Re: Higher cost of Linux system administrators?

On 2008-10-02, Hadron <hadronquark@gmail.com> wrote:[color=blue]
> Ignoramus31561 <ignoramus31561@NOSPAM.31561.invalid> writes:
>[color=green]
>> On 2008-10-02, RonB <ronb02nospam@gmail.com> wrote:[color=darkred]
>>> On Wed, 01 Oct 2008 12:28:40 -0700, Tim Smith wrote:
>>>
>>>> People who can administer two systems make more than people who can only
>>>> do one. Since most Linux administrators fall into this category, the
>>>> average for "Linux administrator" is high.
>>>
>>> Bullcrap. Most shops have had UNIX servers for decades and administering
>>> Linux servers is almost identical to administering UNIX servers.
>>>[/color]
>>
>> Here's how I would install a package on all our Linux servers:
>>
>> for i in $ALL_LINUX_SERVERS; do
>> ssh -t -t -l root $i aptitude -y install wantedpackage
>> done
>>
>> Takes 30 seconds of my time regardless of the number of servers.
>> (though with many servers, it may have to run for a while, usually
>> unattended).
>>
>> Try this on windows. Good luck.[/color]
>
> You probably would not want to.
>
> I think a nicer solution is to either ssh/sync a package file (list of
> packages in it) or to allow the destination machines to pick it up
> themsleves using rsync on a machine specific cron job.
>
> Forcing an update onto remote machines is naive unless you really,
> really now that machine is (a) up and (b) in a position where installing
> SW is a good idea.[/color]

Ubuntu is extremely friendly to doing this, actually, updates in
have never screwed up any running apps. I have been doing this for a
while and had a good luck with it. If some machines are not up, the
same thing can be safey re-run again, the machines that already have
wantedpackage will do nothing.
--
Due to extreme spam originating from Google Groups, and their inattention
to spammers, I and many others block all articles originating
from Google Groups. If you want your postings to be seen by
more readers you will need to find a different means of
posting on Usenet.
[url]http://improve-usenet.org/[/url]

Ignoramus23721 <ignoramus23721@NOSPAM.23721.invalid> writes:
[color=blue]
> On 2008-10-02, Hadron <hadronquark@gmail.com> wrote:[color=green]
>> Ignoramus31561 <ignoramus31561@NOSPAM.31561.invalid> writes:
>>[color=darkred]
>>> On 2008-10-02, RonB <ronb02nospam@gmail.com> wrote:
>>>> On Wed, 01 Oct 2008 12:28:40 -0700, Tim Smith wrote:
>>>>
>>>>> People who can administer two systems make more than people who can only
>>>>> do one. Since most Linux administrators fall into this category, the
>>>>> average for "Linux administrator" is high.
>>>>
>>>> Bullcrap. Most shops have had UNIX servers for decades and administering
>>>> Linux servers is almost identical to administering UNIX servers.
>>>>
>>>
>>> Here's how I would install a package on all our Linux servers:
>>>
>>> for i in $ALL_LINUX_SERVERS; do
>>> ssh -t -t -l root $i aptitude -y install wantedpackage
>>> done
>>>
>>> Takes 30 seconds of my time regardless of the number of servers.
>>> (though with many servers, it may have to run for a while, usually
>>> unattended).
>>>
>>> Try this on windows. Good luck.[/color]
>>
>> You probably would not want to.
>>
>> I think a nicer solution is to either ssh/sync a package file (list of
>> packages in it) or to allow the destination machines to pick it up
>> themsleves using rsync on a machine specific cron job.
>>
>> Forcing an update onto remote machines is naive unless you really,
>> really now that machine is (a) up and (b) in a position where installing
>> SW is a good idea.[/color]
>
> Ubuntu is extremely friendly to doing this, actually, updates in
> have never screwed up any running apps. I have been doing this for a
> while and had a good luck with it. If some machines are not up, the
> same thing can be safey re-run again, the machines that already have
> wantedpackage will do nothing.[/color]

Just think about cron'ing the other machines to get it when they are
ready and are running. No work from you at all. Just a suggestion...

--
"Your Ref header shows bt.com. The "kustomkomputer" troll nymshifted again?"
-- William Poaster <wp@leafnode.amd64.eu> boring people to death with his header compulsion in comp.os.linux.advocacy

"Ignoramus23721" <ignoramus23721@NOSPAM.23721.invalid> wrote in message
news:A66dnWAdzMheUXnVnZ2dnUVZ_o_inZ2d@giganews.com...[color=blue]
> On 2008-10-02, Hadron <hadronquark@gmail.com> wrote:[color=green]
>> Ignoramus31561 <ignoramus31561@NOSPAM.31561.invalid> writes:
>>[color=darkred]
>>> On 2008-10-02, RonB <ronb02nospam@gmail.com> wrote:
>>>> On Wed, 01 Oct 2008 12:28:40 -0700, Tim Smith wrote:
>>>>
>>>>> People who can administer two systems make more than people who can
>>>>> only
>>>>> do one. Since most Linux administrators fall into this category, the
>>>>> average for "Linux administrator" is high.
>>>>
>>>> Bullcrap. Most shops have had UNIX servers for decades and
>>>> administering
>>>> Linux servers is almost identical to administering UNIX servers.
>>>>
>>>
>>> Here's how I would install a package on all our Linux servers:
>>>
>>> for i in $ALL_LINUX_SERVERS; do
>>> ssh -t -t -l root $i aptitude -y install wantedpackage
>>> done
>>>
>>> Takes 30 seconds of my time regardless of the number of servers.
>>> (though with many servers, it may have to run for a while, usually
>>> unattended).
>>>
>>> Try this on windows. Good luck.[/color]
>>
>> You probably would not want to.
>>
>> I think a nicer solution is to either ssh/sync a package file (list of
>> packages in it) or to allow the destination machines to pick it up
>> themsleves using rsync on a machine specific cron job.
>>
>> Forcing an update onto remote machines is naive unless you really,
>> really now that machine is (a) up and (b) in a position where installing
>> SW is a good idea.[/color]
>
> Ubuntu is extremely friendly to doing this, actually, updates in
> have never screwed up any running apps. I have been doing this for a
> while and had a good luck with it. If some machines are not up, the
> same thing can be safey re-run again, the machines that already have
> wantedpackage will do nothing.[/color]


So you think that killing, say, the web server or the database engine while
in the middle of a customer transaction isn't a problem? There are quite
significant differences in how you can admin a server and a workstation
unless you want to be unpopular.

"Chris Ahlstrom" <linonut@bollsouth.nut> wrote in message
news:dX3Fk.43671$vX2.39646@bignews6.bellsouth.net...[color=blue]
> After takin' a swig o' grog, dennis@home belched out
> this bit o' wisdom:
>[color=green][color=darkred]
>>>> are too scared to actually even learn the basics about windows admin.
>>> ^^^^^^
>>>
>>> guffaw[/color]
>>
>> I take it you can only find spell chucker errors to argue about?[/color]
>
> It's not a spell "chucker" (good pun by the way) error. It's a
> deep-structure error ;->
>[/color]

I thought it was quite a good substitution when it was pointed out, maybe a
Freudian slip?[color=blue]
> --
> List at least two alternate dates.[/color]

On 2008-10-02, Chris Ahlstrom <linonut@bollsouth.nut> wrote:[color=blue]
> After takin' a swig o' grog, dennis@home belched out
> this bit o' wisdom:
>[color=green]
>> "Ignoramus31561" <ignoramus31561@NOSPAM.31561.invalid> wrote in message[color=darkred]
>>>
>>> Here's how I would install a package on all our Linux servers:
>>>
>>> for i in $ALL_LINUX_SERVERS; do
>>> ssh -t -t -l root $i aptitude -y install wantedpackage
>>> done
>>>
>>> Takes 30 seconds of my time regardless of the number of servers.
>>> (though with many servers, it may have to run for a while, usually
>>> unattended).
>>>
>>> Try this on windows. Good luck.[/color]
>>
>> Well you wouldn't do it like that under windows.
>> You would load the SW onto a server and then let the windows machines
>> automatically download and update.[/color]
>
> How do you "let" the Windows machines do that?[/color]

Group Policy and WSUS.

--
Tom Shelton

dennis@home wrote:[color=blue]
>
> So you think that killing, say, the web server or the database engine while
> in the middle of a customer transaction isn't a problem? There are quite
> significant differences in how you can admin a server and a workstation
> unless you want to be unpopular.
>[/color]
Why would updating a program be a problem in these circumstances?
Replacing a program is not done by killing a process and restarting it.

Linux and UNIX have been able to manage this "forever."

What happens in this case is that the existing process continues to run. All
new instances of the process use the new version.

At one time, a hard link was made to the existing program and then the
original name of it was removed. That way, no one could start the old one.
But the old process could continue to run. I do not know if that is how they
do it now, but the effect is the same. My system is up 24/7 except when I
reboot it for a new kernel, something I can do whenever I choose; or when I
power it off because the room gets too hot in the summer and it risks
overheating.

--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ PGP-Key: 9A2FC99A Registered Machine 241939.
/( )\ Shrewsbury, New Jersey [url]http://counter.li.org[/url]
^^-^^ 11:45:01 up 7 days, 12:47, 3 users, load average: 4.18, 4.17, 4.27

dennis@home wrote:
[color=blue]
>
>
> "Ignoramus23721" <ignoramus23721@NOSPAM.23721.invalid> wrote in message
> news:A66dnWAdzMheUXnVnZ2dnUVZ_o_inZ2d@giganews.com...[color=green]
>> On 2008-10-02, Hadron <hadronquark@gmail.com> wrote:[color=darkred]
>>> Ignoramus31561 <ignoramus31561@NOSPAM.31561.invalid> writes:
>>>
>>>> On 2008-10-02, RonB <ronb02nospam@gmail.com> wrote:
>>>>> On Wed, 01 Oct 2008 12:28:40 -0700, Tim Smith wrote:
>>>>>
>>>>>> People who can administer two systems make more than people who can
>>>>>> only
>>>>>> do one. Since most Linux administrators fall into this category, the
>>>>>> average for "Linux administrator" is high.
>>>>>
>>>>> Bullcrap. Most shops have had UNIX servers for decades and
>>>>> administering
>>>>> Linux servers is almost identical to administering UNIX servers.
>>>>>
>>>>
>>>> Here's how I would install a package on all our Linux servers:
>>>>
>>>> for i in $ALL_LINUX_SERVERS; do
>>>> ssh -t -t -l root $i aptitude -y install wantedpackage
>>>> done
>>>>
>>>> Takes 30 seconds of my time regardless of the number of servers.
>>>> (though with many servers, it may have to run for a while, usually
>>>> unattended).
>>>>
>>>> Try this on windows. Good luck.
>>>
>>> You probably would not want to.
>>>
>>> I think a nicer solution is to either ssh/sync a package file (list of
>>> packages in it) or to allow the destination machines to pick it up
>>> themsleves using rsync on a machine specific cron job.
>>>
>>> Forcing an update onto remote machines is naive unless you really,
>>> really now that machine is (a) up and (b) in a position where installing
>>> SW is a good idea.[/color]
>>
>> Ubuntu is extremely friendly to doing this, actually, updates in
>> have never screwed up any running apps. I have been doing this for a
>> while and had a good luck with it. If some machines are not up, the
>> same thing can be safey re-run again, the machines that already have
>> wantedpackage will do nothing.[/color]
>
>
> So you think that killing, say, the web server or the database engine
> while in the middle of a customer transaction isn't a problem? There are
> quite significant differences in how you can admin a server and a
> workstation unless you want to be unpopular.[/color]

Why do you think that the application *needs* to be killed when updated?
Since you obviously know nothing about linux, here is a hint: You can update
running apps while they run. You can restart them later, after the update
has taken place, at any time. It is not as with windows where you can't
replace running apps/dlls without stopping them before.
--
We are Linux. Resistance is measured in Ohms.

dennis@home wrote:[color=blue]
>
>
> "Ignoramus23721" <ignoramus23721@NOSPAM.23721.invalid> wrote in message
> news:A66dnWAdzMheUXnVnZ2dnUVZ_o_inZ2d@giganews.com...[color=green]
>> On 2008-10-02, Hadron <hadronquark@gmail.com> wrote:[color=darkred]
>>> Ignoramus31561 <ignoramus31561@NOSPAM.31561.invalid> writes:
>>>
>>>> On 2008-10-02, RonB <ronb02nospam@gmail.com> wrote:
>>>>> On Wed, 01 Oct 2008 12:28:40 -0700, Tim Smith wrote:
>>>>>
>>>>>> People who can administer two systems make more than people who
>>>>>> can only
>>>>>> do one. Since most Linux administrators fall into this category, the
>>>>>> average for "Linux administrator" is high.
>>>>>
>>>>> Bullcrap. Most shops have had UNIX servers for decades and
>>>>> administering
>>>>> Linux servers is almost identical to administering UNIX servers.
>>>>>
>>>>
>>>> Here's how I would install a package on all our Linux servers:
>>>>
>>>> for i in $ALL_LINUX_SERVERS; do
>>>> ssh -t -t -l root $i aptitude -y install wantedpackage
>>>> done
>>>>
>>>> Takes 30 seconds of my time regardless of the number of servers.
>>>> (though with many servers, it may have to run for a while, usually
>>>> unattended).
>>>>
>>>> Try this on windows. Good luck.
>>>
>>> You probably would not want to.
>>>
>>> I think a nicer solution is to either ssh/sync a package file (list of
>>> packages in it) or to allow the destination machines to pick it up
>>> themsleves using rsync on a machine specific cron job.
>>>
>>> Forcing an update onto remote machines is naive unless you really,
>>> really now that machine is (a) up and (b) in a position where installing
>>> SW is a good idea.[/color]
>>
>> Ubuntu is extremely friendly to doing this, actually, updates in
>> have never screwed up any running apps. I have been doing this for a
>> while and had a good luck with it. If some machines are not up, the
>> same thing can be safey re-run again, the machines that already have
>> wantedpackage will do nothing.[/color]
>
>
> So you think that killing, say, the web server or the database engine
> while in the middle of a customer transaction isn't a problem? There are
> quite significant differences in how you can admin a server and a
> workstation unless you want to be unpopular.[/color]

Well thats why there are kill levels. to allow things to shut down
gracefully.

"The Natural Philosopher" <a@b.c> wrote in message
news:1222963476.27441.2@proxy00.news.clara.net...
[color=blue][color=green]
>> So you think that killing, say, the web server or the database engine
>> while in the middle of a customer transaction isn't a problem? There are
>> quite significant differences in how you can admin a server and a
>> workstation unless you want to be unpopular.[/color]
>
> Well thats why there are kill levels. to allow things to shut down
> gracefully.[/color]

You know that and I know that but Peter probably doesn't.
He thinks an update has been done even though the old code is still running.

I wonder if I can find an application that writes its config out when it
exits so it overwrites the new one installed by the update?
Nah, nobody would ever do something that stupid. ;-)

dennis@home wrote:[color=blue]
>
>
> "The Natural Philosopher" <a@b.c> wrote in message
> news:1222963476.27441.2@proxy00.news.clara.net...
>[color=green][color=darkred]
>>> So you think that killing, say, the web server or the database engine
>>> while in the middle of a customer transaction isn't a problem? There
>>> are quite significant differences in how you can admin a server and a
>>> workstation unless you want to be unpopular.[/color]
>>
>> Well thats why there are kill levels. to allow things to shut down
>> gracefully.[/color]
>
> You know that and I know that but Peter probably doesn't.
> He thinks an update has been done even though the old code is still
> running.
>[/color]

It all depends on what you mean by update.

If the program is a child process called by a parent, then when it exist
gracefully of its own accord, the next instance will be the update.

If its a forked copy, then it wont.

If its a complex daemon, you rewrite te start script so that restart
does the correct thing.

The power of *nix is that you CAN rewrite it.

[color=blue]
> I wonder if I can find an application that writes its config out when it
> exits so it overwrites the new one installed by the update?
> Nah, nobody would ever do something that stupid. ;-)[/color]

Microsoft probably would.

It would be considered a top priority bug in Linux.

On 2008-10-02, Jean-David Beyer <jeandavid8@verizon.net> wrote:[color=blue]
> dennis@home wrote:[color=green]
>>
>> So you think that killing, say, the web server or the database engine while
>> in the middle of a customer transaction isn't a problem? There are quite
>> significant differences in how you can admin a server and a workstation
>> unless you want to be unpopular.
>>[/color]
> Why would updating a program be a problem in these circumstances?
> Replacing a program is not done by killing a process and restarting it.
>
> Linux and UNIX have been able to manage this "forever."
>
> What happens in this case is that the existing process continues to run. All
> new instances of the process use the new version.
>
> At one time, a hard link was made to the existing program and then the
> original name of it was removed. That way, no one could start the old one.
> But the old process could continue to run. I do not know if that is how they
> do it now, but the effect is the same. My system is up 24/7 except when I
> reboot it for a new kernel, something I can do whenever I choose; or when I
> power it off because the room gets too hot in the summer and it risks
> overheating.
>[/color]

correct. Most things that are patched, are not actually having their
instances killed. And you CAN remove a file with a running program. It
continues to exist until the program exits.

--
Due to extreme spam originating from Google Groups, and their inattention
to spammers, I and many others block all articles originating
from Google Groups. If you want your postings to be seen by
more readers you will need to find a different means of
posting on Usenet.
[url]http://improve-usenet.org/[/url]

After takin' a swig o' grog, Tom Shelton belched out
this bit o' wisdom:
[color=blue]
> On 2008-10-02, Chris Ahlstrom <linonut@bollsouth.nut> wrote:[color=green][color=darkred]
>>>
>>> You would load the SW onto a server and then let the windows machines
>>> automatically download and update.[/color]
>>
>> How do you "let" the Windows machines do that?[/color]
>
> Group Policy and WSUS.[/color]

Ah, thanks Tom! (As you can tell, I don't do much Windows maintenance,
DFS's demented shriekings to the contrary.)

--
When Dexter's on the Internet, can Hell be far behind?"

"The Natural Philosopher" <a@b.c> wrote in message
news:1222964798.29655.0@proxy00.news.clara.net...[color=blue]
> dennis@home wrote:[color=green]
>>
>>
>> "The Natural Philosopher" <a@b.c> wrote in message
>> news:1222963476.27441.2@proxy00.news.clara.net...
>>[color=darkred]
>>>> So you think that killing, say, the web server or the database engine
>>>> while in the middle of a customer transaction isn't a problem? There
>>>> are quite significant differences in how you can admin a server and a
>>>> workstation unless you want to be unpopular.
>>>
>>> Well thats why there are kill levels. to allow things to shut down
>>> gracefully.[/color]
>>
>> You know that and I know that but Peter probably doesn't.
>> He thinks an update has been done even though the old code is still
>> running.
>>[/color]
>
> It all depends on what you mean by update.
>
> If the program is a child process called by a parent, then when it exist
> gracefully of its own accord, the next instance will be the update.
>
> If its a forked copy, then it wont.
>
> If its a complex daemon, you rewrite te start script so that restart does
> the correct thing.
>
> The power of *nix is that you CAN rewrite it.
>
>[color=green]
>> I wonder if I can find an application that writes its config out when it
>> exits so it overwrites the new one installed by the update?
>> Nah, nobody would ever do something that stupid. ;-)[/color]
>
> Microsoft probably would.
>
> It would be considered a top priority bug in Linux.
>[/color]

The bug reports have several instances.

The Natural Philosopher wrote:
[color=blue]
> It all depends on what you mean by update.
>
> If the program is a child process called by a parent, then when it exist
> gracefully of its own accord, the next instance will be the update.
>
> If its a forked copy, then it wont.
>[/color]
I do not understand the distinction you are making between a "called" child
process and a "forked" process, since they all get started by one of the
fork(2) functions.


--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ PGP-Key: 9A2FC99A Registered Machine 241939.
/( )\ Shrewsbury, New Jersey [url]http://counter.li.org[/url]
^^-^^ 13:30:01 up 7 days, 14:32, 4 users, load average: 4.24, 4.17, 4.35

Jean-David Beyer wrote:[color=blue]
> The Natural Philosopher wrote:
>[color=green]
>> It all depends on what you mean by update.
>>
>> If the program is a child process called by a parent, then when it exist
>> gracefully of its own accord, the next instance will be the update.
>>
>> If its a forked copy, then it wont.
>>[/color]
> I do not understand the distinction you are making between a "called"
> child process and a "forked" process, since they all get started by one
> of the fork(2) functions.
>
>[/color]
I suggest you look at Inetd, for something to prove you wrong.

dennis@home wrote:[color=blue]
> So you think that killing, say, the web server or the database engine while
> in the middle of a customer transaction isn't a problem? There are quite
> significant differences in how you can admin a server and a workstation
> unless you want to be unpopular.[/color]

Jean-David Beyer <jeandavid8@verizon.net> wrote:[color=blue]
> Why would updating a program be a problem in these circumstances?
> Replacing a program is not done by killing a process and restarting it.[/color]

Unfortunately Debian and its derivatives seem trigger happy about
stopping a service before upgrading it. I can't speak for the Redhat
rpm-based distributions, though; can anyone else contribute?

Chris

On 2008-10-02, Chris Davies <chris-usenet@roaima.co.uk> wrote:[color=blue]
> dennis@home wrote:[color=green]
>> So you think that killing, say, the web server or the database engine while
>> in the middle of a customer transaction isn't a problem? There are quite
>> significant differences in how you can admin a server and a workstation
>> unless you want to be unpopular.[/color]
>
> Jean-David Beyer <jeandavid8@verizon.net> wrote:[color=green]
>> Why would updating a program be a problem in these circumstances?
>> Replacing a program is not done by killing a process and restarting it.[/color]
>
> Unfortunately Debian and its derivatives seem trigger happy about
> stopping a service before upgrading it. I can't speak for the Redhat
> rpm-based distributions, though; can anyone else contribute?[/color]

I would expect that this is entirely a matter of the tastes
of the person building the relevant packages.

--
My macintosh runs Ubuntu. |||
/ | \

Posted Via Usenet.com Premium Usenet Newsgroup Services
----------------------------------------------------------
[url]http://www.usenet.com[/url]

dennis@home wrote:[color=blue]
>[color=green][color=darkred]
>>>>
>>>> Try this on windows. Good luck.
>>>
>>> Well you wouldn't do it like that under windows.
>>> You would load the SW onto a server and then let the windows machines
>>> automatically download and update.[/color]
>>
>> How do you "let" the Windows machines do that?[/color]
>
> That is the difference between knowing about something and chucking out
> cr@p statements like some in this group.
>[color=green][color=darkred]
>>> I guess its the difference between someone that knows how to design and
>>> manage a few hundred machines and some cowboy that manages four or five
>>> machines.. one needs to know how to do things the proper way, the[/color][/color][/color]

Get off the dime, wintard, Do you know how to do the update
or not?

I'll believe it when you can write it out in 20 steps or less.


technomaNge
--
The operat~1 system I use at work
unders~1 long filena~1,
at home I use Linux.

dennis@home wrote:
[color=blue]
>
>
> "technomaNge" <piracy@microsoft.com> wrote in message
> news:HUeFk.41356$XT1.28853@bignews5.bellsouth.net...[color=green]
>> dennis@home wrote:[color=darkred]
>>>
>>>>>>
>>>>>> Try this on windows. Good luck.
>>>>>
>>>>> Well you wouldn't do it like that under windows.
>>>>> You would load the SW onto a server and then let the windows machines
>>>>> automatically download and update.
>>>>
>>>> How do you "let" the Windows machines do that?
>>>
>>> That is the difference between knowing about something and chucking out
>>> cr@p statements like some in this group.
>>>
>>>>> I guess its the difference between someone that knows how to design
>>>>> and manage a few hundred machines and some cowboy that manages four or
>>>>> five machines.. one needs to know how to do things the proper way, the[/color]
>>
>> Get off the dime, wintard, Do you know how to do the update
>> or not?
>>
>> I'll believe it when you can write it out in 20 steps or less.[/color]
>
> What do I care what you believe?
> It can be done as others already know.[/color]

It might be. But you certainly don't know /how/
Otherwise you would not evade the question
[color=blue]
> If you want to know you can ask one of those "really stupid" MCSE types to
> educate you.[/color]

Why would he want to know how to run minesweeper? Those MCSE types are not
much better than school dropouts. 99% of them are simply oxygen thiefs
[color=blue]
> You won't get any windows help in a ubuntu group from me, especially as
> you don't really want it.[/color]

Well, I don't think that was what he wanted.
You should provide the outlines of how "simple" (or rather, complicated) the
windows way is, compared to the linux way.

You failed
--
Microsoft software doesn't get released - it escapes, leaving
a trail of destruction behind it.

dennis@home wrote:[color=blue]
>
>
> "technomaNge" <piracy@microsoft.com> wrote in message
> news:HUeFk.41356$XT1.28853@bignews5.bellsouth.net...[color=green]
>> dennis@home wrote:[color=darkred]
>>>
>>>>>>
>>>>>> Try this on windows. Good luck.
>>>>>
>>>>> Well you wouldn't do it like that under windows.
>>>>> You would load the SW onto a server and then let the windows machines
>>>>> automatically download and update.
>>>>
>>>> How do you "let" the Windows machines do that?
>>>
>>> That is the difference between knowing about something and chucking
>>> out cr@p statements like some in this group.
>>>
>>>>> I guess its the difference between someone that knows how to design
>>>>> and
>>>>> manage a few hundred machines and some cowboy that manages four or
>>>>> five
>>>>> machines.. one needs to know how to do things the proper way, the[/color]
>>
>> Get off the dime, wintard, Do you know how to do the update
>> or not?
>>
>> I'll believe it when you can write it out in 20 steps or less.[/color]
>
> What do I care what you believe?
> It can be done as others already know.
> If you want to know you can ask one of those "really stupid" MCSE types
> to educate you.[/color]

Hoity Toity.
[color=blue]
>
> You won't get any windows help in a ubuntu group from me, especially as
> you don't really want it.[/color]

sure thats not because you are talking out yer anus?

seems equally plausible to me.