[lkml]   [2000]   [Oct]   [6]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: evil patents
Martin Pool wrote:
> I worked on something similar to tuxfs a while ago:
> This does clash with the patent claims and it was done only a couple
> of years ago, so it's no direct use as prior art. I do agree about
> software patents: once one starts thinking about the requirements the
> method is fairly obvious.
> In fact I'd be surprised if BSD's write ordering doesn't come close to
> clashing with the method of moving from one consistent state to
> another. I'll read the full text on the weekend.

BSD's write ordering doesn't quite cut it because it doesn't preserve a
complete filesystem state. But the following do:

Auragen (ask Victor Yodaiken)
Nirvana db (my previous work)

> I didn't hear anything from NetApp, though I did talk for a little
> while with somebody from Veritas. I think they were trawling through
> the l-k archives looking for work on filesystems.

It's nice to know we're doing other people's work for them. Oh, wait,
that's what we're supposed to do! Just so long as they remember to put
some cookies back in the jar after they've munched on the ones they

> As it happens I too just had a brush with an evil proprietary company:
> I don't think NetApp would behave this badly.

Soon, all of us will be having these brushes. Patents are a fence
around Linux, trying to hold us in, keep us from using techniques newer
than 17 years old. Of course, we can't accept that - it would be death
for us. Um, I seem to recall Microsoft's Halloween II memo made special
note of exactly that point.

> On a more positive note I'm interested in getting back into this kind
> of development. Are you looking for help?

Yes, absolutely I am:

Tux2 coding is temporarily suspended until after I come back from
Atlanta (ALS) and Miami (Storage conference). After that I'm raring to
go. There are plenty of design notes to read in the tux2-dev mailing
list archives:

We *will* get through this NetApp thing in a satisfactory way, one way
or another. In the meantime I'm really looking forward to fixing up
Tux2 for 2.4.0. I have in mind a really great way to map the metadata
into the page cache - and even into linear memory - using Al Viro's
address_mapping machinery. This will actually make the 2.4.0 code
cleaner than the 2.2.13 version and way more efficient.

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at

 \ /
  Last update: 2005-03-22 12:39    [W:0.017 / U:4.984 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site