@ArtG : Just go to pcbnew and reimport the netlist using timestamps, and make sure ‘Exchange Footprint’ is set to ‘Change’. That will restore all your footprints to their correct references and net names in situ.
If you mean filing a report for the “Create Array” bug then I’ve already done that, check my previous posts for the link. I’m sure any additional information in the bugtracker would be appreciated.
Ah, thanks, I see that! I removed most of my post because it was …unkind ;). But its hardly more than a nuisance really, the create array bug I mean. I don’t have any new information though I might have a a patch, waiting for kicad to build.
1 Like
Wait, what are smileys? Can you please enlighten me?
I guess not using emoticons is the burden I have to live with. I actually grew up in a time when people were reading books as opposed to social media. Back then we actually had to do that crazy thing called inferring authors intent from the context. Dark times…It is a real handicap… you can say a disability… You by chance are not making fun of a disabled person, are you? Why is it you can’t appreciate a disabled person? So not politically correct of you!
[quote=“metacollin, post:23, topic:2368”]
Just go to pcbnew and reimport the netlist using timestamps
[/quote] Thanks! That did work! Never really payed attention to that option. I’m just interested, how does it work? Does it create a time stamp when a footprint is placed? But how does it help to reference to the correct schematic symbol?
As far as your little rant about OpenGL - some of the bigger issues I had with it were submitted as bug reports and feature requests and produced zero results. I’m not going to waste my time on trying to impart common sense to people. I like the improvements you were talking about. What I don’t like is the price we all had to pay for them. It is certainly flashy and looks nice, but all the workflow features went to the trash. I think if you decide to rebuild a program from the ground up, you should at least do a board layout once yourself to understand what is involved.
1 Like
Do you talk about physics/mathematics/engineering books or story books (I prefer scifi operas from Adams, Hamilton, Asimov, etc…)?
The first ones never unveiled anything else than information to me (maybe I did get to read the wrong ones), the latter ones usually described what is going on and do actually give contextual information that does go way beyond emoticons + pure information.
So yeah, putting a smiley behind a comment on a public forum makes sure the other end doesn’t have to guess and infer from 10-20 words how it’s meant. I’m sure you make insightful faces and gestures when you type cynical or sarcastic pieces, but I can’t see them. 
I read recently that only 7% of communication is verbal. So that emoticon is carrying 93% of the weight!
Well, I suggested trying this option almost right away, in the second reply 
:
1 Like
I saw that but for some reason it didn’t connect in my brain that it was an option in the read netlist dialog. I blame it on you. Your non-verbal communication wasn’t very good. Should’ve communicated better with emoticons.
2 Likes
Oh you have no idea…
Mea culpa, mea culpa, mea máxima culpa
Thanks for that explanation! I have been puzzling over the significance of those “Timestamp” values since shortly after I started using KiCAD. (My speculation was that it had something to do with when a symbol or footprint was edited in a library, so you could know whether or not your file used the most recent revision of a symbol or footprint.) After reading your explanation (including those well thought-out examples) I understand the significance of the timestamps and how the concept contributes to the usefulness of KiCAD. This seems like another chunk of information that really needs to find its way into the users’ “Help” function, and the general documentation.
At what point in the design process does the “Timestamp” value get created? From your explanation it sounds like it gets assigned when I click the mouse to create a symbol on an EESchema sheet. And the only constraint on timestamp values is that they must be unique for each component’s symbol in EESchema. (The timestamps don’t try to encode, e.g., the KiCAD build version, the make and model of my computer, the capacity of my harddrive, the last filename opened, my geographic location, the room temperature, my blood type, my wife’s name, or any other information that some programs seem to think is “necessary” for their operation.) Do I have that correct?
If that’s the case it seems that “Select Footprints by Timestamp” would be the preferred method for importing a netlist, based on how the majority of us work. However, the “Import Netlist” dialog presents “Select Footprints by Reference” as the default method. Is there a usage case where “Select Footprints by Reference” would be the preferred method?
Thanks for your help!
Dale
One question for clarification, as I stumbled upon those timestamps a couple of times… if the assumptions of @dchisholm ring true, why is there a (timestamp blabla) field in the .kicad_mod files?
And what happens when I delete it?
Fault is mine. I read your post and used “create array” button. Here we go.
There’s no reason to remove the “Default” canvas although once the OpenGL counterpart supports all the features in Default, at the very least nothing new will be added to Default. Some devs are keen to purge the code but when it gets to that point you can always argue to keep it on as extra baggage.
Sooner or later some new feature will break it and it makes no sense to keep it ‘alive’ after that… I can live with that I guess.
@Andy_P
Thanks for the insight… need to make sure my ‘footprint manager’ doesn’t do this then. With my luck I get the black hole under Texas instead of the pie
2 Likes
It seems Jean-Pierre have committed a fix for this bug: https://bugs.launchpad.net/kicad/+bug/1549231
So we can try it out soon 
1 Like
One side note.
Refdes is crucial for fab layer but silkscreen is just merely cosmetics.
So if you want to look for reference designators you turn on your fab Layer.
IMHO.
@Andy-P - Very insightful and useful explanation - thank you for taking the time to provide that to us. I’ve definitely learned a bunch from this thread, along with the deep-dive technical discussion on the bug report side of the discussion as well.
Great community here. Have one of the last cavendish bananas on me 
1 Like
Why? How hard would it be to enforce the same time stamp for all symbols (units) belonging to the same designator (for example when schematic is annotated the time stamps of all sub-parts are changed to the timestamp of the first unit)? And then revisit the question of why not to have timestamp to be the default mode.
1 Like
I never really understood why timestamps were incompatible with multi-section components, either. I currently have very few instances of multi-section components, so I have adopted the habit of importing netlists by timestamps rather than reference designators.
Someday I’ll almost certainly have to face the problem of multi-section components but for now I’m too cowardly to tangle with a black hole!
Dale