[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [office-comment] ODF still fails to specify scripting properly (ODF 1.2 CD01)
> And for the 0.01% of ODF users who know what an XML processing instruction > is, we could restrict these as well. (Yes, but Microsoft knows about them. And they are specifically in XML to allow unconstrained application-specific data to be inserted despite any schemas.) The point is that the requirement being expressed is something like "There should be a profile of ODF 1.2 which prevents Microsoft from performing the kind of 'embrace, extend and extinguish' associated with it in the previous decade." Whether this embrace and extend would a matter of malfeasance, habit or accident is not relevant. However, the concrete mechanism proposed, merely banning foreign elements, patently fails to meet the requirement. So let us take the requirement seriously. To do it is not impossible with XML and ODF, however, both were designed to be fundamentally open not closed. It involves taking the X out of XML and the O out of ODF! So, without going into MCE (which I as I said, I think is necessary too), what things are needed in order to make it impractical for Microsoft to embrace, extend and extinguish ODF? 1) For a start, clear normative text saying that in a closed minimal ODF document, there should be no use of any mechanism to avoid representing data that could be represented using standard ODF with any other mechanism, nor any obfuscation or tag abuse. This kind of blanket statement closes many loopholes. 2) The ODF ZIP file can only use media and files defined by standards. 3) The ODF ZIP file must contain no places where extension data can be parked: i) No use of foreign elements or attributes in *any* file in the document ii) No use of processing instructions or comments iii) No metadata in any media file iv) No use of any metadata in the ODF or any media, unless it is defined in an (open) standard v) No use of ODF Bookmarks, unless the values are defined in an (open) standard vi) No saving of application settings unless defined by an (open) standard vii) No use of URIs to external files. viii) ...possibly a couple of others. In other words, *no* application data of any kind, *no* non-standard data or data values of any kind, *no* circumvention of any kind. And this applies to all files and formats in the ZIP archive. Cheers Rick Jelliffe P.S. I think my MODUS proposal is better, but it builds on the above, by making a data/metadata distinction and allowing plurality as long as there is at least one vanilla form: http://broadcast.oreilly.com/2009/02/modus---minimum-open-documents.html
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]