[Kicad-developers] [PATCH] eeschema: make footprint preview optional

Hi, I've attached a patch to make the footprint preview and selection functionality in eeschema optional, to address primarily performance concerns some users have, and also to address the handful of planned features still unimplemented. It's a runtime option, default OFF, so users can easily

Re: [Kicad-developers] Debian ngspice package.

I think the wiki page Carsten made on https://wiki.debian.org/KiCad/ngspice is a good overview of that. 2018-01-02 23:43 GMT+01:00 Carsten Schoenert : > Hello Tomasz, > > Am 02.01.2018 um 23:20 schrieb Tomasz Wlostowski: > > Forgive my ignorance, but what's the problem

Re: [Kicad-developers] Debian ngspice package.

Hello Tomasz, Am 02.01.2018 um 23:20 schrieb Tomasz Wlostowski: > Forgive my ignorance, but what's the problem with ngspice's license that > makes it non-free? the license situation for ngspice 27 is much better than before but there are still files or folders licensed by a non free-license from

Re: [Kicad-developers] Debian ngspice package.

On 02/01/18 16:12, Carsten Schoenert wrote: > the recently uploaded new version of ngspice into Debian unstable is > still non-free. By this I can't use it for building KiCad in Debian main. Carsten, Forgive my ignorance, but what's the problem with ngspice's license that makes it non-free?

Re: [Kicad-developers] Debian ngspice package.

Carsten, I think you can also do this using environment variable NGSPICE_ROOT_DIR, which the findngspice.cmake also looks in. Thanks, Dan W On Tuesday, 2 January 2018 20:42:11 GMT Carsten Schoenert wrote: > Hi Wayne, > > Am 02.01.2018 um 16:50 schrieb Wayne Stambaugh: > > Thank you for the

Re: [Kicad-developers] Tool-assisted selection disambiguation

Hi Jeff I was just working on this bug. I was able to fix it by checking if all the selection possibilities in the disambiguation menu had the same net in SELECTION_TOOL::guessSelectionCandidates( ... ) , and not displaying the menu if they where all the same net. This works but sort of breaks

Re: [Kicad-developers] Debian ngspice package.

Hi Wayne, Am 02.01.2018 um 16:50 schrieb Wayne Stambaugh: > Thank you for the update. I did not realize all of this was going on in > order to get a new ngspice package into Debian. I appreciate your > efforts. I will try to build this package, install it on my system, and > build and test

Re: [Kicad-developers] Stable 5 branch status

That's good, I just worry that if people need to read the manual for something as mundane as importing old projects, it might not be ready for release yet. I don't mean to whine at you over this, I understand that this is a complicated thing to get right and there are a lot of corner cases and

Re: [Kicad-developers] Stable 5 branch status

I plan on adding the remapping feature to the Eeschema documentation before the stable 5 release. Hopefully I can find the time to knock it out it the next week or two to give our translators plenty of time. On 1/2/2018 2:59 PM, Chris Pavlina wrote: > Of course. My confusion mostly surrounds how

Re: [Kicad-developers] Stable 5 branch status

Of course. My confusion mostly surrounds how it interacts with rescue and when it does and does not decide to write things to disk (see José's point about it clobbering its own backups). In any case, I don't need you to explain it to me. You and I aren't the only KiCad users, and if people who've

[Kicad-developers] V5 RC and component chooser, one more time

Hi, Just want to get a word in on this before anyone pulls an RC - due to a large number of user complaints about behavior of the footprint selector/preview in the symbol chooser and a number of unimplemented features, I intend to put it behind a default-off runtime option for V5. I don't have

Re: [Kicad-developers] Stable 5 branch status

Let's start at the beginning. Are you clear on why symbols need to be remapped? On 1/2/2018 2:52 PM, Chris Pavlina wrote: > Honestly, with this symbol library stuff it's been hard for me to figure > out what's a bug and what's a confusing design decision. Bug reports for > the latter are rather

Re: [Kicad-developers] Stable 5 branch status

Honestly, with this symbol library stuff it's been hard for me to figure out what's a bug and what's a confusing design decision. Bug reports for the latter are rather frowned upon so I've been refraining from filing many bugs. On Tue, Jan 02, 2018 at 02:51:14PM -0500, Wayne Stambaugh wrote: > I

Re: [Kicad-developers] Stable 5 branch status

I just tested this repo and it failed to remap every symbol so either something has been broken or this project is a corner case that I haven't stumbled across yet. I will keep informed of my progress. In the future, please file bug report when you see something like this. If you just discuss

Re: [Kicad-developers] Stable 5 branch status

Maybe one quick and helpful change would be to make DIALOG_SYMBOL_REMAP::backupProject() add a timestamp or other unique ID to the backup file name, so it won't clobber old backups if run multiple times due to "non-standard" workflows? On Tue, Jan 2, 2018 at 2:42 PM, José Ignacio

Re: [Kicad-developers] Stable 5 branch status

This sequence of actions is what puts the project on a stuffed up state: * Open schematic and be greeted with some symbols that need rescue, rescue a couple symbols * Be greeted by the remap dialog (1) * Regret remapping and quit without save * Open schematic again * Be greeted by the rescue

Re: [Kicad-developers] Stable 5 branch status

On Tue, Jan 02, 2018 at 02:20:10PM -0500, Wayne Stambaugh wrote: > It's exactly as insane as the cache has always been. I respectfully disagree. As I recall from many years of using KiCad before this, nothing about the cache has ever spontaneously deleted all libraries without warning. That's a

Re: [Kicad-developers] Default Canvas for v5

I stand corrected. ‘G’ in the middle of a segment does not work: it moves the closer segment end to the current position and drags the intersection from there. > On 2 Jan 2018, at 17:42, Jeff Young wrote: > > Since segments in OpenGL can be selected, the semantics are a bit

Re: [Kicad-developers] Stable 5 branch status

It's exactly as insane as the cache has always been. The only way I can think of this happening is that the schematic was not saved after the remapping and once all of the libraries are removed from the project file but cache should still be valid if no save was performed. Let me try remapping

Re: [Kicad-developers] Stable 5 branch status

This is doubly compounded if the project has symbols with (newly) illegal names. The cache gets symbols with the original names but the code cant find them anymore, even though the remapper can find it in the library with the slashes, so you have to manually edit the library before opening, or

Re: [Kicad-developers] Stable 5 branch status

I have a slight feeling it should be doing the remap _before_ rescue instead of after, but I'm at work now so I can't really poke around at it. It's stuffing everything up in multiple ways, depending on exactly what I tell it to do and I'm still not sure what I'm doing wrong. In one case, it

Re: [Kicad-developers] Stable 5 branch status

Ah, that explains my question mark case. That's insane. On Tue, Jan 02, 2018 at 01:08:11PM -0600, José Ignacio wrote: > One big problem is that even if everything fails, the project will show up > fine with all symbols, but they wont have the remap done. at the same time > the remapper deletes

Re: [Kicad-developers] Stable 5 branch status

One big problem is that even if everything fails, the project will show up fine with all symbols, but they wont have the remap done. at the same time the remapper deletes all library entries from the project file, which causes it to open with question marks only if you reopen it again after

Re: [Kicad-developers] Stable 5 branch status

You're going to have to give me a bit more than "stuffing everything up". Are the symbols being incorrectly remapped? Even if the symbol library table is empty, the recovery should still be correct as long as the cache library is valid. If as you suggest it's so opaque that users cannot

Re: [Kicad-developers] Stable 5 branch status

Are we really going to do an RC now without any more work to the remap stuff? I just had it ruin another project last night. It behaves very poorly when things also have to be rescued or libraries are missing, it seems. On Tue, Jan 02, 2018 at 09:10:20AM -0500, Wayne Stambaugh wrote: > On

Re: [Kicad-developers] Stable 5 branch status

Please send me a sample project if possible. I have tested this every which way I can think of and it works fine for everything I throw at it. The only time things fall apart are when the cache is corrupted or missing and/or rescues have be ignored. In those cases there is nothing I can do

Re: [Kicad-developers] Stable 5 branch status

Keep in mind my problem is not necessarily that things are _broken_ but perhaps that they are so opaque that they only make sense in the mind of the designer. As I said in the previous message, we've been getting TONS of people on IRC who can't figure out how the hell to drive this thing, and I

Re: [Kicad-developers] Stable 5 branch status

We're currently close to inundated on IRC with people using nightlies wondering what to do because their projects don't load properly. If it's not _broken_ it's definitely so unintuitive that most people seem unable to figure it out. Myself included, it took me way too long to figure out how to

Re: [Kicad-developers] Default Canvas for v5

Since segments in OpenGL can be selected, the semantics are a bit different. Select a segment and a ‘G’ near its end gives you get drag-intersection behaviour while a ‘G’ in its middle gives you drag-segment. ‘M’ works in both cases as well: ‘M’ near an end gives you move-endpoint; ‘M’ in the

Re: [Kicad-developers] Default Canvas for v5

On Tue, Jan 2, 2018 at 11:18 AM, André S. wrote: > Hi everyone, > > I want to add from a users view: > In the current KiCad 4.0.7 (Windows 64 bit) you can't drag traces in > OpenGL/Cairo, only in legacy. (With this I mean: move a segment of a trace > and it stays

Re: [Kicad-developers] Default Canvas for v5

Hi everyone, I want to add from a users view: In the current KiCad 4.0.7 (Windows 64 bit) you can't drag traces in OpenGL/Cairo, only in legacy. (With this I mean: move a segment of a trace and it stays attached to the other segments of the trace (Hotkey: 'G')). Hotkey 'M' moves segments in

Re: [Kicad-developers] Debian ngspice package.

Hey Carsten, On 1/2/2018 10:12 AM, Carsten Schoenert wrote: > Hello Wayne, > > Am 02.01.2018 um 15:33 schrieb Wayne Stambaugh: >> Would one of our Debian developers please prod the Debian ngspice >> package[1] maintainer(s) to enable building ngspice with a shared >> library and a -dev package.

Re: [Kicad-developers] Debian ngspice package.

Hello Wayne, Am 02.01.2018 um 15:33 schrieb Wayne Stambaugh: > Would one of our Debian developers please prod the Debian ngspice > package[1] maintainer(s) to enable building ngspice with a shared > library and a -dev package. A bug report[2] for this was filed almost a > 1 1/2 years ago (aug

Re: [Kicad-developers] Patches consistency

Who are the Mac developers with committer status? Can one of them get this wxWidgets branch setup? (I can help with testing, but I think someone will probably barf if I upload a patch containing the whole source tree. ;) Cheers, Jeff. > On 1 Jan 2018, at 02:03, Simon Wells

[Kicad-developers] Debian ngspice package.

Would one of our Debian developers please prod the Debian ngspice package[1] maintainer(s) to enable building ngspice with a shared library and a -dev package. A bug report[2] for this was filed almost a 1 1/2 years ago (aug 2016) and it still has not be resolved. Carsten was kind enough to even

Re: [Kicad-developers] Stable 5 branch status

On 1/2/2018 8:58 AM, Tomasz Wlostowski wrote: > On 28/12/17 20:24, Wayne Stambaugh wrote: >> There are a few outstanding crash bugs that need to be fixed before we >> can consider branching the stable 5 release. Here is the list of >> unresolved crash bugs that effect the development branch: >>

[Kicad-developers] Tool-assisted selection disambiguation

There’s a bug report[1] which complains that every time you attempt to drag a track corner it asks you which track (when of course they share the same corner). The SELECTION_TOOL shouldn’t know about this kind of stuff, though, so I propose to add a client filter to