Idea picked up from a famous LinuxWorld delegate: wouldn’t it be cool if someone took the relevant bits of code from the OpenOffice.org core, and wrote a silent helper app for converting OpenDocument files into MS formats? Once you installed it, it would register in Windows for the OpenDocument MIME types. Then, when an Office user got an OpenDocument document by email and double-clicked it, the program would silently convert it to MS format, and open Word (or Excel or whatever) on the temporary file. So it would appear to the user as if Word could read OpenDocument docs.
(Yes, you might be able to build the same function into Word itself as an addon, but I suspect that might be quite a bit harder, because you’d need to understand Word’s internal data structures.)
Such a converter would be an important part of the migration path for an organisation moving to OpenDocument. The great thing about it is that it would turn Word into a second class citizen. If everyone’s exchanging OpenDocument documents; OpenOffice.org or KOffice users can both read and write them, but Word users can only read them. What a good reason to upgrade :-)
I wonder how hard it would be to write? It’s basically just a scripted copy of OpenOffice.org, so surely not too hard?
All you need then is a daemonized version of the same tool that watches for new locally-created Word .doc files, and then converts them back to OpenDocument .sxw files, and you have a complete round-trip OpenDocument editing environment using Word.
I’ve got a feeling those “relevant bits of code” add up to a rather large file size.
Not to mention this type of conversion tends to get ugly.
But still would be good to have.
A few minutes of Googling later, it appears that you can run OpenOffice in “-invisible” mode from the command line, and then talk to it as a GUI-less daemon over a UNO connection, which you can then drive from any of a number of different languages. See http://udk.openoffice.org/python/python-bridge.html for more information.
So, the solution seems to be to package a copy of OpenOffice, and a couple of small Python files to drive it.
Does anyone fancy the challenge?
Something like this, perhaps?
Why do you have America?
This would be really cool… maybe have the silent package included with OO.org, and when you send an email, just attach it (or point the recipient to a download site) and then they just click it, click NEXT> a few times and have it done in a few minutes, and viola, OpenDocument opening-able-ness… even tho that’s not a word..
What may be better is a very small program which does what Acrobat Reader does for PDFs: opens OpenDocument documents, allows you to view them, print them, etc. I think this would help spread OD faster since that way you don’t have to own Word plus download a converter program, and you don’t have to download and install OOo/KOffice.
You might even be able to get away with converting the whole shtick into HTML, and then firing up the user’s default browser.
What I would really like is for a Office app that wasn’t as slow and ugly as OOo. OOo seems to be fairly well-behaved in Windows, but in Linux it’s a pain. OOo’s use of the gtk is a fairly buggy in my experience and it’s pretty sluggish as well. I’m currently using abiword (which loads 3-4 times a fast) but it’s import/export of OpenDocument needs help.
OOo is stuck in a catch-22 situation. So long as everyone assumes that their correspondents are using MS Office, they’ll send them MSO files. The recipient meanwhile, on receiving a Word document, will assume that the sender is using Word and will use that format if they need to send documents the other way.
So you could end up with two (or more) people using OOo, all thinking that the others are using MSO and exporting their files accordingly.
Is there a solution to this? How about an export option in OOo which embeds some non-visible tag into the MSO file which the receiving copy of OOo will detect. Opening a Word file in OOo which contains this tag would then somehow notify the user that the file was actually created in OOo, and that they should perhaps contact the author to request a native file.
Otherwise we could conceivably end up with 95% of the world using OpenOffice, but all convinced that everyone else is still using MSO.
Don’t forget the Firefox plugin for OOo documents:
That provides a lighter-weight reader/viewer/printer.
there are a few others out there via google, but that one seems most mature.
Any solution requireing OO.org to be installed is useless. People with OO.org isstalled allready know how to read/save word documents from within OO.
It is the people without OpenOffice that need a solution for when they can’t install OO.org. The real solution is to create an Mso input/output filter
Henrik: any scenario means they have to install something. If you can’t persuade them to change their word processor, you either need the program I propose or the Mso input/output filter plugin. Both have to be installed, and the former (I suggest) is a lot easier to write.