* Re: porting to alternate OSs and relational semantics
2003-11-13 20:19 porting to alternate OSs and relational semantics Phil Frost
@ 2003-11-13 10:24 ` Hans Reiser
2003-11-13 23:06 ` Phil Frost
0 siblings, 1 reply; 4+ messages in thread
From: Hans Reiser @ 2003-11-13 10:24 UTC (permalink / raw)
To: Phil Frost; +Cc: reiserfs-list
Phil Frost wrote:
>My name is Phil Frost and I am looking at ReiserFS as a potential filesystem
>for an OS called Unununium (http://uuu.sf.net/) in the early stages of
>development.
>
Do you have funding from anywhere? Are you private or university or?
> From the beginning, we have planned to make the filesystem an
>integral part of the system as a whole. Thus we came to many of the same
>conclusions as ReiserFS: the filesystem should be an efficient database.
>
>Work had begun on a filesystem of our own making, but after reading about
>Reiser4 I came to realize that ReiserFS wasn't just another filesystem that
>performed better at doing the same useless FS stuff, but rather exactly what
>we wanted, only years advanced :-) So, I am now aiming to get the ReiserFS
>code working for our OS. Any information on how to proceed would be
>appreciated.
>
>Furthermore, because we are developing an entirely new OS, we aren't
>restricted by backwards compatibility or tradition. I would like (and will
>definitely need, in order to sell this idea to the other members of the
>project) to see the database features of ReiserFS develop, in particular,
>relational semantics.
>
>I don't know what progress or ideas have been made in this idea, but as I
>see it, a traditional hierarchal layout is a subset of a relational
>database, only that the only usable relation is "child-of". Our primary goal
>was to allow information to be found quickly by a key other than the name,
>such as finding all music by "Pink Floyd" or all songs from "The Wall".
>
Please read www.namesys.com/whitepaper.html, which describes the
semantics intended for reiser6.
>
>
>To this end I'd like to do whatever I can to assist. I'd rather contribute
>to ReiserFS than produce a redundant effort. Other members of the Unununium
>effort have expressed concerns about the suitability of ReiserFS to our task
>and the time it may take to mature. Any remarks you can make to quell their
>restlessness would be greatly appreciated and may convert a talented
>developer to your cause :-)
>
>
Reiser4 is getting installed on my desktop sometime this week, so I
don't think you should worry about the time to mature, unless you mean
the time required to implement semi-structured semantics on top of the
existing storage layer.
>
>
>
thanks kindly for your interest in ReiserFS
--
Hans
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: porting to alternate OSs and relational semantics
2003-11-13 23:06 ` Phil Frost
@ 2003-11-13 15:57 ` Hans Reiser
0 siblings, 0 replies; 4+ messages in thread
From: Hans Reiser @ 2003-11-13 15:57 UTC (permalink / raw)
To: Phil Frost; +Cc: reiserfs-list
Phil Frost wrote:
>
>
>Yes, that's what I mean. As a new OS with no notions of compatibility or
>tradition we'd rather have the right system initially rather than to regress
>later. I'd like to know if any such development is activly in progress, and
>if so, when it is estimated to be serviceable. Or if you prefer, what would
>be involved time-wise to implement it given the current state? As we are
>already working on such a project, with a little perswasion I'm might get a
>developer in addition to myself working towards that goal.
>
>
>
>
Status is: storage layer is stabilizing now for v4.0, semantics
development will be the next major release, asssuming we get funding for
it....
--
Hans
^ permalink raw reply [flat|nested] 4+ messages in thread
* porting to alternate OSs and relational semantics
@ 2003-11-13 20:19 Phil Frost
2003-11-13 10:24 ` Hans Reiser
0 siblings, 1 reply; 4+ messages in thread
From: Phil Frost @ 2003-11-13 20:19 UTC (permalink / raw)
To: reiserfs-list
My name is Phil Frost and I am looking at ReiserFS as a potential filesystem
for an OS called Unununium (http://uuu.sf.net/) in the early stages of
development. From the beginning, we have planned to make the filesystem an
integral part of the system as a whole. Thus we came to many of the same
conclusions as ReiserFS: the filesystem should be an efficient database.
Work had begun on a filesystem of our own making, but after reading about
Reiser4 I came to realize that ReiserFS wasn't just another filesystem that
performed better at doing the same useless FS stuff, but rather exactly what
we wanted, only years advanced :-) So, I am now aiming to get the ReiserFS
code working for our OS. Any information on how to proceed would be
appreciated.
Furthermore, because we are developing an entirely new OS, we aren't
restricted by backwards compatibility or tradition. I would like (and will
definitely need, in order to sell this idea to the other members of the
project) to see the database features of ReiserFS develop, in particular,
relational semantics.
I don't know what progress or ideas have been made in this idea, but as I
see it, a traditional hierarchal layout is a subset of a relational
database, only that the only usable relation is "child-of". Our primary goal
was to allow information to be found quickly by a key other than the name,
such as finding all music by "Pink Floyd" or all songs from "The Wall".
To this end I'd like to do whatever I can to assist. I'd rather contribute
to ReiserFS than produce a redundant effort. Other members of the Unununium
effort have expressed concerns about the suitability of ReiserFS to our task
and the time it may take to mature. Any remarks you can make to quell their
restlessness would be greatly appreciated and may convert a talented
developer to your cause :-)
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: porting to alternate OSs and relational semantics
2003-11-13 10:24 ` Hans Reiser
@ 2003-11-13 23:06 ` Phil Frost
2003-11-13 15:57 ` Hans Reiser
0 siblings, 1 reply; 4+ messages in thread
From: Phil Frost @ 2003-11-13 23:06 UTC (permalink / raw)
To: reiserfs-list
On Thu, Nov 13, 2003 at 02:24:33AM -0800, Hans Reiser wrote:
> Phil Frost wrote:
>
> >My name is Phil Frost and I am looking at ReiserFS as a potential
> >filesystem for an OS called Unununium (http://uuu.sf.net/) in the early
> >stages of development.
> Do you have funding from anywhere? Are you private or university or?
Hah...I wish! At least we project an image of professionalism :) If there is
anyone reading the list interesting in funding an OS, look no further!
Currently only I and a few other people are involved in the development.
> >From the beginning, we have planned to make the filesystem an integral
> >part of the system as a whole. Thus we came to many of the same
> >conclusions as ReiserFS: the filesystem should be an efficient database.
> >
> >Work had begun on a filesystem of our own making, but after reading about
> >Reiser4 I came to realize that ReiserFS wasn't just another filesystem
> >that performed better at doing the same useless FS stuff, but rather
> >exactly what we wanted, only years advanced :-) So, I am now aiming to
> >get the ReiserFS code working for our OS. Any information on how to
> >proceed would be appreciated.
> >
> >Furthermore, because we are developing an entirely new OS, we aren't
> >restricted by backwards compatibility or tradition. I would like (and
> >will definitely need, in order to sell this idea to the other members of
> >the project) to see the database features of ReiserFS develop, in
> >particular, relational semantics.
> >
> >I don't know what progress or ideas have been made in this idea, but as I
> >see it, a traditional hierarchal layout is a subset of a relational
> >database, only that the only usable relation is "child-of". Our primary
> >goal was to allow information to be found quickly by a key other than the
> >name, such as finding all music by "Pink Floyd" or all songs from "The
> >Wall".
> Please read www.namesys.com/whitepaper.html, which describes the
> semantics intended for reiser6.
> >To this end I'd like to do whatever I can to assist. I'd rather
> >contribute to ReiserFS than produce a redundant effort. Other members of
> >the Unununium effort have expressed concerns about the suitability of
> >ReiserFS to our task and the time it may take to mature. Any remarks you
> >can make to quell their restlessness would be greatly appreciated and may
> >convert a talented developer to your cause :-)
> Reiser4 is getting installed on my desktop sometime this week, so I
> don't think you should worry about the time to mature, unless you mean
> the time required to implement semi-structured semantics on top of the
> existing storage layer.
>
> thanks kindly for your interest in ReiserFS
>
> --
> Hans
Yes, that's what I mean. As a new OS with no notions of compatibility or
tradition we'd rather have the right system initially rather than to regress
later. I'd like to know if any such development is activly in progress, and
if so, when it is estimated to be serviceable. Or if you prefer, what would
be involved time-wise to implement it given the current state? As we are
already working on such a project, with a little perswasion I'm might get a
developer in addition to myself working towards that goal.
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2003-11-13 23:06 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-11-13 20:19 porting to alternate OSs and relational semantics Phil Frost
2003-11-13 10:24 ` Hans Reiser
2003-11-13 23:06 ` Phil Frost
2003-11-13 15:57 ` Hans Reiser
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.