Re: Moving w3m out of standard

[Soren Hansen]

On Mon, Sep 01, 2008 at 02:02:13PM +0100, Matt Zimmerman wrote:
>>> patch
>>> screen
>> This is in desktop, not standard.  A good candidate for the server
>> seed.
> I've added both of these to the server seed now that it exists.

Excellent, thank you.

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Matt Zimmerman]

On Thu, Jun 26, 2008 at 09:57:07AM +0100, Matt Zimmerman wrote:
> > patch
> 
> This doesn't seem to be in standard, though I'm somewhat surprised that
> nothing in there pulls it in through a dependency.
> 
> I think it's a good candidate for the server seed.
> [...]
> > screen
> 
> This is in desktop, not standard.  A good candidate for the server seed.

I've added both of these to the server seed now that it exists.

-- 
 - mdz

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

On Tue, Jul 1, 2008 at 4:50 PM, Michael Hipp <[EMAIL PROTECTED]> wrote:
>
> Our "ultimate goal" IMHO ought to be to have a product that works well for the
> full spectrum of users. This includes those that want a "just enough" install
> as well as those that want a "gimme much" install and everyone in between.
>
> I really don't see what's so controversial about having a standard installer
> that is fairly lean, with a "really bare" option  for those that want it as
> well as "more, more" option(s) for the others.
>
> But if you're unwilling to provide all three of those capabilities (bare, lean
> standard, more-more), then you're saying there's one or more classes of users
> that you don't want to serve. I doubt that's the case. But why fight against
> any of the above three options when there are very valid use cases and lots of
> customers for each?
>
> Michael
>

I tend to think that other sysadmins who are in a similar position as
me, are probably going to see my point: 10 years or so adminning
servers, being in a medium/large server environment (10+/20+ servers),
and being the actual admin (not a lower-level tech or hands-off
department manager)

But I realize not everyone is in that position, and I'm not so selfish
to think Ubuntu has to cater solely to me.  Other people may be
hobbyist setting up a home server for fun or education, or a small
business user with only a single server or two, or in a much larger
environment where you have whole teams of server admins dedicated to a
single service (email, database, etc).  The people in these positions
are going to have very different needs and have very different ideas
on what is "best" or "most useful".

So anyway, my point is that Michael really put it nicely.  I'm just
trying to be a helpful representative of my "class of user" :D as
Michael puts it.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Michael Hipp]

Soren Hansen wrote:
> If "don't be completely useless" is truly all we're aiming for, our
> ultimate goal, I'm in the wrong job. Seriously. This might sound
> pretentious, but I have higher ambitions than that.

Our "ultimate goal" IMHO ought to be to have a product that works well for the 
full spectrum of users. This includes those that want a "just enough" install 
as well as those that want a "gimme much" install and everyone in between.

I really don't see what's so controversial about having a standard installer 
that is fairly lean, with a "really bare" option  for those that want it as 
well as "more, more" option(s) for the others.

But if you're unwilling to provide all three of those capabilities (bare, lean 
standard, more-more), then you're saying there's one or more classes of users 
that you don't want to serve. I doubt that's the case. But why fight against 
any of the above three options when there are very valid use cases and lots of 
customers for each?

Michael


-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Tue, Jul 01, 2008 at 12:40:27PM -0500, James Dinkel wrote:
>> Your reply was something like "I'm not very determined", which
>> doesn't answer my question at all. You want us to remove groff from
>> the standard install, even though you're going to install it
>> yourself.  That sure sounds to me like you find groff (probably
>> actually man-db) to be useful (otherwise, why would you be installing
>> it), but you want to not have it installed by default, thus making
>> the default install less useful.
> It's not that I prefer man-db to be removed by default, because yes I
> do use it.  That doesn't mean everyone uses it though.

I just still fail to see why not using man-db is a more valid use case
than using it.

> because while I think w3m is a complete waste of space, I'm fine with
> just letting it go and moving on to more important issues.

I also still fail to see why w3m is any different than groff+man. 

>>> There seems to be a lot of praise for the "bare install" option in
>>> the installer, and no one has said anything against it.
>> Well, you seem to think that it's not bare enough?
> I think that Ubuntu JEOS is bare enough.

You see, JeOS is an acronym for "Just enough OS". "Just enough OS" in
itself doesn't mean anything, so to make it make any sense at all in my
head, I think I've always understood it as: "Just enough OS to not be
completely useless". 

If "don't be completely useless" is truly all we're aiming for, our
ultimate goal, I'm in the wrong job. Seriously. This might sound
pretentious, but I have higher ambitions than that.

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

Sorry for the very late reply.  I came down with appendicitis, and
then took a trip to New York... anyway...

On Thu, Jun 26, 2008 at 4:52 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
>
> I asked why you were do determined to make the install less useful by
> default.

Well, I do see the value of having some "newbie-friendly" things in
there by default.  This is why I would much prefer to have a "bare
install" option available for old-school admins, rather than be the
default (have a bare-install be opt-in, rather than opt-out).  I'm
just putting my 2 cents in on how _I_ prefer a server to be set up.

> Your reply was something like "I'm not very determined", which
> doesn't answer my question at all. You want us to remove groff from the
> standard install, even though you're going to install it yourself.  That
> sure sounds to me like you find groff (probably actually man-db) to be
> useful (otherwise, why would you be installing it), but you want to not
> have it installed by default, thus making the default install less
> useful.
>

It's not that I prefer man-db to be removed by default, because yes I
do use it.  That doesn't mean everyone uses it though.  I could easily
do without it on my server and save the space by looking up all the
man pages on a workstation, and some other admins may prefer it that
way.  And if it was removed, I wouldn't argue against it because it is
easy enough for me to install it. I say "I'm not very determined"
because while I think w3m is a complete waste of space, I'm fine with
just letting it go and moving on to more important issues.

> I see the usefulness of the minimal install, but this I think this is
> getting ridiculous.  Ubuntu Server is -- after all -- supposed to be for
> human beings, not for embedded systems.
>
>> There seems to be a lot of praise for the "bare install" option in the
>> installer, and no one has said anything against it.
>
> Well, you seem to think that it's not bare enough?
>

I think that Ubuntu JEOS is bare enough.  I'm pretty sure that is what
everyone agrees would be the equivalent of a "bare install" option in
the installer (except without any hardware support stripped out,
keeping the regular -server kernel).  Ubuntu JEOS does not install
man-db, by the way.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Andreas Hasenack]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Ante Karamatic wrote:
> On Thu, 26 Jun 2008 11:44:19 -0400
> Scott Kitterman <[EMAIL PROTECTED]> wrote:
> 
>> Ooh.  Very shiny.  I didn't know about that one.
> 
> Bad thing about it:
> 
> -rwsr-xr-x 1 root root 46084 2008-03-31 06:32 /usr/bin/mtr
> 
> setuid. :(

It shouldn't be hard to create an apparmor profile for it.


- --
Andreas Hasenack
[EMAIL PROTECTED]

-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFIY8F7eEJZs/PdwpARAkpaAJwJVyXPGwcfQXeQ1BB6hiwXKA4cHwCgjJTe
O2i2vPTaV3nZvlbwxqaopBc=
=xd5g
-END PGP SIGNATURE-

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Ante Karamatic]

On Thu, 26 Jun 2008 11:44:19 -0400
Scott Kitterman <[EMAIL PROTECTED]> wrote:

> Ooh.  Very shiny.  I didn't know about that one.

Bad thing about it:

-rwsr-xr-x 1 root root 46084 2008-03-31 06:32 /usr/bin/mtr

setuid. :(

But traceroute isn't:

-rwxr-xr-x 1 root root 41644 2007-11-19 17:23 /usr/bin/traceroute.db

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Scott Kitterman]

On Thursday 26 June 2008 11:03, Ante Karamatic wrote:
> On Thu, 26 Jun 2008 08:27:40 -0400
>
> Scott Kitterman <[EMAIL PROTECTED]> wrote:
> > It might make sense to push tracepath to the desktop seed and add
> > traceroute to the server seed so that traceroute is what we provide
> > for servers.  For me it's certainly one of the first things I install
> > on a server after I set it up.
>
> IIRC, mtr is installed by default and this tool should be enough (I
> know I was the one who started traceroute discussion :).

Ooh.  Very shiny.  I didn't know about that one.

Thanks,

Scott K

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Matt Zimmerman]

On Wed, Jun 25, 2008 at 07:55:28AM -0400, Rich Kulawiec wrote:
> As an die-hard old-school (err, make that, "ancient school") 'nix admin
> I applaud the idea of trying to keep installs minimal, both for space
> and security reasons.
> 
> On the other hand, I recognize the value of having some basic utilities
> handy to administer services, test them, debug them, and fix them.
> Like many other people, I've got a favorite set that I often find
> useful.  I try to keep that set restricted to those that I can run
> on a low-bandwidth ssh connection, because that's often exactly
> what I have to do, and I try to avoid overlapping tools (although
> arguably two of my choices, curl and wget, at least partially do so).
> 
> My suggestion is that discussion take place over (a) whether it makes
> sense or not to bundle a collection of such tools for easy installation,
> and (b) what tools might be good candidates for inclusion.  Here are
> the ones that I've used across a variety of 'nix systems; I don't
> claim that this list is the best or most inclusive or anything, it
> just happens to be the toolset that I've found lets me deal with
> most of the things I've had to deal with.
> 
>   curl

We ship wget in standard, which is close enough in functionality that I
don't think two tools make sense.

>   dig (part of the BIND distribution)

We ship both dig (dnsutils) and host (bind9-host) in standard.

>   lslk

This sounds potentially useful, but is in universe and abandoned upstream.
Doesn't lsof provide this functionality?

>   lsof
>   mutt

Both in standard.

>   nmap

This is in main, but we don't install it by default.  A scanner probably
doesn't belong in every installation.

>   patch

This doesn't seem to be in standard, though I'm somewhat surprised that
nothing in there pulls it in through a dependency.

I think it's a good candidate for the server seed.

>   rsync

standard

>   screen

This is in desktop, not standard.  A good candidate for the server seed.

>   surfraw (I *said* I was ancient-school)

No comment. :-)

>   tcpdump
>   top

Both standard.

>   traceroute

We ship tracepath by default, which isn't setuid.  traceroute is in main,
though, since so many folks are more familiar with it.

>   w3m

Currently in standard, and the subject of this thread.

>   wget

Standard.

-- 
 - mdz

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Ante Karamatic]

On Thu, 26 Jun 2008 08:27:40 -0400
Scott Kitterman <[EMAIL PROTECTED]> wrote:

> It might make sense to push tracepath to the desktop seed and add
> traceroute to the server seed so that traceroute is what we provide
> for servers.  For me it's certainly one of the first things I install
> on a server after I set it up.

IIRC, mtr is installed by default and this tool should be enough (I
know I was the one who started traceroute discussion :).

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Scott Kitterman]

On Thursday 26 June 2008 04:57, Matt Zimmerman wrote:
> On Wed, Jun 25, 2008 at 07:55:28AM -0400, Rich Kulawiec wrote:
...
> > traceroute
>
> We ship tracepath by default, which isn't setuid.  traceroute is in main,
> though, since so many folks are more familiar with it.

I'd like to suggest maybe Desktop and Server should go in different directions 
once we have the server seed.  Tracepath also uses (abuses depending on how 
pedantic one wants to be) UDP instead of ICMP and so there are some real 
functional differences (we discussed this at UDS in one of the server 
sessions).

It might make sense to push tracepath to the desktop seed and add traceroute 
to the server seed so that traceroute is what we provide for servers.  For me 
it's certainly one of the first things I install on a server after I set it 
up.

Scott K

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Mon, Jun 23, 2008 at 11:50:30AM -0500, James Dinkel wrote:
>> ..yet you're ignoring my request for a rationale.
> I have done nothing of the sort.  

I asked why you were do determined to make the install less useful by
default. Your reply was something like "I'm not very determined", which
doesn't answer my question at all. You want us to remove groff from the
standard install, even though you're going to install it yourself.  That
sure sounds to me like you find groff (probably actually man-db) to be
useful (otherwise, why would you be installing it), but you want to not
have it installed by default, thus making the default install less
useful. 

I see the usefulness of the minimal install, but this I think this is
getting ridiculous.  Ubuntu Server is -- after all -- supposed to be for
human beings, not for embedded systems.

> There seems to be a lot of praise for the "bare install" option in the
> installer, and no one has said anything against it.

Well, you seem to think that it's not bare enough?

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Rich Kulawiec]

As an die-hard old-school (err, make that, "ancient school") 'nix admin
I applaud the idea of trying to keep installs minimal, both for space
and security reasons.

On the other hand, I recognize the value of having some basic utilities
handy to administer services, test them, debug them, and fix them.
Like many other people, I've got a favorite set that I often find
useful.  I try to keep that set restricted to those that I can run
on a low-bandwidth ssh connection, because that's often exactly
what I have to do, and I try to avoid overlapping tools (although
arguably two of my choices, curl and wget, at least partially do so).

My suggestion is that discussion take place over (a) whether it makes
sense or not to bundle a collection of such tools for easy installation,
and (b) what tools might be good candidates for inclusion.  Here are
the ones that I've used across a variety of 'nix systems; I don't
claim that this list is the best or most inclusive or anything, it
just happens to be the toolset that I've found lets me deal with
most of the things I've had to deal with.

curl
dig (part of the BIND distribution)
lslk
lsof
mutt
nmap
patch
rsync
screen
surfraw (I *said* I was ancient-school)
tcpdump
top
traceroute
w3m
wget

---Rsk

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Brian McKee]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

I didn't expect this thread to go on like this  But it's good to see
people passionate about the details.  My 2c is basically 'what do I need
 at 3am when the darned thing isn't working and it's got no connection
to the rest of the world.'

That doesn't mean I want every tool under the sun and four extra
kernels but I need a basic set of tools to get it to boot and make a
network connection again.The rest I can explore at my leisure during
working hours with all the choices of the repositories at hand.

I suppose everybody's favourite tools set might be a bit different,  but
that's *my* take on it.

Brian

PS If you remove the man pages, then I hope the gods of unix curse you
to spend the rest of your days trying to convert the bash man page to
the 'info' system, via postscript and latex, by hand, using ed  and
I'm switching to Lindows  :-)
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFIYaDPGnOmb9xIQHQRAgJoAKClxGZzHvzOCoBSEQ1Wk9wijwf69QCgq9c7
WFTjqEPW6dOC3oUQ+Mh/Bew=
=xswK
-END PGP SIGNATURE-

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Server Flavors [Was: Re: Platform vs Product [Was: Re: Moving w3m out of standard]]

[Scott Kitterman]

On Monday 23 June 2008 23:03, Luke L wrote:
> I did it again. Scott, sorry for the reply.

No problem.  You might consider a mail client that has a 'reply to list' 
feature.  I can recommen Kmail/Kontact.

> >> 3) This idea would benefit us greatest if the program (if the
> >> 'program' route were chosen) doing this were somehow cross-platform.
> >> This is not a requirement, but it would be handy.
> >
> > Package install and configuration is so unique that this will be unique
> > to Debian and derived distros.  There is no general solution to such a
> > problem and I'm pretty convinced one isn't feasible.
>
> When I said cross-platform, I meant the program that generates the
> custom ISO, though it seems you have more of the "FAI" and "configure
> IN Ubuntu" idea. In other words, a Windows user could download a
> program that takes the desired specs and creates an ISO to their
> liking so they can install it. However, I see your point, this should
> be more about automated, custom, easy to deploy UBUNTU centric
> software.

The answering the questions part of the process should be cross-platform.  My 
current intent is to steal <-<-<-<- leverage kirkland's VM configuration 
script to get me started so it'll be a javascript thing that should run in 
any browser.  That's the theory anyway.  No actual code has been harmed yet.

Scott K

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Server Flavors [Was: Re: Platform vs Product [Was: Re: Moving w3m out of standard]]

[Luke L]

I did it again. Scott, sorry for the reply.

>
>> 3) This idea would benefit us greatest if the program (if the
>> 'program' route were chosen) doing this were somehow cross-platform.
>> This is not a requirement, but it would be handy.
>
> Package install and configuration is so unique that this will be unique to
> Debian and derived distros.  There is no general solution to such a problem
> and I'm pretty convinced one isn't feasible.
>

When I said cross-platform, I meant the program that generates the
custom ISO, though it seems you have more of the "FAI" and "configure
IN Ubuntu" idea. In other words, a Windows user could download a
program that takes the desired specs and creates an ISO to their
liking so they can install it. However, I see your point, this should
be more about automated, custom, easy to deploy UBUNTU centric
software.

-- 
Luke L.

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Server Flavors [Was: Re: Platform vs Product [Was: Re: Moving w3m out of standard]]

[Scott Kitterman]

On Monday 23 June 2008 20:31, Luke L wrote:
> This sounds like an absolutely terrific idea. The next step (even
> before design) is analysis. What flavors will be offered? What level
> of customization are we talking about?
>
> 1) No matter how implemented, the "form" or install menu would have
> questions on:
>
> VM install / physical machine?
> Minimal tools / Full tools? (Tools like w3m, full vim, etc.)
> Tasks in tasksel (Servers to be installed)

Actually I expect this to go well beyond what is in tasksel.  One of the 
problems with tasksel is that it isn't very scalable.  

> 2) Will this be an install-time set of options that could optionally
> be automated, or will this be in the form of a program that spins a
> custom install ISO?

It could be either.  A lot of this is still TBD depending on the resources 
available to develop it.  The spec is not yet approved.

> While making it an install-time choice might make a tight fit on a CD,
> it would be more flexible to use, IMO.

More flexible for the small shops, but at some point I'd like to integrate 
this into some kind of PXE boot system so that large deployments can easily 
be automated.

> 3) This idea would benefit us greatest if the program (if the
> 'program' route were chosen) doing this were somehow cross-platform.
> This is not a requirement, but it would be handy.

Package install and configuration is so unique that this will be unique to 
Debian and derived distros.  There is no general solution to such a problem 
and I'm pretty convinced one isn't feasible.

> Notes) This customization could go to the extreme (Selection of
> individual packages) if a DVD or multi-CD release were available, like
> FreeBSD. I don't see that being entirely necessary, however. If this
> were made as a "Custom ISO Maker", allowing users to spin install DVDs
> with more info would be a good option.

This goes beyone just selection of packages.  That is easy to do with pre-seed 
and the Debian Installer.  The integration with FAI can provide us with a 
defined set of packages and an integrated configuration.  That's when the big 
win is in my opinion.

> I hope my ideas are of some inspiration, and that I could possibly
> help in the conception of this project.

Thanks.  Assuming this gets approved I'll pick on you for help.

Scott K

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Server Flavors [Was: Re: Platform vs Product [Was: Re: Moving w3m out of standard]]

[Scott Kitterman]

On Mon, 23 Jun 2008 19:10:27 -0500 "Dustin Kirkland" 
<[EMAIL PROTECTED]> wrote:
>On Mon, Jun 23, 2008 at 6:55 PM, Scott Kitterman <[EMAIL PROTECTED]> 
wrote:
>> 3.  Provide a way to scalably provide a lot of different configurations 
for
>> many specific use cases that doesn't require much additional 
configuration
>> and sysadmin time.  I've proposed a spec for one approach to move in this
>> direction:
>>
>> https://blueprints.launchpad.net/ubuntu/+spec/server-flavors
>
>I really like the idea of "canned" server flavors!
>
>Unique CD's per install (as stated in the blueprint), I think, might
>get unwieldy, with perhaps dozens of new server CD's per architecture.
> This is where the hypothetical hosted, vm image builder discussed at
>UDS might be nice.  Fill in a few form values, get a coffee, come back
>and download your new server image.  Perhaps more scalable than unique
>iso for every combination dreamable.
>
Although the spec is written in terms of a new install/ISO, the FAI 
configuration step could also be done on top of an existing base install.  
I should make the explicit in the spec.

Scott K

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Server Flavors [Was: Re: Platform vs Product [Was: Re: Moving w3m out of standard]]

[Luke L]

This sounds like an absolutely terrific idea. The next step (even
before design) is analysis. What flavors will be offered? What level
of customization are we talking about?

1) No matter how implemented, the "form" or install menu would have
questions on:

VM install / physical machine?
Minimal tools / Full tools? (Tools like w3m, full vim, etc.)
Tasks in tasksel (Servers to be installed)

2) Will this be an install-time set of options that could optionally
be automated, or will this be in the form of a program that spins a
custom install ISO?
While making it an install-time choice might make a tight fit on a CD,
it would be more flexible to use, IMO.

3) This idea would benefit us greatest if the program (if the
'program' route were chosen) doing this were somehow cross-platform.
This is not a requirement, but it would be handy.

Notes) This customization could go to the extreme (Selection of
individual packages) if a DVD or multi-CD release were available, like
FreeBSD. I don't see that being entirely necessary, however. If this
were made as a "Custom ISO Maker", allowing users to spin install DVDs
with more info would be a good option.

I hope my ideas are of some inspiration, and that I could possibly
help in the conception of this project.
-- 
Luke L.

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Server Flavors [Was: Re: Platform vs Product [Was: Re: Moving w3m out of standard]]

[Dustin Kirkland]

On Mon, Jun 23, 2008 at 6:55 PM, Scott Kitterman <[EMAIL PROTECTED]> wrote:
> 3.  Provide a way to scalably provide a lot of different configurations for
> many specific use cases that doesn't require much additional configuration
> and sysadmin time.  I've proposed a spec for one approach to move in this
> direction:
>
> https://blueprints.launchpad.net/ubuntu/+spec/server-flavors

I really like the idea of "canned" server flavors!

Unique CD's per install (as stated in the blueprint), I think, might
get unwieldy, with perhaps dozens of new server CD's per architecture.
 This is where the hypothetical hosted, vm image builder discussed at
UDS might be nice.  Fill in a few form values, get a coffee, come back
and download your new server image.  Perhaps more scalable than unique
iso for every combination dreamable.

:-Dustin

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Platform vs Product [Was: Re: Moving w3m out of standard]

[Scott Kitterman]

On Monday 23 June 2008 17:31, Dustin Kirkland wrote:
> On Mon, Jun 23, 2008 at 10:03 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
> > Alternate and the mini ISO will get you the same kernel.
>
> It would be nice if these two had "server" as a possible installation
> target (or "server-minimal" as the case may be), which installed the
> server kernel instead.  As well as an option on the Server install CD
> for a minimal installation (not default, but a target specifiable at
> boot time).
>
> --
>
> Beyond that, I like the way this conversation has turned...a
> productive discussion about the difference between an "Ubuntu Server
> Product", which is true to the spirit of "Linux for Human Beings (who
> happen to also be sysadmins)", as well as an "Ubuntu Server Minimal"
> installation, without the bells and whistles for the uber Ubuntu
> sysadmins (aka old school, die hard).
>
> :-Dustin

The major challenge is that as a product, there is no single Ubuntu Server.  
There are as many servers as there are use cases.  We can deal with this any 
of at least three ways:

1.  Provide the minimal system and tell people to make their own product.  
This will appeal to a certain market segment and is not hard to do.  I think 
as a community there is definite interest in this, but I suspect not a lot of 
Canonical support revenue.  I do think we ought to do this as it can serve 
(pun intended) as the basis for a lot of specific projects.

2.  Provide a mostly right for a number of common use cases, but still 
probabyl not exactly what you want, you'll have to tweak it.  This is, I 
would argue, what we provide today.  It's useful, but also needs some 
extentsion to use beyond a couple of very specific use cases.

3.  Provide a way to scalably provide a lot of different configurations for 
many specific use cases that doesn't require much additional configuration 
and sysadmin time.  I've proposed a spec for one approach to move in this 
direction:

https://blueprints.launchpad.net/ubuntu/+spec/server-flavors

To get to a truly Ubuntu solution in servers, we are going to need something a 
lot more configurable than what we have now.

Scott K

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Platform vs Product [Was: Re: Moving w3m out of standard]

[James Dinkel]

On Mon, Jun 23, 2008 at 4:31 PM, Dustin Kirkland <[EMAIL PROTECTED]> wrote:
>
> Beyond that, I like the way this conversation has turned...a
> productive discussion about the difference between an "Ubuntu Server
> Product", which is true to the spirit of "Linux for Human Beings (who
> happen to also be sysadmins)", as well as an "Ubuntu Server Minimal"
> installation, without the bells and whistles for the uber Ubuntu
> sysadmins (aka old school, die hard).
>
> :-Dustin
>

ooh, and we can call it Uberbuntu.

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Platform vs Product [Was: Re: Moving w3m out of standard]

[Dustin Kirkland]

On Mon, Jun 23, 2008 at 10:03 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
> Alternate and the mini ISO will get you the same kernel.

It would be nice if these two had "server" as a possible installation
target (or "server-minimal" as the case may be), which installed the
server kernel instead.  As well as an option on the Server install CD
for a minimal installation (not default, but a target specifiable at
boot time).

--

Beyond that, I like the way this conversation has turned...a
productive discussion about the difference between an "Ubuntu Server
Product", which is true to the spirit of "Linux for Human Beings (who
happen to also be sysadmins)", as well as an "Ubuntu Server Minimal"
installation, without the bells and whistles for the uber Ubuntu
sysadmins (aka old school, die hard).

:-Dustin

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Luke L]

Sorry, James, damned reply button.

As usual, people look at principle over pragmatism in such discussions.

While a bare-install/full-install option certainly sounds interesting,
Could we look at what we're arguing over? 3MB installed, including
dependencies. It is in the installation already, plenty of people seem
to want it to remain that way, so it seems the only option worth
discussing is whether to have install-time options for "bare/normal"
packages.

Otherwise, it seems it should stay as it is.

On 6/23/08, James Dinkel <[EMAIL PROTECTED]> wrote:
> On Mon, Jun 23, 2008 at 11:06 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
> >
> > ..yet you're ignoring my request for a rationale.
> >
> > --
> > Soren Hansen
>
> I have done nothing of the sort.  If I was unclear on something, maybe
> we could hash it out off-list.  In the end, however, I think we are
> both (and a few others) hard set in our opinions.  There seems to be a
> lot of praise for the "bare install" option in the installer, and no
> one has said anything against it.  Since that will make this all a
> moot point, maybe we should move on and focus on that.
>
> James
>
> --
> ubuntu-server mailing list
> ubuntu-server@lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
> More info: https://wiki.ubuntu.com/ServerTeam
>


-- 
Luke L.

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

On Mon, Jun 23, 2008 at 11:06 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
>
> ..yet you're ignoring my request for a rationale.
>
> --
> Soren Hansen

I have done nothing of the sort.  If I was unclear on something, maybe
we could hash it out off-list.  In the end, however, I think we are
both (and a few others) hard set in our opinions.  There seems to be a
lot of praise for the "bare install" option in the installer, and no
one has said anything against it.  Since that will make this all a
moot point, maybe we should move on and focus on that.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Mon, Jun 23, 2008 at 10:25:20AM -0500, James Dinkel wrote:
>> I'm completely missing why you are so determined to make the server
>> install less useful by default?
> I didn't think I was all that determined.  You asked a question, so I
> gave an answer.  I'm not going to lie about my opinion, and I didn't
> want to ignore you.

..yet you're ignoring my request for a rationale.

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

On Mon, Jun 23, 2008 at 9:23 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
>
> I'm completely missing why you are so determined to make the server
> install less useful by default?
>
> --
> Soren Hansen   |
>

I didn't think I was all that determined.  You asked a question, so I
gave an answer.  I'm not going to lie about my opinion, and I didn't
want to ignore you.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Platform vs Product [Was: Re: Moving w3m out of standard]

[Soren Hansen]

On Mon, Jun 23, 2008 at 09:50:09AM -0400, Jamie Strandboge wrote:
> It's clear to me what is really being discussed is should Ubuntu
> Server be a platform or a product akin to Ubuntu Desktop? Both are
> valid and useful, but *distinct*.

Very good point.

> To quote Soren: Ubuntu is "about making everything more useful by
> default". 

Not surprising, I agree with this :)  I think I should point out that I
don't think just installing more packages is going to make anything more
useful. If you know the name of a command, with command-not-found,
you're even told which command to run to install the package, you're
missing, so this is really the least interesting part of the discussion.

It's true that 99.9% of the changes we can make, we do by making changes
to packages. (The remaing 0.1% would be in CD building process and
such). This makes it easy to get caught up in stuff like how we can make
tiny changes to each package to make it just a few percent more useful.
This is all good stuff, and these changes can get us a long way, but
things don't start to get really interesting until we take a step back
and consider in a more broad perspective what we want to achieve, and
how we can use the tools we have to do it.

For Hardy we tried to get postfix and dovecot to integrate with each
other by default. For various reasons this never came to be, but it's a
rather good example of the kind of stuff we can do because we control
are the suppliers of both pieces of software.

My ssh-login-to-screen idea a few e-mails ago is another example. It
might be silly, but it's an example of thinking outside the confines of
single packages, and considering the software stack as more a whole.

> I absolutely agree with this, and through discussion, things like
> screen and other software, where appropriate and after thorough
> review, should be part of Ubuntu Server's default installation. This
> functionality and integration work is what gets at the heart of what
> Ubuntu is-- "Linux for Human Beings".

I wholeheartedly agree. To a great extent, Ubuntu Server is like Debian,
but with some patches, a release schedule, support options, and
dedicated ressources. This is good, but it could be much, much better.
There's a *lot* of potential for improvement.

> On the other hand, there is a clear need to have an install method for
> server that is minimal, or platform oriented.

Quite true. In our integration efforts we need to also keep in mind that
folks might want to use a different combination of packages, so the
coupling should be rather loose.

> Right now, the mini ISO, the alternate CD, JeOS and the current server
> CD all approach this from different angles, but come close to the same
> end result (though, IIRC, you end up with a different kernel in
> each).

Alternate and the mini ISO will get you the same kernel. JeOS uses the
virtual kernel, and the server CD gets you the server kernel.
Additionally JeOS doesn't install standard, AFAIR, but only minimal, but
yes, their end result is rather similar.

> I suggest we have on the server CD a boot option like 'linux minimal'
> (this has been suggested/implied elsewhere in this thread).  This
> might simply be ubuntu-standard with the server kernel and a note that
> mentions running tasksel for additional packages (but this will need
> to be discussed).

Agreed.

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Platform vs Product [Was: Re: Moving w3m out of standard]

[James Dinkel]

On Mon, Jun 23, 2008 at 8:50 AM, Jamie Strandboge <[EMAIL PROTECTED]> wrote:
> I suggest we have on the
> server CD a boot option like 'linux minimal' (this has been
> suggested/implied elsewhere in this thread). This might simply be
> ubuntu-standard with the server kernel and a note that mentions running
> tasksel for additional packages (but this will need to be discussed).
>
> Using this methodology, we allow for the functionality and integration
> work that defines Ubuntu, but also provide a server platform that is
> useful for those with specialized needs.
>
> Jamie
>
>

ooh, I also like the idea of putting this as a boot option rather than
within the install.  This would keep an "unnecessary" question out of
the installer, leaving less for a newbie to think about (I think this
is very important), while also allowing us "die-hard" sysadmins an
easy way to exclude anything but the bare platform.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Platform vs Product [Was: Re: Moving w3m out of standard]

[Brett Alton]

On Mon, Jun 23, 2008 at 9:50 AM, Jamie Strandboge <[EMAIL PROTECTED]> wrote:
> On Fri, 20 Jun 2008, Michael Hipp wrote:
>
>> I think that's a *good* thing because I consider a server to be a *platform*
>> upon which I add the things (applications) that make it do what I need.
>>
> This is going to echo some of Soren's lastest post, but think it bears
> repeating:
>
> It's clear to me what is really being discussed is should Ubuntu Server
> be a platform or a product akin to Ubuntu Desktop? Both are valid and
> useful, but *distinct*.
>
> To quote Soren: Ubuntu is "about making everything more useful by
> default". I absolutely agree with this, and through discussion, things
> like screen and other software, where appropriate and after thorough
> review, should be part of Ubuntu Server's default installation. This
> functionality and integration work is what gets at the heart of what
> Ubuntu is-- "Linux for Human Beings".
>
> On the other hand, there is a clear need to have an install method for
> server that is minimal, or platform oriented. Right now, the mini ISO,
> the alternate CD, JeOS and the current server CD all approach this from
> different angles, but come close to the same end result (though, IIRC,
> you end up with a different kernel in each). I suggest we have on the
> server CD a boot option like 'linux minimal' (this has been
> suggested/implied elsewhere in this thread). This might simply be
> ubuntu-standard with the server kernel and a note that mentions running
> tasksel for additional packages (but this will need to be discussed).
>
> Using this methodology, we allow for the functionality and integration
> work that defines Ubuntu, but also provide a server platform that is
> useful for those with specialized needs.
>
> Jamie
>
> --
> Ubuntu Security Engineer | http://www.ubuntu.com/
> Canonical Ltd.   | http://www.canonical.com/
>
> -BEGIN PGP SIGNATURE-
> Version: GnuPG v1.4.6 (GNU/Linux)
>
> iD8DBQFIX6oRW0JvuRdL8BoRAjCuAJ9Ovr56BM6ByzdsgRXdXiagpzcM7ACeM4RY
> RoNnGPBE0TrHDGtpjBfFeOw=
> =C+qI
> -END PGP SIGNATURE-
>
> --
> ubuntu-server mailing list
> ubuntu-server@lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
> More info: https://wiki.ubuntu.com/ServerTeam
>

I would like Ubuntu to install a minimal server with the option of
using tasksel. I think that's a great idea! It'll be pretty much JeOS,
correct?

But no matter what, we'll still need a full CD for users who don't
have ethernet connections. Not that they have dial-up but some have to
setup a server with no ethernet connected and need a full server on
CD.

Brett Alton
[EMAIL PROTECTED]

Do you really need to print this email? Help preserve our environment!

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Mon, Jun 23, 2008 at 09:17:47AM -0500, James Dinkel wrote:
>> See, you happen to need to access documentation that's stored in man
>> page format. Others are in a situation where the packages they use
>> store their documentation in html format. Without an html interpreter
>> they're screwed. In this context, groff and w3m are exactly the same,
>> so if we remove one, why not the other?
> I already admitted to pandering to myself.  And if groff is gone, then
> I wouldn't be heart-broken.  I would just include it in my standard
> server set up, or install the package who's man page I need to read on
> one of my handy-dandy test servers.

I'm completely missing why you are so determined to make the server
install less useful by default?

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Platform vs Product [Was: Re: Moving w3m out of standard]

[Jamie Strandboge]

On Fri, 20 Jun 2008, Michael Hipp wrote:

> I think that's a *good* thing because I consider a server to be a *platform* 
> upon which I add the things (applications) that make it do what I need.
> 
This is going to echo some of Soren's lastest post, but think it bears
repeating:

It's clear to me what is really being discussed is should Ubuntu Server
be a platform or a product akin to Ubuntu Desktop? Both are valid and
useful, but *distinct*.

To quote Soren: Ubuntu is "about making everything more useful by
default". I absolutely agree with this, and through discussion, things
like screen and other software, where appropriate and after thorough
review, should be part of Ubuntu Server's default installation. This
functionality and integration work is what gets at the heart of what
Ubuntu is-- "Linux for Human Beings".

On the other hand, there is a clear need to have an install method for
server that is minimal, or platform oriented. Right now, the mini ISO,
the alternate CD, JeOS and the current server CD all approach this from
different angles, but come close to the same end result (though, IIRC,
you end up with a different kernel in each). I suggest we have on the
server CD a boot option like 'linux minimal' (this has been
suggested/implied elsewhere in this thread). This might simply be
ubuntu-standard with the server kernel and a note that mentions running
tasksel for additional packages (but this will need to be discussed).

Using this methodology, we allow for the functionality and integration
work that defines Ubuntu, but also provide a server platform that is
useful for those with specialized needs.

Jamie

-- 
Ubuntu Security Engineer | http://www.ubuntu.com/
Canonical Ltd.   | http://www.canonical.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

On Mon, Jun 23, 2008 at 8:57 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
> On Mon, Jun 23, 2008 at 08:43:28AM -0500, James Dinkel wrote:
>
>> 2) the whole idea that "it's a browser", therefore "it's a huge
>> security hole."
>
> Sorry, I'm having trouble detecting sarcasm over e-mail. Is this a joke
> or a serious comment?
>

Not so much sarcastic, nor serious, as it is a realization of the
subconscious affect of being bombarded with constant IE and Firefox
security warnings.

>
> See, you happen to need to access documentation that's stored in man
> page format. Others are in a situation where the packages they use store
> their documentation in html format. Without an html interpreter they're
> screwed. In this context, groff and w3m are exactly the same, so if we
> remove one, why not the other?
>
>

I already admitted to pandering to myself.  And if groff is gone, then
I wouldn't be heart-broken.  I would just include it in my standard
server set up, or install the package who's man page I need to read on
one of my handy-dandy test servers.

James

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Mon, Jun 23, 2008 at 08:43:28AM -0500, James Dinkel wrote:
>> Do you think we should remove groff (and with it, man-db), too? I'm
>> not trying to be difficult.. This is an actual question.
> No, but I use man all the time.

I forgot to address this bit in my last e-mail..

See, you happen to need to access documentation that's stored in man
page format. Others are in a situation where the packages they use store
their documentation in html format. Without an html interpreter they're
screwed. In this context, groff and w3m are exactly the same, so if we
remove one, why not the other?

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Brett Alton]

On Mon, Jun 23, 2008 at 9:43 AM, James Dinkel <[EMAIL PROTECTED]> wrote:
> On Mon, Jun 23, 2008 at 4:59 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
>>
>> Do you think we should remove groff (and with it, man-db), too? I'm not
>> trying to be difficult.. This is an actual question.
>>
>> --
>> Soren Hansen   |
>>
>
> No, but I use man all the time.  Granted, that may be pandering to
> myself, but man just seems infinitely more useful than a web browser.
> This could be that 1) I have never ever used a web browser, or been in
> a position where it would seem useful, on a server and 2) the whole
> idea that "it's a browser", therefore "it's a huge security hole."
>
> And you may say it's just for looking at local html files, but it
> opens up the thinking that "if they put a web browser on a server,
> then it must be ok to browse the web from a server."  I've seen plenty
> of Windows admins casually browsing the web from a server console
> while waiting on some task to finish in the server room.
>
> But that's just my opinion.
>
> James
>
> --
> ubuntu-server mailing list
> ubuntu-server@lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
> More info: https://wiki.ubuntu.com/ServerTeam
>

You need a web browser to access http://localhost:631 for CUPS, don't you?

Brett Alton
[EMAIL PROTECTED]

Do you really need to print this email? Help preserve our environment!

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Mon, Jun 23, 2008 at 08:43:28AM -0500, James Dinkel wrote:
>> Do you think we should remove groff (and with it, man-db), too? I'm
>> not trying to be difficult.. This is an actual question.
> No, but I use man all the time.  Granted, that may be pandering to
> myself, but man just seems infinitely more useful than a web browser.
> This could be that 1) I have never ever used a web browser, or been in
> a position where it would seem useful, on a server and 

Think it helps immensely if you stop calling it a web browser, and start
calling it an html viewer (which happens to work as a web browser as
well, but that's beside the point).

> 2) the whole idea that "it's a browser", therefore "it's a huge
> security hole."

Sorry, I'm having trouble detecting sarcasm over e-mail. Is this a joke
or a serious comment?

> And you may say it's just for looking at local html files, but it
> opens up the thinking that "if they put a web browser on a server,
> then it must be ok to browse the web from a server."

If people want to use it for that, I'm not going to stop them. I might
point fingers and call them silly, but that's their right.

> I've seen plenty of Windows admins casually browsing the web from a
> server console while waiting on some task to finish in the server
> room.

The difference is that they're using a full featured web browser. w3m
doesn't even do javascript. It's not any more or less of a security hole
than as wget + groff combo would be. (Of course each program might have
its own set of security problems, but the concept of each has no more or
less security implications than the other)

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

On Mon, Jun 23, 2008 at 4:59 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
>
> Do you think we should remove groff (and with it, man-db), too? I'm not
> trying to be difficult.. This is an actual question.
>
> --
> Soren Hansen   |
>

No, but I use man all the time.  Granted, that may be pandering to
myself, but man just seems infinitely more useful than a web browser.
This could be that 1) I have never ever used a web browser, or been in
a position where it would seem useful, on a server and 2) the whole
idea that "it's a browser", therefore "it's a huge security hole."

And you may say it's just for looking at local html files, but it
opens up the thinking that "if they put a web browser on a server,
then it must be ok to browse the web from a server."  I've seen plenty
of Windows admins casually browsing the web from a server console
while waiting on some task to finish in the server room.

But that's just my opinion.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Sat, Jun 21, 2008 at 12:21:10AM -0500, James Dinkel wrote:
> That is that the concern here is not so much about w3m being in the
> server seed, but that if such a useless package can make it into the
> server seed, than what more feature creep will the future bring.  

Do you think we should remove groff (and with it, man-db), too? I'm not
trying to be difficult.. This is an actual question.

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Fri, Jun 20, 2008 at 09:16:11PM -0500, Michael Hipp wrote:
>> If we look solely at the packages I suggested, I don't think 5-6
>> packages constitutes a greatly expanded list of items to be
>> installed.
> Agreed. But I work under the assumption that the list over time will
> grow greatly and not over a very long time at that.

Well, I'm not going to promise anything about the amount of packages.
First of all, it's not all up to me, so I couldn't keep the promise even
if I made it, and second of all, it's not really a useful metric as e.g.
packages get split from time to time, so even if nothing new is
installed, the package count can increase. I think a better promise
would be that I'll do whatever I can to make the set of packages
reasonable and sound. I'm as much against bloat as the next guy.

> If I may expound further (and I think this will speak to several of
> your questions below) I fear this list would become truly ridiculous
> if we pursue this path with anything less than a very hard line
> approach. 

> By way of example: I find it truly amazing, yeah, incomprehensible
> that someone would suggest a text mode browser should be installed by
> default on every server. My mind is unable to grasp the concept.
> Meaning no offense to any who hold this opinion, I just do not know
> how to even think about it.

Look at it this way: I doubt anyone wants w3m on their servers so that
they can read Planet Ubuntu or update their Facebook profiles while
they're in the server room at the console.

To enable our users to read man pages, we install man-db and groff-base
by default. If you're old school enough to know how to use it, groff can
be used as a very full featured typesetting system, which is also not
something that really belongs on all of our servers by default, but
because it's a prerequisite for viewing man pages, we accept it.

Likewise, it's a simple fact of life that some pieces of software have
chosen to provide their documentation in html format. In order to enable
users to read this documentation, we need something that can transform
html into something legible. At the moment, this is w3m.

> >> The rational-defensible approach, IMHO, is to keep the base install
> >> as lean as possible but also make it as easy as possible to layer
> >> on top of that base.
> > What is the goal you are trying to achieve by doing this?
> Thank you for asking. I think I answer this in pieces below.
> 
> I is somewhat scary tho as I thought this was actually a design goal of sorts 
> of Ubuntu (somewhat) and ubuntu-server in particular. 

Up until now, Ubuntu Server has consisted of nothing but the packages in
the standard seed (+ a kernel, of course). Great care is taken to make
sure nothing unnecessary is added to standard, so up until now, you've
actually been right. What I'm suggesting is that we change this policy
slightly to provide more of an integrated platform, based on what we as
a team (which includes anyone who wishes to participate) find to be best
practices.

Package maintainers put great care into making their packages as good as
possible, e.g. adding all the right options when they're compiling
software, putting stuff in the right places, etc. If someone has input,
they can file a bug, discuss it, and if consensus is reached, the
changes can be implemented, so really all packages represent best
practic in terms of packaging and (some) configuration.

However, when it comes to choosing the right packages, and integrating
everything, making it all work together and such, every user has to
start from scratch.  Sure, this makes for a good learning experience,
but I don't think most users are here to learn, but to get stuff done.
Some people refer to this as "freedom of choice", but that's only half
the truth.  There's no reason we couldn't provide the alternatives, so
that people can make different choices than we have, while still giving
them the benefit of being able to simply use what we have determined to
be best practice. Between us, we have probably centuries of sysadmin
experience. We all have our set of packages we always install, a
specific setting we always change, etc., etc.

> The necessity to keep all images on one CD being an example.

This is indeed a goal of every Ubuntu version. At some point in most
release cycles, something is sacrifised, or packages are split, or
others consolidated, etc. to make sure they fit on a CD.  In the server
edition we just happen to have the luxury to be far from the limit. This
doesn't mean that we'll just nonchalantly toss a couple hundred
megabytes of cruft on there, I'm just saying.

>>> The tools that a LAMP stack admin should be using are probably quite
>>> different than what one of my very simple Samba boxes would require.
>>> I don't think either's list should dictate the other.
>> You seem to wilfully miss my point. I'm not trying to determine the
>> /union/ of useful tools for /every possible/ user of Ubuntu Server.
>> I'm trying to deter

Re: Moving w3m out of standard

[Jim Tarvid]

On Sun, Jun 22, 2008 at 6:47 PM, Onno Benschop <[EMAIL PROTECTED]> wrote:

> On 21/06/08 14:53, Eric Hammond wrote:
> >
> > +1 for keeping server minimal or having a bare-bones option
> >
> > I've been maintaining a popular series of Ubuntu images for Amazon EC2
> > (listed at http://alestic.com ).  I like to keep the base install
> > matching the Ubuntu server install as closely as possible so folks know
> > what to expect.
>
> This is an example of a server - use case.
>
> I thought the whole point of the survey we are in the process of getting
> together was to create real end-user data so we can make assessments on
> how to proceed with this question of inclusion vs. exclusion.
>
> --
> Onno Benschop
>
> Connected via Optus B3 at S31°54'06" - E115°50'39" (Yokine, WA)
> --
> ()/)/)()..ASCII for Onno..
> |>>?..EBCDIC for Onno..
> --- -. -. ---   ..Morse for Onno..
>
> ITmaze   -   ABN: 56 178 057 063   -  ph: 04 1219    -
> [EMAIL PROTECTED]
>
>
>
> --
> ubuntu-server mailing list
> ubuntu-server@lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
> More info: https://wiki.ubuntu.com/ServerTeam
>

Hardcore general purpose Internet website server.

Installed Command line installation from Alternate CD.
Used tasksel to install desired server packages
Generated package list with dpkg -l
Pulled list of installed packages on current server.
Ran diff to find those on current server not in command line install +
tasksel.
Pulled those packages which support tasks that could be done on a
workstation.
Generated install script which was fairly short.

tasksel install dns-server
tasksel install lamp-server
tasksel install mail-server
tasksel install openssh-server
apt-get install bash-completion
apt-get install ca-certificates
apt-get install curl
apt-get install cvs
apt-get install cvsd
apt-get install finger
apt-get install imagemagick
apt-get install lftp
apt-get install mailman
apt-get install ntp
apt-get install patch
apt-get install php5-cgi
apt-get install php5-cli
apt-get install php5-curl
apt-get install php5-gd
apt-get install php5-imagick
apt-get install php5-imap
apt-get install php5-mcrypt
apt-get install php5-mysqli
apt-get install phpmyadmin
apt-get install php-pear
apt-get install procmail
apt-get install tree
apt-get install unzip
apt-get install vim
apt-get install vsftpd
apt-get install whois

Easier in some ways than a server install.

Jim Tarvid


-- 
http://ls.net
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Onno Benschop]

On 21/06/08 14:53, Eric Hammond wrote:
>
> +1 for keeping server minimal or having a bare-bones option
>
> I've been maintaining a popular series of Ubuntu images for Amazon EC2
> (listed at http://alestic.com ).  I like to keep the base install
> matching the Ubuntu server install as closely as possible so folks know
> what to expect.

This is an example of a server - use case.

I thought the whole point of the survey we are in the process of getting
together was to create real end-user data so we can make assessments on
how to proceed with this question of inclusion vs. exclusion.

-- 
Onno Benschop

Connected via Optus B3 at S31°54'06" - E115°50'39" (Yokine, WA)
--
()/)/)()..ASCII for Onno..
|>>?..EBCDIC for Onno..
--- -. -. ---   ..Morse for Onno..

ITmaze   -   ABN: 56 178 057 063   -  ph: 04 1219    -  
[EMAIL PROTECTED]



-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Eric Hammond]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1


+1 for keeping server minimal or having a bare-bones option

I've been maintaining a popular series of Ubuntu images for Amazon EC2
(listed at http://alestic.com ).  I like to keep the base install
matching the Ubuntu server install as closely as possible so folks know
what to expect.

The smaller the image the faster a new instance can be loaded.  This is
already an area where the Ubuntu images beat the other distributions
I've tested, and I'd like to keep it that way or even increase the gap.

- --
Eric Hammond
[EMAIL PROTECTED]

-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFIXKV7N9qi28wBkF0RAkohAJsFfLceQkVPLZlQfPmUMiZUB+rvJACfZUgm
3ycVfVAxEfpVoGJHpnCCvG4=
=RzBB
-END PGP SIGNATURE-

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Scott Kitterman]

On Sat, 21 Jun 2008 00:21:10 -0500 "James Dinkel" <[EMAIL PROTECTED]> wrote:
>While there is a lot here that I could debate till my fingers fall
>off, I think one point by Soren actually sums up my concern pretty
>well:
>
>On Fri, Jun 20, 2008 at 7:02 PM, Soren Hansen <[EMAIL PROTECTED]> wrote:
>>
>> Given enough time, any software will prove to be insecure in
>> some way
>>
>
>And that is a very good reason for keeping frivolous packages off the 
server.
>
>I would also reiterate a subject Michael Hipp brought up in his last
>email.  That is that the concern here is not so much about w3m being
>in the server seed, but that if such a useless package can make it
>into the server seed, than what more feature creep will the future
>bring.  Keep in mind that I too am one that jumped ship from RHEL,
>which apparently decided to take after MS and make a nice big leap in
>bloated-by-default with RHEL 5.  SuSe is the same way.  So the thought
>of another distro which I've favored, losing what it is that I favor
>it for, is a little disheartening.
>
You are aware that w3m is part of the existing install and has been since 
approximately forever?  Not removing somethin you feel should be removed is 
not the same thing as adding stuff.

We need a better/more scalable system than we have right now to properly 
accomodate everyone's concerns.  I suspect that is a better thing to expend 
energy on at the moment than exactly what packages should be in or out.

Scott K

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

While there is a lot here that I could debate till my fingers fall
off, I think one point by Soren actually sums up my concern pretty
well:

On Fri, Jun 20, 2008 at 7:02 PM, Soren Hansen <[EMAIL PROTECTED]> wrote:
>
> Given enough time, any software will prove to be insecure in
> some way
>

And that is a very good reason for keeping frivolous packages off the server.

I would also reiterate a subject Michael Hipp brought up in his last
email.  That is that the concern here is not so much about w3m being
in the server seed, but that if such a useless package can make it
into the server seed, than what more feature creep will the future
bring.  Keep in mind that I too am one that jumped ship from RHEL,
which apparently decided to take after MS and make a nice big leap in
bloated-by-default with RHEL 5.  SuSe is the same way.  So the thought
of another distro which I've favored, losing what it is that I favor
it for, is a little disheartening.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Michael Hipp]

Soren Hansen wrote:
> I'm sorry you took offence to what I said. It meant it in the most
> affectionate way imaginable. In fact, I usually consider myself to be in
> that very group. If someone refers to the style with which I attack a
> problem as old school, I blush like a school girl (on the inside, at
> least). I also proudly refer to myself myself as a nerd or geek.

:-)

> If we look solely at the packages I suggested, I don't think 5-6
> packages constitutes a greatly expanded list of items to be
> installed.

Agreed. But I work under the assumption that the list over time will grow 
greatly and not over a very long time at that.

If I may expound further (and I think this will speak to several of your 
questions below) I fear this list would become truly ridiculous if we pursue 
this path with anything less than a very hard line approach. By way of example: 
I find it truly amazing, yeah, incomprehensible that someone would suggest a 
text mode browser should be installed by default on every server. My mind is 
unable to grasp the concept. Meaning no offense to any who hold this opinion, I 
just do not know how to even think about it.

That some folks have cases where such a thing is useful. Certainly. That it 
should be easy to add to a base system. Absolutely. Install it on every box by 
default. Can't imagine it.

I don't mean to pick on that example (or it's proponents), but if such an 
innocuous and (to me) useless thing has a strong chance of being added to the 
list, then I assume we'll shortly have a flood.

If someone wants to say that's an irrational fear, I won't argue the point. But 
having too much history with "kitchen sink" distros like RH/Fedora I just can't 
find it to be embarrassed about fearing to go back to anything similar.

>> The rational-defensible approach, IMHO, is to keep the base install as
>> lean as possible but also make it as easy as possible to layer on top
>> of that base.
> 
> What is the goal you are trying to achieve by doing this?

Thank you for asking. I think I answer this in pieces below.

I is somewhat scary tho as I thought this was actually a design goal of sorts 
of Ubuntu (somewhat) and ubuntu-server in particular. The necessity to keep all 
images on one CD being an example. So you asking is surprising. Apologies if 
I'm reading to much into this. I just thought this had been decided long ago.

>>> The current approach is something like:
>>>
>>> 1. Will more than 95% of our users need it? If yes, install it by
>>> default. If no, go to next question.
>>>
>>> 2. Will more than 80% of our users need it? If yes, include on CD. If
>>> no, go to next question.
>>>
>>> 3. Will more than 10% need it and be completely and utterly screwed
>>> without it? If yes, include it on the CD. If no, go to next question.
>>>
>>> 4. Forget it.
>> This is good stuff.
> 
> ..yet you seem to think that point 1 should read "100%".

Perhaps. But I suspect the difficulty in gathering reliable data about it would 
obscure the difference between 100% and 95%. So your process sounds fine given 
that a certain amount of subjectivity and imprecision is a given.

>> The tools that a LAMP stack admin should be using are probably quite
>> different than what one of my very simple Samba boxes would require. I
>> don't think either's list should dictate the other.
> 
> You seem to wilfully miss my point. I'm not trying to determine the
> /union/ of useful tools for /every possible/ user of Ubuntu Server. I'm
> trying to determine the /intersection/ of useful tools for /most/ users
> of Ubuntu Server.

I don't believe I'm doing that, certainly not willfully. I suppose it's just 
that I'm comfortable that the intersection has already been reached. Not to say 
there aren't things that should be added. But there may also be things that 
could be removed. I truly cannot imagine that such an intersection would 
contain more than a handful of items beyond what is already on the base server 
install.

And, if I may... The seeming conclusion that w3m is a good candidate for being 
added looks to me like the beginnings of a "union" process, not an 
"intersection" process. But I'm probably guilty of making too much of one 
example.

 Here's my list:
>> Evidently my thick sarcasm obscured the fact that this list was
>> intended to show how quickly it becomes ridiculous to include what
>> everyone thinks is a necessity. 
> 
> As we're dealing with tricky bits of set logic, please be careful with
> words like "everyone". Surely, "what everyone thinks is a necessity"
> (i.e. something that *every* user finds necessary) should be installed?

What I meant by that "everyone thinks is necessity" was the idea of taking 
everyone's list, adding them together ("union") and making that the list of 
things we add to the default install. Obviously we won't be doing that. But the 
w3m example looks awfully close to it. (See "irrational fears" paragraph above.)

>> Hence, no matter how bad I want it, it shou

Re: Moving w3m out of standard

[Soren Hansen]

On Fri, Jun 20, 2008 at 03:34:52PM -0500, Michael Hipp wrote:
 We should probably add an install option to the server CD to only
 install the base system, so that the die hard group of old school
 admins can keep their Ubuntu systems as small as possible, though.
> >> I'm not sure if you're trying to spark a flame war or not. 
>> Err.. No, I'm not. I'm not sure a) what would make you say that, and
>> b) why you seem to be taking this so very personal.
> Because you made the statement that anyone who disagrees is "die hard
> old school". That's offensive. 

I'm sorry you took offence to what I said. It meant it in the most
affectionate way imaginable. In fact, I usually consider myself to be in
that very group. If someone refers to the style with which I attack a
problem as old school, I blush like a school girl (on the inside, at
least). I also proudly refer to myself myself as a nerd or geek.

> It directly implies that we're some kind of extremists standing in the
> way of good and holy progress.

By the very definition, I would say that if you are a follower of the
"old school", then you believe in the way things used to be.  In
opposition hereto are the followers of the "new school" which represents
the way things are now (or are perhaps becoming).  That is not to say
that either stance is extreme, good, nor bad. They are simply points of
view.

>> INSWYTISP, but that's really not the case. I'm attempting to start a
>> discussion about what sort of stuff we should put on servers by
>> default.  The operative word here is "should". Not "could".
> Agreed, but many of the comments and your own thoughts seem to be
> leading toward a greatly expanded list of items to be installed.

If we look solely at the packages I suggested, I don't think 5-6
packages constitutes a greatly expanded list of items to be
installed.

> And "should" surely is subjective and situational.

Yes. You will notice that I have not actually made any changes although
I am actually in a position to make the changes in question. I'm trying
to start a discussion about it and gather data points, so that the
decision will be as objectively sound as possible. 

> The rational-defensible approach, IMHO, is to keep the base install as
> lean as possible but also make it as easy as possible to layer on top
> of that base.

What is the goal you are trying to achieve by doing this?

>> The current approach is something like:
>> 
>> 1. Will more than 95% of our users need it? If yes, install it by
>> default. If no, go to next question.
>> 
>> 2. Will more than 80% of our users need it? If yes, include on CD. If
>> no, go to next question.
> > 
>> 3. Will more than 10% need it and be completely and utterly screwed
>> without it? If yes, include it on the CD. If no, go to next question.
>> 
>> 4. Forget it.
> 
> This is good stuff.

..yet you seem to think that point 1 should read "100%".

>> What I'm suggesting is to add an extra step in between 1 and 2.
>> Something like "Is it something most of our users *should* be using?"
>> or "Does using it constitute what we consider best practice?". If so,
>> install it by default.
> This word "should" keeps cropping up and it bothers me. Who is to say
> what I "should" be using. 

Not I. Nor you. That is why we analyse, discuss, and reach conclusions.

> The tools that a LAMP stack admin should be using are probably quite
> different than what one of my very simple Samba boxes would require. I
> don't think either's list should dictate the other.

You seem to wilfully miss my point. I'm not trying to determine the
/union/ of useful tools for /every possible/ user of Ubuntu Server. I'm
trying to determine the /intersection/ of useful tools for /most/ users
of Ubuntu Server.

>>> Here's my list:
> Evidently my thick sarcasm obscured the fact that this list was
> intended to show how quickly it becomes ridiculous to include what
> everyone thinks is a necessity. 

As we're dealing with tricky bits of set logic, please be careful with
words like "everyone". Surely, "what everyone thinks is a necessity"
(i.e. something that *every* user finds necessary) should be installed?

> (Not to advocate those particular packages.) And I don't agree, for
> example, that 'screen' should be installed by default. It's primarily
> useful, I think, to those of us who never sit at the console. 

I sometimes sit at a console and I find screen to be *really* useful.

> Someone who admins from the console would probably never use it.

I think this is a false assumption. I'm not sure though. That's why I
intend to solicit feedback on the issue when we have established a sound
method for evaluating possible inclusion by default.

> Hence, no matter how bad I want it, it shouldn't be installed by
> default.

Why?

>> A guy called Michael Hipp (you may have heard of him) once asked me:
>> "I'm not sure if you're trying to spark a flame war or not."  It just
>> so happens that I'm not, but you sure seem to be.
> And I'm n

Re: Moving w3m out of standard

[Onno Benschop]

On 20/06/08 16:33, Soren Hansen wrote:
> With the advent of the server seed, I think the time is ripe for us to
> be slightly (not much, but just a tiny bit) more promiscuous in our
> choice of what gets installed by default.
>   
In the server meeting on the 4th of June we were discussing ebox and
webmin. During that discussion an interesting comment was made which
indicated that an approach wasn't "ubuntu-like".

The notion that this encompasses speaks to me in ways that other
arguments don't. That is, if we're going to make a server platform seed,
perhaps we could come up with tools that promote best practice, make
life simpler for the administrator and allow for a higher level view of
the administration landscape.

While I'm an ssh and vi kinda admin, I'd appreciate version control for
my configuration files for example. I'd love ssh to come standard and
screen is a useful tool to have available. Disk space is absurdly cheap
and the tools we decide are required are going to be counted in megs
rather than gigs.

It seems that some discussion is already under way and that some
suggestions are already being made.

Perhaps stepping back and looking at each of our own server installs to
see what kind of things are used regularly would be useful.

For me things that come to mind are an annotated server log, that is, on
xyz day Abc installed foo to fix an incompatibility with bar.

There was also a suggestion a while back on the list about logging. A
spec was written and for a while I had a web page open with some actual
logging code, but I've since misplaced it :-(

> We should probably add an install option to the server CD to only
> install the base system, so that the die hard group of old school admins
> can keep their Ubuntu systems as small as possible, though.
>   
Absolutely, for me it's more to do with virtual deployments, that is,
I'd like to get to the point where I can have the one CD and deploy the
server and its virtual machines without needing different versions of
different things.

As for the old-school, once you've been around long enough, everything
old becomes new again :)

Finally, if it all goes down the toilet and a server under my control is
down and I'm sitting at the console needing to make it work *now*, the
last thing I want to do is have to get another internet connection to
fire up my laptop, or to find a spot to balance it while sitting at a
19" rack with a console keyboard on my knees. What I'm saying is that
when we come up with our list, it would be nice to find that on that day
when I'm sitting there, this discussion made my life easier.

-- 
Onno Benschop

Connected via Optus B3 at S31°54'06" - E115°50'39" (Yokine, WA)
--
()/)/)()..ASCII for Onno..
|>>?..EBCDIC for Onno..
--- -. -. ---   ..Morse for Onno..

ITmaze   -   ABN: 56 178 057 063   -  ph: 04 1219    -   [EMAIL PROTECTED]



-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Fri, Jun 20, 2008 at 11:57:29AM -0500, James Dinkel wrote:
 We should probably add an install option to the server CD to only
 install the base system, so that the die hard group of old school
 admins can keep their Ubuntu systems as small as possible, though.
>>> I'm not sure if you're trying to spark a flame war or not.
> I think this is a great idea (I actually even suggested that in the
> last Server Team meeting).  I'm all for keeping the install process as
> simple as possible, which means keeping the questions and options to a
> minimum, but I think an additional option to install a bare system or
> to include a set of recommended tools would be worthwhile.

>>> Here's my list:
>>>
>>> openssh
>>> samba
>>> apache
>>> postfix
>>> dovecot
>>> openvpn
> um, he's not really suggesting making all these part of the standard
> install, is he?  

Well, that was the topic of the discussion.

> Though the original email was off-list, I'm guessing he was being
> sarcastic here.

I wasn't sure, so I went the safe way and assumed that he wasn't and
tried to provide feedback to his suggestions.

> There are certain software packages that go on EVERY linux server I
> set up, such as screen and openssh-server.  However, I refrain from
> suggesting these go in to the server seed, because some people may not
> use them and they are easily, for me, to apt-get them. 

What exactly is the goal we're trying to achieve by refusing to install
something as useful as e.g. screen?

Are you ashamed of the set of packages you install on every server? Are
there ones you really wish you didn't have to install, but still do? If
not, these packages might very well be what we should consider best
practice to install, and I'd very, *very* much like to see your list.

> I'm not preaching to turn Ubuntu into the perfectly customized server
> distro just for me, 

Nor am I. I'm a) trying to have a discussion about the criteria we use
for choosing the software that goes on our server installs by default,
and then b) trying to apply the logic of my suggested changes and see
what the outcome is.

> I'm preaching to keep it efficient and secure, while providing the
> canvas for admins to create the perfectly customized server for
> themselves.

Michael made comments to this affect as well. I'll address this in my
e-mail response to him.

>>> And, excuse me, saying we can just apt-get remove it is surely the
>>> *dumbest* suggestion I've heard on an Internet list anytime
>>> recently.
> While not the dumbest, I do not like this suggestion, though I do
> practice it.  Particularly on my Redhat servers.  I go though the
> running daemons and uninstall any that are stupid to have on a server
> (like pcmcia and bluez services).  I have a "sudo yum remove" line all
> written down in a text file that I just copy, paste, and run on every
> server, right after install.

I'm not suggesting that we install all sorts of stuff that will not be
useful to most people. I'm in fact suggesting that we install stuff that
*will* be useful to most people. I'd like to install the stuff that only
a minority will want to remove. For instance, it's already perfectly
possible to use a system that doesn't provide an editor other than nano
or ed or whatever, but still we provide vim-tiny in our minimal
installs. We do this because a *lot* of people find this useful.

>> "Oh, so maybe we shouldn't even install a coreutils? Or a kernel?
>> Maybe we should make an apt-get remove --ALL option?" (I'm taking a
>> stab at the
>> take-whatever-people-say-and-blow-it-completely-out-of-proportions
>> things. How am I doing?)
> I think there is a pretty big difference between compiling your own
> kernel and running "sudo apt-get install foobar".

I'm not suggesting that people compile their own kernels. I'm just
saying that they're only an apt-get install away anyway, so why bother
installing it by default? As I said, I was deliberately trying to blow
things as much out of proportion as Michael does, in an attempt to get a
point across. The point in question is that we already install lots of
stuff that otherwise would only be an apt-get install away, but the
convenience of their being installed by default by far outweighs the
crappiness of each user having to do it him-/herself.

> I don't believe I've ever used ftp or openssh-client from a server, so
> those could easily go.  I very very rarely have used at and am
> actually kinda surprised to hear it is installed by default.

Well.. it is.  :)

>>> So don't start me out in a mansion when a rustic cabin is adequate
>>> for my needs.
>> To keep to the house analogies, I think that your suggestion is
>> closer to just providing the foundation of the house and leave it up
>> to anyone who actually wants a place to live to build the house
>> itself, install doors, windows, heating facilities, bathrooms,
>> kitchens, etc., because, you know, a very significant percentage of
>> the world's population manages survives

Re: Moving w3m out of standard

[Neal McBurnett]

>  die hard old school sysadmin

As another proud and hardy sysadmin from a venerable institution :) I
think it is helpful to look at the trends, and accept once again that
"linux server" is a very diverse beast, and getting more so all the
time.  So the  suggestion by Soren et al. that there be more flexibility
in the installation of Ubuntu server makes sense to me.

E.g. I'm told there are Gentoo-based JEOS LAMP server images that are
just 10 GB compressed, via use of busybox etc.  And there are
"development" servers that have lots of nice development tools.  Print
servers can be very different from DNS servers or compute servers.  And
there are those that love GUI interfaces too

I like the idea of an easy-to-install minimal server platform: JEOS +
basic hardware drivers.  I also resonate with the idea of encouraging
"best practices", thought I think of that as stuff folks are likely to
*want* to be using as well as what they *should* be using (e.g. for
security).  More doucmentation will slowly help there, as well as more
tasksel options and nicely packaged bundles.

It would be good to have more facts at our disposal.  Are there handy
tools around for asking not just how big a package is, but also how much
space it would take to install it and its dependencies on top of some of
the standard seeds?  And I agree that information on history of
vulnerabilities, open ports, etc are very relevant.  I'd love to see
specific information on uninstall bugs that folks have encountered also
- I can see why they might be more common than we'd like, and the best
way to deal with it is to report them

-- 
Neal McBurnett http://mcburnett.org/neal/



-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

On Fri, Jun 20, 2008 at 3:34 PM, Michael Hipp <[EMAIL PROTECTED]> wrote:
>
> Because you made the statement that anyone who disagrees is "die hard old
> school". That's offensive. It directly implies that we're some kind of
> extremists standing in the way of good and holy progress.
>

As a die hard old school sysadmin, I take offense to you considering
it to be offensive to be called "die hard old school."

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Michael Hipp]

Soren Hansen wrote:
> [Michael told me in a different e-mail that he replied off-list by
> accident, so I'm taking the thread back on the list]
> 
> On Fri, Jun 20, 2008 at 08:07:14AM -0500, Michael Hipp wrote:
>>> We should probably add an install option to the server CD to only
>>> install the base system, so that the die hard group of old school
>>> admins can keep their Ubuntu systems as small as possible, though.
>> I'm not sure if you're trying to spark a flame war or not. 
> 
> Err.. No, I'm not. I'm not sure a) what would make you say that, and b)
> why you seem to be taking this so very personal.

Because you made the statement that anyone who disagrees is "die hard old 
school". That's offensive. It directly implies that we're some kind of 
extremists standing in the way of good and holy progress.

> INSWYTISP, but that's really not the case. I'm attempting to start a
> discussion about what sort of stuff we should put on servers by default.
> The operative word here is "should". Not "could".

Agreed, but many of the comments and your own thoughts seem to be leading 
toward a greatly expanded list of items to be installed. And "should" surely is 
subjective and situational. The rational-defensible approach, IMHO, is to keep 
the base install as lean as possible but also make it as easy as possible to 
layer on top of that base.

> The current approach is something like:
> 
> 1. Will more than 95% of our users need it? If yes, install it by
>default. If no, go to next question.
> 
> 2. Will more than 80% of our users need it? If yes, include on CD. If
>no, go to next question.
> 
> 3. Will more than 10% need it and be completely and utterly screwed
>without it? If yes, include it on the CD. If no, go to next question.
> 
> 4. Forget it.

This is good stuff.

> What I'm suggesting is to add an extra step in between 1 and 2.
> Something like "Is it something most of our users *should* be using?" or
> "Does using it constitute what we consider best practice?". If so,
> install it by default.

This word "should" keeps cropping up and it bothers me. Who is to say what I 
"should" be using. The tools that a LAMP stack admin should be using are 
probably quite different than what one of my very simple Samba boxes would 
require. I don't think either's list should dictate the other.

>> Here's my list:

Evidently my thick sarcasm obscured the fact that this list was intended to 
show how quickly it becomes ridiculous to include what everyone thinks is a 
necessity. (Not to advocate those particular packages.) And I don't agree, for 
example, that 'screen' should be installed by default. It's primarily useful, I 
think, to those of us who never sit at the console. Someone who admins from the 
console would probably never use it. Hence, no matter how bad I want it, it 
shouldn't be installed by default.

> A guy called Michael Hipp (you may have heard of him) once asked me:
> "I'm not sure if you're trying to spark a flame war or not."  It just so
> happens that I'm not, but you sure seem to be.

And I'm not sure why you think that calling us "die hard old school" should not 
be taken personally or with offense. It was. Maybe it shouldn't be, but it was. 
(And it is often said, erroneously perhaps, that perception is reality.)

> Let me offer a take on this.  Say there's a package called foo, which
> 60% of our users would want. If we install it by default, only 40% of
> our users will have to change the default, while 60% will be happy with
> it.  Disregarding all other circumstances, surely that sounds sensible?

No. It doesn't sound sensible. Let me attempt to explain why...

Take the 60% group that needs the package. The only pain they feel is the 
necessity, after install, to type a 30-character apt-get command. (I'm assuming 
it's on the CD or the net is available.) Their system is no worse for it.

But the 40% group must a) allocate partition space for a package they don't 
need, b) possibly answer configuration questions during install for a package 
they don't need, c) remember to uninstall something that is likely 
out-of-sight-out-of-mind, d) type a similar 30-character apt-get command to 
remove the package, e) live with the knowledge and risk of knowing that 
uninstalling packages is far more likely to break something than an install 
would be (from my experience, anyways).

(To use an extreme analogy, it's somewhat like all the "crapware" that comes on 
a PC from a big-name manufacturer. You have to endure the pain of the degunking 
in order to have a machine that runs and responds like it should. Yes, it would 
be way over the top to insinuate this would be anywhere near that bad.)

So a crude economic analysis would put the "cost" of making the 60% group happy 
much higher than going with the wishes of the minority 40%. And please note 
that I place great weight on item 'e' above, uninstalling is far worse than 
installing. Maybe I'm just unlucky.

>> None of them (along with w3m

Re: Moving w3m out of standard

[James Dinkel]

I'll just start by saying that while I disagree with Soren, I in no
way took anything that he said to be flaming or trying to insight a
flame war...

On Fri, Jun 20, 2008 at 10:04 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
> [Michael told me in a different e-mail that he replied off-list by
> accident, so I'm taking the thread back on the list]
>
> On Fri, Jun 20, 2008 at 08:07:14AM -0500, Michael Hipp wrote:
>>> We should probably add an install option to the server CD to only
>>> install the base system, so that the die hard group of old school
>>> admins can keep their Ubuntu systems as small as possible, though.
>> I'm not sure if you're trying to spark a flame war or not.
>

I think this is a great idea (I actually even suggested that in the
last Server Team meeting).  I'm all for keeping the install process as
simple as possible, which means keeping the questions and options to a
minimum, but I think an additional option to install a bare system or
to include a set of recommended tools would be worthwhile.


>
>> Here's my list:
>>
>> openssh
>> samba
>> apache
>> postfix
>> dovecot
>> openvpn
>

um, he's not really suggesting making all these part of the standard
install, is he?  Pretty sure I don't need dovecot or openvpn on my
webserver or any of those but Samba on my file server.  Not even
Windows Server is that sloppy.  Though the original email was
off-list, I'm guessing he was being sarcastic here.

>
>> screen
>
> Agreed.
>
>> vim (full)
>
> vim-full depends on a stack of GUI stuff, but a more full featured vim
> than vim-tiny (like e.g. the "vim" package) would be lovely to have by
> default.
>
>> Just to name a few. And how could anyone possibly object to any of those?
>> Why, they're just basic stuff that I really, really need. Not like it'll
>> hurt anything. So what that ubuntu-server requires a stack of DVDs to
>> install. DVDs are cheap!
>

There are certain software packages that go on EVERY linux server I
set up, such as screen and openssh-server.  However, I refrain from
suggesting these go in to the server seed, because some people may not
use them and they are easily, for me, to apt-get them.  I'm not
preaching to turn Ubuntu into the perfectly customized server distro
just for me, I'm preaching to keep it efficient and secure, while
providing the canvas for admins to create the perfectly customized
server for themselves.

>
>> And, excuse me, saying we can just apt-get remove it is surely the
>> *dumbest* suggestion I've heard on an Internet list anytime recently.
>

While not the dumbest, I do not like this suggestion, though I do
practice it.  Particularly on my Redhat servers.  I go though the
running daemons and uninstall any that are stupid to have on a server
(like pcmcia and bluez services).  I have a "sudo yum remove" line all
written down in a text file that I just copy, paste, and run on every
server, right after install.


>
> "Oh, so maybe we shouldn't even install a coreutils? Or a kernel? Maybe
> we should make an apt-get remove --ALL option?" (I'm taking a stab at
> the take-whatever-people-say-and-blow-it-completely-out-of-proportions
> things. How am I doing?)
>

I think there is a pretty big difference between compiling your own
kernel and running "sudo apt-get install foobar".

> Do you think there are things in the standard seed that doesn't belong
> there? If you truly want to do everything yourself I guess you'd even
> want the server install to not include the standard seed, but only
> minimal? That would remove such completely useless things as psmisc,
> man-db, iptables, ftp, at, cron, file, openssh-client, and wget.
>

I don't believe I've ever used ftp or openssh-client from a server, so
those could easily go.  I very very rarely have used at and am
actually kinda surprised to hear it is installed by default.

>> So don't start me out in a mansion when a rustic cabin is adequate for
>> my needs.
>
> To keep to the house analogies, I think that your suggestion is closer
> to just providing the foundation of the house and leave it up to anyone
> who actually wants a place to live to build the house itself, install
> doors, windows, heating facilities, bathrooms, kitchens, etc., because,
> you know, a very significant percentage of the world's population
> manages survives without most of these things, so who are we to go and
> decide that everyone should have heating facilites installed even though
> they can just choose to not turn them on?
>

I would prefer the foundation so I can customize the size, layout, and
features how I want them.  This is assuming your analogy means I can
just point my finger and say "sudo apt-get install toilet".


>
> I'm sure you'll enjoy installing extra packages over that sort of
> connection.
>

I think he was refering to using gui tools from a remote workstation
over that connection.  The server itself could have been connected to
the internet with a 100mbit fiber line.  Also, everyone keeps bringing
this up that "you'll

Re: Moving w3m out of standard

[Soren Hansen]

[Michael told me in a different e-mail that he replied off-list by
accident, so I'm taking the thread back on the list]

On Fri, Jun 20, 2008 at 08:07:14AM -0500, Michael Hipp wrote:
>> We should probably add an install option to the server CD to only
>> install the base system, so that the die hard group of old school
>> admins can keep their Ubuntu systems as small as possible, though.
> I'm not sure if you're trying to spark a flame war or not. 

Err.. No, I'm not. I'm not sure a) what would make you say that, and b)
why you seem to be taking this so very personal.

> But there's nothing "die hard old school" about not wanting to install
> a bunch of crap that we don't need or want.

Again, I'm not sure why you're taking it so personal. (Henceforth
abbreviated INSWYTISP)

> That's part of what attracted many of us to ubuntu-server.

That's valuable input. Thanks.

> Since this thread has turned into
> let's-add-my-favorite-just-in-case-I-might-need it. 

INSWYTISP, but that's really not the case. I'm attempting to start a
discussion about what sort of stuff we should put on servers by default.
The operative word here is "should". Not "could".

The current approach is something like:

1. Will more than 95% of our users need it? If yes, install it by
   default. If no, go to next question.

2. Will more than 80% of our users need it? If yes, include on CD. If
   no, go to next question.

3. Will more than 10% need it and be completely and utterly screwed
   without it? If yes, include it on the CD. If no, go to next question.

4. Forget it.

What I'm suggesting is to add an extra step in between 1 and 2.
Something like "Is it something most of our users *should* be using?" or
"Does using it constitute what we consider best practice?". If so,
install it by default.

> Here's my list:
>
> openssh
> samba
> apache
> postfix
> dovecot
> openvpn

All of these listen on the network and would violate our
no-open-ports-by-default policy.

> openntpd

I agree that something that makes sure the time on your server is
accurate is needed, which is why I suggested ntp. I'm not familiar with
openntpd. What benefits does it provide over ntp (which is already in
main)?

> no-ip

Is a transitional package for noip2. I think something like noip2 might
make sense to have on the CD, actually, but I wouldn't suggest
installing it by default. It doesn't constitute best practice, and I
don't think it's of great use to the majority of users.

> screen

Agreed.

> vim (full)

vim-full depends on a stack of GUI stuff, but a more full featured vim
than vim-tiny (like e.g. the "vim" package) would be lovely to have by
default. 

> Just to name a few. And how could anyone possibly object to any of those? 
> Why, they're just basic stuff that I really, really need. Not like it'll 
> hurt anything. So what that ubuntu-server requires a stack of DVDs to 
> install. DVDs are cheap!

INSWYTISP. 

> And, excuse me, saying we can just apt-get remove it is surely the
> *dumbest* suggestion I've heard on an Internet list anytime recently.

A guy called Michael Hipp (you may have heard of him) once asked me:
"I'm not sure if you're trying to spark a flame war or not."  It just so
happens that I'm not, but you sure seem to be.

I find that it's sometimes convenient to stop for a second and think
about why you're doing the things you're doing. Simply refusing to
discuss things and reevaluate them is just silly. Any policy that can't
stand being reevaluated once every couple of years is not worth much,
IMO.

Let me offer a take on this.  Say there's a package called foo, which
60% of our users would want. If we install it by default, only 40% of
our users will have to change the default, while 60% will be happy with
it.  Disregarding all other circumstances, surely that sounds sensible?

Now, what if the package is a several hundred megabyte blob of stuff
that would be completely unusable for the 40% (perhaps it's a driver for
some hardware they don't have)?  See, that shifts reality a bit, because
the convenience of the 60% of users who need it doesn't justify the
amount of pain inflicted upon the 40% who has absolutely no use for it.

In Ubuntu I like to think that we take security rather seriously. That's
why I picked checksecurity and chkrootkit as examples of stuff to
install by default. They are tools that at intervals will scan your
system for various things that might represent a security problem.[1]

> (Enough juvenile sarcasm and hate mongering already)

I'm glad you can see it yourself. I'm less glad that you couldn't avoid
it, though.

> Do you see the problem?

I see plenty of problems... so I try to solve them. Where I come from,
this is usually considered a good thing.

> None of them (along with w3m) are in any way essential to get a basic
> server up and running. So why include them?

Because a server that does nothing but boot is useless for anything but
heating your house and increasing your electrical bill?

> Servers are *by 

Re: Moving w3m out of standard

[James Dinkel]

On Fri, Jun 20, 2008 at 7:01 AM, Soren Hansen <[EMAIL PROTECTED]> wrote:
>
> I'm not sure I'm familiar with UCSA?
>
> --
> Soren Hansen   |
>

It's vaporware right now... something that's been tossed around on the
mailing list.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Serge van Ginderachter]

> I'm not sure I'm familiar with UCSA?

Sorry,

  Ubuntu Central Server Administration


serge

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Fri, Jun 20, 2008 at 12:50:32PM +0200, Serge van Ginderachter wrote:
> Maybe a partial parameter to what should be provided could also be
> those packages which are manageable through UCSA 

I'm not sure I'm familiar with UCSA?

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Fri, Jun 20, 2008 at 12:05:37PM +0200, Ante Karamatic wrote:
> > [1]: At least ntp and munin-node do have open sockets, but they're
> > both only bound to the loopback interface, so I don't expect this to
> > be overly controversial.
> There's a huge difference between munin-node and ntp on one side, and
> w3m on the other. 

Certainly. If that wasn't the case, there wouldn't be much to discuss.
:)

> w3m is client side tool, for local users of server (administrators).

Quite right.

> ntp and munin-node are daemons and, when installed, most of the
> administrators share those services on network.

That's also true.  I'm not sure what your point is?

> As you can see, I totally agree with you regarding w3m, but not over
> munin-node and ntp. But, I could get convinced if we start offering
> products, not packages((c) by Rick Clark) :D

I suppose that is sort of what I'm suggesting. The current Ubuntu server
install is not much more than a minimal set of packages to have a
functional system. This is very convenient for some use cases, but at
some point, I'd like for Ubuntu server to be much more than that.

There's lots of simple stuff we could do, such as install some extra
packages by default (chkrootkit, checksecurity, and ntp just to name a
few).

Slightly more involved things to do is to install various monitoring
agents (I'm thinking of munin-node and nagios-plugins, but others might
make sense, too) along with a configuration package to hook them into a
corresponding server (for munin-node this would be as simple as adding
the server's IP to the list of allowed hosts, and for nagios-plugins it
would involve setting up an ssh-server and add a pubkey authenticated
user to call the plugins).  A properly designed directory service could
provide transport for these configuration settings.

I haven't put a whole lot of thought into this, to be honest, but it
seems like an interesting discussion to have.

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Serge van Ginderachter]

"Soren Hansen" <[EMAIL PROTECTED]> wrote:

> This seems to be a common theme when it comes to discussing what goes
> on
> servers by default and what doesn't.
> 
>   "We shouldn't put foo on servers by default, because not everyone
>   needs it, and it's easily apt-get installable if you need it."
> 
> Well, that might be true, but consider the converse:
> 
>   "Some people need it, and it's easily apt-get removable if you
> don't."
> 
> which is also true.
> 
> With the advent of the server seed, I think the time is ripe for us
> to
> be slightly (not much, but just a tiny bit) more promiscuous in our
> choice of what gets installed by default.


What about keeping having the option, to be choosen with some virtual package, 
just like one has with the desktop options?

- ubuntu-server-base
- ubuntu-server-full


> We should probably add an install option to the server CD to only
> install the base system, so that the die hard group of old school
> admins
> can keep their Ubuntu systems as small as possible, though.

Removing the "bloat" would ba as simple as removing the virtual package.
And having the option at installtime to not deploy ubuntu-server-full

Of course, one should be carefull to not extermely bloat ubuntu-server-full :)

But I agree that stuff like smartmontools might be a good candidate for 
ubuntu-server-full.
Maybe a partial parameter to what should be provided could also be those 
packages which are manageable through UCSA 



Serge

 Serge van Ginderachter  http://www.vanginderachter.be/ 

 Kreeg u een "odt" bestand en kan u deze niet openen? Zie http://ginsys.be/odf  

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Ante Karamatic]

On Fri, 20 Jun 2008 10:33:43 +0200
Soren Hansen <[EMAIL PROTECTED]> wrote:

> [1]: At least ntp and munin-node do have open sockets, but they're
> both only bound to the loopback interface, so I don't expect this to
> be overly controversial.

There's a huge difference between munin-node and ntp on one side, and
w3m on the other. w3m is client side tool, for local users of server
(administrators). ntp and munin-node are daemons and, when installed,
most of the administrators share those services on network.

As you can see, I totally agree with you regarding w3m, but not over
munin-node and ntp. But, I could get convinced if we start offering
products, not packages((c) by Rick Clark) :D

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Soren Hansen]

On Mon, Jun 16, 2008 at 01:03:48PM -0500, Michael Hipp wrote:
>> I think an interactive web browser is completely unnecessary on a
>> server.  You should do your browsing from a workstation, and then use
>> wget or sftp to transfer any files you need to the server.
> And w3m or similar tools are only one command away with apt-get if you
> do need them. Let's keep the "clutter" to a minimum.

This seems to be a common theme when it comes to discussing what goes on
servers by default and what doesn't.

  "We shouldn't put foo on servers by default, because not everyone
  needs it, and it's easily apt-get installable if you need it."

Well, that might be true, but consider the converse:

  "Some people need it, and it's easily apt-get removable if you don't."

which is also true.

With the advent of the server seed, I think the time is ripe for us to
be slightly (not much, but just a tiny bit) more promiscuous in our
choice of what gets installed by default.

w3m is a good example of this.  You might not use it every day (or even
week, month, or year), but the convenience of finding it there when you
need it weighed against the space "wasted" for it while you don't makes
it a reasonable candidate for being installed by default, and I'm glad
this decision was made at the server team meeting a few days ago, too.
Something like screen would fall into sort of the same sorting pile, but
I think these are just the starting points.  I for one would certainly
be open to a discussion about additional tools being installed by
default, like chkrootkit, smartmontools, munin-node[1], checksecurity,
bzr, etckeeper (I haven't actually looked into this one, but it seems
like something we might want to push), ntp[1], etc.  This is all stuff
that would (begin to) turn the server install from a very basic system
on top of which you can install stuff to actually make it useful into
something that does that too, but also promotes a set of best practices
and such. So far, our only tools for doing that have been inclusion into
main and sometimes documentation.

We should probably add an install option to the server CD to only
install the base system, so that the die hard group of old school admins
can keep their Ubuntu systems as small as possible, though.

[1]: At least ntp and munin-node do have open sockets, but they're both
only bound to the loopback interface, so I don't expect this to be
overly controversial.

-- 
Soren Hansen   | 
Virtualisation specialist  | Ubuntu Server Team
Canonical Ltd. | http://www.ubuntu.com/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Matt Zimmerman]

On Tue, Jun 17, 2008 at 02:05:09PM -0400, Bud Roth wrote:
> I agree with Dustin's point.  When an install goes flawlessly, a text
> web browser may be superfluous, but when things go crunch and your
> laptop is not handy, a text web browser is a quick way to find solutions
> on the Internet.  Sometimes, it can be as simple a need as looking up a
> VMWARE key or making sure that the server's Apache server is serving
> pages.  Wget doesn't really fit the bill.  Some form of text web browser
> should be easily apt-gettable...

This discussion has no bearing at all on whether a text web browser is
apt-gettable.  There are already three of them in main (w3m, links, elinks),
and even if w3m were moved out of main, it would still be apt-gettable.

My proposal has only to do with removing it from the default installation
on desktops.

-- 
 - mdz

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Mathias Gug]

On Mon, Jun 16, 2008 at 03:31:46PM +0100, Matt Zimmerman wrote:
>  * w3m# we need some text-based html presenter
> 
> I'd like to formally cast doubt on this statement from the standard seed.
> This was originally added a long time ago in order to provide a text-based
> browser for use on servers, at a time when the default server installation
> was a strict subset of the desktop.  It is completely superfluous on a
> modern desktop.
> 
> Now that this is becoming possible with the new server seed[1], I'd like to
> propose that it move to the server seed instead (or even be removed, if the
> server team doesn't feel it's appropriate).

This issue was discussed during today's server team meeting [1]. The
result is that the Server team would like to keep w3m installed by
default when using the ubuntu-server isos for installation.

Once the -server seeds are ready, we can move w3m to one of the new server
seeds when/if w3m is removed from the standard seed.

[1]: https://wiki.ubuntu.com/MeetingLogs/Server/20080617

-- 
Mathias Gug
Ubuntu Developer  http://www.ubuntu.com

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

>
> I also like having a text-based web browser on a server, as others
> have said, for reading documentation (especially when the server is
> offline and you're trying to get it back online) and for local web
> admin tasks.
>
> Space constraints don't seem that big an issue on the server CD.
> And I'd like to consider continuing having it on the server default
> install.
>
> I think the issues are more around security and support and choice.
> w3m is the default now, should that change?  Are there significant
> security risks to text browsing?  Should elinks take is place (noting
> that comment about accessibility recommendations?)  But elinks has
> more dependencies, right?  Detailing the fully-installed footprint
> delta for each choice would help, and knowing security exposures and
> considering features etc.
>
> Neal McBurnett http://mcburnett.org/neal/

I would rather remove it and then give a description of the popular 3
in the documentation.  I suggested in IRC, that the server install
does a very stripped down install by default (akin to JEOS) and then
give the option to add a set of "useful administrative tools".

After I think about it though, all my servers are in virtual machines
anyway, so I'll just use Ubuntu JEOS.

James (Brazen)

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Cody A.W. Somerville]

On Mon, Jun 16, 2008 at 11:31 AM, Matt Zimmerman <[EMAIL PROTECTED]> wrote:

>  * w3m# we need some text-based html presenter
>
> I'd like to formally cast doubt on this statement from the standard seed.
> This was originally added a long time ago in order to provide a text-based
> browser for use on servers, at a time when the default server installation
> was a strict subset of the desktop.  It is completely superfluous on a
> modern desktop.


No complaints from the Xubuntu crew as I don't hear many of our users using
w3m. Furthermore, it is much appreciated that this was brought up as I think
it would be appropriate to note in the release notes (especially so for
Xubuntu at least).

It also appears there is general consensus for this chance so when should we
expect this to occur (as I'm sure the QA team will want to update their
tests/documentation in time for the next batch of testing)? Final
notification that this chance has indeed occurred would be appreciated.




>
> --
>  - mdz 
>


Cheers,

-- 
Cody A.W. Somerville
Software Engineer
Red Cow Marketing & Technologies, Inc.
Office: 506-458-1290
Toll Free: 1-877-733-2699
Fax: 506-453-9112
Cell: 506-449-5899
Email: [EMAIL PROTECTED]
http://www.redcow.ca
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Neal McBurnett]

On Tue, Jun 17, 2008 at 05:00:06PM +0200, Ante Karamatic wrote:
> On Tue, 17 Jun 2008 09:15:17 -0500
> Rick Clark <[EMAIL PROTECTED]> wrote:
> 
> > I think that Matt meant to remove it from standard, so it would free
> > up space on the desktop cd's, we could still put it in the server
> > seed.
> 
> Right, sorry. I got lost in conversation...

I also like having a text-based web browser on a server, as others
have said, for reading documentation (especially when the server is
offline and you're trying to get it back online) and for local web
admin tasks.

Space constraints don't seem that big an issue on the server CD.
And I'd like to consider continuing having it on the server default
install.

I think the issues are more around security and support and choice.
w3m is the default now, should that change?  Are there significant
security risks to text browsing?  Should elinks take is place (noting
that comment about accessibility recommendations?)  But elinks has
more dependencies, right?  Detailing the fully-installed footprint
delta for each choice would help, and knowing security exposures and
considering features etc.

Neal McBurnett http://mcburnett.org/neal/

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Ante Karamatic]

On Tue, 17 Jun 2008 09:15:17 -0500
Rick Clark <[EMAIL PROTECTED]> wrote:

> I think that Matt meant to remove it from standard, so it would free
> up space on the desktop cd's, we could still put it in the server
> seed.

Right, sorry. I got lost in conversation...

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Matt Zimmerman]

On Mon, Jun 16, 2008 at 09:38:41PM -0700, Steve Beattie wrote:
> On Mon, Jun 16, 2008 at 03:31:46PM +0100, Matt Zimmerman wrote:
> >  * w3m# we need some text-based html presenter
> > 
> > I'd like to formally cast doubt on this statement from the standard seed.
> > This was originally added a long time ago in order to provide a text-based
> > browser for use on servers, at a time when the default server installation
> > was a strict subset of the desktop.  It is completely superfluous on a
> > modern desktop.
> > 
> > Now that this is becoming possible with the new server seed[1], I'd like to
> > propose that it move to the server seed instead (or even be removed, if the
> > server team doesn't feel it's appropriate).
> > 
> > Note that wget, which is much smaller, simpler and more generally useful
> > (e.g. in scripts) is already in standard.
> 
> I have no opinion as to where in the seeds w3m exists, so long as I can
> apt-get install it somehow on my apparently archaic desktop, where it
> is not superfluous at all.

The seed lists determine which packages are included in the 'main'
component, from which the official CDs and other end products are built.
They don't affect which packages are available for download, and I'm
certainly not suggesting that w3m should be removed from Ubuntu.

> But do note that wget, curl et al serve slightly different purposes
> than w3m, lynx, elinks et al in that the latter do interpretation and
> presentation of html, not merely just pulling it from the network. For
> example, I use a text based email client and I use "w3m --dump" in my
> mailcap entry[0] to handle html email, so that I can read and respond to
> the processed output, not the raw html. It's particularly effective for
> handling html email that contains tables in it, especially if you need
> to address in your reply to said email specific elements within tables
> and wish to comment inline.

I use both regularly and understand the difference; again, no one will
prevent you from using w3m regardless of its seed status. :-)

-- 
 - mdz

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Martin Pitt]

Matt Zimmerman [2008-06-16 15:31 +0100]:
>  * w3m# we need some text-based html presenter
> [...]
> Now that this is becoming possible with the new server seed[1], I'd like to
> propose that it move to the server seed instead (or even be removed, if the
> server team doesn't feel it's appropriate).

+1 from me. 

(Also, yay for claiming back 1.1 MB on the desktop CDs!)

Martin

-- 
Martin Pitt| http://www.piware.de
Ubuntu Developer (www.ubuntu.com)  | Debian Developer  (www.debian.org)


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Matt Zimmerman]

On Mon, Jun 16, 2008 at 10:30:01AM -0500, Dustin Kirkland wrote:
> But I do see a distinct difference between w3m/lynx/elinks and
> wget/curl.  The latter are useful for scripting the downloading of
> files/content and then acting on it.  The former are more interactive,
> and allow for web searching, following links, reading pages.

Since you're the second person to respond this way, I must not have been
clear enough in my original email.  I did not mean to say that wget provides
the same functionality as w3m.  I only wanted to remind everyone that, even
if w3m is removed from the default install, there are other tools
which implement the relevant protocols for accessing web content
non-interactively, and I don't think that should change.

> We can debate whether an interactive text browser belongs in the
> default server seed, or if we simply document how to install and use
> it in the Server Guide.  I'd like to see an interactive web browser
> remain on the server, though.

I defer to the server team on this point; my main concern is with removing
it from standard.  Note that this saves a megabyte or so on the CDs, and
every bit helps.

-- 
 - mdz

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Rick Clark]

On Tuesday 17 June 2008 09:07:11 Ante Karamatic wrote:
> On Tue, 17 Jun 2008 10:04:35 -0400
>
> "Brian McKee" <[EMAIL PROTECTED]> wrote:
> > I'm all for removing cruft - but a text based browser is a small
> > package with big benefits to many of us.
>
> I agree with Brian.


I think that Matt meant to remove it from standard, so it would free up space 
on the desktop cd's, we could still put it in the server seed.


signature.asc
Description: This is a digitally signed message part.
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Ante Karamatic]

On Tue, 17 Jun 2008 10:04:35 -0400
"Brian McKee" <[EMAIL PROTECTED]> wrote:

> I'm all for removing cruft - but a text based browser is a small
> package with big benefits to many of us.

I agree with Brian.

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Brian McKee]

On Mon, Jun 16, 2008 at 10:31 AM, Matt Zimmerman <[EMAIL PROTECTED]> wrote:
>  * w3m# we need some text-based html presenter
>
> I'd like to formally cast doubt on this statement from the standard seed.
> This was originally added a long time ago in order to provide a text-based
> browser for use on servers, at a time when the default server installation
> was a strict subset of the desktop.  It is completely superfluous on a
> modern desktop.
>
> Now that this is becoming possible with the new server seed[1], I'd like to
> propose that it move to the server seed instead (or even be removed, if the
> server team doesn't feel it's appropriate).
>
> Note that wget, which is much smaller, simpler and more generally useful
> (e.g. in scripts) is already in standard.
>
> [1] 
> https://wiki.ubuntu.com/UDS-Intrepid/Report/Server#head-27b6797bcec023c8714310edd0947ee48caaa8f2
>


Some form of console browser _is_ required - It's a common tool for me
to manage and configure CUPS for example, or browse html documentation
(and no, I don't want to have to install a server package on a desktop
just to look at the documentation - I might be off-site, or even stuck
using some other OS at the time)

I'm all for removing cruft - but a text based browser is a small
package with big benefits to many of us.

Brian

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Steve Beattie]

On Mon, Jun 16, 2008 at 03:31:46PM +0100, Matt Zimmerman wrote:
>  * w3m# we need some text-based html presenter
> 
> I'd like to formally cast doubt on this statement from the standard seed.
> This was originally added a long time ago in order to provide a text-based
> browser for use on servers, at a time when the default server installation
> was a strict subset of the desktop.  It is completely superfluous on a
> modern desktop.
> 
> Now that this is becoming possible with the new server seed[1], I'd like to
> propose that it move to the server seed instead (or even be removed, if the
> server team doesn't feel it's appropriate).
> 
> Note that wget, which is much smaller, simpler and more generally useful
> (e.g. in scripts) is already in standard.

I have no opinion as to where in the seeds w3m exists, so long as I can
apt-get install it somehow on my apparently archaic desktop, where it
is not superfluous at all.

But do note that wget, curl et al serve slightly different purposes
than w3m, lynx, elinks et al in that the latter do interpretation and
presentation of html, not merely just pulling it from the network. For
example, I use a text based email client and I use "w3m --dump" in my
mailcap entry[0] to handle html email, so that I can read and respond to
the processed output, not the raw html. It's particularly effective for
handling html email that contains tables in it, especially if you need
to address in your reply to said email specific elements within tables
and wish to comment inline.

And that's not even mentioning w3m-img, which lets it display graphical
elements within an X terminal just peachy.

(I picked w3m a couple of years ago because it handled tables and other
formatted html layouts the best of it, lynx, and elinks. But which one
does this best may have changed.)

[0] Sorry, Kees. I don't even run it wrapped in an apparmor profile,
though I know w3m has had a few security issues in the past.

-- 
Steve Beattie
<[EMAIL PROTECTED]>
http://NxNW.org/~steve/


signature.asc
Description: Digital signature
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Chuck Short]

If this does happen then the qa testing guidelines will have to be updated.

chuck

Dustin Kirkland wrote:
> On Mon, Jun 16, 2008 at 1:08 PM, James Dinkel <[EMAIL PROTECTED]> wrote:
>   
>> I would read the documentation on another system then, assuming the
>> system referred to is a production server and not a test server.  Even
>> it there is a viable need for an interactive web browser, I think the
>> use case would be so rare and added to the fact that people will have
>> differing preferences on which text browser is best, that including
>> any one in the default seed is a bad idea.
>> 
>
> This sounds reasonable to me.  As mentioned above, I think it can be
> handled in the Ubuntu Server Guide document with a list of viable text
> browser alternatives and a pointer to the apt-get command.
>
>
> :-Dustin
>
>   


-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

RE: Moving w3m out of standard

[Miano, Steven M.]

>> I think an interactive web browser is completely unnecessary on a
>> server.  You should do your browsing from a workstation, and then use
>> wget or sftp to transfer any files you need to the server.
>
> There are cases where the documentation that is packaged with typical
> server applications only comes as HTML (e.g. Wordpress). You won't get
> far with 'less' here, I always installed lynx for these cases.

I would have to whole heartedly agree on less is more when it comes to servers. 
The vast majority of server applications will not require an interactive web 
browser. If one really needed an interactive web browser they would surely 
already have a net connection and be able to use aptitude to add additional 
packages as they wish.

Just my .02,

~Steve

CONFIDENTIALITY NOTICE: This e-mail may contain information that is privileged, 
confidential or otherwise protected from disclosure. If you are not the 
intended recipient of this e-mail, please notify the sender immediately by 
return e-mail, purge it and do not disseminate or copy it.

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Dustin Kirkland]

On Mon, Jun 16, 2008 at 1:08 PM, James Dinkel <[EMAIL PROTECTED]> wrote:
> I would read the documentation on another system then, assuming the
> system referred to is a production server and not a test server.  Even
> it there is a viable need for an interactive web browser, I think the
> use case would be so rare and added to the fact that people will have
> differing preferences on which text browser is best, that including
> any one in the default seed is a bad idea.

This sounds reasonable to me.  As mentioned above, I think it can be
handled in the Ubuntu Server Guide document with a list of viable text
browser alternatives and a pointer to the apt-get command.


:-Dustin

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

On Mon, Jun 16, 2008 at 12:31 PM, Andreas Schildbach
<[EMAIL PROTECTED]> wrote:
>
>> I think an interactive web browser is completely unnecessary on a
>> server.  You should do your browsing from a workstation, and then use
>> wget or sftp to transfer any files you need to the server.
>
> There are cases where the documentation that is packaged with typical
> server applications only comes as HTML (e.g. Wordpress). You won't get
> far with 'less' here, I always installed lynx for these cases.
>

I would read the documentation on another system then, assuming the
system referred to is a production server and not a test server.  Even
it there is a viable need for an interactive web browser, I think the
use case would be so rare and added to the fact that people will have
differing preferences on which text browser is best, that including
any one in the default seed is a bad idea.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Michael Hipp]

James Dinkel wrote:

> I think an interactive web browser is completely unnecessary on a
> server.  You should do your browsing from a workstation, and then use
> wget or sftp to transfer any files you need to the server.

Well said.

And w3m or similar tools are only one command away with apt-get if you do need 
them. Let's keep the "clutter" to a minimum.

Michael

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Andreas Schildbach]

On Mon, 2008-06-16 at 11:17 -0500, James Dinkel wrote:

> I think an interactive web browser is completely unnecessary on a
> server.  You should do your browsing from a workstation, and then use
> wget or sftp to transfer any files you need to the server.

There are cases where the documentation that is packaged with typical
server applications only comes as HTML (e.g. Wordpress). You won't get
far with 'less' here, I always installed lynx for these cases.

Best regards,

Andreas




-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[James Dinkel]

On Mon, Jun 16, 2008 at 10:30 AM, Dustin Kirkland
<[EMAIL PROTECTED]> wrote:
>
> Matt, et al.-
>
> I agree 100% that w3m is completely unnecessary on a modern Ubuntu
> desktop machine.  No argument there.
>
> On a server without X, however, I think that some form of an
> interactive text web browser is still useful.  I know of 3 in Ubuntu
> main, (w3m, lynx, elinks).  I can't speak for the state of upstream
> development of any of those.  They may be other alternatives.
>
> But I do see a distinct difference between w3m/lynx/elinks and
> wget/curl.  The latter are useful for scripting the downloading of
> files/content and then acting on it.  The former are more interactive,
> and allow for web searching, following links, reading pages.
>
> We can debate whether an interactive text browser belongs in the
> default server seed, or if we simply document how to install and use
> it in the Server Guide.  I'd like to see an interactive web browser
> remain on the server, though.
>
>
> :-Dustin
>

I think an interactive web browser is completely unnecessary on a
server.  You should do your browsing from a workstation, and then use
wget or sftp to transfer any files you need to the server.

James

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Dustin Kirkland]

On Mon, Jun 16, 2008 at 9:31 AM, Matt Zimmerman <[EMAIL PROTECTED]> wrote:
>  * w3m# we need some text-based html presenter
>
> I'd like to formally cast doubt on this statement from the standard seed.
> This was originally added a long time ago in order to provide a text-based
> browser for use on servers, at a time when the default server installation
> was a strict subset of the desktop.  It is completely superfluous on a
> modern desktop.
>
> Now that this is becoming possible with the new server seed[1], I'd like to
> propose that it move to the server seed instead (or even be removed, if the
> server team doesn't feel it's appropriate).
>
> Note that wget, which is much smaller, simpler and more generally useful
> (e.g. in scripts) is already in standard.

Matt, et al.-

I agree 100% that w3m is completely unnecessary on a modern Ubuntu
desktop machine.  No argument there.

On a server without X, however, I think that some form of an
interactive text web browser is still useful.  I know of 3 in Ubuntu
main, (w3m, lynx, elinks).  I can't speak for the state of upstream
development of any of those.  They may be other alternatives.

But I do see a distinct difference between w3m/lynx/elinks and
wget/curl.  The latter are useful for scripting the downloading of
files/content and then acting on it.  The former are more interactive,
and allow for web searching, following links, reading pages.

We can debate whether an interactive text browser belongs in the
default server seed, or if we simply document how to install and use
it in the Server Guide.  I'd like to see an interactive web browser
remain on the server, though.


:-Dustin

-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam

Re: Moving w3m out of standard

[Steven Harms]

This is a great idea, as almost all administration tasks using w3m are
generally accomplished by wget.  Also I believe the upstream is either dead
or near dead on this one, I was unable to get any response on bugs.

- sharms

On Mon, Jun 16, 2008 at 10:31 AM, Matt Zimmerman <[EMAIL PROTECTED]> wrote:

>  * w3m# we need some text-based html presenter
>
> I'd like to formally cast doubt on this statement from the standard seed.
> This was originally added a long time ago in order to provide a text-based
> browser for use on servers, at a time when the default server installation
> was a strict subset of the desktop.  It is completely superfluous on a
> modern desktop.
>
> Now that this is becoming possible with the new server seed[1], I'd like to
> propose that it move to the server seed instead (or even be removed, if the
> server team doesn't feel it's appropriate).
>
> Note that wget, which is much smaller, simpler and more generally useful
> (e.g. in scripts) is already in standard.
>
> [1]
> https://wiki.ubuntu.com/UDS-Intrepid/Report/Server#head-27b6797bcec023c8714310edd0947ee48caaa8f2
>
> --
>  - mdz
>
> --
> ubuntu-devel mailing list
> [EMAIL PROTECTED]
> Modify settings or unsubscribe at:
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
>



-- 
GPG Key ID: C92EF367 / 1428 FE8E 1E07 DDA8 EFD7 195F DCCD F5B3 C92E F367
-- 
ubuntu-server mailing list
ubuntu-server@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server
More info: https://wiki.ubuntu.com/ServerTeam