It doesn’t address the core problem. Most OSS companies have a business model that revolves around support. If a large hosting provider like Amazon comes in and provides an “as a service” version, that cuts off a primary revenue stream. If said hosting provider doesn’t produce improvements to the codebase then AGPL doesn’t matter.
I thought AGPL is specially forged to prevent that. Or do you mean that Amazon recreated their own version from scratch?
AGPL says you have to release improvements. It doesn’t make you contribute to the community.
If a community is getting a lot of financial support from a company like Redis Labs paying for core open source work, a company like Amazon can come along and do an as a service version and contribute nothing. AGPL does nothing for that.
The issue is many projects are pushed forward by commercial offerings that rely on support/services as a means to provide financial support. Our open source licenses provide no protection for that model.
Perhaps the model is flawed and we need something better. But there is no protection from parasitic behavior in that case.
It extends further though, in general, there’s no way for an open source community to develop a means to financially support itself and not rely on free labor that is free of concerns. But that’s another topic.
The world has changed around free and open source. They haven’t adjusted to that change beyond AGPL being created to address some issues.
I personally don’t think that commons clause is the right solution but I understand the problem they are looking to solve.
Apologies for any typos. I answered this from my phone.
Companies dual-license under both GPL and AGPL. So, it could be done AGPL with cloud vendors paying a license. There’s a lot of FOSS developers that oppose the AGPL, though.
It absolutely is. Read the FAQ section on the AGPL, it’s very unclear. ‘Many features of the AGPL…’ kind of language. What features? It’s not the Linux kernel, it’s a license, it’s pretty small, just say what these supposed features are.
Of course the reason they don’t is that it’s a smokescreen: the AGPL is of course fine, but their goal isn’t to make the software free, it’s to profiteer off it.
Yes, Redis Labs is in the business of paying people to work on Redis and the Redis ecosystem and needs to make money to do that. The business model for companies such as that is based on support. If someone cuts off that revenue stream the money falls apart. We can as a community accept that such companies will need to build protections for themselves (licenses like common cause or having some closed source components) or accept a world in which there are no companies that exist to support specific products that could be turned into as “as a service” by a large player.
The AGPL does nothing to stop someone like AWS from taking what Redis Labs does and making money off of it and wrecking the Redis Labs business model (which is shared by a number of companies). I commend them for trying an approach that leaves the module source available and even “open” for some segment of the user base. The alternatives are “new business model”, “go out of business”, or starting to make more and more of their offerings closed source.
The AGPL does nothing to stop someone like AWS from taking what Redis Labs does and making money off of it and wrecking the Redis Labs business model (which is shared by a number of companies).
Nonsense. AWS wouldn’t touch an AGPL redis with a ten foot barge pole.
AGPL/commercial dual licensing is actually open source.
I commend them for trying an approach that leaves the module source available and even “open” for some segment of the user base. The alternatives are “new business model”, “go out of business”, or starting to make more and more of their offerings closed source.
Calling this open is literally telling a lie.
why link to undeadly instead of directly to the tutorial? https://github.com/bob-beck/libtls/blob/master/TUTORIAL.md
that I asked myself afterwards too. :/ - sorry for the noise (would have avoided the duplicate as well)
Duplicate from 3 days ago: https://lobste.rs/s/qnmrs2/happy_bob_s_libtls_tutorial
indeed. I’ve used ejabberd as well as openfire in the past. Cert-Handling (as an example) was hell with openfire.
Surely there is some good solution by now for the mess that is (was?) certificate trust stores on JVM based languages?
Serious Cryptography - J.P. Aumasson - https://www.goodreads.com/book/show/36265193-serious-cryptography
(German) Darm mit Charme - https://www.goodreads.com/book/show/20755203-darm-mit-charme
Alright, I changed my suggestion to your proposal. Did you suggest the same? I think the only reason why this article is not on the top is because the title is so misleading and boring.
If an article gets multiple identical suggestions the title will be changed automatically. So it would make sense.
thanks for adjusting the subject. I was not sure, when I submitted, but initially had a different title.
Would love to see some more infos about build quality, battery life, touchpad performance, how many nits can the display deliver and so on.
A friend bought a 2015 model (I believe) and he was not happy with the overall build quality. But I had the chance to have the newer InfinityBook model in my hands for a short moment and I have to say that it felt much better (build quality-wise).
Glad to see more Linux-first devices. Tuxedo seems to be a smaller German manufacturer. Is this CLEVO hardware? Do they support fwupd?
Vendor is considering adding firmware to the LVFS
Thanks for the feedback. Since I got quite a few hardware-detail related questions, I will write a follow-up blogpost covering those. I’ve also approached the vendor to see whether there are more details that can be covered.
Definitely interested in a follow up on this.
A colleague of mine had a Tuxedo notebook but this thing looked rather Chinese than German. (I don’t know what version it was, though.)