Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Lucio De Re]

On 12/3/19, ori@eigenstate.org <ori@eigenstate.org> wrote:
>
> Unrelatedly, would there be interest in adding the `$split{cmd} syntax
> from 9atom to 9legacy?  I think it's currently the only reason that
> git9 doesn't work out of the box there, and it's very nice syntax.
>
I'm quite obsessed with converging the very best of all 9worlds into a
single 9heritage version as an on-going project, but I am not very
good at staying focused. Updating rc I think is both a very good start
and a potential curse: the divergence in 9front is quite significant.

Plus, we need git9 and better for the community to contribute to a
neutral project (my opinion, of course).

What feels right to me is to focus on git9 before going any further
and make it as 9friendly as possible,  to all 9fans. Would that
conflict with anyone's pet use of Plan 9?

If it makes sense, the git9 project could supply an upgraded rc as an
additional incentive.

But most of all, I think projects like these need participation and a
record of decisions taken. I know Go have a proposal system, could we
attempt to create something similar? Ori did point out in his
presentation that /sys/doc has great value in the Plan 9 ecosystem and
I could not agree more.

Lucio.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[ori]

> On 12/3/19, ori@eigenstate.org <ori@eigenstate.org> wrote:
>>
>> Unrelatedly, would there be interest in adding the `$split{cmd} syntax
>> from 9atom to 9legacy?  I think it's currently the only reason that
>> git9 doesn't work out of the box there, and it's very nice syntax.
>>
> I'm quite obsessed with converging the very best of all 9worlds into a
> single 9heritage version as an on-going project, but I am not very
> good at staying focused. Updating rc I think is both a very good start
> and a potential curse: the divergence in 9front is quite significant.
>
> Plus, we need git9 and better for the community to contribute to a
> neutral project (my opinion, of course).
> 
> What feels right to me is to focus on git9 before going any further
> and make it as 9friendly as possible,  to all 9fans. Would that
> conflict with anyone's pet use of Plan 9?

*shrugs* Sure, if it doesn't work the way you want it to, feel free to
send in patches.

I prefer you send changes as emails with inline patches that I can
'git/import'. You can generate that by committing and running
'git/export' with git9, or 'git format-patch HEAD^' on unix.

> But most of all, I think projects like these need participation and a
> record of decisions taken. I know Go have a proposal system, could we
> attempt to create something similar? Ori did point out in his
> presentation that /sys/doc has great value in the Plan 9 ecosystem and
> I could not agree more.

I thought that was patch(1).

The missing part doesn't seem to be the mechanism, but the people
using it.  Very few people are sending them, and nobody is building a
release with an "OFFICIAL" stamp on it with appropriate patches
integrated, and moved to the applied directory.

Though, I suppose it says something when the version of patch in
9legacy still tries to submit to sources.cs.bell-labs.com.

I'll leave it to the 9legacy folks to figure out how they want to
handle that, but I'd be happy to see patch notifications ending up on
9fans.  Patches aren't high volume enough at the moment to warrant
their own list, and 9fans seems like a good way to both let observers
know what's going on, and get eyeballs for review.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Richard Miller]

> and nobody is building a
> release with an "OFFICIAL" stamp on it

What exactly would you want such a stamp to mean?

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Lucio De Re]

On 12/3/19, Richard Miller <9fans@hamnavoe.com> wrote:
>> and nobody is building a
>> release with an "OFFICIAL" stamp on it
>
> What exactly would you want such a stamp to mean?
>
>
If I may hi-jack the question, I think one needs a starting point that
everyone can use as reference.  What we have is the exact opposite and
that "damages the brand".

It would be ideal if we actually aimed for a small, common Plan 9 core
with the best of all versions (take 9atom, for example: it is the only
version that can reboot my workstation, even 9front freezes when
pressing ctrl-alt-del) and good enough for all improvements to be
capable of being rebased upon it.

It would be a target, not a starting point, one I would think we can
all be inspired by?

Lucio.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[hiro]

we have no brand, but our kernel panics are better.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Skip Tavakkolian]

[-- Attachment #1: Type: text/plain, Size: 399 bytes --]

Our kernel panics aren't backward.

On Tue, Dec 3, 2019, 7:27 AM hiro <23hiro@gmail.com> wrote:

> we have no brand, but our kernel panics are better.
>
> ------------------------------------------
> 9fans: 9fans
> Permalink:
> https://9fans.topicbox.com/groups/9fans/T861dcd95691a032b-M485e7e8e1250ffc3861ba982
> Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
>

[-- Attachment #2: Type: text/html, Size: 932 bytes --]

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[hiro]

out of my pov there is no big problem, my quick summary of the history:
1) it *seemed* hard to get patches in when the gatekeepers were still
around (i didn't try)
2) respected gatekeepers have left the building (geoff was last iirc)
3) lots of stuff get fixed in 9front
4) nobody on 9fans realizes bec. the main developer doesn't care about
credit or trolling as much as discussing boring technical (results in
killfile on 9fans) issues.
5) 9front attracts lots of trolls that are able to use the system
efficiently and are now contributing patches back to the effort.
6) skip continues to protect the doormat somebody forgot in front, but
he never dared to look up at the long demolished building.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Skip Tavakkolian]

[-- Attachment #1: Type: text/plain, Size: 499 bytes --]

I can't tell if this is a similitude or an allegory.

On Tue, Dec 3, 2019, 12:32 PM hiro <23hiro@gmail.com> wrote:

> 6) skip continues to protect the doormat somebody forgot in front, but
> he never dared to look up at the long demolished building.
>
> ------------------------------------------
> 9fans: 9fans
> Permalink:
> https://9fans.topicbox.com/groups/9fans/T861dcd95691a032b-M2955e39c4e8b1fd29c25b4e5
> Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
>

[-- Attachment #2: Type: text/html, Size: 1134 bytes --]

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[hiro]

in your words it would be an insult, no?

seriously, i hear only complaints here, meanwhile 9front is doing just
fine. you are deciding against it for non-technical reasons (or
probably misunderstandings).

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[andrey mirtchovski]

i resubscribed a week ago but i don't think i'll last another week
here. sp9sss is where the action's at.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Skip Tavakkolian]

[-- Attachment #1: Type: text/plain, Size: 584 bytes --]

On Tue, Dec 3, 2019, 2:47 PM hiro <23hiro@gmail.com> wrote:

> in your words it would be an insult, no?
>

Of course! How did I miss that?!

seriously, i hear only complaints here, meanwhile 9front is doing just
> fine. you are deciding against it for non-technical reasons (or
> probably misunderstandings).
>

Is this irony?


> ------------------------------------------
> 9fans: 9fans
> Permalink:
> https://9fans.topicbox.com/groups/9fans/T861dcd95691a032b-Ma77821383571b0f9de61f05c
> Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
>

[-- Attachment #2: Type: text/html, Size: 1647 bytes --]

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[sstallion]

> seriously, i hear only complaints here, meanwhile 9front is doing just
> fine. you are deciding against it for non-technical reasons (or
> probably misunderstandings).

I have no interest in using 9front for both technical and
non-technical reasons.

The "community" is atrocious; anyone willing to turn a blind eye
toward bad behavior due to technical merit is sophomoric if not
abusive.  ISTR calling this "elitist fuckery" in the past, I still
stand by that.

Second, the gatekeeping that was done while the labs was still active
was tough, but usually fair.  Geoff did a nice job of keeping the
spirit of plan9 alive and often rejected larger changes that could
have upset the balance.  David's work on 9legacy also follows the same
spirit; it's a patch queue that pulls from a number of different
sources but strives to keep the original spirit intact.  I struggle to
see the same mindset held by 9front.

That said, there is work being done outside of 9front.  It's not as
public or as polarizing, but there are still people hacking away.

Steve

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Alex Musolino]

> I have no interest in using 9front for both technical and
> non-technical reasons.

What are some of the technical reasons?

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Lucio De Re]

The world has really moved along since 2000, even since 95/98 (all
clear Microsoft milestone years, in case one misses the connection).
That genie ain't going back in any bottle. Well, is there really more
than one, or is it just one, multiheaded Hydra?

I feel at home on Plan 9 and continue to curse, mostly the resource
greed of Linux, but also the continual "enhancements" it insists on
foisting on me. Android is worse by a few orders of magnitude, it's
even taken to remind me to restart my mobile instrument because it's
been accumulating spiderwebs for more than a week!

But then there is Go which is pretty much as bad, and yet I do not
feel the same dislike, something about efforts towards discipline seem
to avoid ruffling my feathers.

The analogy with the South African socio-political situation is
irresistible: the notable difference is that there isn't a power play
in action with real people's lives being threatened. But the same
dismissive attitude on one side that the other side uses to create an
unjustified boundary is pretty much identical.

But I would need to explain much I wish no one else will ever need to
learn about, to avoid being insulting rather than encouraging. I would
like the two main communities (despite Stallion's hints that we ought
to look outside of these) to define their differences and bring to a
table only those that one can build on. And that means a lot of
rejections that neither side is going to appreciate having to swallow.
But they are realities, essential if we are to move on.

The significant thing, though, that we seem to overlook, is that 9fans
is still active. For all the friction, we are still talking.

Lucio.

On 12/4/19, Alex Musolino <alex@musolino.id.au> wrote:
>> I have no interest in using 9front for both technical and
>> non-technical reasons.
>
> What are some of the technical reasons?
>
> ------------------------------------------
> 9fans: 9fans
> Permalink:
> https://9fans.topicbox.com/groups/9fans/T861dcd95691a032b-M7705c3f507b9b7dbb8a8800d
> Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
>


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

Ph.: +27 71 471 3694
Cell: +27 83 251 5824

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Alex Musolino]

> What are some of the technical reasons?

To clarify, I'm asking about technical reasons for not using 9front
as opposed to some other Plan 9 flavour; not as opposed to some other
operating system(s) altogether.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[hiro]

> I have no interest in using 9front for [...] non-technical reasons.

as i said, probably some misunderstanding.

> The "community" is atrocious; anyone willing to turn a blind eye
> toward bad behavior due to technical merit is sophomoric if not
> abusive.  ISTR calling this "elitist fuckery" in the past, I still
> stand by that.

this is the kind of stuff that *seems* so self-evident it's rare i
hear it spelled out. i'm happy we finally get to this point, because i
partly agree with the statement in theory.

bad behavior is bad, and passive acceptance of bad behavior in your
own social circles is just as bad.

i just don't see how people would still apply it to 9front. we have
tried quite hard to relativize whatever might have once inspired the
thought of calling 9front development elitist fuckery. and i would
appreciate it if we can fully clear it up for whoever still feels
offended in this way. no community is ever perfect and the difficulty
in real consensus naturally rises with size of the community and the
complexity of the tackled problems.
pointing out how we should change is important, even if it comes from
an outsider. but there always have to be people changing things from
the inside, about oneself and the people nearby oneself. some are
willing to work on whatever real issues there are. in my personal
experience there are always issues. always new things to develop,
learn, fix, also about oneself. please keep up such constructive
criticism so we can all develop together.

> Second, the gatekeeping that was done while the labs was still active
> was tough, but usually fair.  Geoff did a nice job of keeping the
> spirit of plan9 alive and often rejected larger changes that could
> have upset the balance.

i also have no personal complaints. though i have heard complaints
second-hand (as i hinted in my previous mail). and given the
alternative, most people, even the ones having problems getting
through geoff's selection process, would probably agree it was still
better to have geoff and a working bell-labs lab and website around
than whatever state we are in now. even if for some the only positive
outcome is that thousands of html links would at least not be broken
now.

David's work on 9legacy also follows the same
> spirit; it's a patch queue that pulls from a number of different
> sources but strives to keep the original spirit intact.  I struggle to
> see the same mindset held by 9front.

i agree it uses the same contrib technology. the mindset of 9legacy
indeed didn't change much, but mainly because not much at all got
changed in the first place.
i'm not sure this is a spiritual difference or rather a difference of
plain volume.

also, we were not the ones going around and adding syscalls and what have you.
9front tried much harder to preserve backwards compatibility. and that
takes quite some skill considering how many more lines of code have
been touched!

so what direction do you think did the "9front mindset" wander off to?
all i see is more coherency, not a new direction.
btw, in case you didn't notice, many changes in 9front were fixes and
completing already existing features.

probably the only big regret that all 9front users are still suffering
from is not rewriting nupas from scratch. 9legacy has never moved
fully to nupas so it didn't manage to run into this error.

and i hope we can remove python and hg some day soon. that would be
really awesome. just like openssl got removed from python before. that
was one of those great days in 9front history that i will always look
back to with joy :)

> That said, there is work being done outside of 9front.  It's not as
> public or as polarizing, but there are still people hacking away.

sure, some form of IT slavery is how most of us are surviving.
thankfully they pay for most of this non-plan9 work, even though i can
not philosophically identify with that industry as much.

it's not like without plan 9 we couldn't have fun, too. but with plan
9 we can also get peace.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[Nicolas S. Montanaro]

An aside: as someone who has only dipped their toes into Plan 9 over the last couple of years, the quality of discussion on this mailing list is quite refreshing to lurk… and at times quite entertaining. The smaller size of the community and the self-evident passion and skill of the contributors is something that certainly lights a fire under my ass, making me want to spend more time programming and understanding the system.

Merry Christmas and thanks for all the work that has been, and continues to be, done.

Re: [9fans] upgrading the 9legacy shell (was: Gmail vs Upas)

[ori]

>> and nobody is building a
>> release with an "OFFICIAL" stamp on it
> 
> What exactly would you want such a stamp to mean?
> 

Largely 2 things:

	1) That this is the "up to date plan 9" that the
	   community around it believes should be used.

	   This seems to have improved since the last
	   time I looked -- 9legacy.org no longer calls
	   itself 'an experimental patch queue'. Good!

	2) That the update mechanisms documented will
	   bring you up to date and incorporate those
	   changes. It seems that the documented way
	   to update your system is still replica/pull.
	   I haven't checked that it still works.