Date: prev next · Thread: first prev next last
2015 Archives by date, by thread · List index

Hi :)
I agree except there are at least 4 approaches that i know of.  My post
avoided going down the LTS approach on the grounds that it had been
rejected out of hand already, even though it works well for some.

The Redhat approach is to have 2 distinctly different distros with separate
names and branding.  One tests ultra-new technologies often before any
other distros.  The other, their flagship one called "Redhat", stays set
for years before getting upgraded.  There is a big fuss and much publicity
in the run-up to the upgrade.

Then there's the Debian style, which is roughly what we use.  The new
branch has all the exciting experimental stuff in it.  Once it's been out
in the wild on real-world machines and on enough bare-metal to shake a
stick at and received plenty of patches and updates the community
eventually decides that in a year or so it can be considered what they call
"Stable branch".  Of course when their branch is very fresh and new it is
also stable in the developers way of thinking because it's been tested in
all the ways they reasonably can and is not crashing or anything like that
- so perhaps "Stable" is a bit misleading but it makes intuitive sense to
users so Debian goes with it.  Then they have a new "Development branch"
which gets used by pretty much everyone anyway with pretty much all of them
appreciating the opportunity to work with something more advanced than the
standard.  It's special and a bit edgy so they feel privileged to use it
and it makes them feel like they are possibly more geeky than they really
are.  The 'older' branch, now called "Stable" branch continues to get
security updates and such.

SliTaZ does much the same except they call their newer branch their
"cooking branch" but they are French so it suits them well.  Beats mucking
around with horses, right?

These are all not-quite the same as LTS.  Everything for the LTS release
has to conform to a MUCH stricter set of rules.  So that is all versions of
all packages and modules have to be up to a certain standard otherwise they
risk being left out entirely, perhaps in favour of a competitor.  Evolution
couldn't consistently make the standard nor meet any deadlines so for a
long while it consistently got in as an older version but eventually got
ditched in favour of Thunderbird and, i think, Lightning.  Even though
Evolution would be more ideal as it's more like a drop-in replacement for
Outlook (hence why it was given so many chances) but it just became
untenable to continue having it as the default.

The LTS does tend to have new features and some extra "wow" factor(s) above
and beyond what could be expected for a normal release.  This draws
attention from the press and others who eagerly speculate and anticipate
what may or may not be in it.  Discussions rage.  But those new features
have often been well and truly tested well in advance but in something
approaching secrecy so that only a few people really know what is going to
be in it.

So with the LTS release it's not just about it getting longer term support
after it's been released.  That is, of course, crucial but it's not the
main thing.  The main thing is that it's substantially better quality on
it's release date than any other release is on their release date - even
the subsequent next couple of releases.  So people often choose the LTS
even after there have been a few more recent releases = because they know
they get better quality.

To me that is substantially better than just using it because it's old!!
It's a huge 'seasonal' boost to their marketing - much the same speculation
and anticipation as before the release of a new iPhone!!

We don't get anything like that level of excitement before the release of a
new branch.  We get a bit of fresh interest at each new branch's release
but less and less each time.  Maybe it might be possible to learn something
from people who make it to the next plateau up - or perhaps we are more
like Microsoft in being unable to admit that others may have a point.  But
perhaps i am wrong and Microsoft, Apple, Google and Ubuntu are really
clueless morons that we have nothing to learn from.

Thanks for the run-down on rolling releases.  I knew there had to be
something but i had no idea what.  I'd also guessed that people wouldn't
really run into those problems for a couple of years.  Right now people are
having the usual teething problems of adapting to a new layout and
presumably the typical problems of using an MS product before "Service Pack
1" - and without the confidence of knowing roughly when the equivalent
might be.

Thanks and regards from
Tom :)

On 11 October 2015 at 21:42, Jay Lozier <> wrote:

On 10/11/2015 03:42 PM, Tom Davies wrote:

Hi :)
"Thanks for the flowers"/approval which i've snipped.  It's a shock to
finally agree on something! :))

The LTS approach was a new way of dealing with an old problem.  The old
still current problem is that projects are pulled in 2 opposing
1.  exciting and new developments, fashion, bling
2.  stay with something familiar and see it mature.  NOT having to
constantly work at it.

That is probably why Redhat and Debian (and family) and many others (even
[shudders] Microsoft and to a lesser extent Apple) provide a version that
basically stays the same for years.  Heck, many places grumble about
'having to' upgrade from Xp because it 'only' lasted 10 years!  Some
organisations happily pay millions per year extra purely in order to be
able to stay with the same old Xp and STILL haven't developed a strategy
for upgrading.

Arch and others attempt to deal with the problem by doing rolling releases
- which brings it's own set of problems - as Windows 10 users and
will doubtless be learning afresh over the next couple of years.  Arch has
already long ago grokked this so MS could learn valuable lessons from them
but i think we all know they can't learn wisdom from outside, unless they
really have changed.

Both approaches have problems with either needing to maintain security
releases for old versions (LTS) or with system stability/breakage
(Rolling). The first appears safe because the system is relatively stable
but older OSes may not support easily newer technologies. This can be
problematic as the OS ages. Also, security releases and bug fixes must be
maintained over several version of a library. Rolling releases can be have
stability issues with being too close to the bleeding edge but they are
likely to support the latest technologies. Also, there are fewer library
versions to be maintained.

Having used both, I recommend LTS releases for most users knowing every x
years their system must be upgraded to the current supported release.

My fear with W10 is MS does not truly understand the nature of a rolling
release and their users are not at all familiar with the quirks of a
rolling release. I have found one needs to pay closer attention to update
issues as they occur with a rolling release and it helps to have a good
grasp of how a computer works. Windows users are not used to more active
update management and often have a very poor understanding of how a
computer works. IMHO, the potential for a disaster about 6 - 12 months from
initial release is very high with W10.

So in answer to your question to Alex; "Yes".  Many places would
updates rather than to keep demanding their Sys. Admins have to keep
re-installing new upgrades.

It'd also be great if there were some sort of "Super Still" branch, like
Debian, or Redhat (and many others) that kept getting updates for 3-4
years.  So that organisations could install the Super Still branch on new
systems in complete confidence that they wouldn't need to touch the system
again for a couple years.

There are other cases where people don't have broadband for downloading
full upgrades but could do with having a system they could rely on for
years.  European city-dwellers might not quite realise what it's like
without broadband.

I think it's interesting that the super-rich share a problem in common
the desperately isolated and cut-off.  One which is largely addressed by
almost all of Gnu&Linux but not by LibreOffice.
Regards from
Tom :)

<snip />

Also you might add that TDF does not offer LTS because TDF is not a
business and therefore has no incentive in a LTS version which only makes
sense if you monetize it. The poster example of this is Canonical and
Ubuntu LTS.  Canonical makes money on LTS and is only able to do so
the LTS itself is a profitable business.  Otherwise you would not even
of it. Businesses looking for something very similar to a LTS version of
LibreOffice can contact our certified developers and their companies



<snip />

To unsubscribe e-mail to:
Posting guidelines + more:
List archive:
All messages sent to this list will be publicly archived and cannot be

To unsubscribe e-mail to:
Posting guidelines + more:
List archive:
All messages sent to this list will be publicly archived and cannot be deleted


Privacy Policy | Impressum (Legal Info) | Copyright information: Unless otherwise specified, all text and images on this website are licensed under the Creative Commons Attribution-Share Alike 3.0 License. This does not include the source code of LibreOffice, which is licensed under the Mozilla Public License (MPLv2). "LibreOffice" and "The Document Foundation" are registered trademarks of their corresponding registered owners or are in actual use as trademarks in one or more countries. Their respective logos and icons are also subject to international copyright laws. Use thereof is explained in our trademark policy.