Installing Libreoffice in Ubuntu

users
#20

Hi :slight_smile:
I suspect that's it's mostly only the "Fresh" branch.

LibreOffice has an extremely fast-paced development. Each new branch
introduces tons of new functionality and deals with legacy problems. This
is partly "catch up", partly recoding to reduce the wide number of
different programming languages into just C(++?) and Python, partly
increasing compatibility with constantly changing proprietary formats,
partly dealing with revolutionary changes in IT and radically new ways of
thinking.

"Sun" achieved a stability by preventing almost anyone from making any
changes at all. This meant that although there were tons of problems these
were all "known problems" that people knew to either work-around or
accept. They seldom worked on bug-reports or fixed anything, apparently.

With each change, even if it's 'just' patching some problem there is a
potential for unexpected side-effects. OpenOffice has a tradition (thanks
to Sun and Oracle) of not dealing with problems. So OO remains "stable"
even in a wide-eyed-end-user type of way. LibreOffice marches fearlessly
on, deals with issues, adds new stuff to the "Fresh" branch and then
patches as many of the new problems as it can with subsequent releases in
that branch even after it becomes the "Still" branch and a new "Fresh"
branch has been released.

So it really isn't any surprise that LO has issues that OpenOffice
doesn't. The same is true the other way around too! Both still have some
long-running issues, just as any long-running project has. Just as MS
Office has. Some may never be resolved but it's more likely to be fixed
some day in LibreOffice precisely because so much more work is going on and
because of the faster pace of development.

This all happened long ago too ...
Back before TDF and LibreOffice, before Oracle, various companies such as
Novell, SuSE, Redhat, Gnome, Debian, Canonical (Ubuntu) and many more got
together and developed a project called Go-oo that added a ton of stuff to
each new release of OO.o making it faster and with better compatibility.
However this often left the original branding in place so that many people
thought they were using OpenOffice.org as released by "Sun" when they were
actually using Go-oo. This was the case with almost all Gnu&Linux
distros.

IBM created their own additions and changes, eschewing the Go-o changes and
just doing it all their own way to create "IBM Lotus Symphony" to sell to
corporate organisations and others.

Some Mac people did their own thing to OO.o to create NeoOffice.

So lots of people were a little unhappy with the stagnation of OpenOffice
under Sun and created their own forks or bunch of additions/changes to the
infrequent OO.o releases.

When Oracle acquired Sun and refused to communicate with the community at
all, apparently seized assets, bank accounts and funds (almost entirely
raised by the community and meant for the community's usage) a small group
of (imo) heroes (incl Charles Schulz, Sophie, Italo, Micheal Meeks (of
Novell) and about 16 more) pushed through the plan that had been developed
years earlier to break away and create an independent organisation purely
for the OpenOfice.org office suite.

Luckily, under Sun, some communities had created independent "local"
community organisations. Famously one in Germany, one in France, one in
Brasil and so on. This made it easier for those places to set-up events
and respond to "local" situations faster - without needing to run and ask
"Sun" if every 'little' thing was ok with them. The Brazilian one had
added some of their own coding making BrOO, yet another fork(ish?). Mostly
these properly registered independent organisations were able to hang onto
their own assets. The German one had enough funds, and enough expertise,
to lend "The Document Foundation" enough funds to "start-up" as a "new"
charity/business/organisation - and to give outsiders confidence in the
legitimacy and future prospects of the new charity/business/organisation.
Presumably that has long since been repaid or become irrelevant and TDF has
shown that it is excellently well managed and gained a strong reputation in
it's own right!

Oracle kept the name and the branding so the newly formed "The Document
Foundation" had to develop something new, even if it seemed like it would
only be temporary and that Oracle would do the sensible thing and just give
the name and branding back to the community. Some people stayed with OO.o,
even under Oracle, and went through a ton of hardships there. The press
and media blew it up as a fight between them and us when really it was
still one community with 2 slightly diverging office suites and 2 different
organisations "in charge".

Go-oo quickly (well, in under a year) merged their changes into the main
branch. Almost everyone in Go-oo was already heavily involved in
LibreOffice anyway. For a year or so afterwards their website carried a
really sweet and somewhat triumphant message saying that they had closed
down and gone to TDF and LibreOffice. It's gone now but the domain is
still hosted somewhere.

Almost all Gnu&Linux distros switched to having LibreOffice as the default
office suite quite quickly, even many that had previously used KOffice
(which also forked at around that time) as their default office suite.

I think NeoOffice and BrOffice also merged back into main-branch of
LibreOffice too along with other less-well-known forks and extra projects
although i've not kept track of what they have been doing.

Oracle attempted to claim their OpenOffice was the superior by suddenly
working frantically towards a new release, which they numbered just
slightly higher than LibreOffice's numbers at the time but by then they had
already lost the impetus and their paid devs weren't familiar with the OO.o
code-base so their 'new' version didn't have anything like as much polish
or new features. TDF responded to the challenge by simply re-numbering the
releases they were already working on at the time and swept in some new
features they had been going to leave for the next "Fresh" branch. So TDF
got their version out days earlier than Oracle and with it looking much
better too according to all the reviews at the time. By the time Oracle
released their 3.4 (or was it 5?).0 a few days later it was tooo late and
unimpressive so few, if any, articles appeared about it - except to compare
it against the LO one that 'everyone' had already been using.

Oracle finally seemed to wake up to the fact they weren't going to be able
to compete and weren't going to be able to split the program/suite up in
order to make an "enterprise" or "professional" version to profit from.
They seemed to see it as a "mill-stone around their neck".

At the time they were in court fighting against Apache. IBM allegedly
managed to convince them to 'give' OpenOffice away. Better a millstone
around an opponents neck than around your own when you are trying to fight
someone, right?

Since then OpenOffice has really flown. Their community, along with some
new people from Apache, have done some amazing good work. With their new
owner just letting them "get on with it", rather than constantly fighting
against them, made a huge difference! It then became much easier for
people to be in both communities and to some extent share work across both
projects. Sadly by then so much development had already gone into
LibreOffice that the two projects really had diverged from each other so
sharing code is often not possible any more - but that hasn't stopped
people working in both or sharing ideas across both or just helping each
other personally. IBM eventually 'gave' "Lotus Symphony" to Apache
OpenOffice so they could merge.

Hagar helped me with one of our wiki-pages by basically letting me just
copy&paste one of his help-pages from their forum - and helped me with some
of the changes it needed too. The head of the LO Documentation Team spent
a lot of time heading up their Documentation Team too. All across the
projects there are people working together quietly.

So throughout the history of this forking and so-called fragmentation it's
actually been a case of merging and consolidation - with even the 2 main
apparent 'rivals' working together to a much greater extent than outsiders
would understand.

Regards from
Tom :slight_smile:

#21

Hi :slight_smile:
The "Published Guides" for OpenOffice or LibreOffice are often the best
documentation for AOO or LO - at least for English readers. The LO ones,
including incomplete books and archived versions, can be found at;

https://wiki.documentfoundation.org/Documentation/Publications

for free or bought from the Lulu bookstore as proper paper-back books. I
bought a few of these and I'm really glad i did even though i haven't read
all of it yet! I think i managed to time it right so that Lulu were
offering one of their frequent discount weekends or something.

http://www.lulu.com/shop/libreoffice-documentation-team/getting-started-with-libreoffice-42/paperback/product-21682463.html

The guides are also available in various App Stores or online book-stores
for reading on-screen; such as the Apple store, the Ubuntu one and maybe
others. They usually cost a bit but not much. Enough to cover the costs
of publishing them in those places and a little more to allow a bit of
expansion in their distribution.

Also available on the official LibreOffice website (for free of course!);
http://www.libreoffice.org/get-help/documentation/

For non-English languages it is usually the 'in-built' help that is far
better.

There are also on-line videos such as the excellent range at Spoken
Tutorials;
http://spoken-tutorial.org/
although they cover a much wider range of OpenSource projects so you'd have
to hunt for LibreOffice in that lot :slight_smile: Also while English is excellent
they also do many languages from around the Indian basin as it's a
non-profit organisation largely funded by the Indian Government.
Regards from
Tom :slight_smile:

#22

Would you be so kind as to tell us which aspects got worse?

This one is a tiny example but symptomatic for the whole development
process:

https://bugs.documentfoundation.org/show_bug.cgi?id=94557

New "feature" without specification nor reason plus bad implemention.
I'm grateful that this "feature" will be reverted because "my users"
suffer from it every day. Well, "my users" preferred the green one over
the blue one.

#23

I never install any fresh brunch. If I install LO, then I install some
version that works acceptably well. LO 5.0 was not even Alpha because
large parts of the Calc component remained untested.

#24

Hi :slight_smile:
I think that is one thing many of us on this list agree on. I think most
of us have said it at some point or other over the years.

I think it's a travesty that "Fresh" appears to be the only option on the
official downloads page to first-timers and that getting the more solid,
reliable "Still" requires a lot more geekiness. To my mind it 'should' be
the other way so that people have to dig a bit for the shiny new features -
with the majority of new-comers getting the reliable "Still", at least
until the curiosity get piqued.

So often the answer to a noobs problem is to simply try the "Still" branch
and then trying to explain why and that it's not "old". As soon as they
try it that is usually enough to fix almost any noobs problem.

It's the same with a lot of articles out there. When an article claims
that LO is not mature and had obscure problems it's almost invariably
because they were using "Fresh" rather than "Still".

"Fresh" is exciting but it should be more like an "easter egg" in a movie
or game that has to be discovered or some challenge over-come in order to
find it.

Regards from
Tom :slight_smile:

#25

Hi :slight_smile:
Looks like a teaching lesson where the dev learned a lot about practical
and real-world coding that maybe hadn't been taught in his largely
theoretical studies.

Looks like they had the suitable qualifications on their CV but lacked a
bit of "street"-wisdom.

Also looks like most of their commit was smart and useful and has been
retained while only a small element at the start is being held back for
them to re-write with their newly learned wisdom.

So while it was annoying to one of us (and his clients) most of us wouldn't
have even noticed it and that it got fixed pretty quickly. Top marks for
posting the bug-report and helping them with it. That has helped us all
even before most of is realised there was a problem :slight_smile:

It's a good example of why it's good for most of us and other experienced
users to try "Fresh" before foisting it on our clients or on noobs.
Regards from
Tom :slight_smile:

#26

Please avoid entering in a discussion with a well know enemy of
LibreOffice. Andreas Saeger aka Villeroy has been spreading FUD about
LibreOffice since forever. People happy with a dead and buggy software -
aka Apache OpenOffice - should avoid commenting on LibreOffice mailing
lists.

#27

Hold on: you suggest that people with opposing views should be silenced? Most civilised democracies value the expression of opposing views, even at the highest level. Oh, but despots and dictators don't.

Brian Barker

#28

People happy with ... Apache OpenOffice - should avoid commenting on LibreOffice mailing lists.

Hold on: you suggest that people with opposing views should be silenced?

suggesting someone avoids commenting is not the same as silencing them.

but like you, I want to hear different perspectives.

however, it seems the rhetoric, and heart rates, are rising.

the furniture, gentlemen, mind the furniture! we can have an interesting conversation without breaking it.

f.

#29

suggesting someone avoids commenting is not the same as silencing them.

It's asking for their silence.

but like you, I want to hear different perspectives.

Good-oh!

Brian Barker

#30

Hi.
For one, printing documents with EPS. I still have to retain a copy of AOO for printing. It is the only thing I have to revet to AOO for.
Steve

#31

Would you be so kind as to tell us which aspects got worse?

Would you be so kind as to tell us how to enter some numbers into a
Writer table?
LibreOffice 4.4
menu:File>New>Text
Ctrl+F12 [Insert]
menu:Table>Number Recognition is checked
Locale is English(UK)

09 3.14
1999-12-31 5.87%
23:45:59 True

Everything is text.
Oh, the number format code of my new table is "@". May be this is the
reason. Change it to "General" since I have entered various subtypes and
"General" is the number format to accept and represent any subtype.
To my surprise, changing the number format attribute converts the text
data to this:

9 3.14
36525 0.06
0.99 1

Everything is converted and rounded to integer and 1/100th decimals.
23:45:59 --> 0.9900000

OK, let's append a new table and apply number format "General" _before_
typing in numerals. Wow, this time it seems to work.
But why is
=SUM(<B1:B3>) = 4.1987000 (correct)
and
=COUNT(<B1:B3>) = 4.1987000 (expected value is 3)

And why is
=SUM(<B1:B3>) = 3,14
after reloading the same file with a German locale?
Cell B1 has still the "General" number format, B2:B3 is text.
Almost the same in column A. A1 is "General" A2:A3 is "@" but the
forumal cell in A4 still shows the old result cut down to 2 decimals.
When I hit F2 to inspect the formula, the formula bar shows an equal
sign only.

By the way: Did anybody notice that the glorious "date acceptance
patterns" do not work at all?

https://bugs.documentfoundation.org/show_bug.cgi?id=94727

Just watch the screen shot and the spreadsheet. Every spreadsheet in the
world accepts 2/10/ or 2/10 or 2/ (this month's 2nd) on the num pad.
With German LibreOffice locale it is impossible to enter any date on the
num pad without typing a full ISO date.
You messed this up. The dumb users are happy because they get what they
type as literal text. The educated users (normal users, no experts) are
wondering why there are any "date acceptance patterns" if they get
invalid text data anyway.

#32

Suffice it to say that Andreas is a vociferous participant in this
discussion list, but that doesn't make his criticisms any less justified
or relevant. What he dislikes is badly implemented change for change's
sake, and that is an inherent problem in LibreOffice's development. The
project from the start has sacrificed behavioural stability with regard
to the end user for feature creep. We are quite clearly in the "bazaar"
mode of the cathedral and bazaar dichotomy, where no overlying
dictatorship (benevolent or otherwise) exists to govern the direction
code development should take. This has positive and negative effects -
the positive being that people can just turn up and work on the thing
they want to implement - the negative being the law of unintended
consequences, or collateral damage, i.e. bugs newly introduced that
change long standing behaviour to which users have become accustomed.

Fortunately, there are still people like Andreas to call the code
contributors out on those decisions.

I would suggest putting yourself in an admin's place where they have
probably invested long hours in developing a turnkey
OpenOffice/LibreOffice solution for their group of users, then finding
one day that that longstanding behaviour has changed because someone
else has not thought through a code change due to the tentacular nature
of the code base with no one having an overarching knowledge of it all,
and you will perhaps understand Andreas' frustration (which I happen to
share and have voiced it on the mailing lists in the past).

At present, long term support (bug fixes, security updates) for older
versions is to my knowledge only available on Linux and only with regard
to certain distributions. If you are not on Linux, then you are stuck
playing catch up with versions that successively introduce new bugs or
behaviours that don't get fixed for at least several point releases, or
for certain OSes, over multiple major version releases. Steve's mention
in this thread of EPS support and printing is just yet another
illustration of a change that was made that has a huge impact on
non-Linux OSes - all because someone thought it would be a good idea to
make that change without providing a solution for all platforms. Video
support in Impress is yet another issue that got significantly worse
with the move to the 4.x branch. What was the message we gave to our
users ? "Suck it up." There is only so much of that that users and their
admins are prepared to do, and in the end, it won't be surprising if
people switch to another product that offers them greater longterm
stability where such changes are less invasive or devastating to the
day-to-day running of the organisation.

Alex

#33

+20

#34

Macro coders depend on either MRI (Python extension) or XRay (Basic
library). Both tools can be linked to the alphabetic index of the IDL
reference for OpenOffice, however LibreOffice changed everything and
there is no index that can be accessed by these tools. But hey, the
OpenOffice reference is still fine because many of the subtile LO
"improvements" are not visible in the API anyway.
Oh, and most of my Python stuff broke due to the switch from "stable"
Python 2.x to "fresh" 3.x.

... to be continued ... (disregarding Mr. Vitrioli from Italy)

#35

Alex,

I do not know where you got that support and security updates are only available on Linux. That is factually wrong and serious bullshit. Get your facts straight: support is the same for the three officially supported platforms: Windowslinux and OS X. Remember that many code contributors have customers too.

As for calling developers on their responsibility that is quite easy especiaIly when that call takes an oracular form: doing it in such a way is one of the things defining a troll. I wonder if Andreas does the same for AOO ? Something tells me that is not the case but I could be wrong.

Best,

Charles.

#36

Hey Andreas,

Quick question: when was Python 3.x released ? Several years ago - at least 5 if I am not mistaken. So the ecosystem is moving after all this time so such migrations should be expected.

Best,

Charles.

#37

Charles,

I do not know where you got that support and security updates are only available on Linux. That is factually wrong and serious bullshit. Get your facts straight: support is the same for the three officially supported platforms: Windowslinux and OS X. Remember that many code contributors have customers too.

I would be interested in learning where longterm support for "older"
versions of LibreOffice on OSX which involves security updates and bug
fixes is generally available. It would certainly help me explain to
people on bugzilla using older versions of LibreOffice on OSX why there
is no hope of bug xyz being fixed unless they upgrade both their OS and
install the latest version of LibreOffice.

If I look here :

https://www.libreoffice.org/get-help/system-requirements/#Apple

Either I can't read properly or else the information at that place is
incorrect. As I was unable to find anywhere on the LibreOffice website
that indicates that there is support available for "older" versions of
LibreOffice on OSX < 10.8, then may I assume that my facts are indeed
right after all ?

Alex

#38

Hi all,

Suffice it to say that Andreas is a vociferous participant in this
discussion list, but that doesn't make his criticisms any less justified
or relevant.

Indeed. Maybe I do not have the same opinion as Andreas always, but his
technical contributions are useful and IMO much appreciated.

[...] The
project from the start has sacrificed behavioural stability with regard
to the end user for feature creep.

I fully understand your concern on development, and partly share it.
I seriously don't think you can uphold the opinion that all major and so
much needed improvements are 'feature creep'.
Italo is simply correct when he refers to AOO as a dead and buggy code
project. The technical debt that LibreOffice started to fight in 2010
was one of maybe already a decade or more old. New features in
OpenOffice.org in the last years were even implemented in Java
extensions, so that the core code could be avoided..
But doing that work is a pain and needs many hands.

The good thing is that there is now a wonderful amount of people helping
on the code. Still more welcome, but we may feel happy with where we are.
The sad thing is that on the QA side less people contribute and things
can be further improved. This affects potentially all users, of whom
many could be valid contributors :slight_smile:

Thus it is not without reason that LibreOffice has two branches. And
depending on your work, skills, time etc. there is a wide range to chose
from the last updates in the still branch, to early in the Fresh line
and even daily builds. The latter are great for most of the work, but of
course not for professional deployments.
And there are supported LTS versions - see below.

Fortunately, there are still people like Andreas to call the code
contributors out on those decisions.

And he is not alone in that.
But note that the ESC and the board are working (after all that already
has been done) on more ideas to further improve:
https://wiki.documentfoundation.org/Development/RegressionIdeas

Mind that not changing and improving compared where we are and were five
years ago, is no option.

At present, long term support (bug fixes, security updates) for older
versions is to my knowledge only available on Linux and only with regard
to certain distributions.

To my knowledge bot Collabora and CIB offer versions with much longer
support. Quite affordable and really valuable for organizations.

Cheers,
Cor

#39

Mind that not changing and improving compared where we are and were five
years ago, is no option.

Agreed, but not, and I am expressing my own opinion, to the detriment of
dropping a software behaviour nuke on the users/admins.

At present, long term support (bug fixes, security updates) for older
versions is to my knowledge only available on Linux and only with regard
to certain distributions.

To my knowledge bot Collabora and CIB offer versions with much longer
support. Quite affordable and really valuable for organizations.

Sure, but both of these focus on at least LO 4.4.8, from my reading of
their respective websites. That is not an "old" version, it is the
latest version from the Still branch.

Alex