| |||||||||||||
|
| |||||||||||||
freeopenofficefreeopenofficePosted Sep 29, 2004 9:35 UTC (Wed) by tzafrir (subscriber, #11501)In reply to: Marketing OpenOffice.org by mdekkers Parent article: Marketing OpenOffice.org
What would you need to start a free (LGPL-only) for of OpenOffice.org?
If any such fork gets off the ground and get substantail code contributions, linux/BSD distros will adopt them on the count of being sane to build and more useful. Thus you have a distribution channel that bypasses OpenOffice.org .
Are theere any independent distributors of OOo for win32? (For OSX there is probably fink)
(Log in to post comments)
freeopenoffice Posted Sep 29, 2004 15:49 UTC (Wed) by mdekkers (guest, #85) [Link] money and a heap of developers - don't underestimate the complexity of the codebase - forking ooo is probablygoing to be just a bit harder then forking the kernel.
freeopenoffice Posted Sep 30, 2004 1:35 UTC (Thu) by Duncan (guest, #6647) [Link] No kidding. A recent article compared code base sizes. OOo is said tohave more code than /all/ /of/ /KDE/ put together. I can't verify that claim as I've not checked (I don't need or use any office suite), but as I run Gentoo, it's well known that compiling OOo requires /five/ /gigs/ of free temp space to build and merge the main OOo ebuild, which source is pretty much straight up. That's certainly far more than any Linux kernel would be, even in my wildest nightmares. Personally, I consider that simply far to code bloated. Of course, it should be said that much of the code size owes to the fact that OOo pretty much implements its own UI and widget sets from the ground up. However, it's still comparable to QT and all of KDE in size, and that's an entire desktop environment, not just an office suite. The Ximian OO edition is rather smaller, requiring "only" ~3G of free temp space for compiling. That sounds rather more reasonable to me, certainly in comparison to 5G. Of course, it's probably the MSWormOS compatibility stuff that they cut among other things, so I doubt it's as cross-platform. That brings up a question I've been wondering about for a few weeks, tho. If we are talking a fork, how far from that is Ximian? I think they intend to remain synced to the main version, but would forking from it, if a fork is being considered, be more reasonable? If I'm right about the MSWormOS compatibility stuff above, and MSWormOS compatibility remains a goal, than I suppose I've answered my own question. However, it's something those considering a fork may wish to look at, anyway. Duncan
freeopenoffice Posted Sep 30, 2004 11:57 UTC (Thu) by tzafrir (subscriber, #11501) [Link] Disclaimer: this is all talk and no code. So feel free to ignore it on those grounds.
OOo seems to suffer from a huge code bloat. The code includes a thick cross-platform compatibility layer. OOo is unable to share code with other projects due to licensing issues. So it built everything on its own. Right to a complete fonts rendering system. That fonts rendering system was later replaced on linux in basically every linux distro.
It seems to me that if one is weeling to do the immense work required, it will turn out that much of the code of Abiword or of Mozilla could be reused. The result would be a much smaller OOo, with much faster start-up times.
Besides, the acronyms are too good to waste on a minor project.
|
|
||||||||||||