info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
OpenSuse -vs- Centos/RHEL
39 views
Skip to first unread message
MegaWattz
Feb 21, 2021, 10:06:11 PM2/21/21
to
RedIBM is in the process of screwing over Centos users. A lot
of people rely on Centos to affordably run big server setups,
clusters or distributed clusters sometimes with distributed
storage options.
A lot more use it for more simple, "conventional", servers for
web or mail or NAS applications.
But now it is going away, to be replaced with glitchy alpha code.
Forget "reliable" ... not even suitable for Granny's home box.
A few are stepping up to create various "Neo-Centos" distros
but they will not have the cooperation of RedIBM and, except
for one, they're all still pretty much alpha. Could be a year or
more before they're up and running right and can prove their
mettle.
Which leaves us looking for alternatives. Plain-Jane Debian can
be good for almost anything, but it is not aimed straight at the
needs of biz or bigger-biz. Ubuntu is about the same. This means
highly customized installs which are often hostage to the ONE
tekkie, who glued it all together with Algol-68 and "B" just
for kicks, understands how all the bits work together. Better pay
Real Real Good, or else ....
On the RPM side, we have OpenSuse and Tumbleweed. Of the
two Tumbleweed has one of the prized features of Centos - an
extremely long lifetime. In theory it could be updated forever
without a single dreaded from-scratch re-install.
SUSE is a sort of "enterprise Linux" as is RHEL. This means it
SHOULD be easier to score tested code/apps to do some of
those more complicated biz things. OSTW is very easy to work
with too - YAST is actually *helpful* and can seriously slash
the time/error equations for setting up systems and software.
OSTW is also fairly compatible ... you can use a lot of RHEL
and Fedora RPMs straight out of the box, so to speak.
So - what is the general feeling on using Tumbleweed as a Centos
replacement ? Rather than trying to rebuild Centos from scratch
(and likely flack from RedIBM from time to time) might it be better
to add a LITTLE more sophistication to OSTW instead ??? This
is becoming a real time/money/reliability issue now.
If Granny just wants to surf FB and Zoom the grandkids on her
garage-sale box then most anything will serve nicely - even Antix
is more than enough and MX is pure luxury. But when you have
an office full of people, or officeS, that need to be tied together,
where downtime is unacceptable, well, we need some good backup
plans PDQ. There is a gap between needing a free OS and being
able to pay for RHEL that a LOT of businesses and orgs fall into.
of people rely on Centos to affordably run big server setups,
clusters or distributed clusters sometimes with distributed
storage options.
A lot more use it for more simple, "conventional", servers for
web or mail or NAS applications.
But now it is going away, to be replaced with glitchy alpha code.
Forget "reliable" ... not even suitable for Granny's home box.
A few are stepping up to create various "Neo-Centos" distros
but they will not have the cooperation of RedIBM and, except
for one, they're all still pretty much alpha. Could be a year or
more before they're up and running right and can prove their
mettle.
Which leaves us looking for alternatives. Plain-Jane Debian can
be good for almost anything, but it is not aimed straight at the
needs of biz or bigger-biz. Ubuntu is about the same. This means
highly customized installs which are often hostage to the ONE
tekkie, who glued it all together with Algol-68 and "B" just
for kicks, understands how all the bits work together. Better pay
Real Real Good, or else ....
On the RPM side, we have OpenSuse and Tumbleweed. Of the
two Tumbleweed has one of the prized features of Centos - an
extremely long lifetime. In theory it could be updated forever
without a single dreaded from-scratch re-install.
SUSE is a sort of "enterprise Linux" as is RHEL. This means it
SHOULD be easier to score tested code/apps to do some of
those more complicated biz things. OSTW is very easy to work
with too - YAST is actually *helpful* and can seriously slash
the time/error equations for setting up systems and software.
OSTW is also fairly compatible ... you can use a lot of RHEL
and Fedora RPMs straight out of the box, so to speak.
So - what is the general feeling on using Tumbleweed as a Centos
replacement ? Rather than trying to rebuild Centos from scratch
(and likely flack from RedIBM from time to time) might it be better
to add a LITTLE more sophistication to OSTW instead ??? This
is becoming a real time/money/reliability issue now.
If Granny just wants to surf FB and Zoom the grandkids on her
garage-sale box then most anything will serve nicely - even Antix
is more than enough and MX is pure luxury. But when you have
an office full of people, or officeS, that need to be tied together,
where downtime is unacceptable, well, we need some good backup
plans PDQ. There is a gap between needing a free OS and being
able to pay for RHEL that a LOT of businesses and orgs fall into.
stepore
Feb 22, 2021, 12:44:10 AM2/22/21
to
On 02/21/2021 07:06 PM, MegaWattz wrote:
> <snip> But when you have
hundreds of Enterprise servers.
But if you want to stick with RHEL/rpm, then why not Oracle Linux? It's
a RHEL clone, production-ready and free.
> <snip> But when you have
> an office full of people, or officeS, that need to be tied together,
> where downtime is unacceptable, well, we need some good backup
> plans PDQ. There is a gap between needing a free OS and being
> able to pay for RHEL that a LOT of businesses and orgs fall into.
>
>
>
Personally, I don't see what's wrong with Ubuntu, I've deployed it to
> where downtime is unacceptable, well, we need some good backup
> plans PDQ. There is a gap between needing a free OS and being
> able to pay for RHEL that a LOT of businesses and orgs fall into.
>
>
>
hundreds of Enterprise servers.
But if you want to stick with RHEL/rpm, then why not Oracle Linux? It's
a RHEL clone, production-ready and free.
Carlos E.R.
Feb 22, 2021, 7:04:09 AM2/22/21
to
On 22/02/2021 04.06, MegaWattz wrote:
...
> On the RPM side, we have OpenSuse and Tumbleweed. Of the
> two Tumbleweed has one of the prized features of Centos - an
> extremely long lifetime. In theory it could be updated forever
> without a single dreaded from-scratch re-install.
To be precise, you have openSUSE in two versions:
...
> On the RPM side, we have OpenSuse and Tumbleweed. Of the
> two Tumbleweed has one of the prized features of Centos - an
> extremely long lifetime. In theory it could be updated forever
> without a single dreaded from-scratch re-install.
Leap (stable, classical)
Tumbleweed (rolling)
Leap derives directly from the commercial product from SUSE, SLE. It
takes the kernel, gnome, and most of the core. Meaning that the major
version number in Leap (15) corresponds to the same number in SLE, and
those core packages change little in several years (5?).
However, the kernel which seems old in version number is heavily
patched, thus supporting more hardware that it would seem on a quick look.
Other packages are community generated, like KDE/Plasma, and thus have
much more modern versions.
Tumbleweed is a rolling version where development is done, but tested
automatically and thoroughly by openQA before actual release. SLE
derives from TW at some point where they take the photo, and Leap
"copies" from SLE (same source code packages).
> SUSE is a sort of "enterprise Linux" as is RHEL. This means it
> SHOULD be easier to score tested code/apps to do some of
> those more complicated biz things. OSTW is very easy to work
> with too - YAST is actually *helpful* and can seriously slash
> the time/error equations for setting up systems and software.
> OSTW is also fairly compatible ... you can use a lot of RHEL
> and Fedora RPMs straight out of the box, so to speak.
maintained by the community which are not in the base system, or in
newer versions.
Don't forget the OBS:
<https://en.wikipedia.org/wiki/Open_Build_Service>
«The Open Build Service (formerly called openSUSE Build Service)[1] is
an open and complete distribution development platform designed to
encourage developers to compile packages for multiple Linux
distributions including SUSE Linux Enterprise Server, openSUSE, Red Hat
Enterprise Linux, Mandriva, Ubuntu, Fedora, Debian, and Arch Linux.[2]
It typically simplifies the packaging process, so developers can more
easily package a single program for many distributions, and many
openSUSE releases, making more packages available to users regardless of
what distribution they use. Also, product and appliance building is
supported by OBS.
The Build Service software is published under the GPL.[3] In an
acknowledgement of its usefulness to the wider Linux community, the
Linux Foundation has announced that the project will be added to the
Linux Developer Network (LDN).[4] Also, various companies,[5] MeeGo
project and Tizen are using it for developing their distribution.
It also delivers a collaboration environment, allowing developer groups
to build and submit changes to other projects.»
--
Cheers, Carlos.
MegaWattz
Feb 23, 2021, 12:09:20 AM2/23/21
to
the "who owns what" aspects more carefully. Oracle
plans to make money too after all and could easily follow
the RH/IBM route at any time.
As I said, Debian and Ubuntu Server are not "bad" per-se,
but they are initially set up to be desktop systems. You
can add and configure (at great expense of time) them to
do anything RHEL can do, but I do not favor overly
customized systems. They tend to be hack-jobs and
are hell for anyone else who follows after. If you want
"job security", well, convince your pointy-haired bosses
this is the way to go ..... they won't know any better,
until later ......... but that's unethical ................
I wasn't kidding about "glued together with Algol-68
and "B" just for kicks" either - done that myself :-)
Hey, if you want to mystify Gen-Z, write in Pascal.
Used to be common - and it's beautiful in its way -
but now is largely abandoned. Can't find a good
Modula 2/3 compiler anymore alas. How about Prolog ?
There used to be a "Turbo Prolog" for the masochists -
all the lambda you could ever want. I have a 5 1/4 floppy
somewhere. An emulated Micro-Vax perhaps ? VMS was
a GREAT system, way ahead of its time, did everything
you'd torture Ubuntu into doing right out of the box.
The BHD Unixes are a good biz foundation, but they are
just enough "different" to throw off a lot of your usual
Linux drudges. You'd need an expert, or polymaths, and
they cost more and are harder to find. Most of the
IT polymaths these days are in their 60's ... seen/done
it all.
MegaWattz
Feb 23, 2021, 12:23:37 AM2/23/21
to
On Mon, 22 Feb 2021 13:04:00 +0100, "Carlos E.R."
<robin_...@es.invalid> wrote:
>On 22/02/2021 04.06, MegaWattz wrote:
>
>...
>
>> On the RPM side, we have OpenSuse and Tumbleweed. Of the
>> two Tumbleweed has one of the prized features of Centos - an
>> extremely long lifetime. In theory it could be updated forever
>> without a single dreaded from-scratch re-install.
>
>To be precise, you have openSUSE in two versions:
>
> Leap (stable, classical)
> Tumbleweed (rolling)
I've TW running on several boxes at the moment, but I'm
<robin_...@es.invalid> wrote:
>On 22/02/2021 04.06, MegaWattz wrote:
>
>...
>
>> On the RPM side, we have OpenSuse and Tumbleweed. Of the
>> two Tumbleweed has one of the prized features of Centos - an
>> extremely long lifetime. In theory it could be updated forever
>> without a single dreaded from-scratch re-install.
>
>To be precise, you have openSUSE in two versions:
>
> Leap (stable, classical)
> Tumbleweed (rolling)
wondering how far it can be pushed as a Centos replacement.
Leap is perfectly OK for your usual desktops, but you only
want to set up a server ONCE under ideal conditions.
Hell, I even installed TW on a R-Pi last week. A bit pokey, but
perfectly functional :-)
In any case, the clock is ticking on Centos. I'd set up a number of
servers with C8 over the past year, now I have to re-do them all
with something else. VERY annoying. I'm too old for this shit ....
If not TW, then it's Ubuntu Server 20 - with a handy LXDE desktop
added, my favorite combo. That should be good for several years.
Had CRAP luck updating U18 to U20 though ... three of three tries
crashed halfway through the upgrade. Fortunately I had more than
one terminal open, managed to force more updates and dist-upgrades
in there to cover what the Ubuntu upgrader missed - an extra two
hours work before being confident to do a reboot !
Carlos E.R.
Feb 23, 2021, 7:12:07 AM2/23/21
to
On 23/02/2021 06.23, MegaWattz wrote:
> On Mon, 22 Feb 2021 13:04:00 +0100, "Carlos E.R."
> <robin_...@es.invalid> wrote:
>
>> On 22/02/2021 04.06, MegaWattz wrote:
>>
>> ...
>>
>>> On the RPM side, we have OpenSuse and Tumbleweed. Of the
>>> two Tumbleweed has one of the prized features of Centos - an
>>> extremely long lifetime. In theory it could be updated forever
>>> without a single dreaded from-scratch re-install.
>>
>> To be precise, you have openSUSE in two versions:
>>
>> Leap (stable, classical)
>> Tumbleweed (rolling)
>
>
> I've TW running on several boxes at the moment, but I'm
> wondering how far it can be pushed as a Centos replacement.
>
> Leap is perfectly OK for your usual desktops, but you only
> want to set up a server ONCE under ideal conditions.
I don't see the problem. :-?
> On Mon, 22 Feb 2021 13:04:00 +0100, "Carlos E.R."
> <robin_...@es.invalid> wrote:
>
>> On 22/02/2021 04.06, MegaWattz wrote:
>>
>> ...
>>
>>> On the RPM side, we have OpenSuse and Tumbleweed. Of the
>>> two Tumbleweed has one of the prized features of Centos - an
>>> extremely long lifetime. In theory it could be updated forever
>>> without a single dreaded from-scratch re-install.
>>
>> To be precise, you have openSUSE in two versions:
>>
>> Leap (stable, classical)
>> Tumbleweed (rolling)
>
>
> I've TW running on several boxes at the moment, but I'm
> wondering how far it can be pushed as a Centos replacement.
>
> Leap is perfectly OK for your usual desktops, but you only
> want to set up a server ONCE under ideal conditions.
Leap is derived directly from SLE, shares the source code, and SLE is
the commercial SUSE version for servers.
You mean because upgrades?
The upgrades from, say, 15.1 to 15.2 are really minor. You don't have to
do again the server, just upgrade it, about once a year or so. Then,
after some years there is one major upgrade; 15.0 releases on 2018-05,
version 15.3 is being prepared now. Next may be 15.4 or perhaps 16.0.
That would be 2022 the earliest for the next major release.
In contrast, TW is upgrading every week, very minor things but very
often. Now an then, there are drastic changes on this or that. Many
people have production machines with TW and are very happy, but some
times there are breakages. Just follow the mail lists to know about things.
> Hell, I even installed TW on a R-Pi last week. A bit pokey, but
> perfectly functional :-)
> If not TW, then it's Ubuntu Server 20 - with a handy LXDE desktop
> added, my favorite combo. That should be good for several years.
> Had CRAP luck updating U18 to U20 though ... three of three tries
> crashed halfway through the upgrade. Fortunately I had more than
> one terminal open, managed to force more updates and dist-upgrades
> in there to cover what the Ubuntu upgrader missed - an extra two
> hours work before being confident to do a reboot !
>
Cheers, Carlos.
MegaWattz
Feb 23, 2021, 10:55:39 PM2/23/21
to
On Tue, 23 Feb 2021 13:10:56 +0100, "Carlos E.R."
2022 ... is less than a year away. No, no, no .....
I've had it with versioned systems. There is just too much
stuff involved in setting up any kind of server - too many
system files to re-tweak, too many custom services to
re-install, too many well-tuned scripts that have to be
re-tuned. It's just not worth it.
Give me something that needs minimal work for FIVE
years or TEN years - let's say until well after I retire.
C8 was supposed to be that system, but now ......
As I mentioned to someone here, three recent attempts to
go from US18.04 to US20.04 ALL crashed out badly about
halfway through. Lucky I had multiple terminal windows
open and could force all the final updates a bit at a time
before daring to reboot. This is NOT inspirational. 16->18
went much smoother.
>In contrast, TW is upgrading every week, very minor things but very
>often. Now an then, there are drastic changes on this or that. Many
>people have production machines with TW and are very happy, but some
>times there are breakages. Just follow the mail lists to know about things.
I always expect minor issues, and am never disappointed :-)
But it's still not like a full version re-install from scratch on
any Linux. The more boxes you administer, the worse the
time/effort equation becomes.
>> Hell, I even installed TW on a R-Pi last week. A bit pokey, but
>> perfectly functional :-)
>
>You are not the only one, I hear :-)
Tried to put US20.04 on a Pi-4 the other day. It starts up just
fine. But once you do the initial updates it hangs up forever on
a large number of services (a lot of them Canonical crap you
don't really want anyway). Disable one and the next one will
get you - and no timeout limits on most. So, forget it. Raspbian
works perfectly. Poor results with Mageia and Alpine too.
I wonder if there's an Antix port - something super light ?
Taz ??? :-)
But TW did work. Alas even with XFCE the system is a bit
"heavy" for a Pi. Oh well, it is what I call a "Cadillac" system.
Maybe with no, or a super-minimal, GUI it'd be snappier.
Sorry, I'm not a command-line fascist, GUIs do make lots
of stuff a lot easier. I was tired of command-line when the
VIC-20 was shiny new. I may still have that in a box somewhere.
Yet, I still wind up in nano a dozen times a day ......
About Centos, or that I'm getting too old for this shit ? :-)
I've had it with versioned systems. There is just too much
stuff involved in setting up any kind of server - too many
system files to re-tweak, too many custom services to
re-install, too many well-tuned scripts that have to be
re-tuned. It's just not worth it.
Give me something that needs minimal work for FIVE
years or TEN years - let's say until well after I retire.
C8 was supposed to be that system, but now ......
As I mentioned to someone here, three recent attempts to
go from US18.04 to US20.04 ALL crashed out badly about
halfway through. Lucky I had multiple terminal windows
open and could force all the final updates a bit at a time
before daring to reboot. This is NOT inspirational. 16->18
went much smoother.
>In contrast, TW is upgrading every week, very minor things but very
>often. Now an then, there are drastic changes on this or that. Many
>people have production machines with TW and are very happy, but some
>times there are breakages. Just follow the mail lists to know about things.
But it's still not like a full version re-install from scratch on
any Linux. The more boxes you administer, the worse the
time/effort equation becomes.
>> Hell, I even installed TW on a R-Pi last week. A bit pokey, but
>> perfectly functional :-)
>
>You are not the only one, I hear :-)
fine. But once you do the initial updates it hangs up forever on
a large number of services (a lot of them Canonical crap you
don't really want anyway). Disable one and the next one will
get you - and no timeout limits on most. So, forget it. Raspbian
works perfectly. Poor results with Mageia and Alpine too.
I wonder if there's an Antix port - something super light ?
Taz ??? :-)
But TW did work. Alas even with XFCE the system is a bit
"heavy" for a Pi. Oh well, it is what I call a "Cadillac" system.
Maybe with no, or a super-minimal, GUI it'd be snappier.
Sorry, I'm not a command-line fascist, GUIs do make lots
of stuff a lot easier. I was tired of command-line when the
VIC-20 was shiny new. I may still have that in a box somewhere.
Yet, I still wind up in nano a dozen times a day ......
The Natural Philosopher
Feb 24, 2021, 9:02:46 AM2/24/21
to
On 24/02/2021 03:55, MegaWattz wrote:
>> It is a shame.
>
> About Centos, or that I'm getting too old for this shit ?:-)
>
>
I think the issue is that most old software doesn't age. And on old
>> It is a shame.
>
> About Centos, or that I'm getting too old for this shit ?:-)
>
>
hardware is probably as good. I have a VPS that is now around 10 years
old running headless debian so ancient that I am not sure I could
upgrade it.
And why should I? If it gets hacked or crashes I have a nightly backup
and could rebuild a more modern version if I had to.
But its never been hacked - access is extremely limited - it is in the
and mail and web only - so why would I bother to upgrade it?
My current home server is also old But its job is even less onerous and
its even more protected
--
No Apple devices were knowingly used in the preparation of this post.
Carlos E.R.
Feb 24, 2021, 12:36:08 PM2/24/21
to
I have recently upgraded several machines from 15.1 to 15.2 and I had to
modify nothing.
>
> Give me something that needs minimal work for FIVE
> years or TEN years - let's say until well after I retire.
> C8 was supposed to be that system, but now ......
>
> As I mentioned to someone here, three recent attempts to
> go from US18.04 to US20.04 ALL crashed out badly about
> halfway through. Lucky I had multiple terminal windows
> open and could force all the final updates a bit at a time
> before daring to reboot. This is NOT inspirational. 16->18
> went much smoother.
>
>
>> In contrast, TW is upgrading every week, very minor things but very
>> often. Now an then, there are drastic changes on this or that. Many
>> people have production machines with TW and are very happy, but some
>> times there are breakages. Just follow the mail lists to know about things.
>
> I always expect minor issues, and am never disappointed :-)
machines :-)
(It is upgrade every week and reboot)
...
--
Cheers, Carlos.
Eli the Bearded
Feb 24, 2021, 3:54:33 PM2/24/21
to
In comp.os.linux.misc, The Natural Philosopher <t...@invalid.invalid> wrote:
> I think the issue is that most old software doesn't age. And on old
> hardware is probably as good. I have a VPS that is now around 10 years
> old running headless debian so ancient that I am not sure I could
> upgrade it.
You probably can't then.
> I think the issue is that most old software doesn't age. And on old
> hardware is probably as good. I have a VPS that is now around 10 years
> old running headless debian so ancient that I am not sure I could
> upgrade it.
> And why should I? If it gets hacked or crashes I have a nightly backup
> and could rebuild a more modern version if I had to.
for your old way might no longer exist, or the config files have
sufficiently changes, or ...
> But its never been hacked - access is extremely limited - it is in the
> and mail and web only - so why would I bother to upgrade it?
not notice it? To get updated software that will work with the rest of
the Internet? Eg, certificate authorities need updating, and SSL
libraries might not support newer encryption as older ones get depcated
for security issues.
https://www.theregister.com/2020/11/06/android_encryption_certs/
Websites accepting credit cards stopped accepting TLS1.1 a couple of
years ago. Other sites followed suit, even if they don't accept credit
cards. There are older browsers which do not support the TLS1.2. People
still run them because "why would I bother to upgrade it?"
Elijah
------
has never crashed a car, so why bother to upgrade to one with seat belts?
MegaWattz
Feb 24, 2021, 11:45:44 PM2/24/21
to
On Wed, 24 Feb 2021 14:02:44 +0000, The Natural Philosopher
<t...@invalid.invalid> wrote:
>On 24/02/2021 03:55, MegaWattz wrote:
>>> It is a shame.
>>
>> About Centos, or that I'm getting too old for this shit ?:-)
>>
>>
>I think the issue is that most old software doesn't age. And on old
>hardware is probably as good. I have a VPS that is now around 10 years
>old running headless debian so ancient that I am not sure I could
>upgrade it.
>
>And why should I? If it gets hacked or crashes I have a nightly backup
>and could rebuild a more modern version if I had to.
It depends on what it's used for ,the environment in which
<t...@invalid.invalid> wrote:
>On 24/02/2021 03:55, MegaWattz wrote:
>>> It is a shame.
>>
>> About Centos, or that I'm getting too old for this shit ?:-)
>>
>>
>I think the issue is that most old software doesn't age. And on old
>hardware is probably as good. I have a VPS that is now around 10 years
>old running headless debian so ancient that I am not sure I could
>upgrade it.
>
>And why should I? If it gets hacked or crashes I have a nightly backup
>and could rebuild a more modern version if I had to.
it exists. There are applications where an '85 PC-XT running
PC-DOS would be considered "good leveraging of hardware
and software". However something exposed to the internet,
well, better have the latest security patches. NOT wise to
run a web server on something that hasn't been updated
for years, or decades.
>But its never been hacked - access is extremely limited - it is in the
>and mail and web only - so why would I bother to upgrade it?
>
>My current home server is also old But its job is even less onerous and
>its even more protected
Somewhere deep down in the pile I have a very very old
version of Red Hat - I think it was the first to sport a GUI.
Took awhile to get proto-X to understand my mouse,
monitor and keyboard though. As I recall it was kind of
the original Gnome (still hate it BTW). Now while it was
a perfectly good and functional Linux, I would not dare
use it as an inet-facing server these days.
MegaWattz
Feb 25, 2021, 12:11:41 AM2/25/21
to
On Wed, 24 Feb 2021 18:33:31 +0100, "Carlos E.R."
17 years old with a closet full of Jolt Cola you found
on E-Bay ??? :-)
>I have recently upgraded several machines from 15.1 to 15.2 and I had to
>modify nothing.
That''s not my complaint. It's the 15 --> 16, or 18, that's
the time-wasting pain.
You CAN update Stretch to Buster - I somehow managed to
do it ACCIDENTLY on one box while trying to get the latest
version of some program. But it was not without any downstream
issues. That was a trivial box however, not something the company
and dozens of people relied on every hour of every day. For some
it will be companieS, and thousands of people and Big money.
Can't play around. It's a reason why, in the age of cheaper
hardware, to stay away from VMs. One job, one box (and a
mirror box). I've had horrible luck with VMs over the year - and
one bad chip means you've just lost the equivalent of two,
five, ten, servers all at once. No $$$ today ...
>> Give me something that needs minimal work for FIVE
>> years or TEN years - let's say until well after I retire.
>> C8 was supposed to be that system, but now ......
>
>IMO, TW is not that.
Well, that's the kind of input I was looking for.
Now Centos isn't "no work" either ... the question is
whether for most common purposes TW is much
worse, or seriously less capable.
Thing is, I have to start making some serious decisions
like Real Soon Now. Lots of people do. Centos is going
down the garbage chute. Its future is writ. Some are
working on their own RHEL clones, but they are still
basically alpha, no better than what Centos is becoming.
I wouldn't dare host an intra-office mail server on them.
OS/TW is far more stable and developed and can use
most of the familiar RPMs out there.
>> As I mentioned to someone here, three recent attempts to
>> go from US18.04 to US20.04 ALL crashed out badly about
>> halfway through. Lucky I had multiple terminal windows
>> open and could force all the final updates a bit at a time
>> before daring to reboot. This is NOT inspirational. 16->18
>> went much smoother.
>>
>>
>>> In contrast, TW is upgrading every week, very minor things but very
>>> often. Now an then, there are drastic changes on this or that. Many
>>> people have production machines with TW and are very happy, but some
>>> times there are breakages. Just follow the mail lists to know about things.
>>
>> I always expect minor issues, and am never disappointed :-)
>
>This week, people are reporting upgrades of 3000 packages on their TW
>machines :-)
>
>(It is upgrade every week and reboot)
So long as the upgrades WORK.
SO far, no real TW issues worth mentioning.
But pretty quick my C8 "upgrade" will be an upgrade
to tons of alpha code. Can't go there.
on E-Bay ??? :-)
>I have recently upgraded several machines from 15.1 to 15.2 and I had to
>modify nothing.
the time-wasting pain.
You CAN update Stretch to Buster - I somehow managed to
do it ACCIDENTLY on one box while trying to get the latest
version of some program. But it was not without any downstream
issues. That was a trivial box however, not something the company
and dozens of people relied on every hour of every day. For some
it will be companieS, and thousands of people and Big money.
Can't play around. It's a reason why, in the age of cheaper
hardware, to stay away from VMs. One job, one box (and a
mirror box). I've had horrible luck with VMs over the year - and
one bad chip means you've just lost the equivalent of two,
five, ten, servers all at once. No $$$ today ...
>> Give me something that needs minimal work for FIVE
>> years or TEN years - let's say until well after I retire.
>> C8 was supposed to be that system, but now ......
>
>IMO, TW is not that.
Now Centos isn't "no work" either ... the question is
whether for most common purposes TW is much
worse, or seriously less capable.
Thing is, I have to start making some serious decisions
like Real Soon Now. Lots of people do. Centos is going
down the garbage chute. Its future is writ. Some are
working on their own RHEL clones, but they are still
basically alpha, no better than what Centos is becoming.
I wouldn't dare host an intra-office mail server on them.
OS/TW is far more stable and developed and can use
most of the familiar RPMs out there.
>> As I mentioned to someone here, three recent attempts to
>> go from US18.04 to US20.04 ALL crashed out badly about
>> halfway through. Lucky I had multiple terminal windows
>> open and could force all the final updates a bit at a time
>> before daring to reboot. This is NOT inspirational. 16->18
>> went much smoother.
>>
>>
>>> In contrast, TW is upgrading every week, very minor things but very
>>> often. Now an then, there are drastic changes on this or that. Many
>>> people have production machines with TW and are very happy, but some
>>> times there are breakages. Just follow the mail lists to know about things.
>>
>> I always expect minor issues, and am never disappointed :-)
>
>This week, people are reporting upgrades of 3000 packages on their TW
>machines :-)
>
>(It is upgrade every week and reboot)
SO far, no real TW issues worth mentioning.
But pretty quick my C8 "upgrade" will be an upgrade
to tons of alpha code. Can't go there.
The Natural Philosopher
Feb 25, 2021, 3:31:14 AM2/25/21
to
On 24/02/2021 20:54, Eli the Bearded wrote:
> In comp.os.linux.misc, The Natural Philosopher <t...@invalid.invalid> wrote:
>> I think the issue is that most old software doesn't age. And on old
>> hardware is probably as good. I have a VPS that is now around 10 years
>> old running headless debian so ancient that I am not sure I could
>> upgrade it.
>
> You probably can't then.
>
>> And why should I? If it gets hacked or crashes I have a nightly backup
>> and could rebuild a more modern version if I had to.
>
> Good luck. Sometimes that's trivial, sometimes not. Necessary packages
> for your old way might no longer exist, or the config files have
> sufficiently changes, or ...
>
>> But its never been hacked - access is extremely limited - it is in the
>> and mail and web only - so why would I bother to upgrade it?
>
> To know that you can? To stop a hack before it happens, since you might
> not notice it?
If the hack has no effect its arguably not a problem
> In comp.os.linux.misc, The Natural Philosopher <t...@invalid.invalid> wrote:
>> I think the issue is that most old software doesn't age. And on old
>> hardware is probably as good. I have a VPS that is now around 10 years
>> old running headless debian so ancient that I am not sure I could
>> upgrade it.
>
> You probably can't then.
>
>> And why should I? If it gets hacked or crashes I have a nightly backup
>> and could rebuild a more modern version if I had to.
>
> Good luck. Sometimes that's trivial, sometimes not. Necessary packages
> for your old way might no longer exist, or the config files have
> sufficiently changes, or ...
>
>> But its never been hacked - access is extremely limited - it is in the
>> and mail and web only - so why would I bother to upgrade it?
>
> To know that you can? To stop a hack before it happens, since you might
> not notice it?
To get updated software that will work with the rest of
> the Internet? Eg, certificate authorities need updating, and SSL
> libraries might not support newer encryption as older ones get depcated
> for security issues.
>
> https://www.theregister.com/2020/11/06/android_encryption_certs/
>
> Websites accepting credit cards stopped accepting TLS1.1 a couple of
> years ago. Other sites followed suit, even if they don't accept credit
> cards. There are older browsers which do not support the TLS1.2. People
> still run them because "why would I bother to upgrade it?"
>
> Elijah
> ------
> has never crashed a car, so why bother to upgrade to one with seat belts?
>
You might find the philosophical 'problem of induction ' edifying.
It reminds us that all knowledge is uncertain, unprovable, and
conjecture isn't even dignified by repetitive failure to refute its
hypotheses.
--
The New Left are the people they warned you about.
Carlos E.R.
Feb 25, 2021, 7:56:09 AM2/25/21
to
If that's an issue, buy SLES with expanded maintenance, good for 7 or
more years, I don't know exactly.
>
> You CAN update Stretch to Buster - I somehow managed to
> do it ACCIDENTLY on one box while trying to get the latest
> version of some program. But it was not without any downstream
> issues. That was a trivial box however, not something the company
> and dozens of people relied on every hour of every day. For some
> it will be companieS, and thousands of people and Big money.
> Can't play around. It's a reason why, in the age of cheaper
> hardware, to stay away from VMs. One job, one box (and a
> mirror box). I've had horrible luck with VMs over the year - and
> one bad chip means you've just lost the equivalent of two,
> five, ten, servers all at once. No $$$ today ...
>
>>> Give me something that needs minimal work for FIVE
>>> years or TEN years - let's say until well after I retire.
>>> C8 was supposed to be that system, but now ......
>>
>> IMO, TW is not that.
>
> Well, that's the kind of input I was looking for.
>
> Now Centos isn't "no work" either ... the question is
> whether for most common purposes TW is much
> worse, or seriously less capable.
It has many people that love it. Specially developers.
TW is very capable. People are using it on production. But it is a
rolling distro on the edge. Mind the later, it is the development
version. It is where changes are applied first. Tested, yes, but applied.
It follows the Linux mantra of "update often" (cathedral and the bazaar,
was it?)
So, every week you get a "release" of sorts. Mostly minor changes, but
sometimes this or that will suffer a mayor change. Some configuration
changes. Something will migrate to /usr or whatever. You have to be
paying attention and following the mail list to be aware of what will
change next week.
>
> Thing is, I have to start making some serious decisions
> like Real Soon Now. Lots of people do. Centos is going
> down the garbage chute. Its future is writ. Some are
> working on their own RHEL clones, but they are still
> basically alpha, no better than what Centos is becoming.
> I wouldn't dare host an intra-office mail server on them.
> OS/TW is far more stable and developed and can use
> most of the familiar RPMs out there.
>
>
>>> As I mentioned to someone here, three recent attempts to
>>> go from US18.04 to US20.04 ALL crashed out badly about
>>> halfway through. Lucky I had multiple terminal windows
>>> open and could force all the final updates a bit at a time
>>> before daring to reboot. This is NOT inspirational. 16->18
>>> went much smoother.
>>>
>>>
>>>> In contrast, TW is upgrading every week, very minor things but very
>>>> often. Now an then, there are drastic changes on this or that. Many
>>>> people have production machines with TW and are very happy, but some
>>>> times there are breakages. Just follow the mail lists to know about things.
>>>
>>> I always expect minor issues, and am never disappointed :-)
>>
>> This week, people are reporting upgrades of 3000 packages on their TW
>> machines :-)
>>
>> (It is upgrade every week and reboot)
>
> So long as the upgrades WORK.
that somebody thought and wrote a test. There are many many tests. You
may hit something new that has not been found yet and you are the one
that reports it. Thank you for reporting it, by the way ;-)
If postfix get a major change upstream that changes the main.cf file to
be called now "principal.cf" because the developer is now Spanish, well,
you will get that change :-p
And the next week, apache does a major version change, by default it
serves from /server, not /srv.
Then, the next week it is systemd which does a minor change that changes
when timers trigger.
Don't forget libc changes. Or kernel major versions.
I don't know you, but I prefer to suffer configuration changes when I
want to, not any random week.
> SO far, no real TW issues worth mentioning.
>
> But pretty quick my C8 "upgrade" will be an upgrade
> to tons of alpha code. Can't go there.
>
Cheers, Carlos.
Eli the Bearded
Feb 25, 2021, 1:58:12 PM2/25/21
to
In comp.os.linux.misc, The Natural Philosopher <t...@invalid.invalid> wrote:
> On 24/02/2021 20:54, Eli the Bearded wrote:
>> To stop a hack before it happens, since you might
>> not notice it?
> If the hack has no effect its arguably not a problem
You not noticing != no effect.
> On 24/02/2021 20:54, Eli the Bearded wrote:
>> To stop a hack before it happens, since you might
>> not notice it?
> If the hack has no effect its arguably not a problem
>> has never crashed a car, so why bother to upgrade to one with seat belts?
> Rather, Elijah, has never seen a car crash in the neighbourhood at all.
Elijah
------
enquiring minds
MegaWattz
Feb 25, 2021, 11:14:11 PM2/25/21
to
On Thu, 25 Feb 2021 13:55:20 +0100, "Carlos E.R."
Well, we can all become slaves to MS .... everything will
be better then .......... :-)
Life is always compromises. If it's some box I own then
I can take more chances. If, as a supposed 'professional',
I switch to a crap OS for employers, that's BAD. They
are never forgiving or understanding. They know exactly
dick about computers and operating systems. Their stuff
is supposed to work ALL the time. That's what they think
they're paying for. In real life, well, but the idea is to keep
things fairly close to what those Lords Of The Paychecks
expect.
Anyhow, I'm looking for the best Centos replacement, and
now having to do from-scratch reinstalls every year or two
is as major an issue as reliability. The Liniverse does not
offer anything "perfect", but for a give application there will
be a "best fit".
Oh well, in five more years almost everybody will be
prisoners of the CloudyVerse. Oh, they sell, the great
and wonderful CloudyVerse solves ALL your problems.
You don't need to hire anymore weirdo I.T. people who
lurk around the coffee machine anymore ! You don't
need to buy expensive servers, just thin clients. Your
apps and data will be perfectly and reliably handled
by invisible elves somwhere "out there". Well, until
Bejing pushes the button .......
No, it won't all go blank at once, it will be a "pesty"
constant stream of issues, undermining confidence
yet leaving few options. Your databases won't work
for a day or two, chunks of your data will vanish
forever or be mysteriously corrupted. Then the mail
and VPN will get balky. Your Direct Deposit checks
will be directly deposited to /dev/null. Your Zoom
conferences will break at random intervals. The DB
will come back for awhile, but then get all weird again
and ...
And only those "survivalist" companies with their own
internal alternatives and people to keep them working
will be left. Alas, if 95% of everybody else goes wonky
there will be almost nobody to do business with. Over
the years we've sophisticated ourselves into a deadly
corner. "Told Ya So" won't mean much.
be better then .......... :-)
Life is always compromises. If it's some box I own then
I can take more chances. If, as a supposed 'professional',
I switch to a crap OS for employers, that's BAD. They
are never forgiving or understanding. They know exactly
dick about computers and operating systems. Their stuff
is supposed to work ALL the time. That's what they think
they're paying for. In real life, well, but the idea is to keep
things fairly close to what those Lords Of The Paychecks
expect.
Anyhow, I'm looking for the best Centos replacement, and
now having to do from-scratch reinstalls every year or two
is as major an issue as reliability. The Liniverse does not
offer anything "perfect", but for a give application there will
be a "best fit".
Oh well, in five more years almost everybody will be
prisoners of the CloudyVerse. Oh, they sell, the great
and wonderful CloudyVerse solves ALL your problems.
You don't need to hire anymore weirdo I.T. people who
lurk around the coffee machine anymore ! You don't
need to buy expensive servers, just thin clients. Your
apps and data will be perfectly and reliably handled
by invisible elves somwhere "out there". Well, until
Bejing pushes the button .......
No, it won't all go blank at once, it will be a "pesty"
constant stream of issues, undermining confidence
yet leaving few options. Your databases won't work
for a day or two, chunks of your data will vanish
forever or be mysteriously corrupted. Then the mail
and VPN will get balky. Your Direct Deposit checks
will be directly deposited to /dev/null. Your Zoom
conferences will break at random intervals. The DB
will come back for awhile, but then get all weird again
and ...
And only those "survivalist" companies with their own
internal alternatives and people to keep them working
will be left. Alas, if 95% of everybody else goes wonky
there will be almost nobody to do business with. Over
the years we've sophisticated ourselves into a deadly
corner. "Told Ya So" won't mean much.
Carlos E.R.
Feb 26, 2021, 8:28:09 AM2/26/21
to
On 26/02/2021 05.14, MegaWattz wrote:
> On Thu, 25 Feb 2021 13:55:20 +0100, "Carlos E.R."
> <robin_...@es.invalid> wrote:
>
>> On 25/02/2021 06.11, MegaWattz wrote:
>>> On Wed, 24 Feb 2021 18:33:31 +0100, "Carlos E.R."
>>> <robin_...@es.invalid> wrote:
...
> On Thu, 25 Feb 2021 13:55:20 +0100, "Carlos E.R."
> <robin_...@es.invalid> wrote:
>
>> On 25/02/2021 06.11, MegaWattz wrote:
>>> On Wed, 24 Feb 2021 18:33:31 +0100, "Carlos E.R."
>>> <robin_...@es.invalid> wrote:
> Well, we can all become slaves to MS .... everything will
> be better then .......... :-)
>
> Life is always compromises. If it's some box I own then
> I can take more chances. If, as a supposed 'professional',
> I switch to a crap OS for employers, that's BAD. They
> are never forgiving or understanding. They know exactly
> dick about computers and operating systems. Their stuff
> is supposed to work ALL the time. That's what they think
> they're paying for. In real life, well, but the idea is to keep
> things fairly close to what those Lords Of The Paychecks
> expect.
>
> Anyhow, I'm looking for the best Centos replacement, and
> now having to do from-scratch reinstalls every year or two
> is as major an issue as reliability. The Liniverse does not
> offer anything "perfect", but for a give application there will
> be a "best fit".
For decades now :-)
--
Cheers, Carlos.
MegaWattz
Feb 26, 2021, 10:49:13 PM2/26/21
to
On Fri, 26 Feb 2021 14:26:58 +0100, "Carlos E.R."
Not ALL distros are upgrade-friendly. You can be sneaky or
forceful and somehow get it done on almost any Linux, but
it's a labor-vs-rewards equation. A mix of old/new kernels
and software meant for them gives Linux indigestion.
Sometimes it IS just better/easier to reinstall from scratch.
But I don't like it.
forceful and somehow get it done on almost any Linux, but
it's a labor-vs-rewards equation. A mix of old/new kernels
and software meant for them gives Linux indigestion.
Sometimes it IS just better/easier to reinstall from scratch.
But I don't like it.
MegaWattz
Feb 26, 2021, 11:01:01 PM2/26/21
to
On 26 Feb 2021 21:08:00 GMT, Stéphane CARPENTIER <s...@fiat-linux.fr>
wrote:
>Le 25-02-2021, Carlos E.R. <robin_...@es.invalid> a écrit :
>>
>>
>> If postfix get a major change upstream that changes the main.cf file to
>> be called now "principal.cf" because the developer is now Spanish, well,
>> you will get that change :-p
>>
>> And the next week, apache does a major version change, by default it
>> serves from /server, not /srv.
>>
>> Then, the next week it is systemd which does a minor change that changes
>> when timers trigger.
>>
>> Don't forget libc changes. Or kernel major versions.
>
>In my experience, every change like that is taken care by the distro.
>when it's a change at a time, it's OK. When there are a lot of changes,
>the issues begin.
>
>For me centos is deprecated by design and at work the time taken to
>install a full application with the server made the first upgrade almost
>never done. Then the second upgrade wasn't done neither. Then comes a
>time when there is a large gap between the last version and the
>installed version. And a big project is launched to take care of all the
>very deprecated servers.
>The issue is how painful it is to do. The fact is: with Ubuntu it has
>always been difficult when with Archlinux it has always been smooth.
>
>I'm really not ready to switch back to a versioning distro.
Uh oh, an Arch masochist ! :-) Great system, but it makes you WORK.
I do agree that versioned systems get worse and worse over time. By
the time you're a full version, or two, behind no upgrade is going
go well. Rolling releases might have a little alpha-level revision
code in them, but in my experience they work pretty well. The
nature of the upgrading process means you're more likely to do
them more often. I did a "zypper dup" on a TW box just today.
Took about ten minutes and then all was well. But try to go
from US18 to US20 and, in three tries, I've NEVER had it go
well. Always crashes halfway through, leaving a Frankenstein
system you have to bully into finishing all the updates while
hoping your terminal doesn't freeze up.
Hmm ... while I've never messed much with the BSD's, I'm
going to look into how well they upgrade - minor and major
versions. Unix was intended for Big Stuff that was NEVER
supposed to go down.
wrote:
>Le 25-02-2021, Carlos E.R. <robin_...@es.invalid> a écrit :
>>
>>
>> If postfix get a major change upstream that changes the main.cf file to
>> be called now "principal.cf" because the developer is now Spanish, well,
>> you will get that change :-p
>>
>> And the next week, apache does a major version change, by default it
>> serves from /server, not /srv.
>>
>> Then, the next week it is systemd which does a minor change that changes
>> when timers trigger.
>>
>> Don't forget libc changes. Or kernel major versions.
>
>when it's a change at a time, it's OK. When there are a lot of changes,
>the issues begin.
>
>For me centos is deprecated by design and at work the time taken to
>install a full application with the server made the first upgrade almost
>never done. Then the second upgrade wasn't done neither. Then comes a
>time when there is a large gap between the last version and the
>installed version. And a big project is launched to take care of all the
>very deprecated servers.
>
>> I don't know you, but I prefer to suffer configuration changes when I
>> want to, not any random week.
>
>For me, the issue is not when I choose or when my distro choses for me.
>> I don't know you, but I prefer to suffer configuration changes when I
>> want to, not any random week.
>
>The issue is how painful it is to do. The fact is: with Ubuntu it has
>always been difficult when with Archlinux it has always been smooth.
>
>I'm really not ready to switch back to a versioning distro.
Uh oh, an Arch masochist ! :-) Great system, but it makes you WORK.
I do agree that versioned systems get worse and worse over time. By
the time you're a full version, or two, behind no upgrade is going
go well. Rolling releases might have a little alpha-level revision
code in them, but in my experience they work pretty well. The
nature of the upgrading process means you're more likely to do
them more often. I did a "zypper dup" on a TW box just today.
Took about ten minutes and then all was well. But try to go
from US18 to US20 and, in three tries, I've NEVER had it go
well. Always crashes halfway through, leaving a Frankenstein
system you have to bully into finishing all the updates while
hoping your terminal doesn't freeze up.
Hmm ... while I've never messed much with the BSD's, I'm
going to look into how well they upgrade - minor and major
versions. Unix was intended for Big Stuff that was NEVER
supposed to go down.
Carlos E.R.
Feb 27, 2021, 8:00:10 AM2/27/21
to
pros of openSUSE ;-)
> You can be sneaky or
> forceful and somehow get it done on almost any Linux, but
> it's a labor-vs-rewards equation. A mix of old/new kernels
> and software meant for them gives Linux indigestion.
> Sometimes it IS just better/easier to reinstall from scratch.
>
> But I don't like it.
>
Cheers, Carlos.
MegaWattz
Feb 28, 2021, 12:02:29 AM2/28/21
to
On 27 Feb 2021 13:05:02 GMT, Stéphane CARPENTIER <s...@fiat-linux.fr>
wrote:
>Yes. And as I don't like the mouse when I don't need it, you could say
>that for a lot of applications I'm using.
>ubuntu, agreed. But compared to what I had to do to install slackware
>more than twenty five years ago, it's really a piece of cake.
Oh yea ... I remember early Slack ! Almost put me off Linux
forever and ever :-)
If MS had stopped with Win2k and just lightly buffed and
polished it thereafter I might have stuck with them (despite
finding the NSA backdoors in the registry). They still had
a few old guys who understood the WHOLE THING and
knew if you changed 'X' then it would affect "Y" and "Z".
>For the management part, it's really the other way around. Since
>everything is as I want, it runs smoothly. The new stuff I want to add
>fits perfectly well and easily.
Well, for *a* box that's great. Now think ten or twenty of
the things, each running one or two dedicated jobs like
mail, GIS, web or database servers. Suddenly the time/effort
in getting everything "perfect",well, it's more an exponential
sort of challenge ...........
This is when a more "general purpose", not-so-highly-tuned,
distro with at least a minimal GUI becomes attractive.
I always start with the "minimal" system and then add LXDE
(LXQT still sux) without the "recommends". Then add just
the necessary support packages. This gives a server that's
easy/quick to use and debug without unnecessary bulk.
Ubuntu, Debian, Centos, OSuse - hardly matters. You can
use VNC to see it anywhere (with proper precautions of
course).
On the whole, vanilla Debian is my fave, but it
doesn't do rolling distributions. And no, I do *not* use any
of those all-purpose "administer all your servers" software.
That means the bad people can administer all your servers
as well. Make it HARD for them and they'll go for softer
targets .....
wrote:
>Yes. And as I don't like the mouse when I don't need it, you could say
>that for a lot of applications I'm using.
>
>> Great system, but it makes you WORK.
>
>For the installation part, compared with what I had to do to install
>> Great system, but it makes you WORK.
>
>ubuntu, agreed. But compared to what I had to do to install slackware
>more than twenty five years ago, it's really a piece of cake.
Oh yea ... I remember early Slack ! Almost put me off Linux
forever and ever :-)
If MS had stopped with Win2k and just lightly buffed and
polished it thereafter I might have stuck with them (despite
finding the NSA backdoors in the registry). They still had
a few old guys who understood the WHOLE THING and
knew if you changed 'X' then it would affect "Y" and "Z".
>For the management part, it's really the other way around. Since
>everything is as I want, it runs smoothly. The new stuff I want to add
>fits perfectly well and easily.
Well, for *a* box that's great. Now think ten or twenty of
the things, each running one or two dedicated jobs like
mail, GIS, web or database servers. Suddenly the time/effort
in getting everything "perfect",well, it's more an exponential
sort of challenge ...........
This is when a more "general purpose", not-so-highly-tuned,
distro with at least a minimal GUI becomes attractive.
I always start with the "minimal" system and then add LXDE
(LXQT still sux) without the "recommends". Then add just
the necessary support packages. This gives a server that's
easy/quick to use and debug without unnecessary bulk.
Ubuntu, Debian, Centos, OSuse - hardly matters. You can
use VNC to see it anywhere (with proper precautions of
course).
On the whole, vanilla Debian is my fave, but it
doesn't do rolling distributions. And no, I do *not* use any
of those all-purpose "administer all your servers" software.
That means the bad people can administer all your servers
as well. Make it HARD for them and they'll go for softer
targets .....
MegaWattz
Feb 28, 2021, 12:20:49 AM2/28/21
to
On Sat, 27 Feb 2021 13:56:13 +0100, "Carlos E.R."
RHEL/Centos is tuned more towards the biz environment.
They WANT smooth upgrades. RH would rather you pay
your tech-consulting fee, but never use it.
Most other distros are far more oriented towards the
desktop user. Yes, you can make good servers out of
them without TOO much effort, but they are not nearly
so much biz-oriented.
OSuse, like its parent, is MORE biz/enterprise-ready than,
say, Debian. Not as much as RHEL, but relatively "better".
Hmm ... there used to be a distro called "Scientific Linux".
It was another RHEL clone, customized by, I think, CERN
and the Argonne National Laboratory. It was pretty nice
actually - all the good stuff with a much smoother interface.
Trust the IQ-200+ crowd. Alas it seems to have dropped
off the update radar. Them's the breaks ....
They WANT smooth upgrades. RH would rather you pay
your tech-consulting fee, but never use it.
Most other distros are far more oriented towards the
desktop user. Yes, you can make good servers out of
them without TOO much effort, but they are not nearly
so much biz-oriented.
OSuse, like its parent, is MORE biz/enterprise-ready than,
say, Debian. Not as much as RHEL, but relatively "better".
Hmm ... there used to be a distro called "Scientific Linux".
It was another RHEL clone, customized by, I think, CERN
and the Argonne National Laboratory. It was pretty nice
actually - all the good stuff with a much smoother interface.
Trust the IQ-200+ crowd. Alas it seems to have dropped
off the update radar. Them's the breaks ....
0 new messages