Building a GNU/Hurd Cluster
HTML-код
- Опубликовано: 20 сен 2024
- Lecture at Catholic University, January 17, 2018.
Abstract:
The GNU Hurd is a UNIX-like operating system based on Mach, an
operating system kernel developed at Carnegie Mellon University and designed to support multi-node clusters, which have become ubiquitous in high performance computing. This talk will highlight current trends in cluster computing, outline the design of Mach and Hurd, and summarize the speaker's work to realize Hurd's potential as a POSIX-compatible operating system that can leverage the power of an entire cluster.
Thanks for a lecture, possibilities are so exciting, can't wait when I will have some free time to play with GNU Hurd.
When will Hurd be user ready ?
It runs; you can download and install a Hurd distribution. It's device driver support is limited, so it's best to run it in a virtual machine. The big question is what can it do that you can't do with Linux, and the answer is not much. When will it achieve something like the functionality I anticipate in the video? I don't know. I'm not working on Hurd anymore due to FSF's legal requirements for submitting code to the Hurd repository. They want all copyright assigned to them, for all past *and future* work on the project, and I'm not willing to do that.
@@BrentBaccala Sad FSF needs such stringent rules , this further more will delay the development as more n more developers feel the same way. Appreciate
the reply thanks !
Can I just clarify something here. It looks like you are saying you want to leverage all the work done by all the people who came before you to produce Hurd, who *were* willing to grant FSF perpetual right to their work on Hurd code to ensure Hurd remains free, but you don't want your work on the Hurd to be free too. Which suggests you want to use that work for some kind of gain for yourself yet still leveraging all the free work done by others - I'm sure you'll correct me if I'm wrong. I'm not trying to shame you or anything, I absolutely appreciate that the time spent on this kind of thing is very significant, but surely you can add your contribution to all the others within the Hurd code then build your work on top of Hurd once it is mature enough. That would not require any handover of rights to FSF and you can freely profit from that. Then others in the future can benefit as you have from the lessons to be learned from Hurd, including your contributions to it. That is in the spirit of the GNU project and FSF as I see it. Just my thoughts on it as it seems a real shame that you ended up abandoning it all like that.
@@captainblood9616 My major issue with the FSF license grant is their insistence on language that states "all past and future work" on the program. Emphasis (mine) on "all future work". I'm willing to assign copyright on a lot of the work that I've already done, but I'm not handing them that kind of blank check. I want to retain copyright on some of it not to profit off of it, but because my personal vision of free software is "no strings attached", and the GPL is a major string. I'd be willing to dual-license my code to other open source projects that didn't use the GPL, but FSF won't do that, and that's why I want to limit my copyright assignment to them.
@@BrentBaccala That's a real shame