Slashdot Mirror
History of Software Forks Favors LibreOffice
jfruhlinger writes "The forking of LibreOffice from the OpenOffice.org project, followed by Oracle's donation of OpenOffice.org to the Apache Software Foundation, has been something of a bumpy road. But if history is any guide, it's the fork, LibreOffice, that might have the brighter future."
LibreOffice has already undertaken massive cleanups of OpenOffice.org code. It's pretty obvious which one will survive. Also one doesn't have a stupid TLD in the name (although the other is a bit freetard for my tastes).
1. If they were going to release it into the wild at the end, they should have done so at the beginning.
2. They fail to understand the advantage that MS Office integration brings in MS's SQL Server and other server strategy.
OpenOffice is the one thing that MS sales reps really hate. A few million investment can have a big impact on MS's bottom line.
I'm not a lawyer, but I play one on the Internet. Blog
It's a fork
NB: The message above might reflect my opinion right now, but not necessarily tomorrow or next year.
I think every company that acquires an open source project could learn something from how Oracle handled openoffice.org
The uncertainty and the lack of commitment by Oracle practically forced the community to fork the project. And even after that, Oracle had a chance of do the right think and donate the name to the Open Document Foundation, but they just sat down and done nothing, LibreOffice became a strong fork, and in the end they realized an "asset" that they bought from Sun was basically worthless.
Congraturations, you get award: Reading Fail. TFA's title = "Fork history does not favor OpenOffice.org" OPENOFFICE, not LIBREOFFICE. No contradiction. A fork in this context is when someone takes the code from a project and "forks" it on a different development path, like a "fork" in the road. OO.o went one way, Libre went the other. Have a nice day. :)
Here's a cookie... *psst* it's MAGIC
Apache will provide the LibreOffice folks with a copy of the OpenOffice code base that is under a license which removes any possible obligation they might ever have to Oracle regarding the code, unless they do something incredibly stupid (like failing to attribute or reproduce the license at all as Katzer did in Jacobsen v. Katzer). LibreOffice can choose to use that code base or not.
If we really want to lay blame, it's not just Oracle's. Sun Microsystems didn't ever achieve a viable community for OpenOffice. There were operational and technical reasons, but the one that might have been most important was the requirement to sign your copyright over to a company that might take the work private the next day, with no quid-pro-quo at all.
In 1999 or so, Danise Cooper called me to explain what Sun would do with OpenOffice. I explained at that time that they needed to have some sort of quid-pro-quo for code donors, even if it was only a covenant that Sun would keep their own development available under a free software license for some time or remove the contribution from their version. This was not implemented. It was difficult for independent developers to see a reason to work with Sun.
Bruce Perens.
Selective quoting of history can be used to predict whatever future a magazine thinks will sell the most ads.
LibreOffice is superior to Open Office in my experience. It is faster, It opens complex M$ Office documents and complex power point presentations more cleanly (assuming you have fonts installed.) It is a definite upgrade from OO.org. One problem. OO.org has brand recognition. Big time. It established itself as a market force. LibreOffice will need to establish that all over again.
XFree86 died when the community got up and left. Even with free hosting, the remaining XFree86 partisans couldn't keep it alive and lost interest.
Before that, the X Consortium - backed by the might of industry - died when no-one could be found to participate in it ... because XFree86 was where the action (i.e., community) was.
Citizendium forked from Wikipedia, recruited a pile of academics, then Larry Sanger drove them away. (And then the cranks moved in.) When someone said "chaps, CZ is dead" and tried another fork, they called him ... a "traitor". This from the project that was a fork itself.
XOrg is under the MIT X11 licence, but seems to get plenty of contributions back - because it's where the community is. An open source licence with centripetal force from the gravitational pull of the community.
Wayland's lead developers and all the people pushing for it in Fedora are X.Org developers. They're not "traitors" to X, they're people with their eye on the target: a good open source desktop.
EGCS won by the community getting up and leaving GCC.
LibreOffice won when the community got up and left Oracle. Oracle and IBM's approach in trying to claw it back is gibberingly, hilariously misconceived. (And Rob Weir blew his cred irretrievably lying about what the FSF had said and directing abuse at the FSF rep who tried to correct his lie. Once a shill equals a shill.)
OOo=XFree86 with a sponsor. Yay sponsors. Can IBM employ enough contributors to single-handedly make up for the enthusiasm to be found at LibreOffice? I really doubt it.
http://rocknerd.co.uk
So how about Emacs/XEmacs, which was arguably the first great open source fork? Both projects are still around, but I don't get the impression that the fork (XEmacs) has run away with the ball by any stretch of the imagination.
Virtually serving coffee
In my corner of our (large) corporation, we are required to use MS Office for documentation and presentations. The target audience for user documents are engineers and physicists. This leads to much frustrated screaming - not all of it silent - by the authors of those documents, who all have at least MSc and almost half have a PhD. In truth, we'd all rather use LaTeX, but that's not considered to be group-editable by cretins located around the world. The corporate requirement for group-editing is bogus in our case anyway, as nobody else ever touches our documents, so LaTeX to PDF should be adequate. Actually, judging from the Word documents I receive daily there are damn few of the >10^5 employees who can either structure a Word document logically, or format it consistently.
Powerpoint presentations for sales also include equations; otherwise the customers would ask for the equations and physics underlying the devices we make. All our sales and marketing people have to be able to handle equations and physical models to the level they are presented (or at least to come over as understanding them, even if any query is deflected with "I better check that with R&D, and get back to you"). Most of the sales & marketing people have technical degrees, often at MSc level. Equations in Powerpoint are slightly less problematic than in Word, as the issues of inserting objects all over a slide are less consequential. There is still a strong need for consistency in naming and presentation of variables, of course - and the screaming starts again...
It really galls me that OpenOffice made the same stupid design mistake as Microsoft Office. And the error still persists despite many requests for in-line equations in the text in the style of LaTeX. They can keep their lousy math editor "feature"; just let me use LaTeX style math formatting.
Those who can make you believe absurdities can make you commit atrocities. - Voltaire
The history of OpenOffice has been utterly depressing in so far as they've just aped Microsoft Office. Maybe that had to happen if OO was to gain any of MSO's user base, but I hope that LO will break out of the cloud of crap that is MSO.
For example, there is no reason whatsoever to default to throwing away your work. The convention of opening up a new document and THEN having to save it is utterly ridiculous! In fact, there is no reason to have a "save" command at all. All user input should be sacred, and every keypress should be saved. Another example is the crazy arrangement of menu items (made worse by the "ribbon bar" in MSO) that attempts to cram every command into a menu structure. A word processor is for most people a tool of reasonably frequent use, yet even after many years of using MSO, I still can't remember where the word-count is, because I use it hardly ever. I also constantly forget how to bring up the styles library or insert a picture. Instead of a labyrinthine menu, why not have a search (with command completion), and leave the menu for those who want to browse?
This comment is way OT though, so I'll stop.
"And the meaning of words; when they cease to function; when will it start worrying you?"