Reiser4 File-System Now Works For Linux 3.13 Kernel

[cjcox]

But you know that Reiser4 was backed by a vendor (namely Namesys but also by DARPA and Linspire) ? The problem is that Hans Reiser went to jail and as a result of this his company Namesys died. AFAIK the people still working on Reiser4 are ex Namesys employees. Anyway, all this has nothing to do with university student vs. vendor, Linus vs. Hans or whatever you wanted to tell us.

Uh... so you're making a point againts FOSS? Ok. I disagree.

Hans, a murdering genius is no longer involved with the project. So, we automatically sacrifice the other FOSS programmers working on it because it does not have large (and often times the source of much closed source and proprietary software) corporate backing? Tell me what I'm missing? I think this has everything to do with whatever you meant (and trust me, you are much more cryptic than I).

[V10lator]

Uh... so you're making a point againts FOSS?

No, I don't.

Hans, a murdering genius is no longer involved with the project. So, we automatically sacrifice the other FOSS programmers working on it because it does not have large (and often times the source of much closed source and proprietary software) corporate backing?

We don't sacrifice anybody. It's just a matter of fact that with the end of Namesys the manpower for Reiser4 was hardly reduced. There where more than just one review for kernel inclusion, guess why it never happened. Also this article here:

updating the kernel compatibility from Linux 3.12.6 to Linux 3.13.1

Why wasn't it updated for any kernel in between? Again this has nothing to do with being corporation backed or not but with manpower, so I still fail to see the point you wanted to make ("Now if it's not backed by a "vendor" (i.e. Red Hat) then it's dangerous crapola.").

[nils_]

I don't even remeber how long ago I used Reiser4 ... must be at least 5 years now. It'd be interesting how it compares to current filesystems like btrfs, ext 4, zfs or xfs (I always mingle those two up).
Other than that ... it doesn't seem to make much sense to keep it around for other purposes than accessing old data. And I can't imagine many people still use it.

Has it ever been production ready? I remember having some legacy systems running reiser3 but I've never seen reiser4 in the wild.

[DrYak]

I remember the days when a university student could just surprise the world with some code and that was good enough. Now if it's not backed by a "vendor" (i.e. Red Hat) then it's dangerous crapola.

The thing with university students, is that after a while they graduate and then move to other projects or gets hired by the industry.
Yup, the university genius did amaze the world with some batshit-crazy-fast filesystem he wrote for Linux 2.6.3 as a master thesis or as a side project.
The problem is that after a not so long time, the kernel has moved to 3.20.4, but the filesystem hasn't be updated since then.
Because the former university genius is now either employed at a big company where he's in charge of optimizing the infiniband handling code on a complex cluster application, and his new preferred side projected is writing an amazing application using the latest 3D virtual headset.
And meanwhile there this other new university genius coming to the scene. Except that, she's not interested in maintaining the old filesystem. Instead she prefers to amaze the world with her own new insanely wonderful creation.

The lack of corporate backing isn't a since of dangerous crap. The lack of corporate backing is a sign you won't be sure if 5 years down the line there will still be some around to keep taking care of the project.

Yes, free software make it possible for solo geniuses to create wonders. But unless there's a huge community interest picking up from them and insuring that the project will be kept alive, there's nothing certain about the future of a small wonder-project.
Opensource world is littered with small abandonned project. That's not a bad thing *per se*: it's a sign that reserch and experimentation *are* happening, which mean that overall, the platform moves forward, and which also means that eventually, some small project will attract enough interest to become sustainable (usually, in the form of some corporate backing paying salaries so that the initial genius can still hack his/her project while still having food on the table).

Disclaimer: in my day job (research), I'm regularly confronted with small projects, that were done as master thesis or as PhD project, and which are currently abandonned because the author has moved to other places and interests.

So, we automatically sacrifice the other FOSS programmers working on it because it does not have large (and often times the source of much closed source and proprietary software) corporate backing?

At some point of time these other (albeit wonderful) FOSS programmers will need to eat. That means they'll need to get jobs. That means they won't have that much time to spend keeping an old project alive.
Case in point: Edward Shishkin is more or less the last guy taking care of Reiser4, and he's doing it in he's spare time, as the company which initially paid the development of ReiserFS and Reiser4 - a.k.a. NameSys - is no more.

Compare this with BTRFS. This is file system is next "future of filesystem", it is heralded to one day bring all the improvement and features, that things like reiser4 used to promise bringing back when they got started. But the big difference, is that there are companies throwing money at it (Oracle is paying development, SuSE did help develop critical tools, etc.) Meaning that the developer will be around and will get paid to work on it.