Comments on: Novell and Microsoft Make an ‘Impartial’ Crowd http://techrights.org/2007/11/29/xml-2007-lies/ Free Software Sentry – watching and reporting maneuvers of those threatened by software freedom Tue, 07 Feb 2012 14:00:35 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: Stephane Rodriguez http://techrights.org/2007/11/29/xml-2007-lies/comment-page-1/#comment-3936 Stephane Rodriguez Fri, 30 Nov 2007 12:01:53 +0000 http://boycottnovell.com/2007/11/29/xml-2007-lies/#comment-3936 I know it's trendy these days to call out Miguel and his friends. I think it's deserved, he's a moron. Just take a look at his blog, it's 100% Microsoft related these days. I think Miguel is just a Microsoft guy who takes a pride not to be on their payroll (he told me privately a while back that he made a lot of money and that he did not need to work anymore). But to be constructive, I'll add what seems to me more important on the OOXML debate. It's a little subtle. The real question is not about supporting this file format or not. Market forces dictate that any vendor of an Office suite will have to support it sooner or later or will lose customers. (As an aide, MS Office 2009 beta is just around the corner, it will be interesting to know what kind of changes are being made to the file format.) The real question is about whether a vendor takes the memory structure to implement the main functions out of OOXML files (read, write, render, calc, ...) to exactly match the one from OOXML, or if it's a superset, or something different. First an anecdote, ever wondered why Jody Goldberg is boasting support for some of OOXML file format? Clearly because Gnumeric is a decade-old project whose internal memory structure was DESIGNED TO MATCH the one from Excel. And since OOXML spreadsheets is just angle brackets around the old Excel, it automatically follows that for someone with all the stack implemented already, it does not take years to implement some of the OOXML file format. Note that, to be accurate, Gnumeric won't probably support any of the new theming/DrawingML/charting stuff introduced in Excel 2007 for at least two years, and that won't stop him from claiming suport anyhow. Now someone must ask himself, what time is it going to take for someone starting from scratch now? 2 years? 5 years? 10 years? The only reason why Jody Goldberg does not support ODF, or not so much, is because the memory structure of Gnumeric is too tightly coupled with Excel's, therefore it can't adapt. What we need is a vendor whose memory structure is so versatile it can adapts to any Office related file format. As a matter of truth, Microsoft could have made this task much easier if they understood XML. The way the Excel team uses XML only proves they don't understand XML at all : in essence, they use it as a binary stream instead of a text stream. It's only the result of Microsoft own incompetence that implementing OOXML and interoperating with other file formats is going to be so much taxing for everyone out there. For this reason alone, OOXML should be rejected at ISO, until it gets significantly improved. (I have no problem Microsoft getting the ISO timestamp if the file format has real merit, which it lacks right now). I know it’s trendy these days to call out Miguel and his friends. I think it’s deserved, he’s a moron. Just take a look at his blog, it’s 100% Microsoft related these days. I think Miguel is just a Microsoft guy who takes a pride not to be on their payroll (he told me privately a while back that he made a lot of money and that he did not need to work anymore). But to be constructive, I’ll add what seems to me more important on the OOXML debate.

It’s a little subtle. The real question is not about supporting this file format or not. Market forces dictate that any vendor of an Office suite will have to support it sooner or later or will lose customers. (As an aide, MS Office 2009 beta is just around the corner, it will be interesting to know what kind of changes are being made to the file format.)

The real question is about whether a vendor takes the memory structure to implement the main functions out of OOXML files (read, write, render, calc, …) to exactly match the one from OOXML, or if it’s a superset, or something different.

First an anecdote, ever wondered why Jody Goldberg is boasting support for some of OOXML file format? Clearly because Gnumeric is a decade-old project whose internal memory structure was DESIGNED TO MATCH the one from Excel. And since OOXML spreadsheets is just angle brackets around the old Excel, it automatically follows that for someone with all the stack implemented already, it does not take years to implement some of the OOXML file format. Note that, to be accurate, Gnumeric won’t probably support any of the new theming/DrawingML/charting stuff introduced in Excel 2007 for at least two years, and that won’t stop him from claiming suport anyhow. Now someone must ask himself, what time is it going to take for someone starting from scratch now? 2 years? 5 years? 10 years?

The only reason why Jody Goldberg does not support ODF, or not so much, is because the memory structure of Gnumeric is too tightly coupled with Excel’s, therefore it can’t adapt.

What we need is a vendor whose memory structure is so versatile it can adapts to any Office related file format. As a matter of truth, Microsoft could have made this task much easier if they understood XML. The way the Excel team uses XML only proves they don’t understand XML at all : in essence, they use it as a binary stream instead of a text stream. It’s only the result of Microsoft own incompetence that implementing OOXML and interoperating with other file formats is going to be so much taxing for everyone out there. For this reason alone, OOXML should be rejected at ISO, until it gets significantly improved. (I have no problem Microsoft getting the ISO timestamp if the file format has real merit, which it lacks right now).

]]>