Copper zones won't fill

[Simon]

First they were there, then i assigned them to a net, now I only see the outline. Yes I have pressed "B", yes I have clicked on the filled view mode, a zone on the front mask behaves, why do the copper zones do different?

[Simon]

This now seems to affect the abilitf to via stitch as without the zones filling the via's have no net to lock onto as it's not there.

[NivagSwerdna]

The first time I used a filled zone in KiCad they disappeared... in my case this was because the Cu layer where I was creating the fill was not visble.

By the time I figured that out I had many filled zones

 

[Simon]

My layers are all visible, all 3 of them that have unpoured copper fills. If I make them have no net they pour, if I attach them to a net they vanish, just the outline remains.

[thinkfat]

My layers are all visible, all 3 of them that have unpoured copper fills. If I make them have no net they pour, if I attach them to a net they vanish, just the outline remains.

IIRC, if you have a net that is not connected to any pad, the zone will not fill. Also if the zone is not connected to any pad.

[Simon]

Yes I have them working now. But I am at a loss as to why there is this "rule". What I want to do is not insane from a manufacturing point of view, but apparently KiCad knows better than me.

[thinkfat]

Well, at least that rule is somewhat sensible and you might be able to remember it.

At the very least it is more sensible than the completely obscure feature in pcbnew with selecting components: Did you know it makes a difference whether you open the selection box left-to-right vs. right-to-left? In the first case the selection includes only items within the box, in the second case it includes also items that intersect with the box.

WTF were they smoking??? 

[Simon]

Well, at least that rule is somewhat sensible and you might be able to remember it.

At the very least it is more sensible than the completely obscure feature in pcbnew with selecting components: Did you know it makes a difference whether you open the selection box left-to-right vs. right-to-left? In the first case the selection includes only items within the box, in the second case it includes also items that intersect with the box.

WTF were they smoking??? 

No that rule on copper zones is barmy and is messing with my design.

The selection behavior you describe is very sensible and a historic feature in CAD packages like autocad. It makes it easier to selectively select things but not others. If you drag right you have to fully encompass the part, this means that you can pass over other items that you have to pass over but don't want but they won't be selected. If you drag left you can select in a smaller area avoiding surrounding parts or including them just by partially selecting them but avoiding encompassing parts you don't want. It's actually quite useful but you need to make it work for you.

[NivagSwerdna]

WTF were they smoking??? 

I actually like that feature, let's you start or extend a selection with just the mouse.

[JackJones]

Did you know it makes a difference whether you open the selection box left-to-right vs. right-to-left? In the first case the selection includes only items within the box, in the second case it includes also items that intersect with the box.

WHAT!? Selecting things outside the box I draw has been annoying me for so long, and now you're telling me all I had to do was reverse my selection order? 

[Simon]

Yup, been an autocad feature for decades so logically carried over to other cad packages.

[Monkeh]

No that rule on copper zones is barmy and is messing with my design.

For what reason would you want a copper pour which is assigned to a net but never connects to it?

[Simon]

No that rule on copper zones is barmy and is messing with my design.

For what reason would you want a copper pour which is assigned to a net but never connects to it?

Initially there was not even a net, but without a net you can't via stitch but then without parts you can't have a net. This layer is connected to the case chassis and shielding gaskets. Apparently this is a feature.

[thinkfat]

No that rule on copper zones is barmy and is messing with my design.

For what reason would you want a copper pour which is assigned to a net but never connects to it?

Initially there was not even a net, but without a net you can't via stitch but then without parts you can't have a net. This layer is connected to the case chassis and shielding gaskets. Apparently this is a feature.

Yeah, that's indeed a bit dumb. You'll have to provide a dummy part connected to a chassis-ground net so that the pours are filled. But once you remember it, it's easy enough to work around it.

The odd selection feature - that's completely obscure and entirely un-discoverable. I start dragging my selections from where it's convenient, i.e. where I have space on the board not covered by components. That might be right or left of the desired group. In fact this selection feature only makes sense if you checked "Prefer selection to dragging" in the Pcbnew preferences. If you didn't, for whatever reason - tough luck, the selection logic can not be changed. 

[Simon]

The selection logic makes sense, as i said it's not a feature of KiCad but an industry standard that all CAD packages use.

[thinkfat]

The selection logic makes sense, as i said it's not a feature of KiCad but an industry standard that all CAD packages use.

It doesn't "make sense". If it made sense, it would be completely obvious how it works. Arguing that "all CAD packages do it like this" (a bold, highly disputable claim) does not create "sense".

[Simon]

well it's undocumented like my issue was. but it's a feature I use all the time.

[thinkfat]

I guess I should have watched Chris Gammells Kicad 5.0 introduction videos with more than half an eye... I could have learned something new.

Still, I did a quick survey among some friends in the business:
- Eagle (up to version 7.7 at least): Nope. It may have changed now, because Autodesk.
- CircuitMaker: Nope
- Altium Designer: Nope
- Nemetschek (Architecture-CAD completely unknown to me): Nope
- EasyEDA: YES!

The survey additionally yielded two "WTF?" and one "I already hated KiCad anyway" response 

[Wolfram]

I guess I should have watched Chris Gammells Kicad 5.0 introduction videos with more than half an eye... I could have learned something new.

Still, I did a quick survey among some friends in the business:
- Eagle (up to version 7.7 at least): Nope. It may have changed now, because Autodesk.
- CircuitMaker: Nope
- Altium Designer: Nope
- Nemetschek (Architecture-CAD completely unknown to me): Nope
- EasyEDA: YES!

The survey additionally yielded two "WTF?" and one "I already hated KiCad anyway" response 

Altium Designer works like this as well, as demonstrated in this video:

[PlainName]

The odd selection feature - that's completely obscure and entirely un-discoverable.

Yes, you're right. But how could it be discoverable?

I'm a fan of the feature (in other applications), and in pretty much all of them I either stumbled across it or discovered it late (after seeing someone use it and having it explained). I presume that's because it's the kind of sub-feature that would quick-start tutorials tediously long if they got mentioned, never mind detailed. I think it was mentioned in one quick-start but, being a sub-feature detail, it just got lost in the noise of every other sub-feature detail bottlenecking my poor brain at the time.

So, how could it be discoverablel? In Altium it is one colour for right drag, another colour for left drag. Obvious once you know, but it's the first knowing, obviously. Having a pop-up ("Hey, did you know you can drag this the other way for a different effect? Try it now!") every time you do it wouldn't be a good idea. Maybe a tiny icon attached to the pointer as you drag, but a) tiny icons usually mean nothing at all because you can't figure out what they are, and b) how could you display what it is and what it could be in a tiny icon?

Sometimes things are difficult to do. Doesn't make them bad, just difficult.

[delfinom]

Still, I did a quick survey among some friends in the business:
- EasyEDA: YES!

I find it funny that people accept using a cloud service run out of China for their intellectual property....

/siderant

[Simon]

Why would you trust any cloud based system? the same can be said for circuit maker. I went banana's when I found that KiCad by default wanted to use online libraries that I had no control over. Fortunately it's now clearer, I don't know if it still does that but it was a piece of cake setting my libraries up properly this time.

[Watth]

Well, at least that rule is somewhat sensible and you might be able to remember it.

At the very least it is more sensible than the completely obscure feature in pcbnew with selecting components: Did you know it makes a difference whether you open the selection box left-to-right vs. right-to-left? In the first case the selection includes only items within the box, in the second case it includes also items that intersect with the box.

WTF were they smoking??? 

This kind of selection method is not used only by pcbnew, and is very helpful.

[Bassman59]

Well, at least that rule is somewhat sensible and you might be able to remember it.

At the very least it is more sensible than the completely obscure feature in pcbnew with selecting components: Did you know it makes a difference whether you open the selection box left-to-right vs. right-to-left? In the first case the selection includes only items within the box, in the second case it includes also items that intersect with the box.

WTF were they smoking??? 

This kind of selection method is not used only by pcbnew, and is very helpful.

I did not realize that this was how the selection box thing worked ... I just tested it and it's actually very useful.

[Simon]

the thread is over 1 year old and KiCad 6 is now out!

[Cerebus]

the thread is over 1 year old and KiCad 6 is now out!

Well it's lucky you're not trying to learn how to use zone fills for the first time in KiCad 6.

4.4. Working with footprints
NOTE
TODO: Write this section - covers footprint properties, updating from library, etc.
4.5. Working with pads
NOTE
TODO: Write this section - covers pad properties
4.6. Working with zones
NOTE
TODO: Write this section

Funny, but I have a very different idea of "It's finished" as in "It's finished, ship it" than the KiCad developers obviously do. 

[Simon]

why do you think the developers think it is finished?

[Cerebus]

why do you think the developers think it is finished?

Well, call me old fashioned, but when you've gone through two release candidates, then you bump the version number and make a release announcement it's becuase you've got a complete, finished fully functional product. Sure, you may have plans for the next version, but that shouldn't include "complete the help files for the previous version".

[Simon]

Yea, I guess so. But these are people spread across the globe so maybe someones timing slipped by they didn't want to mess up the xmas day release. Is it that different from 5? If not I guess not too much sweat.

I'm just glad it exists and is now pretty good for a free product having done the round of all the affordable ones and seen that even the god software is no better supported than the low end. And if all I need to do to keep it supported is may sure all work prototypes go through Aisler with a donation that means I am getting work to sort of pay something towards it too. Last set I put through my boss went down the list of stuff before paying and got to the donation bit which I justified and he was like yea sure why not. Cheapest software I've ever used with the best support.

[eugene]

why do you think the developers think it is finished?

Well, call me old fashioned, but when you've gone through two release candidates, then you bump the version number and make a release announcement it's becuase you've got a complete, finished fully functional product. Sure, you may have plans for the next version, but that shouldn't include "complete the help files for the previous version".

You're free to continue using V5.1 until you've decided that V6 is finished, no matter how you define 'finished.'

EDIT: That sounded snarky, and to be completely honest, I might have wanted it to. But, you have been arguing that the developers gave a version of the software the title of "released" when (in your opinion) they should have waited until a later version. OK, well, if that's not just semantics, then it's not far from just semantics. So, if you disagree with the developers that's your prerogative. Just wait until we get to the version that you would have released and start using that.

Seriously, it's FREE software. You should be supporting the developers. Start a college fund for their kids, or maybe just be kind and show appreciation for the hundreds of hours that many of them have individually contributed. It's OK to scoff off the retort "if you don't like it then fix the code yourself." But here we're just talking about a situation where you think they bumped the revision number before they should have. C'mon, really?

[PlainName]

Seriously, it's FREE software

That's a can of worms to open!

I don't think reality is allowed to be distorted by something being free. Sure, one is also free to not buy it, but if it sucks dead donkeys then being free only means there is little point in trying to get your money back. And being free doesn't usually prevent someone from taking pride in their work.

However, in the specific case of Kicad, it is targeted to be as good as commercial offerings. Users expect a certain amount of professionalism from the developers, and the developers would be silly to disregard that - look at how it and other packages languished in the foothills of hobbydom until some higher purpose (aka an entity with funds and a need) either took over or steered the direction. Without CERN, Kicad wouldn't be the high quality product it is today. And it's important to recognise that it is the quality that makes Kicad desirable, not the free-ness (you can get Altium for free if you really want).

Open-source developers deserve our thanks and some leeway, but if they can only turn out good stuff when they are paid developer rates, perhaps they should be looking for a more suitable project.

[LazyJack]

Well, yes and no.
Most projects work in a constrained resource environment (unless it's Moon landing or Manhattan project). So they need to prioritize on where they allocate the resources. I think the KiCAD project made a good decision on focusing on actual development than writing help files. Why?
First while you correctly state that user documentation is part of software development, but it is not done by the same people in any serious project. I think you get better bang for the buck by spending CERN donation on software developers than on technical writers. Being an open source project it is accepted if not even expected to have a community based help and support. So I don't think much is lost by some unfinished help files.
Second by actually having the TODO in the help file, I think the development team acknowledges that there are thinks still to be done, instead of just sweeping things under the carpet. Would it be any better if instead of "TODO: write the help" you would have "On this page you set the parameters for zone fills." Like we see too often in professional software where there is a one sentence description for a window with 4 dozen fields and buttons.
Should they have not released 6.0 until all TODOs are removed? No. It's much better for the momentum of the package to go forward with features, usability and reliability  by focusing resources on them.
At least this is my opinion. YMMV.

[eugene]

I do have different requirements and expectations. For years I used Altium (paid for by employer.) If something was missing from the documentation then I called them. We paid some thousands of USD per year for support and I felt entitled to support.

But I know that documentation is always the last thing to get done. The developers could have held off an official release of V6 until the docs were ready, but they decided not to wait. Ideally, someone with experience in technical writing would have volunteered to work on the help files while the code was being written, but evidently there was no such person. Instead, as is often the case with software like this, the coders are going to end up writing the help files, but they've been busy.

So, yeah, cut them some slack. These days as an independent consultant I use Kicad professionally. It's not perfect (neither is Altium) but I use it and am grateful that it works as well as it does. FWIW, I installed V6 on the day it was released. I've spent some time with it, but continue to use V5.10 to complete projects that I was already working on.

[Cerebus]

why do you think the developers think it is finished?

Well, call me old fashioned, but when you've gone through two release candidates, then you bump the version number and make a release announcement it's becuase you've got a complete, finished fully functional product. Sure, you may have plans for the next version, but that shouldn't include "complete the help files for the previous version".

You're free to continue using V5.1 until you've decided that V6 is finished, no matter how you define 'finished.'

EDIT: That sounded snarky, and to be completely honest, I might have wanted it to. But, you have been arguing that the developers gave a version of the software the title of "released" when (in your opinion) they should have waited until a later version. OK, well, if that's not just semantics, then it's not far from just semantics. So, if you disagree with the developers that's your prerogative. Just wait until we get to the version that you would have released and start using that.

Seriously, it's FREE software. You should be supporting the developers. Start a college fund for their kids, or maybe just be kind and show appreciation for the hundreds of hours that many of them have individually contributed. It's OK to scoff off the retort "if you don't like it then fix the code yourself." But here we're just talking about a situation where you think they bumped the revision number before they should have. C'mon, really?

All the usual FOSS fanboy tropes, yet no acknowledgement that software without documentation is not finished and not worthy of being regarded as of release quality. Why oh why, do the FOSS fanboys always counter any criticism, no matter how valid with "It's free, you should be grateful." and characterise any criticism as moaning?

By the way, "semantics" doesn't mean what you thinks it means.