I use git nowadays to also maintain my PCB designs. I also work from 2 computers. However sometimes I forget to push a commit and I cannot continu to work on my other PC on the same project. Slight inconvenience. Since I work alone I fear no merge conflicts.
I was thinking about moving my git repositiories to my dropbox folders. I would still have the git features like commits, branches and version control n such. I just would no longer have to keep pushing and pulling my repo’s continously because I work on two platforms. I can just push occassionally to github for deploying releases as well as safe keeping purposes.
If you would see my commits, you will see some useless entries like: “Commit for push” which I sometimes make when I am halfway a change or so.
I have all source not only locally on two computers but also on github and dropbox. So even if both PCs and dropbox all die on the same day… I am still good.
Actually that’s not necessarily true. Cloud based doesn’t have to mean the data is elsewhere. Take for example the online markdown editor dillinger.io. Your documents are saved in the browser’s local storage.
The problem with getting just the app from the cloud and working with your own data is that the app is sandboxed because it comes from another domain. It’s assumed that you have verified the app and trust the publishers, just like you trust the KiCad packages. So there will be strict controls over what can be accessed. For a complex app like KiCad that uses many resources this is problematic. Database libraries can allievate some of that, but web apps are constrained to the browser’s part of the filesystem.
And even when you have overcome all of that, you essentially have to make the software cater for another platform. I don’t see that happening soon.
Android tried instant apps for a while but now it’s back to installing apps first. So I think you will be installing KiCad packages for a while yet, tempting though it might be to be able to approach a fresh browser and start designing.
This is also an appropriate solution I agree with you…just on Dropbox or OneDrive and/or on similar don’t forget to use this for example: Boxcryptor Security for Your Cloud
Having worked for AWS as sysadmin for tons of years, I know what you can see from an open (not encrypted) S3, say… a lot
Having the files on your device only part of you control your data
Being able to use said data again is the other half of it
Dillinger works fine because it’s markdown and the entire world has reimplemented markdown and fragmented it a billion times now.
KiCad is very niche in the broader world, sure there can be another kicad_pcb, kicad_sch supporting editor and there are right now ones that support it, but perhaps not always will be. Who knows what the future holds.
At least with KiCad you get this nice binary that’ll work until the end of human civilization on the OS+machine it originally did.
Sure, and of course you’d want the cloud based app to be open source like KiCad is, with an appropriate license that prevents hoarding enhancements, which was why the Affero GPL came about. If it’s derived from KiCad then it would have to be open source too.
Wait, this just came in on my crystal ball from the forum in year 2032: Hi, I’m still using KiCad 14, must I upgrade to KiCad 16? I like how it works and don’t want to change my habits.
We actually once had a AGPL dependency that caused corporate donors to KiCad to balk because it meant they couldn’t use it themselves. Meaning we can’t go the AGPL route unless we want to start charging for KiCad to fund instead.
The other thing from experience, cloud/server apps are still not ideal here. A desktop app will just run out of the box after you install it. The user doesn’t need to be technically skilled. Self-hosting a cloud/server based app means you now need that skillset. Sure maybe someone takes up the task for hosting it for others but depending on the goodwill of others, not to mention IP protection concerns among others isn’t that good there.
Sure, the pendulum swings back and forth. Some might remember the hoopla over running Java apps everywhere but in the end it ended up as server side technology. So the browser as universal client dream has been scaled back to more interactivity and now there’s a distinct app for every task on your phone. So apps for desktops will be with us for a long time.
If it wasn’t already clear I don’t see a case for a cloud based KiCad in my lifetime but then I won’t live forever.
My experience using VDI’s for other software has always been less than pleasant. Clunky, barely works, etc. granted, it has been probably 3 years since I tried so maybe the tech is way better now.
But if cloud based were made an option, not a requirement, then how would it affect you? Just don’t turn it on, and use KiCad like you always do.
I don’t think anyone is suggesting that it be a cloud-only thing. The two aren’t mutually exclusive.
I totally understand the $$ aspect. Makes sense that it wouldn’t be a viable thing to implement because of financial concerns.
What I’m not understanding is it seems some of the comments are assuming that the question I posed (as the OP) somehow implied that hypothetically developing a cloud based KiCad would make its use instantly be mandatory. As I replied to another poster, if it were ever something that was developed, couldn’t it simply be an option? If someone is appalled by cloud based software, they could just keep that option turned off and continue using it like they always have.
Realistically? No, why would someone put all that effort in and then maintain two different versions, the cloud version and the non-cloud version? Someone would need to be paying a lot of money for that to ever happen…
Well it greatly depends on the quality of your internet.
Anyway, i think the two people editing at same time, if each person is in their own block. could be made possible by a version control system. This would also handle the file syncing across network… All we really need is version control hooks in file menu…
In my understanding, a cloud solution for kicad is not the work of an application in a browser or on an organization’s server … But the ability of a group of engineers to track changes in a project in parts with general synchronization, this also applies to libraries … A certain weaver online and version control of changes
Syncthing works really well to keep files in sync between two computers – I highly recommend it. It has replaced dropbox for most of my needs. You can still use Git in parallel to backup and version the files, or collaborate with others, but it is no longer necessary to git commit/pull when moving between machines.
An option! I would suggest that cloud based, multiple user access would be the domain solely of business. Why not make that type of option a paid for service to supplement the coffers of the development team. It’s not NEEDED by the vast majority of individuals using KiCAD.
In theory fine option, but in practice it eats up development resources. So for example in exchange of having this feature in 3 years. Are you wiling to have no bug fixes and no upgrades on KiCad?
So nothing stops you from making a project that does this, you can then put your effort towards this and try to get others onboard. This way the main project does not suffer any delays, but your project also has minimal resources.
The primary problems with this kind of work is that programmers enjoy solving problems not translating code. So it can be hard to find people willing to do this. AI perhaps?
