To view parent comment, click here.
To read all comments associated with this story, please click here.
Others are easily fixable (typos) or revisable (eliminating VML from the next version of the spec).
If Office 2007 produces documents with VML in it (as a result of legacy Office documents being opened and saved) then on a practical level, this is absolutely meaningless.
Regardless of whether you allow KParts or Bonobo or anything else, you still need a viewer for that embedded element and the document wouldn't work on the alternate platform anyway. The same is true of ODF or any other format which allows embedding data and rendering instructions from outside apps.
The difference is that ODF references standards that have largely been implemented elsewhere in the open source world, such as SVG. Anyone implementing OOXML will need to recreate large amounts of technology that only runs on Windows.
This doesn't make an actual difference for interoperability, because one would only presume docs to be interoperable if they stay within the standard and are produced solely by the standardized implementation...
Errrrr, right, yer. Take a look at what Office 2007 produces:
http://www.codeproject.com/cs/library/office2007bin.asp
What was that about interoperability again?
Do you have something specific here from these comments, or is there some set of specific complaints that you think will make OOXML unimplementable by outsiders?
I would suggest some serious reading before asking that question, because it has been done to death.
(Incidentally, check out Jody Goldberg's post on dealing with SpreadsheetML for Gnumeric:
Have you seen how utterly basic that example file is? People have macros, charts, formulas etc. etc. in the real world. Jody also tells us this:
"Brian’s example of Numbers reading an OOX file written by Gnumeric could just as easily been an XLS file.....In contrast XLSX may be ugly, but it’’s concepts were very familiar from XLS. We already had much of the code required to handle it."
So in order to handle a naively simple spreadsheet file, they already largely had what they needed to implement it having mucked about with XLS for several years? Well that's just brilliant, and not exactly a ringing endorsement for OOXML.
I would advise reading Stephane Rodriguez's comments in that article for a fuller picture.
Have you seen Stephane Rodriguez's comments?? He's not exaclty an unbiased or particularly polite commenter. He goes around calling people "bitches."
What you've shown is that there are features in Office which do not fall under the purview of the OOXML standard. So what? If you want to interchange documents, then don't use those features. If you're writing an app which has special processing instructions (say, for instance AppleScript from iWork), you can include your stuff in the container as well without any problems. Can you explain why you think the BIN parts are a problem?





Member since:
2006-01-02
I looked through the first 30-40 comments in that Danish reply. Some of them are unnecessary ("make this change to do it like ODF") because ODF and OOXML are different specs and it's okay for them to diverge in minor ways (they are wildly different in overall structure anyway). Others are easily fixable (typos) or revisable (eliminating VML from the next version of the spec). The OLE comments in there are interesting: if you have some data linked into the document and you don't have OLE on your platform, what are you going to do? Regardless of whether you allow KParts or Bonobo or anything else, you still need a viewer for that embedded element and the document wouldn't work on the alternate platform anyway. The same is true of ODF or any other format which allows embedding data and rendering instructions from outside apps. This doesn't make an actual difference for interoperability, because one would only presume docs to be interoperable if they stay within the standard and are produced solely by the standardized implementation (i.e. not using non-standard pieces, like OLE embeddables).
Do you have something specific here from these comments, or is there some set of specific complaints that you think will make OOXML unimplementable by outsiders? Keep in mind that implementing an Office Suite is hard and that the file format isn't usually the biggest issue. (Incidentally, check out Jody Goldberg's post on dealing with SpreadsheetML for Gnumeric: http://blogs.gnome.org/jody/2007/09/10/odf-vs-oox-asking-the-wrong-...).