Guest post by The Document Foundation (TDF) has announced the release of version 3.4.2 of the open source LibreOffice office suite. According to TDF Steering Committee member Italo Vignoli, the third release in the 3.4.x branch of LibreOffice is now ready for both individuals and enterprises, and can be used for production.
Vignoli says that 3.4.2 is the result of the “combined activity of 300 contributors having made more than 23,000 commits, with the addition, deletion or modification of around five million lines of code”. SUSE and Oracle have each provided approximately 25% of the total commits. New individual volunteers accounted for another 25%, while 20% came from RedHat; the remaining contributions came from Canonical and other companies.
Further information about the release, including a full list of new features, can be found in the release notes and on the 3.4 New Features and Fixes page.
The fork of OpenOffice was a good thing for those who switched to LibreOffice. OpenOffice remains frozen with an incompletely released version 3.3. While I assume IBM is facilitating some development, Oracle clearly isn’t, as there are no newer release candidates on the OpenOffice site.
Meanwhile LibreOffice cleans up the code and adds features at a great pace. In particular, I appreciate the improved file format support, and the better on-screen display for some fonts, such as Free Serif.
I really appreciate all this effort, and think it is marvelous that so many new people are helping out.
Sounds like OpenOffice is basically dead in the water now and Libre is the ‘new thing’ ?
OpenOffice is not dead in the water, it is now an Apache project, and Oracle (and I presume IBM) are still contributing to it.
The thing is, because of the Apache license, LibreOffice can also incorporate any changes Oracle makes to OpenOffice.
http://www.h-online.com/open/news/item/LibreOffice-3-4-2-is-enterpr…
Oracle have ended up providing about 25% of the changes that have gone into LibreOffice.
By the same token, this same picture conveys the fact that there are roughly four times as many changes/improvements going into LibreOffice as are going into OpenOffice.
Edited 2011-08-09 02:27 UTC
Seems kind of dumb to have these two projects running in parallel, but whatever …
Ask IBM … IBM apparently wanted to be able to take the OpenOffice codebase and add proprietary bits to it and release the result as closed source (Symphony Office). The Apache license allows this.
http://en.wikipedia.org/wiki/IBM_Lotus_Symphony
License: Proprietary (registerware)
The open source community, however, wanted a codebase that would remain open forever, that everyone could use at low cost to their mutual benefit, that no one party could exclusively benefit from, and a codebase they could contribute to but still retain the copyrights to their own work. The GPL license ensures this.
There are a lot more contibutors in the latter group than the former.
Edited 2011-08-09 03:22 UTC
Why did this get a downvote, for goodness sake!
It was a perfectly factual post.
Apache’s OpenOffice hasn’t really taken off yet: it’s easy to underestimate the huge amount of resources that SUN/Oracle had into this and that now have to be migrated. Of course the trademark is valuable but there’s a lot more in the scope of the grant that Oracle is making to Apache.
It’s just my personal position, but I think it will be very difficult for LibreOffice to enforce the (L)GPL given that:
-The vast majority of their code is now under a liberal license. Despite their efforts, anything they take from the new Apache OO and IBM (which has spent many man-years on this) will further dilute the remaining copyleft content.
– TDF doesn’t really own any code at all and the new copyright owners are not likely to sue anyone for copyright infringement.
So.. perhaps Apache OO is the place to look at after all.
Perhaps this is a slightly skewed way of looking at it.
The GPL only applies to code under the GPL. That is, code which the authors place under the GPL. As you say, much of LibreOffice is now also under a liberal Apache license.
The point is that it is also under a GPL license, and there are a lot of contributions now being made (via TDF) that are available ONLY under the GPL license.
“Enforcing” the GPL in this context means only that any of the code which is available ONLY under the GPL is always re-distributed under the GPL. There is no “enforcement” to be made concerning code which is also available under the Apache license. That simply doesn’t make any sense.
Of course it is
.
The objective of copylefting some part of the code is that it is supposed to taint all the code that is linked to it. In this case it’s the LGPL so the scope is limited but nevertheless the license is unenforceable.
Suppose I take massive code from “LibreOffice” and make it into “MyOffice”: I have no interest to share my code and new features with TDF. First TDF will have a tough time proving that I took some features from their code and not from OpenOffice.
TDF is not requesting copyright assignment so then, assuming the LGPL’d code is significant enough that it can be traced they also have to make sure the code was not submitted elsewhere under a different license, they will have to find the copyright owner (Maybe Novell if it still exists) to force me to comply.
If the LGPL’d contribution is just a small patch I may just get with it. If the LGPL’d contribution is big (which hasn’t happened yet) I can just make it into a library and comply .. iff they do get to sue me.
The license “war” looks pretty much irrelevant.
Whoever writes the code is the copyright owner.
You cannot take someone else’s code and then release it under a new license and thereby place restrictions on it, or indeed lessen restrictions on it … you simply don’t have that right if you don’t have “copyrights”.
Hence, the TDF can only have any influence at all, and any right to sue, over code that is either authored by the TDF, or code for which the author assigns copyrights to the TDF.
None of the original OpenOffice codebase falls under this description. The TDF cannot sue anybody at all over this code.
As for the inane contention that the GPL is somehow meant to “infect” other code … utter nonsense. The GPL applies ONLY to code released by its authors under the GPL.
The only occasion that LibreOffice code can engender a lawsuit is when someone takes code unique to LibreOffice (newly written just for LibreOffice, and licensed by its author under the GPL) and incorporates THAT code in some other proprieatry program.
Even then, LibreOffice does not have ownership rights over the other program … the new program (incorporating in small part some of LibreOffice new code) is what is known as a “derivative work” under copyright law. Ownership of the derivative work is shared between the authors of the parts of the derivative work according to the relative proportion of each bit.
The GPL is frequently referred to as “viral” for a reason. What you described is CDDL not GPL.
We are in agreement, however it’s only the copyright owner that can sue: TDF has no “teeth” and is unable to enforce it’s own license, simply because it doesn’t own any code.
The same License Agreement that they so much criticized from SUN/Oracle is now their weakness, but then .. the code was never theirs to start with.
It is, and that reason is FUD. The GPL is not viral.
No, what I described is copyright law. Copyright law says who has ownership of the copy rights, and who does not. Not the CDDL, not the GPL, not proprietary FUDsters or anyone else, but rather the law of the land.
Not only is the original OpenOffice codebase not the TDF’s code, the GPL license that applies to it is not the TDF’s license.
It was Sun who released the OpenOffice codebase under the GPL.
The code was never whose to start with? The code certainly did belong to Sun. Sun acquired this code when the bought a company called StarDivision in 1999.
http://en.wikipedia.org/wiki/StarOffice
The source code of the StarOffice suite was released in July 2000, creating a free, open source office suite called OpenOffice.org.
http://en.wikipedia.org/wiki/OpenOffice.org
License: GNU Lesser General Public License v3 (OpenOffice.org 2 Beta 2 and earlier are dual-licensed under the SISSL and LGPL)
Here is the story regarding ownership of that code:
http://en.wikipedia.org/wiki/OpenOffice.org#Ownership
It now belongs to the Apache Software Foundation.
However, the nature of the GPL license allows re-distribution of the code by anyone. Hence TDF can take the code, develop it, and re-distribute the result also under the GPL license, and repeat that cycle for as long as they want to.
Since the TDF do not own the codebase, what TDF may NOT do is start to re-distribute LibreOffice under any other license. TDF must always distribute LibreOffice under the GPL. This same rule, which comes from copyright law, applies also to anyone who receives LibrOffice code. If they take LibreOffice code, and modify it and then re-distribute it, they to do not own that code, and they must re-distribute it under the GPL.
The GPL license does not transfer ownership, it only transfers permissions to use, modify and re-distribute the code. The code still belongs to the originator.
Since Sun, and then Oracle did own the codebase, this is what allowed Oracle to give it over to the Apache Foundation to release it under the Apache license. Because they owned the code, Sun releasing the OpenOffice code under the GPL did not mean that Sun could not re-release it to Apache, or indeed make a proprieatry release of it also.
This is because the GPL is not viral.
Edited 2011-08-12 04:47 UTC
Also, one of the things that I have never seen reported in the mainstream media is the performance boost of LibreOffice that has come from this massive code cleanup.
Anecdote: I own an under-powered netbook, an Acer Aspire One 522, which uses an AMD Ontario APU (combined CPU and GPU). The CPU is a dual core AMD64 part, but it is clocked at only 1GHz, and the machine has only 1GB of RAM. I have both Windows 7 and Kubuntu 11.10 Alpha 3 installed (dual boot) on this machine. On this hardware, LibreOffice loads and runs noticeably faster under Kubuntu than MS Office 2010 loads and runs under Windows 7.
Edited 2011-08-09 02:29 UTC
Probably, because it’s not sure at all that there is a performance boost, the code cleanup is for improving maintainability, it may also improve performance but that’s just a side benefit..
As for your anecdote, sorry but it’s quite meaningless, to show that the code cleanup improves performance you should compare OpenOffice to LibreOffice on the same OS, not two difference office suite on two different OS!
There has been a very long-standing claim the OpenOffice was a long way behind the performance of MS Office, especially OpenOffice on Linux.
By comparing LibreOffice under Linux to MS Office under Windows 7, I am supposedly penalising LibreOffce!
When I next boot my netbook to Windows 7, I will install LibreOffice 3.4.2 on Windows 7 and compare it in speed to MS Office 2010. This way I won’t be accused of giving LibreOffice an advantage, or a penalty for that matter, by running it under my normal OS.
This may be a while, however, since I don’t have cause to boot Windows very often at all.
You have the reason of running the speed test.
Actually, this is something I’ve thought about and keep running into two questions:
– what affect does MS Office integration with Windows have on the findings? Office is using much of the already loaded IE widget set.
— expect shorter load times since much of Office is loaded with the OS.
— expect lower resource usage increase until large saved files are opened (an Office app with blank document is essentially displaying what is already in memory)
– what affect does the LibreOffice “load on startup” function provide to counter-balance Office pre-loaded benefits and how affective is it?
I like the New icons introduced by LibreOffice.
The only problem is they make the rest of the suite’s UI look very dated.
They also need to drop the “AA_A_” Icons used instead of “BIU” for Bold Italics and Underline – the guidelines supporting them are totally weird.
And if you don’t speak English, BIU will not be meaningful.
Well, then they could be replaced with the appropriate characters for the language in question, as long as there aren’t any duplicates. I’d hazard to guess that english, german, spanish and french are the four most popular languages for Libre. Would it really be tough to do it just for those four and revert to the other connotation for other langauges? I’m guessing the A based indicator may not make sense for non latin alphabets.
True, B[old] I[talics] C[ursive] is not so meaningful is you don’t speak English.
On the other hand, my native tongue is Dutch, and I would still prefer BIC because it correlates with keyboard shortcuts.
For whatever reason, the Norwegian version of MS Office uses FKU (Fet, kursiv, understreket) instead of BIU … and they’ve changed the shortcuts to match. (And yes, that means Ctrl+F isn’t search … they swapped it with Ctrl+B.)
At least they’re consequent.
BIU replaced the AAA in the English UI. The letters are localized.
It’s a shame that Powerpoint compatibility bug mentioned in the Release Notes wasn’t fixed*, as it’s the only thing keeping me from trying out the 3.4 line. I was hoping it’d make it into this release, but I guess I’ll have to wait another month…
*: Or rather, it was fixed early/mid July, but didn’t get merged for some reason:
https://bugs.freedesktop.org/show_bug.cgi?id=32709
Edited 2011-08-09 00:41 UTC
LibreOffice set up recently a compelling update calendar (a minor one every month, a bigger one every six months). I installed it directly on a Linux box. However, if I wish to enjoy the benefit of a new update, I must download it fully.
This is a self-defeating purpose. I think LibreOffice should set up delta-replacing only updates. Without this, I only update when I cannot bear anymore some pesky bug. If not, I just hang on…
Edited 2011-08-09 06:34 UTC
If you are using Fedora, you get delta updates by default. So not a problem here
http://www.fedoraproject.org
(just to chime in on the list)
Debian as libreoffice in the Squeeze-backports so you should be kept up to date easily that way also.
It would be nice if LibreOffice could sort out their Linux packaging – it has so many issues, it’s completely embarrassing. Here’s a list:
* The downloads are three separate .tar.gz files containing RPMs, but two of the three could be merged (the langpack and helppack – both are locale-specific and the latter only has one RPM inside the .tar.gz!).
* The RPMs have the X.Y version number in the package field which is an admin nightmare – rpm -Uvh the 3.4 packages and the older packages (e.g. 3.3.X) remain on your system and you get two sets of everything! Who, other than developers, would want to install two minor versions side by side? This mess probably also explains why there isn’t an RPM repo for Libreoffice either.
* You unpack the core .tar.gz, to be confronted with literally dozens of RPMs, when they only need about 6 (common code and an RPM per application).
* The en-US RPMs are dubiously included in the core .tar.gz, when they should be a separate langpack. In my en-GB case, I have to delete all the en-US RPMs from the core and replace them with equivalents from en-GB.
* For some inexplicable reason, the core .tar.gz includes Spanish and French dictionary RPMs, which is frankly a waste of space – these should be in the appropriate langpack .tar.gz’s surely?
* The testtool RPM is included – is that really needed for a production install?
* The desktop integration RPMs are in a subdir in the core .tar.gz (easily missed) – also wouldn’t it be better to have one such RPM that worked out your distro and installed in the appropriate place rather than a separate RPM per distro?
* After all this, I have to install 52 (!) RPMs (possibly removing the old version manually as I said before) – not good at all.
A pretty dire situation overall then, but I still persevere because it’s clearly now the best Office suite on Linux.
Just install it from your distribution’s repository using the package manager. Easy peasy.
> Just install it from your distribution’s repository using the package manager. Easy peasy.
Not so easy peasy if you’ve got a work desktop on a long-term distro (no, it’s not sensible to run Ubuntu or Fedora on work desktops). We just moved from CentOS 5 to 6 on our work desktops and that doesn’t even have Libreoffice in its repos! Instead, you’re stuck with OpenOffice.org 3.2.1 – all the more reason that it would be nice for an official Libreoffice repo to be available.
Edited 2011-08-09 16:12 UTC
That’s you problem, right there.
As a side note, I find it funny how people argue that you can’t use Ubuntu or Fedora on work desktops and then go ahead and use CentOS. You know, the distro that is RHEL without any of the actually useful enterprise parts.
Being supported for significant periods of time and Enterprise Software compatibility (backup tools, unified authentication, etc. that only ship as RPM) are the ‘Enterprise features’ people running CentOS are worried about. Not so much on the global file systems and whatnot.
Because that’s not available on any other distro…
Seriously, the only thing RHEL have that other distros do not is the support contracts and RHnetwork and such.
That’s why you get RHEL and why CentOS us pretty much pointless.
You can’t have long term releases with the latest stuff except when the distro has just been released. Over time, long term releases will basically backport selected bug fixes and security fixes.
Sounds like your screwed unless someone does a CentOS third party repository or better conslidates Libre’s direct RPM packaging.
(I’ve not looked at how they did the DEB packages)
Though you may not be in a possition to muck with the workstations, other distributions do a better job. But then, CentOS is much more focused on the server side versus a more general purpose distribution that has more focus on the workstation/desktop side.
OK, I clearly should have said: “Just install it from your desktop distribution’s repository using the package manager. Easy peasy.”
LibreOffice is, after all, clearly software for a desktop.
Since this is a very recent release, and most distributions are not rolling releases, then it may be necessary to identify and enable your desktop distributions “backports” repository, or the equivalent.
They did this because when 3.4 was released, it had more significant internal changes than 3.3, and was considered less stable. You kept two versions on your system to do real work with the older one, and to test the newer one. 3.4.2 is the first of the 3.4 series that the Document Foundation is calling stable. I didn’t find it hard to remove the older version.
I’m going to assume that the RPM version is packaged similarly to the DEB version I downloaded. I think that’s the case, because I go back to the days when only the RPM version was available, and I had to convert the RPMs to DEBs before installing. As an English speaker I issue two commands as root, or prefaced by sudo:
dpkg –force-all -i *.deb
After doing that once in the DEBS directory, I change to the desktop-integration directory, and press the up arrow and Enter to repeat the command. An install script that essentially does this would be nice, but its not necessary.
It sucks to be a native English speaker.
The command line is awesomely efficient. Here’s a question for you. I also use Synaptic from time to time. Does your graphical package manager offer search capabilities? You could do a search for libobasis3.3 and delete them all. If your package manager tracks dependencies, and I think it does, then it will take care of the rest of the LibreOffice 3.3 install as well.
Two things: LibreOffice is the best thing out there, and I suspect with their new lease on life, they might actually respond to some of your suggestions. Second, you persevere, but it need not be that big a challenge.
OpenOffice was always slow and clunky on my hardware. But I never had stability problems. It was rock solid. But unfortunately, it appears that LibreOffice is streamlining the codebase and implementing new features at the cost of stability. In my anecdotal experience, LibreOffice has introduced many new bugs, some of which I have reported, and some of which have been fixed. But it seems like every new LibO 3.x release fixes one bug and introduces several other new ones. In my experience, I have had many more full crashes of the suite after a few months with LibO than I ever did with OOo over many years.
So, here’s hoping that LibO developers just concentrate on stability, with speed as a secondary goal and features a distant third. I exclusively use LibreOffice, I don’t plan on going back to OpenOffice, and I don’t even own/use Microsoft Office, so this comment is intended to offer constructive criticism. I spend many hours in front of LibreOffice almost daily, so I sincerely hope that it improves in the future, especially on the stability front.
Edited 2011-08-10 05:19 UTC
LibreOffice 3.4.0 introduced new features, and it was not claimed to be stable by the TDF on release. It was for developers who wanted to help debug the code.
http://blog.documentfoundation.org/2011/06/03/the-document-foundati…
“The first release of the 3.4 series, LibreOffice 3.4.0, is targeted to community members and power users, and should not be implemented in a corporate environment.”
LibreOffice 3.4.1 introduced NO new features, it was a bug-fix release, and it was claimed by the TDF on release that this version was still not fully suitable for production, but it was now suitable for early adopters as well as developers who wanted to help further debug the code.
LibreOffice 3.4.2, which is what this thread is about, introduced NO new features, it was a second bug-fix release, and it was claimed by the TDF on release that this version was now suitable for production. This release is claimed to be “very stable”.
LibreOffice 3.4.x is not claimed to become “rock solid” until 3.4.3.
http://tdfsc.files.wordpress.com/2011/07/libreoffice-versions.png
A picture is worth a thousand words.
Please constrain your comments on the stability / suitability of this release for production only on the just-released LibreOffice version 3.4.2, which is the only version of LibreOffice 3.4.x claimed by the TDF to be suitable for the use to which you claim to be putting it.
Edited 2011-08-10 06:27 UTC
Even then, it isn’t a justification. I installed the then-current version of LibreOffice two weeks ago and have been dismayed to see that the basic “Undo” function didn’t work at all in Writer, messing up the actions stack, removing styles and text that hadn’t been changed in the current work session, etc. I put up with that for about 10 minutes then reverted to OO 3.3. That was the single time I ever came across a program that had the Undo/Redo feature f–ked up.
Unfortunately, I don’t see any mention of that in the fixes page. So I’ll pass on this version.
As a side note, the picture you linked to shows 3.4.1 in the “stable” row so I think the original poster has a valid point as he/she is discussing stability issues and not “rock solidness” issues.
So how come the release notes page (in the news teaser) is titled “New features and fixes” and it shows NEW features? For instance, ALL Writer items on that page are new features. Unless someone made a mistake somewhere, I find it hard to believe that this release if only made of bug fixes.
Last, I remember you personally advertising LibreOffice right here on osnews in comments pages of previous news items and now you say those releases were meant for developers and early adopters? That restriction was nowhere to be found back then.
Edited 2011-08-10 09:17 UTC
The “then-current” version you installed was obviously not the “then-stable” version, because the “then-stable” version of LibreOffice doesn’t do that at all.
Edited 2011-08-10 12:02 UTC
LibO_3.4.1_Win_x86_install_multi.exe; 192539 KB, MD5 F2B782D7AFE3A7D71F60A6F745B359B1. The torrent file finished downloading on 14/07/2011, that’s just a month ago. So no, it wasn’t a mysterious build that I dug out of some weird deep folder.
Actually, Writer 3.4.2 is also f–ked up and still has the same problem that Undo is totally incapable of putting back the text that has just been removed and instead, it keeps on deleting text that hasn’t been edited since the document was opened. The non-working Undo/redo feature makes the software totally unusable. 3.4.2 isn’t ready for everyday use either.
Edited 2011-08-13 13:10 UTC
Hi again. I’d just like to clarify a few points. I have been trying to use LibreOffice since it was first announced. However, I couldn’t use the 3.3.x series due to several showstopper bugs. Since then, the 3.4.x series fixed the showstopper bugs. Now, 3.4.2 is claimed to be generally suitable for some enterprise use. Nevertheless, in the short time that I’ve been on 3.4.2 I have experienced at least one full crash resulting in loss of my unsaved changes. (If I’m not mistaken it was actually an undo action that brought it down.) I am not an enterprise user, just a heavy personal user. The majority of my LibreOffice usage isn’t heavily complex or abnormal; it just involves time-consuming work for me like writing and data-entry.
There are three open bug reports related to undo functionality in LibreOffice.
Bug 37022 – Writer crash when Undo-Redo editing tables. (can’t be reproduced by developers)
https://bugs.freedesktop.org/show_bug.cgi?id=37022
Bug 38745 – redlining / undo document breakage / corruption (a CONFIRMED regression, is not present in LO 3.3.3)
https://bugs.freedesktop.org/show_bug.cgi?id=38745
Bug 39844 – EDITING: Crash while Undo action (Still needs info, UNCONFIRMED)
https://bugs.freedesktop.org/show_bug.cgi?id=39844
The first relates to undo in tables, the other two relate to undo when track changes is active.
If none of these describe your issue(s), you can perhaps get your issue resolved by submitting a new bug report of your own:
http://wiki.documentfoundation.org/BugReport
This would be a better thing to do, both for you as a LibreOffice user and for everybody else as well, rather than simply complaining about it on an Internet forum such as OSAlert.
I think in general the point is that for some people it is not certain the LO 3.4.x releases are ready for real “enterprise” or stable use.
Compared to what? I have encountered quite a number of issues with MS Office or OpenOffice versions crashing at various times as well, especially in relation to change tracking.
There is even a feature of Office suites which is called “document recovery”, which is really a euphamism for “recovering documents which were munged by the last software crash”.
Edited 2011-08-11 06:32 UTC
Compared to nothing, all software has bugs. I’m just commenting on the fact that people in this thread have a reasonable gripe with some LO features in 3.4.
You seem to be trying to invalidate their comments by first stating that they must talk about 3.4.2 and then when it was clear this was already the case, you said that they were better not posting here, but filing bug reports.
As you say, all software has bugs.
What exactly would be the point of someone uncovering a bug in a new software release and not telling the developers about it, but rather bleating about it on an Internet forum? Considering that all software has bugs, what exactly is the point of attempting to disparage in public the “stability” of the higher-performing, open and cost-free LibreOffice while remaining mute about the exactly similar failings of the closed, slow, bloated, legal-risk-encumbered and expensive commercial alternatives?
Who exactly is served by such behaviour? Certainly not the “average Joe” public.
Edited 2011-08-11 23:17 UTC
I do not disagree that filing bugs is important, of course it is.
Ideally every comment (positive or negative) about LibreOffice would be put in context agains perhaps MS Office, the Apple Pages/Numbers/etc, perhaps Calligra.
However, it is biased to only demand such context when talking only about the negatives (bugs).
… and it is equally biased to bring up bugs in the context of a new stable version of LibreOffice in an apparent effort to disparage it as being “unstable”, when it is no more unstable than any other typical office suite.
Furthermore this type of behaviour is particularly unhelpful to ordinary people, who have no particular ability to assess the technical attributes of software they might want to use. This is especially a problem in light of the fact that there is a huge commercial interest from the incumbent near-monopoly to keep said near-monopoly intact and thereby impose significant costs on ordinary people for no good reason.
This is a rip-off, pure and simple. It isn’t a help to anyone to try to mislead them that they still need to run MS Office, and thereby have to endure its significant expense, lower performance, onerous EULA terms, and requirement to run under MS Windows and face the attendant exposure to malware that that entails.
Such attempts to misdirect ordinary people, and thereby rip them off, are well worth nipping in the bud.
Edited 2011-08-12 03:17 UTC
There’s bias in not talking about MS Office in a LibreOffice thread? You have a strange definition of bias.
You just don’t want people to say bad things about LibreOffice, praise is OK, criticism is bad. I’m glad you’re not an actual developer on the project, you’d be marking all those bugs as “Works for me: closed”. In any case, and I’m saying this as a free software contributor, they really do not need your brand of “help”.
You’re not nipping anything in the bud, you’re jumping at shadows.
Not at all. In this context, bias means a favouring of one over another.
In your original context, since all software has bugs, your notion of “stability” necessarily implies that one is less stable than another. Your notion of stability has no context other than as a comparison with other office suites.
LibreOffice is no more or less stable than other office suites. It has bugs, and it can crash, but so can othe office suites. To claim otherwise would be biased.
If you did not want comparisons with other office suites brought into the discussion, then you should not mention the stability or lack thereof of LibreOffice. The same is true for other comparative evaluations, such as for example bloat, speed, performance, expense, interoperability, functionality or encumberances on use. All of these have meaning for an office suite only as a measure relative to other office suites.
Ugh, bias is indeed giving different treatment to one than the other. But if I talk about why I don’t like ham sandwiches, do I really have to mention cordon bleu? Am I biased in that case? No.
Bias would be to hold two products to different standards, that’s not happening here. Maybe in your mind, but not in the original poster’s comment.
Compared to OpenOffice. OK, rant mode officially on and don’t try to Spock me by being ‘logical’ – it’s software Jim, but not as we have known it, unfortunately. I drank all the ‘LibreOffice Heroes’ kool-aid and the dead-in-the-water opining of pundits regarding OO, and all it left me with on OS X 10.6 is a whole heap of instability issues; I have lost more work in the past few months on LB that I have in years with OO. Yes, I know, file a bug report, join in, do your bit, etc, etc. etc. But this spat amongst developers has cost me hours and late nights! Grr.
OK, rant over (some of this fuelled by UK post-riot stress disorder, I am not usually, I hope, this ill-mannered :-s)
Hi there. With all respect, I mentioned in my first comment that I have filed some bug reports. But the crash I experienced on 3.4.2 was too random and I was too busy to remember what the exact circumstances were surrounding it.
the “open” office is as immature as the open desktop. it is forever just a few years away. like fusion!
where is “open cloud office” to give us something meaningful to talk about. at least that one will have an ipad app.