Linked by Thom Holwerda on Sat 6th May 2006 22:05 UTC
Microsoft A group of software developers have created a program to make Microsoft Office work with files in the OpenDocument format, a move that would bridge currently incompatible desktop applications. Gary Edwards, an engineer involved in the open-source OpenOffice.org project and founder of the OpenDocument Foundation, on Thursday discussed the software plug-in on the Web site Groklaw.
Permalink for comment 122060
To read all comments associated with this story, please click here.
RE[2]: Worse news yet
by dsmogor on Sun 7th May 2006 22:00 UTC in reply to "RE: Worse news yet"
dsmogor
Member since:
2005-09-01

The point is that it's f*&*&#^$*ing hard to design a document format that will be:
1. can intuitively and easily manipulated without burdening user with many dtp concepts
2. is rock stable when it comes to rendering on different codebases/configuraions/media/dpi... even in case of lamest of lame documents edited without a basic knowledge of dtp concepts
3. offers enough power and flexibility for long, structured documents, unusuall applications and dtp people
4. fallback bracefully in case of interoperability with simplier software, also nondamaging back and forth interoperability with simplier formats such as docbook commes here.
5. is transparent and enables meanigfull extraction of information for cooperating systems
6. enables interoperability with non wordprocessing office applications (browsers, spreadsheets, presentations), mainly in terms of preserving meaningfull formatting information.

None of existing formats score in all of those categories. In fact every of them sucks badly in at least one.

1. Commes from requirement that documents developed in glofied typewriter style are as easily edited in all supporting suites. MS seemed to do well here even optimising their application for this style. Whether it is function of a single codebase and their desperate efforts to emulate all the stupid special cases to preserve compatibility will remain a mistery. All the failed attempts to reproduce MS rendering engine in other codebase despite years of rev eng seem to confirm that. The fact that docxml is havily based on old doc, looks like the way not to dismiss years of such plumbing and the only sensible techncal solution to predictably convert from doc to docxml and the reverse.
ODF is havily based on html which escapes the problem by explicitly discouraging space padding and such. This may bode ill with interoperability with independent suites.
2. This requires a very strict and and unabigous specification and wery, wery broad and complex test suite and a certification process. Doc may fail miserably here as MS goal of zero interoperability for all the past years has been clearly contradicting with this requirement. ODF in turn is based on mature and well tested web standards. On the other hand Sun is still too shy about test suite and certification.It will take years to develope one. Actually the only format that passed realword test here is PDF.
3. ODF has a clear egde her and it shows when working with long documents in both test suites.
4. It is almost rarely talked about and I think only doable with well stuctured documents. ODF has some chance to inherit some html qualities here.
5. this is the very reason why xml was choosen above binary data. other than that its hard to say, MS seem's to cook some aliances to strenghten docXML in this regards. ODF counts on the power of standardization itself.
6. MS has years of experience and pushing envelope here. ODF is based on well established web standards but they are quite complex.

Reply Parent Score: 3