OpenOffice 2.3 Released

ClickOnThis writes "Surely I'm not the only one who noticed that OpenOffice.org has announced the release of version 2.3! From the website: 'Available for download now, OpenOffice.org 2.3 incorporates an extensive array of new features and enhancements to all its core components, and protects users from newly discovered security vulnerabilities. It is a major release and all users should download it. Plus: It is only with 2.3 that users can make full use of our growing extensions library.' You can download it but be kind and use a P2P client instead, such as bittorrent."

Comments

I wonder

[Virtual_Raider]

When will they focus on usability and speed rather than adding features. It may or may not be feature-complete (whatever that is) but it certainly is not yet quite as easy and streamlined to use even as some early nineties suites... Just my $0.02, don't bite my head off =)

Re:I wonder

[cyborch]

Speed has a lot to say about user experience. It feels slow, which makes it fell like a huge bloated application. Somehow that needs to change, at least before I personally would feel better using it.

Then egain, I may really just be in the market for a .pages -> .odt converter, rather than a full OOo suite...

Re:I wonder

[temcat]

You're too optimistic. For the issues that I filed (being pretty important stuff like absence of Normal/Draft mode and shitty notes implementation), "a while" has already taken five years - and still counting.

Re:I wonder

[Anonymous Coward]

Give it more RAM to make it as fast a MS Word. Tools > Options> Memory

YMMV.

Re:I wonder

[mhall119]

Your issues require either brand new features or major re-writes of features. The "adapt row height" slowness is more likely a bug fix or simple optimization, which wouldn't take as long.

When complaining about missing features

[pembo13]

as is inevitable, it might help if you give details, and leave out things like "doesn't act exactly like Word"

Re:When complaining about missing features

[Anonymous Brave Guy]

It needs to act like Word in one particular case. It must load Word documents and format them exactly like Word, and when it saves them there must be no way of telling whether it was done in Word or OpenOffice. [...] One of the distribution channels for OpenOffice must be by sneaking in as a faster replacement for Word

I respectfully disagree. OO is never going to be a better Word than Word, nor is it realistic to expect perfect preservation of complex formatting when moving between different software packages that use different models for the data and different file formats to store them. This is a battle that cannot be won, and it is a waste of resources trying to fight it.

In any case, we can readily see that perfect document interchange is not a priority for most users. After all, people open Word documents that were laid out for US Letter paper in A4-friendly Europe and vice versa, even though this typically affects pagination. It's the content that matters for most people, not the round trip, which means you need to be able to import and export readable files but the odd blemish isn't catastrophic. For in-house people, you'll typically be using the same software across an organisation anyway, so round-tripping isn't a problem if you need to do it. And if you really do need exact reproduction for an external source, for example to send to a print shop or for a downloadable brochure, then it's better to use a format such as PDF or PostScript that is designed for that purpose. But this is usually a one-way trip, so that's not a problem.

Of course there will always be exceptions, where people want to round-trip with perfect formatting between different packages. But to be brutally honest, that is unrealistic, and it always has been. Once you can import the content accurately and a good approximation of the formatting, you rapidly get diminishing returns trying to get the corner cases with complex page layouts and the like. Personally, once you've reached that point, I'd rather see the developers of other word processors try to do things in their own way that is better than Word. Fighting for every extra last ounce of .doc compatibility can yield a Pyrrhic victory at best.

Re:Error bars - woohoo!

[antiknijn]

If you're serious about your charting, why not look into good alternatives like R, Octave or GNUPlot? These all come for free and offer vastly superior charting possibilities than Excel.

Source Code Cleanup

[mdm42]

What I really, really want from OOo is a cleanup of the code to the point where merely-mortal developers like myself can actually do something useful with it. As it is, the codebase is just this great big hairy ball of stuff -- completely unapproachable unless you have someone willing to fork out a paycheck for you to bang on it full time.

Far too many open-source projects miss the point that one of their major "features" is clean code, design and architecture documentation; a big part of the "user base" are the people who might want to live (sometimes) inside the code. That means you have to keep the barrier to entry low for the programmer who is a noob to your codebase. (We could talk about how some OS projects lack developers who are clued enough to actually write clean code or design decently, but we won't go there ;-)

Until a real and deep codebase cleanup happens OOo is "open-source" in name only as far as I am concerned.

Re:Source Code Cleanup

[JohnFluxx]

I just have to defend OSS here :)

OpenOffice.org is a horrible mess _because_ it was developed in house with paid developers etc.

Look at the koffice code instead - it's beautiful. It uses KDE parts, the Qt library, the general KDE spelling framework, and so on. It's modular and reusable. The formula thing (one part that I happen to know about) it used koffice, but also has it's own program for standalone math editing, and is also used by another program that uses it as frontend to math engine (maxima etc).

I know reuse isn't proof of clean code, but it's evidence of such :)

Re:Source Code Cleanup

[jeremyp]

I fail to see how being proprietary necessarily makes code horrible. I've seen good and bad open source code and good and bad proprietary code. I think the quality of the code is more a function of how well the project is run than whether it is proprietary or not.

Re:Source Code Cleanup

[Abcd1234]

I fail to see how being proprietary necessarily makes code horrible.

Well, to play devil's advocate (I happen to agree with you, actually), it may be true that, given the scale of large open source projects, and the nature of the collaboration (anonymous over the intarwebs), it may simply be necessary for developers to write cleaner code, as that code is one of the primary forms of communication between the various team members. After all, it's not like koffice developer #1 can just walk down the hall and ask koffice developer #2 what the hell he was thinking when he wrote some piece of code.

Re:Not compatible ?

[Anonymous Brave Guy]

The problem with your slightly flippant reply is that some of us did once file and vote for bugs, but after seeing some of the most popular bugs in the whole system stay dormant for literally years and given that the OO bug reporting system is ludicrously overcomplicated for casual users, we don't generally bother any more.

Re:Sign the damn installer (Windows)

[Anonymous Brave Guy]

You're missing the point. Paying a fee to sign the installer is just a tax on legitimate software developers. Everyone else manages without it, and if all it takes is a $100 bill to get a certificate, then that is exactly what a certificate is worth (and deflation will take place the first time a major trojan is installed by signed software).

Re:Sign the damn installer (Windows)

[tomknight]

Certification isn't just about paying $100, it's about meeting a standard. Here's an html version of the MS doc saying what a package must be/do to be certifiable (as 'twere).

I've had to deal with crappy installers (I've created a few of them...) and know that it's much easier to deal with a good one, especially when supporting a large number of machines. That bit of certification can help give a sysadmin confidence that this installation isn't going to be a PITA when it comes to upgrading/removing/conflicts/reboots over a large number of machines.

Does that help at all?