A Tale of Two Office Suites

users
#21

Hi :slight_smile:
I think corruption is a lot rarer than might seem from these lists.  This list is available all around the world and people of all skill levels post here.  So we often get questions from people that have absolutely no experience of running OpenSource things on their system and that seems to be the vast majority of people that first approach these lists.  (or they might have used Firefox and sometimes Thunderbird without realising they are OpenSource too).

Of the upwards of 60 million users using LO we have had just a few score questions involving a corrupted User Profile.  Making about 0.0002% of users.  So the odds are about 99.9998% that you wont have a problem.  If you are that unlucky then you already know how to fix it easily.  Also you already know how to create a back-up before running into trouble.

Regards from
Tom :slight_smile:

#22

Sorry, but no.

First, how do you compute number of LO users? Some time ago Rob Weir (sadly he
is biased towards LO) that there are some uncertainties on how TDF counts LO
downloads.

Second, count out all LO users who don't speak English (that is, how many? No
one really knows and it would be really hard to get this number right). They
are not capable of commenting LO bugs.

Third, we don't really know what percentage of English-speaking LO users take
time to report or comment bugs they have encountered. There are - perhaps
majority - of users who encounter bug but try to work around them themselves.
Or just dump application and use another one.

And, last but not least, bugs tend to be deterministic - they will ALWAYS
happen under certain circumstances (which may be uncommon). So if you want to
talk about "odds", you must tell what are the odds that certain user
encounters certain circumstances that trigger bug. And this is place where
things are getting really complicated.

[0] http://www.robweir.com/blog/2012/10/libreoffices-dubious-claims-part-i-
downloads.html

#23

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I'm obviously not communicating very well.

I haven't had any "conflict between configuration profiles or during

normal usage that leads to a crash or data loss" from running all three
programs, so I currently have no reason to file any bug reports.

So, again, my question.

Why is it not advisable to run LO and AOO on the same system? What

harm does it do to my system?

Because you may have a "conflict between configuration profiles or
during normal usage *that leads to a crash* or data loss".

/**** THAT LEADS TO A CRASH *** <--- POSSIBLE HARM TO YOUR SYSTEM*/ (ie.
software/OS normal operation)

If your data isn't critical, this won't matter much. I run similar
configurations on a Windows test system for example.

I believe in your very first post you also indicated the forks will
differ so much in the future that this won't remain practical/compatible
enough.

F.

#24

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

[...] Of the upwards of 60 million users using LO we have had just a few score questions involving

a corrupted User Profile. Making about 0.0002% of users. So the odds are
about 99.9998% that you wont have a problem. If you are that unlucky
then you already know how to fix it easily. Also you already know how to
create a back-up before running into trouble.

When this happened, it hit me hard. I lost data. I had advised a bunch
of people to use LibO which was crashing and loosing data, 100% of the time.

This improves over time, but I generally prefer not increasing the odds
of that happening - such as running a beta concurrently with stable/AOO
version.

You make your luck too :wink:

F.

#25

Are you using this extension?

Compose Special Characters - extension - Compose Special Characters lets you type two or three characters and use a keyboard shortcut to convert them into a single accented or special character. You can also compose unicode characters using its 4 character unicode value.

#26

...

Please think about this: "Don't say/write everything that you think, but
think everything that you say/write"

...

I *highly* recommend that you follow your own 'advise'.

#27

There is no reason to not use LO and AOO on the same system. They use different user profiles and are completely separate programs.

#28

...

I also have that version installed:
$ locate soffice.bin
/opt/libreoffice3.4/program/soffice.bin
/opt/libreoffice3.5/program/soffice.bin
/opt/libreoffice3.6/program/soffice.bin
/opt/openoffice.org3/program/soffice.bin
/usr/lib/libreoffice/program/soffice.bin <== Ubuntu version 3.5.4.2

LibreOffice 3.5.4.2
Build ID: 350m1(Build:2)
This product was supplied by The Document Foundation, Debian and Ubuntu.

Interesting, mine does not.
Lowercase (c) = (c)
Uppercase (C) = ©

Note: I've also tested the same on my Windows versions.

I wonder if perhaps it may be due to the lanuguage/local.
$ locale
LANG=en_US.UTF-8
LANGUAGE=
LC_CTYPE="en_US.UTF-8"
LC_NUMERIC="en_US.UTF-8"
LC_TIME="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_MESSAGES="en_US.UTF-8"
LC_PAPER="en_US.UTF-8"
LC_NAME="en_US.UTF-8"
LC_ADDRESS="en_US.UTF-8"
LC_TELEPHONE="en_US.UTF-8"
LC_MEASUREMENT="en_US.UTF-8"
LC_IDENTIFICATION="en_US.UTF-8"
LC_ALL=

#29

LibreOffice and Apache OpenOffice use different user profiles. They are completely separate programs and there is no conflict.

#30

(c) does not, (C) does. LO 3.6.0.2 on Ubuntu 12.04....

#31

Should be: Rob Weir (sadly he is biased against LO).

Tom pointed it out off the list; thanks!

#32

(c) does not, (C) does. LO 3.6.0.2 on Ubuntu 12.04....

...Same on Ubuntu Raring LO 3.6.2.2

#33

Don't misunderstand me.

I'm not upset about having to delete the "(c) to copyright symbol" option in
my autocorrect feature. I appreciate that many people prefer this behavior.
I get it that developers have to make choices and they make those choices on
the basis of what a majority of users want. That much I get.

Understand. The auto insert symbol most likely goes back to this 2002
OOo bug:
<https://issues.apache.org/ooo/show_bug.cgi?id=4579>
[Special Character Shortcuts]

The initial definitions for the autocorrect are located in the /autocorr
directory. They are simple ziped .dat files (DocumentList.xml is the
actual file) & probably could be modified & then copied across like a
standard template.

What I don't like is having to do it twice because neither AOO nor LO has
all the features I need to get my work done, and that is because, for
whatever reasons, the developers of the two office suites either can't or
won't combine their efforts.

You might want to look at the archives of the AOO dev list. The early
posts regarding this issue provide an interesting read.

#34

Running LO v 3.6.0.1 (Build ID: 360m1(Build:101)) on 64-bit Ubuntu 12.04, I
see what NoOp describes above. However, when I want to get © rather than
avoid it, I just perform «Alt Gr + c», which is simpler (this on a standard
105 key Scandinavian keyboard)....

Henri

#35

Hi :slight_smile:
Errr, sharing your User Profile between AOO and LO sounds like a really BAD idea after all.  I thought it would be jolly clever to try it but if there is risk of DATA LOSS then please do AVOID it!!

(Another advantage of a mailing list / forum is that mistakes hopefully get noticed before causing problems.  Data loss is a pretty serious problem)

Apols and regards from

Tom :slight_smile:

#36

Virgil,

Thanks! It can't be said better!

Isn't the statement that competition helps to improve not simply an excuse to not being force to work on an attempt that both (LibO & AOO) teams can work together again?

Isn't MSO a good competitor, which helps improve an OpenSource Office (the combination of LibO and AOO) suit being available for those who just cannot afford MSO?

I really would like to understand what attempts have been made to get both teams together and why the attempts failed? And when will the next attempt be made?

I German is a phrase which I found being translated at LEO into English as: "Constant dripping wears away the stone". The Japanese say: "Until the ears hurt".

I hope the responsible persons of LibO and AOO keep talking to each other until they find a way to cooperate again as one team to create the best Office Suit available.... and affordable for those with less financial resources. That is the real challenge and worthwhile to go for it.

ROSt52

#37

Hi all,

Interesting, mine does not.
Lowercase (c) = (c)
Uppercase (C) = ©

Unfortunately, I've noticed this too on Mac, with FR as the default language / locale, and it is rather annoying when typing lists within a paragraph. Mind you, it has been like that for a very long time, even when I was previously using OpenOffice.org.

Alex

#38

Hi :slight_smile:
If there is no preceding open-bracket the specific problem doesn't happen?  So a

c)
doesn't get changed but a

(c)

does?  Also using proper "Bullets and Numbering" doesn't change the (c).  Hmmm, tricky.

With txting on my phone i find if i type a string of upper-case it switched to entirely upper-case but if i type a space or . between each character it retains the sentence-case that i generally prefer.  So, what happens if you type ( c ) or (c ) and then go back and delete the extra space(s)?

Regards from

Tom :slight_smile:

#39

Hi Tom,

I have to leave a space either after the opening bracket, or before the
closing bracket, in order for (c) not to be converted into the copyright
symbol. If I enter a space, close the bracket and then delete the space,
then leave that and start typing the next word, it automatically
converts that sequence back into the copyright symbol.

From reading the French mailing lists, it might be something you can

turn off in the autocorrect dialogues.

Alex

#40

Try this folks:

1. Type (
2. Type C
3. Type )
4. Type space
5. Previous three characters become a copyright sign automatically
6. Type ctrl-z and step 5 is undone
7. Type ctrl-z and the space is deleted.

This seems to be general behaviour throughout LO.
Automatic changes are treated as a virtual keystroke and can be undone.

Has anyone else had similar experience?

Barry

PS I've just done it again to check. - B