Re: [9fans] 9P or better file services for multiple platforms

Re: [9fans] 9P or better file services for multiple platforms

[cinap_lenrek]

this has little chance to get into linux imho. theres nobody
in charge.

supporting foreign protocols in plan9 is doable as done with
ntlm for example.

the best authentication infrastructure linux has is ssh with
ssh-agent imho. so we might as well let linux users authenticte
against plan9 using ther existing ssh rsa keys (for stuff like
9p mounts, drawterm).

--
cinap

Re: [9fans] 9P or better file services for multiple platforms

[Brian L. Stuart]

On Sat, 9/1/18, Lucio De Re <lucio.dere@gmail.com> wrote:
> I'm trying to arrive at the most elegant solution to the following
> problem that does not sacrifice a great deal of efficiency. And, maybe
> I need to state this, the final result must be as robust or more
> robust than what I have in place currently, which has yet to let me
>
> My system mixes up Plan 9 and Linux platforms, none of them
>
> My hope is to provide a central file server that fulfills reliable
> file services to both Plan 9 and Linux as seamlessly as possible. I am

I'm not going to make any guarantees about the reliability,
but I do have a file system running on Plan 9, that natively
provides NFS service as well as 9P.  I also run it with a
snapshot device under it and get the type of history we
expect in a Plan 9 world.  To make the *nix side happy,
it does support symbolic links.  (Reading a symlink in Plan 9
just results in the path name string that the link points to.)
And to make things really fun, it also serves AOE.

I've been running it now on my home system for several
years.  I honestly don't use the NFS capability all that often,
but I did test it a fair amount back at Coraid.  Recently, I
added a little feature to the snapshot device to allow me
to easily migrate to a larger disk.  As a matter of fact, I read
your e-mail in acme on a 9vx which was taking its root from
this file system.

I'm sure there are plenty of nits that people could pick with
it and the details of its design, but it was an interesting
approach to experiment with and it's been serving me
well for about 4 or 5 years.  The file system itself runs in
user space on vanilla Plan 9, and the snapshot device
can be added to the kernel very easily.

Although there is a version of both the snapshot device
and the file system on contrib, if anyone's interested, I
can get you the most recent code to play with.

BLS

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, Brian L. Stuart <blstuart@bellsouth.net> wrote:
> On Sat, 9/1/18, Lucio De Re <lucio.dere@gmail.com> wrote:
>> [ ...  ]
>> My hope is to provide a central file server that fulfills reliable
>> file services to both Plan 9 and Linux as seamlessly as possible. I am
>
> I'm not going to make any guarantees about the reliability,
> but I do have a file system running on Plan 9, that natively
> provides NFS service as well as 9P.

That's definitely a good start.

> I also run it with a
> snapshot device under it and get the type of history we
> expect in a Plan 9 world.  To make the *nix side happy,
> it does support symbolic links.  (Reading a symlink in Plan 9
> just results in the path name string that the link points to.)

Does stat report anything fancy in the case of symbolic links? That
may be all that's needed to handle special applications that are
willing and able... But I will take more than a peek, find there the
answers I seek.

> And to make things really fun, it also serves AOE.
>
I wish...

> I've been running it now on my home system for several
> years.  I honestly don't use the NFS capability all that often,
> but I did test it a fair amount back at Coraid.  Recently, I
> added a little feature to the snapshot device to allow me
> to easily migrate to a larger disk.  As a matter of fact, I read
> your e-mail in acme on a 9vx which was taking its root from
> this file system.

But 9vx (which I nearly worship, but have been reluctant to get
attached to since I've been running 64-bit Linux) uses 9P, so that is
purely of anecdotal interest - I love anecdotes, myself. It's been a
long time, long enough for me to forget whether 9vx will execute Go
binaries or not - now, that's embarrassing!

>
> I'm sure there are plenty of nits that people could pick with
> it and the details of its design, but it was an interesting
> approach to experiment with and it's been serving me
> well for about 4 or 5 years.  The file system itself runs in
> user space on vanilla Plan 9, and the snapshot device
> can be added to the kernel very easily.
>
What nits would you pick, though, being the best qualified to do that?
I have a feeling for your abilities, this may be something 9fans (the
people, not the list) may be able to add much value to, if given some
direction, even I may be able to chip in.

> Although there is a version of both the snapshot device
> and the file system on contrib, if anyone's interested, I
> can get you the most recent code to play with.
>
Not Venti? I like Venti even more than I like 9vx, although
nit-picking there would yield a rich crop :-).

tl;dr: Please upload you efforts to github (or...) unless you have
extreme reasons not to. I'd love to be part of a Plan 9 revival and we
do have a presence on github, let's exploit the time github has got
left.

(GIT is the main reason for my begging here: I "pull" the latest Go to
my workstation, then "archive" to Plan 9 (fossil). But fossil is slow,
too slow to compete even with cross-compiling to plan9_386. Part of
the problem is needing to flush the "archive" target in case bits have
been removed and "export" does not delete them on the target - that
works well, though, with fresh releases, like go1.11, of late. To be
honest, I have the slack to use Plan 9 on a low-powered
server/workstation combination and even cross-produce Linux-64
executables for production; but building the Go distribution isn't
really worth the trouble - I make a special effort to do that, not
often enough.)

Lucio.

PS: I prefer not to be controversial, but Plan 9 needs a bit of
religious fervour to keep it alive: if I need to start a fire under
anyone's buttocks to help along, trust me, I'll do it. :-)

Re: [9fans] 9P or better file services for multiple platforms

[Bakul Shah]

> On Sep 2, 2018, at 2:25 AM, Lucio De Re <lucio.dere@gmail.com> wrote:
> 
> (GIT is the main reason for my begging here: I "pull" the latest Go to
> my workstation, then "archive" to Plan 9 (fossil). But fossil is slow,
> too slow to compete even with cross-compiling to plan9_386. Part of
> the problem is needing to flush the "archive" target in case bits have
> been removed and "export" does not delete them on the target - that
> works well, though, with fresh releases, like go1.11, of late. To be
> honest, I have the slack to use Plan 9 on a low-powered
> server/workstation combination and even cross-produce Linux-64
> executables for production; but building the Go distribution isn't
> really worth the trouble - I make a special effort to do that, not
> often enough.)

I now run 9front on FreeBSD under bhyve on a 10 year old
athlon64 machine. Initially I cross compiled go but now a native
compile doesn’t take all that long, using a previously compiled
Go as the bootstrap compiler. I’m using 9front’s new filesystem,
not fossil.

The “dgit” program (Dave McFarlane, with assists from Chris
McGee) works well enough now for “go get” or “git pull”.

I too want a unified filesystem that all my machines feed off of but
so far I have not found a solution. Local file systems are still faster.

[9fans] 9P or better file services for multiple platforms

[Lucio De Re]

I'm trying to arrive at the most elegant solution to the following
problem that does not sacrifice a great deal of efficiency. And, maybe
I need to state this, the final result must be as robust or more
robust than what I have in place currently, which has yet to let me
down, partially due to the very duplication I'd like to eliminate. I
think such robustness is possible without the excessive redundancy of
the current system.

My system mixes up Plan 9 and Linux platforms, none of them
particularly modern, in various fashion. Linux runs on workstations
(and laptops) and Plan 9 has one, sometimes more, CPU/fileservers and
a main workstation, plus a small secondary one.

My hope is to provide a central file server that fulfills reliable
file services to both Plan 9 and Linux as seamlessly as possible. I am
willing to sacrifice a few Unix features, such as file links, in that
file server, if I can dedicate it to a narrower role than to support
the full Linux environment. In Plan 9 parlance, I only need file
services, not computing capabilities and the file server is allowed to
limit some of the computing needs involved (like, say, graphics, any
multimedia stuff, even mouse use).

The question, then, is what file service will satisfy these needs,
including access control, automatic backup as provided by default
under Plan 9, etc. I am not very fond of Linux's propensity to need
daily upgrades, but Plan 9 has quirks of its own, which I would be
hard pressed to enumerate here, but we are all aware of.

If I could run the file server on a modern (or even an ancient)
version of NetBSD, I'd be even happier as NetBSD is the Unix flavour I
highly favour. But that is a bonus, not an essential.

I welcome questions that may help me give this a more concrete shape.
I'm hoping there's a definitive answer out there and all proposal will
receive my attention, I do suspect it will be a compromise solution
I'll need to consider, but that is perfectly understandable.

A big thank you to anyone who is willing to contribute.

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Rui Carmo]

[-- Attachment #1: Type: text/plain, Size: 127 bytes --]

I myself have similar needs and recently bookmarked this: https://github.com/chaos/diod (but had no time to test it yet).

R.

[-- Attachment #2: Type: text/html, Size: 342 bytes --]

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/1/18, Rui Carmo <rui.carmo@gmail.com> wrote:
> I myself have similar needs and recently bookmarked this:
> https://github.com/chaos/diod (but had no time to test it yet).
>
Thank you, Rui, that looks pretty exciting, I'll be happy to look into it.

It does rather look like Plan 9 itself may have to be of the 9front
variety, but that is a minor obstacles and in fact more an incentive
for me to familiarise myself with it. At the moment only 9legacy is
readily available to me.

That's just my gut reaction, I may be totally wrong.

I really appreciate your response.

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Emery Hemingway]

I don't think you can find better than u9fs for unix.

I've tried to use diod once or twice, but it is some weird overengineered
linux shit.

Emery

On Saturday, September 1, 2018 3:33:54 PM CEST, Lucio De Re wrote:
> On 9/1/18, Rui Carmo <rui.carmo@gmail.com> wrote:
>> I myself have similar needs and recently bookmarked this:
>> https://github.com/chaos/diod (but had no time to test it yet).
>> 
> Thank you, Rui, that looks pretty exciting, I'll be happy to look into it.
>
> It does rather look like Plan 9 itself may have to be of the 9front
> variety, but that is a minor obstacles and in fact more an incentive
> for me to familiarise myself with it. At the moment only 9legacy is
> readily available to me.
>
> That's just my gut reaction, I may be totally wrong.
>
> I really appreciate your response.
>
> Lucio.
>
>
>

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/1/18, Emery Hemingway <ehmry@posteo.net> wrote:
> I don't think you can find better than u9fs for unix.
>
I tend to use that as a norm, but the backing Plan 9 server is kind of
in the wrong "key". OK for Plan 9, but too slow for Linux. Still, that
sounds like a warning that better that u9fs is not out there.

> I've tried to use diod once or twice, but it is some weird overengineered
> linux shit.
>
I quickly built it and deployed it on my Linux Mint (32-bit, 1.18.1 or
some such), and it builds OK, skips a lot of tests, but fails none. As
"root" the tests jam, that can't be good.

Trying it out, it fails to find "attach" and there is no clue where
that should come from. It did strike me as complex, but if it serves
an NFS filesystem, that is probably adequate.

I'll wait to pass judgement for after I have it actually serving
anything at all.

Thanks for your comments, in any case.

Lucio.

PS: I suppose the Plan 9 problem is that it is too many different
things to too many different people and it has yet to find one niche
purpose that it serves better than any other OS. It's a sad destiny,
really sad.

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/1/18, Lucio De Re <lucio.dere@gmail.com> wrote:
>
> Trying it out, it fails to find "attach" and there is no clue where
> that should come from. It did strike me as complex, but if it serves
> an NFS filesystem, that is probably adequate.
>
> I'll wait to pass judgement for after I have it actually serving
> anything at all.
>
I have made some progress, /tmp/9/ (the missing link, it seems) and
/mnt/ now look the same, so presumably something is going on. Maybe
the overengineering isn't too top-heavy, I need to check.

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Joseph Stewart]

[-- Attachment #1: Type: text/plain, Size: 894 bytes --]

This thread got me searching and I found MJL's guide for running a plan9
network on a *nix system using u9fs.

Hope this helps:

https://www.ueber.net/who/mjl/plan9/plan9-obsd.html

I'm gonna tinker with this myself.

-joe

On Sat, Sep 1, 2018 at 8:20 AM Lucio De Re <lucio.dere@gmail.com> wrote:

> On 9/1/18, Lucio De Re <lucio.dere@gmail.com> wrote:
> >
> > Trying it out, it fails to find "attach" and there is no clue where
> > that should come from. It did strike me as complex, but if it serves
> > an NFS filesystem, that is probably adequate.
> >
> > I'll wait to pass judgement for after I have it actually serving
> > anything at all.
> >
> I have made some progress, /tmp/9/ (the missing link, it seems) and
> /mnt/ now look the same, so presumably something is going on. Maybe
> the overengineering isn't too top-heavy, I need to check.
>
> Lucio.
>
>

[-- Attachment #2: Type: text/html, Size: 1453 bytes --]

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/1/18, Joseph Stewart <joseph.stewart@gmail.com> wrote:
> This thread got me searching and I found MJL's guide for running a plan9
> network on a *nix system using u9fs.
>
> Hope this helps:
>
> https://www.ueber.net/who/mjl/plan9/plan9-obsd.html
>
> I'm gonna tinker with this myself.
>
That authsrv9 looks very promising. Definitely a gem worth knowing about.

What would be necessary for some bod to curate (as David to a great
extent does with the 9legacy stuff, better than may seem justified,
for his pay grade :-) ) a broader range of Plan 9 offerings, so it's
easier to keep the community informed?

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

what form of curation are you imagining? we have a generic place for
plan9-related news at http://ninetimes.cat-v.org/, but perhaps we
haven't looked out far enough around 9front?

there used to also be a planet/rss aggregation, but nobody alive knows
how to get the used software behind this to run on plan9 again. and
most people's blogs in that list shifted from plan9-related to
go-related activities, so i'm not sure it's even worth updating any
more.

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

i used to think it's good to put plan9 authentication in the linux
kernel, and perhaps it's true, because linux is so big, this little
more doesn't even hurt that much.

but it might be better to extend the pseudo-plan9-kernels in
drawterm/inferno in a way so that the linux kernel can talk plain 9p
to them, so that less code has to depend on redhat^Wlinux
patch-refusal monarchy.

already cinap is supporting dp9ik in drawterm, so...

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, hiro <23hiro@gmail.com> wrote:
>
> already cinap is supporting dp9ik in drawterm, so...
>
That's subversive in the most practical sense. Is academia aware of it
and its import? That is what curatorship (a friend from past days was
and may still be the curator of the South African National (Art)
Gallery, I am thinking of the role he and his predecessors play in the
Arts landscape): inform potentially interested parties of the
existence of something valuable.

Of course, it is clearly not only academia that needs to know and
curatorship would extend beyond that, clearly. Sponsorship is probably
even more important. A curator may be able to justify a salary from a
successful crowd-funding effort, for example.

Am I being unreasonably optimistic?

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

On 9/2/18, Lucio De Re <lucio.dere@gmail.com> wrote:
> On 9/2/18, hiro <23hiro@gmail.com> wrote:
>>
>> already cinap is supporting dp9ik in drawterm, so...
>>
> That's subversive in the most practical sense. Is academia aware of it
> and its import? That is what curatorship (a friend from past days was
> and may still be the curator of the South African National (Art)
> Gallery, I am thinking of the role he and his predecessors play in the
> Arts landscape): inform potentially interested parties of the
> existence of something valuable.
>
> Of course, it is clearly not only academia that needs to know and
> curatorship would extend beyond that, clearly. Sponsorship is probably
> even more important. A curator may be able to justify a salary from a
> successful crowd-funding effort, for example.
>
> Am I being unreasonably optimistic?
>
> Lucio.
>
>

I'm not aware of adademia, but funds and code are always welcome.

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

The way I inform myself of valuable contributions to plan 9 these days
is by watching 9front commit logs and the #cat-v irc channel.

If there are any valuable commits in David's repository that we should
apply, please inform us.

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, hiro <23hiro@gmail.com> wrote:
> The way I inform myself of valuable contributions to plan 9 these days
> is by watching 9front commit logs and the #cat-v irc channel.
>
> If there are any valuable commits in David's repository that we should
> apply, please inform us.
>
I was waiting for the real Hiro to pop out <grin>.

Of course, you're sadly right. On the merit side, 9legacy is
frighteningly stable (ask me, I ain't changed a thing in years), so it
is eminently ready to absorb 9front's commits, if only the effort did
not reduce 9legacy to a stuttering idiot.

Small as Plan 9 is, bringing divergent developments together requires
that scarce resource called humility on all sides. I can only beat
humility into others, no risk of swallowing a dose of it for my own
edification.

No, I am joking, which is another side of Plan 9 I cherish: few, if
anyone at all have their livelihood dependent on Plan 9. I actually
can fake humility better than I can fake just about anything else and
people do accept it as real. That's good enough.

To return to 9legacy versus 9front, I'm totally incompetent at this
point, but I'm hoping to familiarise myself with 9front fairly soon,
it is too inviting. But my loyalty to the original Plan 9 source is
very strong, it's hard to tell how it will pan out.

In the meantime, I do have answers I need to pursue, and this is as
good an opportunity as any to say thanks to all who have reply and I
look forward to hearing even the smallest tidbits that may illuminate
the way forward. Unknown to the majority, we are privileged to be
acquainted in some detail with a precious thing, we need, as a
community, to appreciate whatever it is that we can do to nurture it.

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

> Of course, you're sadly right.

I don't know what's so sad to you.

apart from all other developers having left and work for google now.

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

> bringing divergent developments together

one big diversion one should not forget is inferno.
i would welcome if some efforts were put into synchronizing some of
the inferno and plan 9 changes and perhaps apply acquired wisdom that
stems from either project.

otherwise 9front doesn't try to divert at all. even some bell-labs
changes that were frowned upon and broke backwards compatibility were
imported into 9front bec. of the wish to stay compatible.

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

I think your reply touches most of the sore spots I am familiar with.
There's no doubt that 9legacy is missing out badly from the absence of
cinap's contributions to 9front and I'm the first to believe that a
one and true Plan 9 cannot afford to omit such pertinent enhancements,
just to list one, pretty massive example.

In a way, David IS the Plan 9 curator, but like all volunteers, his
focus has to be restricted to his interest scope, at the expense of
features that may spell life and death (in a metaphoric sense) for the
Plan 9 effort.

If one considers how much has been spent on Plan 9 so far and how much
it has served, irrespective of the emotional and aesthetic appeal that
we as 9fans presumably feel for it, it would be a shame to let slip,
as no doubt has happened already in the past, opportunities to give
Plan 9 a better grip in its context to ensure its survival against
more popular brands.

I guess it's a Darwinian issue: is there a flavour of Plan 9 capable
of earning sufficient traction where it matters (minds, more likely
than physical hosts) to ensure that its many still under-utilised
unique features are not lost, particularly to our very own community?
Can that flavour of Plan 9 be created, if necessary, from the
available "corpus" and what would the cost a committed community like
this one would need to bear to make that possible?

Such an effort would be the curatorship I have in mind, if you can call it that.

Lucio.


On 9/2/18, hiro <23hiro@gmail.com> wrote:
> what form of curation are you imagining? we have a generic place for
> plan9-related news at http://ninetimes.cat-v.org/, but perhaps we
> haven't looked out far enough around 9front?
>
> there used to also be a planet/rss aggregation, but nobody alive knows
> how to get the used software behind this to run on plan9 again. and
> most people's blogs in that list shifted from plan9-related to
> go-related activities, so i'm not sure it's even worth updating any
> more.
>
>


--
Lucio De Re
2 Piet Retief St
Kestell (Eastern Free State)
9860 South Africa

Ph.: +27 58 653 1433
Cell: +27 83 251 5824
FAX: +27 58 653 1435

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, hiro <23hiro@gmail.com> wrote:
>
> there used to also be a planet/rss aggregation, but nobody alive knows
> how to get the used software behind this to run on plan9 again

I used to be a debugging whiz, in happier, more youthful times, maybe
I can give that a try (it seems a challenge, rather than a really
useful idea). Would you care to make that available?

I do think there are many here who would be better at this, so if you
say "no one alive" it does sound totally hopeless. But I presume
you're exaggerating :-).

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

i think the original planet software ran on linux. right now cat-v.org
is maintained by sl, and on 9front, not linux.

and it might indeed be best to concentrate on creating software of
actual value, as opposed to administrating even more third-party
systems that don't share our style of approach...

i suppose you could check the individual blogs, possibly in an
automated way by writing some one-liner rc and hget script and publish
the outcome, plus keep it updated. then perhaps you can figure out if
this is the kind of information currently lacking.

On 9/2/18, Lucio De Re <lucio.dere@gmail.com> wrote:
> On 9/2/18, hiro <23hiro@gmail.com> wrote:
>>
>> there used to also be a planet/rss aggregation, but nobody alive knows
>> how to get the used software behind this to run on plan9 again
>
> I used to be a debugging whiz, in happier, more youthful times, maybe
> I can give that a try (it seems a challenge, rather than a really
> useful idea). Would you care to make that available?
>
> I do think there are many here who would be better at this, so if you
> say "no one alive" it does sound totally hopeless. But I presume
> you're exaggerating :-).
>
> Lucio.
>
>

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, hiro <23hiro@gmail.com> wrote:
>
> i suppose you could check the individual blogs, possibly in an
> automated way by writing some one-liner rc and hget script and publish
> the outcome, plus keep it updated. then perhaps you can figure out if
> this is the kind of information currently lacking.
>
The book, the book! I think that's what I'll look into. We have Nemo's
commentary and of course Forsyth's (9) man pages, maybe a start would
be to condense knowledge in a "Tiger's" commentary of 9front. I can
already see the two columns of 9legacy and 9front code side by side
(the kernel isn't going to be enough, though, you guys have been way
too busy).

And, while I have your attention, I was a fan of Tanenbaum and Minix.
Is anyone here familiar with Minix v3? Thinking about, isn't a book
precisely what it takes to guarantee an OS eternal life?

Thanks for the thought, Hiro, in my dotage, when evening comes, my
thinking slows down beyond useful pondering, blogs may be just what I
need instead of television. That will almost certainly improve the
quality of my last years.

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Ethan Gardener]

On Sat, Sep 1, 2018, at 11:29 AM, Rui Carmo wrote:
> 
> I myself have similar needs and recently bookmarked this: https://github.com/chaos/diod (but had no time to test it yet).

diod's readme states it speaks 9p2000.L.  Isn't that incompatible with Plan 9?  I recall reading something like, "9p maps Plan 9 syscalls to messages, 9p2000.L maps Linux syscalls to messages."

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

no, 9p2000.L or Linux syscalls are not supported by plan9.

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/1/18, hiro <23hiro@gmail.com> wrote:
> no, 9p2000.L or Linux syscalls are not supported by plan9.
>
>
So Ethan is right, this is a "lark", if an interesting one. 9P is
getting quite a few "takers". I seem to recall a paper on adding Plan
9 authentication to the Linux kernel, for purposes beyond the Plan 9
scope? That also needed 9P features.

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, Lucio De Re <lucio.dere@gmail.com> wrote:
> On 9/1/18, hiro <23hiro@gmail.com> wrote:
>> no, 9p2000.L or Linux syscalls are not supported by plan9.
>>
>>
> So Ethan is right, this is a "lark", if an interesting one. 9P is
> getting quite a few "takers". I seem to recall a paper on adding Plan
> 9 authentication to the Linux kernel, for purposes beyond the Plan 9
> scope? That also needed 9P features.
>
Found it <https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/34433.pdf>:

"This paper discusses the extension of Linux authentication
mechanisms to allow the use of the Plan 9 approach with
existing Linux applications in order to reduce the security
risks mentioned earlier. It describes the port of the Plan 9
capability device as a character device driver for the Linux
kernel. It also describes the port of the Plan 9 authentication
server and the implementation of a PAM module which
allows the use of these new facilities. It is now possible to
restrain processes like login and su from the uncontrolled setuid
bit and make them run on behalf of an unprivileged user
in Linux."

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[hiro]

On 9/2/18, Lucio De Re <lucio.dere@gmail.com> wrote:
> On 9/2/18, Lucio De Re <lucio.dere@gmail.com> wrote:
>> On 9/1/18, hiro <23hiro@gmail.com> wrote:
>>> no, 9p2000.L or Linux syscalls are not supported by plan9.
>>>
>>>
>> So Ethan is right, this is a "lark", if an interesting one. 9P is
>> getting quite a few "takers". I seem to recall a paper on adding Plan
>> 9 authentication to the Linux kernel, for purposes beyond the Plan 9
>> scope? That also needed 9P features.
>>
> Found it
> <https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/34433.pdf>:
>
> "This paper discusses the extension of Linux authentication
> mechanisms to allow the use of the Plan 9 approach with
> existing Linux applications in order to reduce the security
> risks mentioned earlier. It describes the port of the Plan 9
> capability device as a character device driver for the Linux
> kernel. It also describes the port of the Plan 9 authentication
> server and the implementation of a PAM module which
> allows the use of these new facilities. It is now possible to
> restrain processes like login and su from the uncontrolled setuid
> bit and make them run on behalf of an unprivileged user
> in Linux."
>
> Lucio.
>
>

sounds like a linux-centric approach, with little or even no novelty.
once you have to deal with stuff like PAM you can give up altogether.

Re: [9fans] 9P or better file services for multiple platforms

[Ethan Gardener]

I had a thought pertaining to the original topic.

On Sat, Sep 1, 2018, at 6:21 AM, Lucio De Re wrote:
> The question, then, is what file service will satisfy these needs,
> including access control, automatic backup as provided by default
> under Plan 9, etc. I am not very fond of Linux's propensity to need
> daily upgrades, but Plan 9 has quirks of its own, which I would be
> hard pressed to enumerate here, but we are all aware of.
>
> If I could run the file server on a modern (or even an ancient)
> version of NetBSD, I'd be even happier as NetBSD is the Unix flavour I
> highly favour. But that is a bonus, not an essential.

Regarding the automatic backup, FreeBSD's UFS (the default filesystem) provides snapshots much like CWFS's dump.  I think you can enable it for the whole filesystem with a mount option, but if not, I think it's quite easy to set up anyway.

FreeBSD has ZFS too, which of course offers snapshots, but it has so many options that I found it a bit too much.  It seems well documented and the interface seems reasonable for the feature set, but it's a big feature set.  I have a habit of trying to understand and use too much at once, I guess.

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, Ethan Gardener <eekee57@fastmail.fm> wrote:
> I had a thought pertaining to the original topic.
>
[ ... ]
>
> FreeBSD has ZFS too, which of course offers snapshots, but it has so many
> options that I found it a bit too much.  It seems well documented and the
> interface seems reasonable for the feature set, but it's a big feature set.
> I have a habit of trying to understand and use too much at once, I guess.
>
A personality analysis of 9fans I'm sure will reveal some traits that
make us very much alike. A desire to get to the bottom of things would
be one of those, is my guess.

Of course, it's simplistic. Thing is, if our community were to find a
sufficiently capable leader, we'd be able to contribute immensely.
It's unfortunate that such coordination doesn't naturally occur
whenever the resources become available. But maybe the necessary
catalyst will eventually materialise, before it's too late.

Here's to that hope.

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Ethan Gardener]

On Sun, Sep 2, 2018, at 7:02 PM, Lucio De Re wrote:
> On 9/2/18, Ethan Gardener <eekee57@fastmail.fm> wrote:
> > I had a thought pertaining to the original topic.
> >
> [ ... ]
> >
> > FreeBSD has ZFS too, which of course offers snapshots, but it has so many
> > options that I found it a bit too much.  It seems well documented and the
> > interface seems reasonable for the feature set, but it's a big feature set.
> > I have a habit of trying to understand and use too much at once, I guess.
> >
> A personality analysis of 9fans I'm sure will reveal some traits that
> make us very much alike. A desire to get to the bottom of things would
> be one of those, is my guess.

Probably! :)

> Of course, it's simplistic. Thing is, if our community were to find a
> sufficiently capable leader, we'd be able to contribute immensely.
> It's unfortunate that such coordination doesn't naturally occur
> whenever the resources become available. But maybe the necessary
> catalyst will eventually materialise, before it's too late.
>
> Here's to that hope.

I don't know.  cinap tried to teach me.  He's a good teacher, but he couldn't help me.  Things are looking up now I'm trying to improve my life generally, dealing with sources of stress instead of trying to block them out with a million distractions.  I'm cutting back to one hobby.  Unfortunately, it's not Plan 9, but it will share some ideas and I intend to make it run on Plan 9 when it runs at all.  Anyway, one hobby means anyone trying to get me to work on Plan 9 would be up against some stiff competition. :)

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

You're here. Sometimes an audience is all the artist needs as the
stimulus. How does it go? "They also serve...".

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Ethan Gardener]

On Sun, Sep 2, 2018, at 7:24 PM, Lucio De Re wrote:
> You're here. Sometimes an audience is all the artist needs as the
> stimulus. How does it go? "They also serve...".

I probably shouldn't talk about all I've done for the sake of an audience. XD  I tell myself I'm doing my latest project primarily for myself though.

Re: [9fans] 9P or better file services for multiple platforms

[Skip Tavakkolian]

[-- Attachment #1: Type: text/plain, Size: 1837 bytes --]

Have you considered using AoE (Coraid)? It would require dedicated fossil,
NFS CIFS servers, but they'd all be sharing the storage -- Coraid supports
ext4 and NTFS. Most servers have multiple NICs, which makes a dedicated LAN
for AoE traffic easy.

Regarding authentication and access control, I think the only *standard*
option for a mixed OS environment (Plan 9, Linux/*BSD, Windows) is Kerberos.

It would be great if factotum could handle Kerberos 5. There is a pure Go
package (https://github.com/jcmturner/gokrb5) that could help. Of course,
this would take some work.

This setup would require 4 (or potentially only 3) machines:
  * AoE block storage
  * the KDC (MIT, Heimdal, or MS Active Directory)
  * Plan 9 file server (fossil, but also CIFS and NFS servers)
  * Linux or NetBSD file server for CIFS and NFS

-Skip

On Fri, Aug 31, 2018 at 10:22 PM Lucio De Re <lucio.dere@gmail.com> wrote:

>
>
> My hope is to provide a central file server that fulfills reliable
> file services to both Plan 9 and Linux as seamlessly as possible. I am
> willing to sacrifice a few Unix features, such as file links, in that
> file server, if I can dedicate it to a narrower role than to support
> the full Linux environment. In Plan 9 parlance, I only need file
> services, not computing capabilities and the file server is allowed to
> limit some of the computing needs involved (like, say, graphics, any
> multimedia stuff, even mouse use).
>
> The question, then, is what file service will satisfy these needs,
> including access control, automatic backup as provided by default
> under Plan 9, etc. I am not very fond of Linux's propensity to need
> daily upgrades, but Plan 9 has quirks of its own, which I would be
> hard pressed to enumerate here, but we are all aware of.
>
> Lucio.
>
>

[-- Attachment #2: Type: text/html, Size: 2304 bytes --]

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, Skip Tavakkolian <skip.tavakkolian@gmail.com> wrote:
> Have you considered using AoE (Coraid)? It would require dedicated fossil,
> NFS CIFS servers, but they'd all be sharing the storage -- Coraid supports
> ext4 and NTFS. Most servers have multiple NICs, which makes a dedicated LAN
> for AoE traffic easy.
>
I have, many times, Skip.

I was told once or twice that is money is the only obstacle between me
and success, then there is no obstacle at all, but my ability to grow
debt or at least keep it at the permanent level my bank tolerates from
me, is at least one factor against having any Coraid kit in my space.

I think that is the greatest regret I have, by far. Coraid kit should
be listed as a non-negotiable human right.

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/2/18, Skip Tavakkolian <skip.tavakkolian@gmail.com> wrote:
>
> Regarding authentication and access control, I think the only *standard*
> option for a mixed OS environment (Plan 9, Linux/*BSD, Windows) is
> Kerberos.
>
Is that still actively used (I mean, outside of Microsoft's attempted
hi-jacking)? In my Linux-prone wider environment, the name is never
uttered.

> It would be great if factotum could handle Kerberos 5. There is a pure Go
> package (https://github.com/jcmturner/gokrb5) that could help. Of course,
> this would take some work.
>
Factotum deserves a research institute of its own. Every time I have
to change a password that has expired or have to look in my secstore
for one I have not used in a while, I wish I could add the few modules
I find lacking in factotum.

So much Plan 9 that comes just short of perfect, where everything else
is at best mediocre!

Lucio.

Re: [9fans] 9P or better file services for multiple platforms

[Kurt H Maier]

On Sun, Sep 02, 2018 at 08:09:55PM +0200, Lucio De Re wrote:
> On 9/2/18, Skip Tavakkolian <skip.tavakkolian@gmail.com> wrote:
> >
> > Regarding authentication and access control, I think the only *standard*
> > option for a mixed OS environment (Plan 9, Linux/*BSD, Windows) is
> > Kerberos.
> >
> Is that still actively used (I mean, outside of Microsoft's attempted
> hi-jacking)? In my Linux-prone wider environment, the name is never
> uttered.

Yes, it's extremely common in many business and government
environments.  All of linux's weird-ass authentication systems are
poorly-reinvented kerberos implementations, with the primary limitations
and pain points directly stemming from unix tropes.  Generally someone
comes up with a bad idea, everyone adopts it, and then that bad idea
slowly evolves as closely as it can to being kerberos.

Most commonly, someone will mandate two-factor authentication, and
kerberos tickets (usually via GSSAPI) are the back-end, regardless of
which security tokens (RSA SecurID, smart cards, yubikeys, etc) are
chosen.

khm

Re: [9fans] 9P or better file services for multiple platforms

[Lucio De Re]

On 9/3/18, Kurt H Maier <khm@sciops.net> wrote:
> [ ... ]
>
> Most commonly, someone will mandate two-factor authentication, and
> kerberos tickets (usually via GSSAPI) are the back-end, regardless of
> which security tokens (RSA SecurID, smart cards, yubikeys, etc) are
> chosen.
>
Thanks, Kurt, I knew 9fans was a fountain of knowledge with, mostly,
just the right sprinkling of skepticism. Plus the humour, of course.

Lucio.

PS: I suffered a server failure, way back when, and returning to the
9fans fold took some determination. I'm not sure I'll be able to do
without 9fans any time soon.