, CA
www.kipro-pcb.com    in...@kipro-pcb.com


--
You received this message because you are subscribed to the Google Groups "KiCad Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.

Rafał Pietrak

unread,
Feb 29, 2024, 1:15:36 PMFeb 29
to dev...@kicad.org
Seth,

I really do regard my phone number as private information and even if I
share it with my physician without hesitation, giving it out to a global
corpo is out of the question for me. I do understand, that they may have
some internal policies based on some reasoning, and I'm not going to
argue with that. Let them have it. Only I'm not going to accept that.

Then again. I do UNDERSTAND, that I won't be able to contribute to the
project. I really do get it. My intention in writing the massage was
only that:
1. the core developers may not be aware, that this sort of "additional
confirmation" could be unfriendly to people. (no matter argumentation).
2. and I wanted to tell Oliver that at this point I won't be
contributing to this "multiboard docs" project. I'm sorry.

Anyway, thenx everybody for wonderful KiCAD and see you somewhere else.

bye.

-R

W dniu 29.02.2024 o 18:24, Seth Hillbrand pisze:
> GitLab is where all development happens.  If their requirements are too
> onerous then, while we are sorry to lose you, there are extremely
> limited ways that you'll be able to be involved in KiCad at all.
>
> You can still offer translations at
> https://hosted.weblate.org/projects/kicad/
> <https://hosted.weblate.org/projects/kicad/> but that is all.
> Everything else happens on GitLab.  We do not accept patches through the
> mailing list.
>
> It is interesting to me that you would consider phone numbers to be
> private information.  We used to print huge books with everyone's name,
> address and phone number in them.  Then everyone got a copy of the
> book.  Of course, I put my phone number in my e-mail signature, so maybe
> I'm missing something.  Regardless, GitLab isn't using your phone number
> for nefarious purposes, they offer a free service with free CI time on
> their shared machines.  Before they asked for verification information,
> cryptobros would abuse this for bitcoin mining and everyone suffered.
> Now, projects like KiCad are able to use the CI machines for testing
> builds and finding bugs.
>
> Seth
>
> KiCad Services Corporation Logo
> Seth Hillbrand
> *Lead Developer*
> +1-530-302-5483
> Long Beach, CA
> www.kipro-pcb.com <https://www.kipro-pcb.com/> in...@kipro-pcb.com
> <mailto:in...@kipro-pcb.com>
> <mailto:devlist%2Bunsu...@kicad.org>.
> To view this discussion on the web visit
> https://groups.google.com/a/kicad.org/d/msgid/devlist/a76d8ad8-7505-40f2-9108-67b2c48f4889%40electric-sheep.eu <https://groups.google.com/a/kicad.org/d/msgid/devlist/a76d8ad8-7505-40f2-9108-67b2c48f4889%40electric-sheep.eu>.
>
> --
> You received this message because you are subscribed to the Google
> Groups "KiCad Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to devlist+u...@kicad.org
> <mailto:devlist+u...@kicad.org>.
> To view this discussion on the web visit
> https://groups.google.com/a/kicad.org/d/msgid/devlist/CAFdeG-oLhUNkkdVFeHsxHEP%2B2r-jGrSYg2DKUi%3Dk1U83fGcwOw%40mail.gmail.com <https://groups.google.com/a/kicad.org/d/msgid/devlist/CAFdeG-oLhUNkkdVFeHsxHEP%2B2r-jGrSYg2DKUi%3Dk1U83fGcwOw%40mail.gmail.com?utm_medium=email&utm_source=footer>.

White Fox

unread,
Feb 29, 2024, 2:38:32 PMFeb 29
to dev...@kicad.org
Hi guys

That's sad that the topic goes this way.
@Rafal: I have never tried it and do not want to advertise this service,
but maybe this (or something similiar) is an option for you?
https://quackr.io/

best regards
Oliver

Rafał Pietrak

unread,
Feb 29, 2024, 3:39:30 PMFeb 29
to dev...@kicad.org
Oliver,

Thenx. If it works, it's fine by me. I'll give it a try in three days
(my username is used/reserved at gitlab for that duration).

-R

W dniu 29.02.2024 o 20:38, 'White Fox' via KiCad Developers pisze:

Eeli Kaikkonen

unread,
Mar 1, 2024, 2:53:21 AMMar 1
to dev...@kicad.org
On Thu, Feb 29, 2024 at 7:25 PM Seth Hillbrand <se...@kipro-pcb.com> wrote:
GitLab is where all development happens.  If their requirements are too onerous then, while we are sorry to lose you, there are extremely limited ways that you'll be able to be involved in KiCad at all.
On Thu, Feb 29, 2024 at 4:08 AM 'Rafał Pietrak' via KiCad Developers <dev...@kicad.org> wrote:
Hmmm... I just tried to setup an account there (on gitlab), and it asked
me for a phone number "for confirmation" - that's unacceptable invasion
of privacy. So, (regrettably) I'm not be able to join you there.

BR.

Rafał

This seems to  be a new direction for gitlab, and people aren't happy with it. See for example https://www.reddit.com/r/gitlab/comments/182orq6/cannot_create_new_account_without_phone_number/. This affects KiCad and many other projects. New people won't even write bug reports because of this. I don't know what to think about this, my first reaction is anger and a suggestion to find another alternative for the KiCad project, no matter how difficult it is.

Eeli Kaikkonen

Mark Roszko

unread,
Mar 1, 2024, 7:16:55 AMMar 1
to dev...@kicad.org
GitLab, like any other platform, has been suffering from cryptocurrency abuse which has been resulting in more and more lock down of signups. 
There's a new bull wave in crypto so of course the abuse has ramped up.


However, I just confirmed that the phone number verification is risk assessment based like the other verification they have. 
If it prompts you for a phone number or even credit card to validate, it is because it thinks you are a high risk signup.

I just made a new account with absolutely zero phone number verification prompt. Heh.


There really is no good solution here because other "OSS platforms" completely ban CI from users that aren't verified projects due to cryptocurrency abuse, so forks don't get CI for example.


GitHub otoh attracts a very toxic userbase which we don't want anything to do with. We already have to suffer the occasional drive-by fling of insults on Gitlab, but GitHub lowers the barrier to toxicity to the floor.
Not to mention it's UI is very poor for large orgs and repos.


--
You received this message because you are subscribed to the Google Groups "KiCad Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.

White Fox

unread,
Mar 12, 2024, 4:17:55 AMMar 12
to dev...@kicad.org
Hi Rafal, did you managed to join gitlab?

Is there somethign new on this front side?

greetings
Oliver Lenz

Rafał Pietrak

unread,
Mar 12, 2024, 7:25:02 AMMar 12
to dev...@kicad.org
Ho Oliver,

Yes, I've tried, and with numbers from US, from swiss and even from
pakistan - all that to no avail. gitlab declined verification almost
immediately, so I won't be able join the work there. Sorry.

bye.

-R

W dniu 12.03.2024 o 09:17, 'White Fox' via KiCad Developers pisze:

Jon Evans

unread,
Mar 12, 2024, 10:15:34 AMMar 12
to dev...@kicad.org
For what it's worth, I'm guessing that if you are trying phone numbers from around the world, you are getting these phone numbers from VOIP services and they aren't actually linked to a physical phone that you control, which is kind of exactly what GitLab is trying to block...

I don't claim that GitLab's system is perfect or ideal, but the point is that they are giving anyone who signs up free resources, and they have a lot of trouble with abuse of their resources.  In order to get access to those resources, you need to be willing to disclose enough information to them for them to feel like you are a low risk (because, if you started abusing their system, you'd be "burning" the good will of your identity).

-Jon

--
You received this message because you are subscribed to the Google Groups "KiCad Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.
To view this discussion on the web visit https://groups.google.com/a/kicad.org/d/msgid/devlist/3aa37107-fe4b-473e-9e69-a58b1b7e143e%40electric-sheep.eu.

Rafał Pietrak

unread,
Mar 12, 2024, 10:45:59 AMMar 12
to dev...@kicad.org
Hi,

W dniu 12.03.2024 o 15:15, Jon Evans pisze:
> For what it's worth, I'm guessing that if you are trying phone numbers
> from around the world, you are getting these phone numbers from VOIP
> services and they aren't actually linked to a physical phone that you
> control, which is kind of exactly what GitLab is trying to block...

The numbers from that site look like being able to receive activation
codes for verity of services (including google). The receive box is
publicly visible, so I could browse the history and see many such codes
there. So, even if they are VoIP, they get the SMS from senders.

But there is a limited count of those numbers, so they are easy to
enumerate on blacklists, and that's what gitlab probably did.

>
> I don't claim that GitLab's system is perfect or ideal, but the point is
> that they are giving anyone who signs up free resources, and they have a
> lot of trouble with abuse of their resources.  In order to get access to
> those resources, you need to be willing to disclose enough information
> to them for them to feel like you are a low risk (because, if you
> started abusing their system, you'd be "burning" the good will of your
> identity).

I have a different opinion:
1. I haven't started accessing their resources (unless you count putting
a phrase into an input-text field as such "access")
2. They've started abusing my patience.

BTW: the only problem they could have with abusing their resources is
through their free CI/CD - and this can be prevented easily by allowing
only "thoroughly verified" users to commit code to the builds. Or they
could disabled CI/CD for the "free" projects and be 100% on the safe
side. The didn't. They have overdone their "security".

So I'm very sorry, but that's a nogo for me.

-R
PS: this is out off the topic here, but since I'm probably be signing
off the list in near future, I'd like to ask a question regarding KiCAD
design. After years of doing the ordinary sequence of: draw-sch,
numerate, footprint assign, export to pcb ... I've looked up the files
and to my surprise found out, that the footprints are actually stored in
the PCB, not in the SCH. So my question is: is the footprint assignment
BEFORE netlist export to PCB a "design feature", or is it implemented
this way because of some legacy code from the past? (the question
originates from pondering of multiboard/multi-pcb-variant design options)
> >>> https://quackr.io/ <https://quackr.io/>
> >>>>> www.kipro-pcb.com <http://www.kipro-pcb.com>
> <https://www.kipro-pcb.com/ <https://www.kipro-pcb.com/>>
> in...@kipro-pcb.com <mailto:in...@kipro-pcb.com>
> >>>>> <mailto:in...@kipro-pcb.com <mailto:in...@kipro-pcb.com>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> On Thu, Feb 29, 2024 at 4:08 AM 'Rafał Pietrak' via KiCad
> Developers
> >>>>> <dev...@kicad.org <mailto:dev...@kicad.org>
> <mailto:devlist%2Bunsu...@kicad.org>.
> To view this discussion on the web visit
> https://groups.google.com/a/kicad.org/d/msgid/devlist/3aa37107-fe4b-473e-9e69-a58b1b7e143e%40electric-sheep.eu <https://groups.google.com/a/kicad.org/d/msgid/devlist/3aa37107-fe4b-473e-9e69-a58b1b7e143e%40electric-sheep.eu>.
>
> --
> You received this message because you are subscribed to the Google
> Groups "KiCad Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to devlist+u...@kicad.org
> <mailto:devlist+u...@kicad.org>.
> To view this discussion on the web visit
> https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCA-%3DAsZk2x0qyZnPT%3DajyREsW%2B98fGqH5rQC%3D00C71Qkw%40mail.gmail.com <https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCA-%3DAsZk2x0qyZnPT%3DajyREsW%2B98fGqH5rQC%3D00C71Qkw%40mail.gmail.com?utm_medium=email&utm_source=footer>.

White Fox

unread,
Mar 12, 2024, 3:24:15 PMMar 12
to dev...@kicad.org
Well, that's sad.

I also avoided to start new registrations some months ago, so I
understand your point, but nevertheless, I would like to continue.
Maybe some other people can invite me to the project, or what ever must
done before I can start to cloak some conception scratchs in.

In the meanwhile, I would start to write a first shot down. Are txt
files fine, or is another file format prefered?

best regards
Oliver Lenz

Ladislav Láska

unread,
Mar 12, 2024, 8:50:22 PMMar 12
to dev...@kicad.org
Hi,

I currently have a few disposable numbers I use for local internet in Hongkong. I'll let them expire in a few months, but if someone would like the number for gitlab registration, reach out to me personally and we can figure out how to proceed.

I assume the number is not necessary after initial registration and different 2FA is set up.

Cheers,
Ladislav
--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

Mark Roszko

unread,
Mar 13, 2024, 11:01:47 PMMar 13
to dev...@kicad.org
>  So my question is: is the footprint assignment
> BEFORE netlist export to PCB a "design feature", or is it implemented
> this way because of some legacy code from the past? (the question
> originates from pondering of multiboard/multi-pcb-variant design options)

It is an intended design feature.

Schematics often depend on the footprint. Many components, even modern microcontrollers may come in multiple packages simultaneously, a designer may actually lose pins and features depending on package. Etc.

Not to mention when you are designing electronics, you almost always want to size your resistors and capacitors at the schematic stage where you can capture critical design parameters such as max currents and voltages that drive the selection of components.

> that the footprints are actually stored in the PCB, not in the SCH. 

The actual footprint is stored in the PCB as a cache to the footprint that was on disk when it was loaded.
The footprint assignment is stored in both. 

To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.
To view this discussion on the web visit https://groups.google.com/a/kicad.org/d/msgid/devlist/eaa99e2e-869a-4fdb-9e00-e571c74a78c3%40electric-sheep.eu.

Rafał Pietrak

unread,
Mar 14, 2024, 3:11:02 AMMar 14
to dev...@kicad.org
Hi Mark,

Your explanations are most surprising to me.

In particular, from the fact, that "many components come in multiple
package variant simultaneously" I'd rather draw a conclusion, that the
SCH/PCB tools should let the designer maintain that flexibility; and
with the same SCH allow him/her to work out different variants of PCB
for different variants of component packaging (or for that matter:
currents/voltages/etc) "rooted" at the same "logic", laid down on that SCH.

But thenx for clearing it up for me. It's a design foundation, I get it.

-R
BTW: following an advice, I've registered myself at github (no phone
number required), and attempt to sign-in to gitlab through that other
service - regretably that doesn't work either. gitlab gave me a 402
error page on that attempt.


W dniu 14.03.2024 o 04:01, Mark Roszko pisze:
> <http://www.kipro-pcb.com <http://www.kipro-pcb.com>>
> >     <mailto:devlist%2Bunsu...@kicad.org
> <mailto:devlist%252Buns...@kicad.org>>.
> >     To view this discussion on the web visit
> >
> https://groups.google.com/a/kicad.org/d/msgid/devlist/3aa37107-fe4b-473e-9e69-a58b1b7e143e%40electric-sheep.eu <https://groups.google.com/a/kicad.org/d/msgid/devlist/3aa37107-fe4b-473e-9e69-a58b1b7e143e%40electric-sheep.eu> <https://groups.google.com/a/kicad.org/d/msgid/devlist/3aa37107-fe4b-473e-9e69-a58b1b7e143e%40electric-sheep.eu <https://groups.google.com/a/kicad.org/d/msgid/devlist/3aa37107-fe4b-473e-9e69-a58b1b7e143e%40electric-sheep.eu>>.
> >
> > --
> > You received this message because you are subscribed to the Google
> > Groups "KiCad Developers" group.
> > To unsubscribe from this group and stop receiving emails from it,
> send
> > an email to devlist+u...@kicad.org
> <mailto:devlist%2Bunsu...@kicad.org>
> > <mailto:devlist+u...@kicad.org
> <mailto:devlist%2Bunsu...@kicad.org>>.
> > To view this discussion on the web visit
> >
> https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCA-%3DAsZk2x0qyZnPT%3DajyREsW%2B98fGqH5rQC%3D00C71Qkw%40mail.gmail.com <https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCA-%3DAsZk2x0qyZnPT%3DajyREsW%2B98fGqH5rQC%3D00C71Qkw%40mail.gmail.com> <https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCA-%3DAsZk2x0qyZnPT%3DajyREsW%2B98fGqH5rQC%3D00C71Qkw%40mail.gmail.com?utm_medium=email&utm_source=footer <https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCA-%3DAsZk2x0qyZnPT%3DajyREsW%2B98fGqH5rQC%3D00C71Qkw%40mail.gmail.com?utm_medium=email&utm_source=footer>>.
>
> --
> You received this message because you are subscribed to the Google
> Groups "KiCad Developers" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to devlist+u...@kicad.org
> <mailto:devlist%2Bunsu...@kicad.org>.
> To view this discussion on the web visit
> https://groups.google.com/a/kicad.org/d/msgid/devlist/eaa99e2e-869a-4fdb-9e00-e571c74a78c3%40electric-sheep.eu <https://groups.google.com/a/kicad.org/d/msgid/devlist/eaa99e2e-869a-4fdb-9e00-e571c74a78c3%40electric-sheep.eu>.
>
> --
> You received this message because you are subscribed to the Google
> Groups "KiCad Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to devlist+u...@kicad.org
> <mailto:devlist+u...@kicad.org>.
> To view this discussion on the web visit
> https://groups.google.com/a/kicad.org/d/msgid/devlist/CAJjB1qJyWLag3%2BoAkPEn46MNZPjGsVbFMugp_Z%2BDn2R9V_GbSw%40mail.gmail.com <https://groups.google.com/a/kicad.org/d/msgid/devlist/CAJjB1qJyWLag3%2BoAkPEn46MNZPjGsVbFMugp_Z%2BDn2R9V_GbSw%40mail.gmail.com?utm_medium=email&utm_source=footer>.

Leah Leen

unread,
Apr 9, 2024, 4:18:54 PMApr 9
to KiCad Developers, Rafał Pietrak
Hi Rafał,

Don't give up just because that one didn't work. Why wouldn't you try out another one?
I would recommend using this service -  https://anonymsms.com/
Their numbers work for my verification purposes. You can rent or get one for free, so you decide

Best Regards
Leah

Rafał Pietrak

unread,
Apr 11, 2024, 9:57:39 AMApr 11
to Leah Leen, KiCad Developers
Leah,

Yes, I gave up - from my perspective it's a wast of time.

But since you give me this much undeserved support, I tried this service
a moment ago ... one (US) number failed for gitlab, and on another
attempt gitlab asked me for a credit card number.

I'm plain speachless.

-R
BTW: I'm currently testing 8.99, and quite nice ReplicateLayout plugin
doesn't work anymore. The author says it's a KiCAD problem, not his ...
so I think I'll have some hacking to do instead of trying to force
gitlab doors open.
> >>> https://quackr.io/ <https://quackr.io/>
> >>>>> +1-530-302-5483 <tel:(530)%20302-5483>‬
> >>>>> Long Beach, CA
> >>>>> www.kipro-pcb.com <http://www.kipro-pcb.com>
> <https://www.kipro-pcb.com/ <https://www.kipro-pcb.com/>>
Reply all
Reply to author
Forward
0 new messages