Monetary needs and all that. If it’s a startup with VC then there is either not enough people paying or not enough private users supporting by other means like bug fixing, support, etc. Or greed by VC. But in general, what do we expect where the time comes from people spend on coding for us? See openssl that is essential and was about to loose its main dev. Pay or donate or contribute. Preferably to a non VC project.

Me too …

That’s a great game! That ja for the pointer

I am not saying that ntfy is responsible or even the cause but since I tried it my notifications on Android are beyond screwed. I removed it and still … But it would be nice to have and since a lot of people are using it successfully I am sure it’s a viable alternative.

Oh wow.

I can’t seem to find it. I want to know what their product is.

Thank you!

How does the a380 impact your power consumption? If you have ever measured it.

Wondering if you have and insight on power usage with the a310 in the system while idling. I built a sub 25w server and don’t want to mess that up.

Sometimes it’s a power consumption issue.

In my case I could use a pi in terms of CPU and ram but I need a lot of SATA. After putting in a perfect energy saving MB and a picoPSU I am idling at 16 watts. But id preferred a pi regarding power.

But yeah sff PCs are cheaper and much more powerful.

Ah didnt even realize you are the dev. Thanks for the work you put Into this!

I get that ocr clashes but for me it is mostly scanned documents that are not really useable without ocr. I imagine a lot of people face the same situation. Either way I understand your position on this.

Either way, great project!

It’s hungry in resources. But most of all it’s complex. It’s more suited for medium sized company’s or at least it feels like that.

Sorry typo. Find

I am considering switching from paperless but I can’t find anything about ocr. Does anybody know?

Okay let’s talk terminology

I meant the useable application as a for example chat bot.

The model itself is useful as far as it enables me to understand the underlying logic of the whole thing right? But without the data and I guess a training period with it I can’t use it if I understood it correctly.

If I am right with the above this wouldnt make it useful as in I can reproduce it as a functioning application and I wouldn’t call that opensourve in my understanding of the word.

Honest Questions Can a model be open source? Dont we need the data it was trained on to truly be able to reproduce, understand and modify it?

I absolutely could. That doesn’t change my stance on this software.

The code hidden in xz was also publicly available and didn’t get caught. So much for open source making all things safe just by being open source. And that was a high value target. Imagine what happens (or could) on a smaller scale.

Honestly now, are you reading code of a nice project You want to spin up in a docker to try out? I don’t. I check the project and the stars/engagement. If it goes any further I check who is involved in it and that’s about it.

It looks fishy. Happy to see somebody putting something out there but if I am not reading the code I have to trust reputation and the age and number of stars on GitHub (although those can be bought too) are not doing it for me.

When will there be the next xz attack thingy?

That looks like a fantastic project! Thanks for the effort

That’s strange. It’s working with the IP and port and explicitly removing the s from https im the setup screen. It is sneaky about the s.