LibreOffice 4.0: The big changes will be under the hood | Hacker News Reader

LibreOffice 4.0: The big changes will be under the hood

zdnet.com

82 pointsCrankyBear47 comments

The (in-progress) 4.0 changelog is at https://wiki.documentfoundation.org/ReleaseNotes/4.0

It's nice to see a software project keeping with a sane version number system (major.minor.release; major number change = architectural changes), and it's nice to see them working on fixing a pile of bugs and trying some new approaches in their architecture, instead of just introducing a meaningless UI change.

Major version number = architectural change doesn't make a whole heap of sense from a consumer perspective.

Sure it does.

I signals major change that could cause breakage and new bugs.

That only makes sense for users who are programmers. Non-technical users have got no idea what you're talking about. All they see is the UI and the presented feature sets.

What kind of version numbering would make more sense for non-technical users?

Microsoft Office "The version to buy" edition. Followed by Microsoft Office "The version to buy, now for real, don't even think about getting anything else" edition.

Year-Index.

LibreOffice 2012-1 is the first release in 2012, 2012-2 is the second release, 2013-1 is the first in 2013, etc.

Depending on release schedule, the second number can be the month (like Ubuntu) or dropped altogether.

OK, sure. That seems smart, I hadn't thought of that. It would still be nice (for me) to know how carefully I need to read the changelog and whether it's a 30 second or half-day upgrade, but for most people, I can see how year-release would work better.

Microsoft was doing that with their consumer software for a while. The biggest issue is that people think the software is out of date as soon as the next year comes along.

With an actively developed and rapidly changing piece of free software, version 2012 actually would be outdated (and possibly insecure) by mid-2013. If the perception of outdatedness causes more users to upgrade in a timely fashion, I think it might be a good thing. Especially since you don't need to pay anybody $300+ for the new version.

Having said that, LibreOffice seriously needs a more seamless update experience for consumers. Re-downloading a 100MB installer and re-installing every time a minor update is published gets tired rather quickly.

Breakage (and bugs) for users.

  Windows 7
  Windows 8
  iPhone 3
  iPhone 3GS (minor update)
   etc.

The Apple S models generally keep the outer shell similar, but introduce completely new innards.

And notice that none of them break compatibility upon incrementing their major version number. They sure don't follow semantic versioning.

Architectural changes in a consumer product almost always bring along a new set of requirements and user-facing features: platform/OS, performance, UI design, file formats, updating policy, etc. as well as undesired bugs and regressions.

I would've really liked them to introduce a new more modern UI with this release. I like these UI mock-ups here, with the left sidebar. I think that's a good idea (the symbols above should be in white, too, though):

http://pauloup.deviantart.com/gallery/28216273#/d37dx4a

The light one looks good, too:

http://pauloup.deviantart.com/art/LibreOffice-UI-Mock-up-lig...

If I were them I'd try to release the new UI in the same time with the release for mobile, and try to make it a bit more unified, at least in style, since tablets and PC's might need their UI's optimized for each.

I think that they are being very smart, fixing the plumbing so they can then make fundamental UI changes.

However, I don't agree with the UI mockups. They look very, very confusing.

The screenshots at https://wiki.documentfoundation.org/Development/WidgetLayout show pretty poor results with the new layout. Hopefully things will get better before it ships.

Also: they use GTK tools for designing the UI, but are they actually using GTK to implement it? Is its Mac support good enough now?

Yep, it looks like there is no padding around the widgets yet.

I have no real knowledge of this situation, but I guarantee they are not using GTK everywhere. Cross platform support (especially with GTK3) is not nearly good enough.

My understanding of LibreOffice was that it existed as a fork purely as a response to Oracle exerting excessive control over the project. Now that Oracle has relinquished OpenOffice.org, can anyone shed some light on what the stated goals of LibreOffice actually are? How are they distinct from OpenOffice.org?

This isn't an answer, but you could equally ask the opposite question about what makes OpenOffice.org distinct from LibreOffice, given that LibreOffice was a community-led project before OpenOffice.org was.

I guess you stopped after the first four paragraphs. The fifth reads:

"In other words, LibreOffice is becoming more than just an OpenOffice fork, but an independent office suite in its own right. At the same time, OpenOffice has been struggling. OpenOffice makes no bones that "Volunteers [are] needed in all areas"."

Veering off the question of LibreOffice's motivation, that paragraph struck me as bizarre. I've yet to see an open source project that didn't want volunteers in all areas.

So I clicked through to check the citation for "OpenOffice has been struggling" is even more bizarre, as it's a link back to Zdnet with info supposedly about IBM withdrawing support...whose only support is another link back into Zdnet, and it's rebutted below by OpenOffice's Rob Weir, whose comments were then incorporated into the article.

It does seem like Zdnet are trying really hard to manufacture controversy here.

The giant red plea for volunteers on openoffioce.org is a little odd.

Well, basically, LibreOffice forked while OpenOffice was languishing under Oracle and no one knew what would happen. Since then, it has actually generally had more development than OpenOffice; there was a long period of time in which OpenOffice development halted while it was being moved over to Apache, and relicensed under the Apache license. In the meantime, LibreOffice was receiving active development and new features, and many distros started shipping it because it was actually shipping and ready while Apache OpenOffice was still being held up in bureaucracy.

Now that Apache OpenOffice is starting to ship and receive development (much in the form of porting features from Lotus Symphony, which has also been donated to the Apache project), they are being developed pretty much independently. There's a certain amount of acrimony between the projects, which is unfortunate. But there's also the unfortunate case that due to licens issues, it would be hard to merge the projects. LibreOffice is licensed under the LGPLv3, while OpenOffice has been relicensed under the Apache License (Oracle allowed their portions to be relicensed, and some of what remained from third-parties was thrown out or rewritten). This means that LibreOffice can incorporate code from OpenOffice, while OpenOffice cannot incorporate code from LibreOffice (at least, without work to turn it into separate libraries or changing the license back to LGPL).

So, due to the acrimony between the projects, and licensing issues, they are pretty much developing independently now, with some code moving between them when possible.

We'll see. If revision tracking finally works properly (ie. interchange of Word 97 .doc files with revision tracking doesn't just end up corrupting horribly) then it might be useful. Until then, it's not a usable substitute for Word. I've ended up bringing work home to edit on Windows, because LibreOffice trashed my document.

Of course it would be nice if we didn't use Word format (and obsolete Word format at that), in this way, but that's life in corporateville. Engineering shall toe the party line.

I just wish they would spend a release just on the import / export filter for Microsoft Word. It is still pretty horrible overall.

It's impossible (without an AGI :) ) to reproduce all the features and bugs in something as complex as MS Word, so import/export will never be perfect. Just have a look at the "standard" docx format specification.

Well, nobody's expecting 100% perfection anyway. Even if they just squashed a dozen well-known bugs such as revision tracking, that would be a major accomplishment.

How to switch on revision tracking in LO: http://ask.libreoffice.org/en/question/3961/track-changes/

Now if you're saying it's buggy as hell, that's a different matter :-)

Any document you can actually release, file a bug with a copy of the document - they actually care about fixing this stuff, and treat it as a fun game to do so. This link shows what they need: https://wiki.documentfoundation.org/Bug_Report#Detailed_expl...

For widest good results, and if you have the time, checking against the latest Apache OpenOffice dev build https://cwiki.apache.org/confluence/display/OOOUSERS/Develop... and filing a bug there too strikes me as likely to get your class of bug dealt with quicker.

I'm sure you've filed an issue in their bug tracker, right? Because you can't expect them to fix an issue that they may not know exists...

This doesn't bode well for the new release.

New releases after a redesign always have bugs, and since there are no user visible changes, it's all negative and no positive for the users.

I'll be skipping this release for at least half a year.

This is the thing the project actually recommends: use the shiny new version if you want new and shiny, use the previous version if you want stability.

My prediction: The Apache OpenOffice project will die. Seriously, if you had to pick between the two, which would you choose?

Thought it's dead already. :)

>they'll be using a dual licensed approach with LGPL 3.0 and the Mozilla Public License (MPL) Version 2.0.

>On Linux, however, LibreOffice will continue to be under the LPGLv3.

I am confused, are the licenses (LGPL & MPL) for source code? the small part of linux specific code is LGPL-licensed (not MPL)?

I love the feature list at https://wiki.documentfoundation.org/ReleaseNotes/4.0

Seriously, LibreOffice is starting to feel very advanced! I'm wondering how long it will take before it will surpass Microsoft Office!

This aint gonna happen.

Why is that?

IMHO there are two reasons this is unlikely:

1. MS has always been the king of "on paper" features. That was an important reason back in the 80s and 90s for Word, Excel, PP to win the market.

2. MS Office is by far the most profitable software product in the world, while Libre Office is not the coolest open source project. Which means a lot more developers on the MS side.

I didn't say "the most popular" Office product, I said "the most advanced".

The logo (i.e. turtle graphics) integration seems really cool (in a geeky way):

http://libreoffice.hu/2012/11/29/logo/

http://libreoffice.hu/2012/12/12/logo-for-desktop-publishing...

LibreOffice: Going to be very a very good desktop office-suite real soon now, just need to sort out these technical issues. Call back in a year, will you?

Ofcourse, in a year, you'll all be using web-based office suites anyway.

I switched from MS Office to LibreOffice for lightweight document work and I'm very satisfied with it.

I see no advantage in web-based office suites. LibreOffice + Dropbox is all our small office needs.

Good for you. I see more people moving from on-premise Office solutions to the cloud than the other way around though.

If you are doing any sort of collaboration across offices, Google docs documents just seems to spring up by themselves, because nobody is able to manage an email-stream of documents in various states of updates.

Personally I've moved most (not all) my stuff to Google docs, and apart from Excel, which still rules the spreadsheets, I find it making me much more productive.

I can access and update my documents on all my computers and on all my mobile devices on the go.

No desktop office + dropbox solution makes that painless enough to be viable.

Reply on news.ycombinator.com