June 09, 2003

Summary of Slashdot Comments

This is a summary of the ideas raised by the 200+ comments in the slashdot discussion [6/4/03]. For lack of anything better, I use the comment number as the link to specific comments.

The one big issue that the sample law did not address was formats that have some intellectual property strings attached to them. They are covered by a patent, they are licensed by a standards body, etc. This was brought up several times (6120347, 6120617, 6120387). One person suggested simply prohibiting formats that require royalty payments (6120467), another suggested three rules for granting an exception to releasing formats, using Oracle as an example (6120410).

There was also some discussion of the openness of Adobe's Portable Document Format (PDF), since the reader is free but the official writing tools are not. One writer pointed out that there are other tools that write PDF (6121600), and another mention that an International Standards Organization (ISO) subset of PDF is being defined (6120729).

This is an issue that needs to be discussed further. One comment stated "I have yet to see a patent that prohibits decoding of file formats, even in USA." (6120920), implying that you could read (but not write) a patented formula without licensing it, but I do not know if this is true.

Other suggestions for improvements to the bill included tightening the language to prevent companies from weaseling out of it by storing their data as an executable (6120634, also posted as a comment on this site, 6120915); defining "useful life" more precisely (6120367); specifying how long before a program is released its formats have to be documented publicly (6133353); and using a Wiki to draft the sample bill (6120419).

Some suggestions on expanding the law will likely not be considered: requiring network access to data (6122275), open network protocols (6120550, 6122467), and not allowing governments to charge for access to public data (6121492).

Several comments expressed misunderstanding about what ODFI was trying to do, despite the fact that I tried to be very clear about this. One person claimed ODFI would force companies to migrate to new software (6120459), and two complained about forcing companies to adopt standards (6124115, 6124612). This shows that we need to keep emphasizing that these are not part of the goals of ODFI. Then again, one person who thought we were requiring standards expressed approval for that (6124730).

Two others believed that ODFI was requiring that all data formats be human-readable (6120589, 6120890). This is probably a misunderstanding of the statement that the documentation itself must be human-readable.

There were some comments against the bill. Some people simply don't want more laws (6121801, 6123346), while someone else asked isn't the current situation OK because programs will always be able to read old formats (6120561)--an argument that was quickly rebutted. One person mentioned that companies would complain that open data formats would hurt "security through obscurity", admitting that it was a bogus argument but still one that would come up in the debate (6120439).

A few people took the pragmatic approach; common sense would be better, but sometimes legislation is needed (6120407, which started a long tangent about seatbelt laws), and an open data format bill may be dogmatically inferior to an open source bill, but it is more likely to pass (6120614).

Several writers said that a program which by default used a proprietary format should be allowed as long as it had an option to export/save its data in a standard format. (6120867, 6121837, 6120585, 6129107). All of these were shouted down by others explaining why people would still tend to use the default format, and many of the standard formats are not as rich as the default.

There were many expressions of support for ODFI. One person wrote "This idea seems too obvious, too clear, too intuitive, and far too easy to implement for any respectable lawmaker to consider it for even a single nanosecond." (6120329). Another pointed out that "the argument for open data formats is the unassailable bastion of open source bills" (6121680). People liked ODFI because it would force more competition (6124468), or remedy Microsoft's monopoly (6126110).

Some quotes from comments include "People acting toward their own individual self interest often create an undesirable result. A little enforced cooperation benefits all." (6121045), "Wouldn't it just be better to have an open market for politicians and officials with an auction process?" (6121740), "I pushed for a similar 'rule' at the local government offices I work at" (6120530), and "You lose freedom when you lose the information that the government (which belongs to everyone) produces/compiles." (6123517).

A few final comments: One person said that many open source programs don't have their formats written down anywhere, so allowing open source as a good alternative would not necessarily work (6120882); another mentioned that a standards-based format for office documents was being proposed (6121464); and finally someone said that the Gnu Public License allowed companies to restrict access to their source code to only those who had binary copies (6121339). This comment was preceded with "IIRC", and wading through the GPL, it's not clear that the poster did recall correctly. It seems the GPL may prevent State Government A from simply taking code that is used by State Government B, but it would allow State Government B to give a copy to State Government A, so I still think government software vendors would be leery of releasing their code under the GPL.

Posted by Adam Barr at June 9, 2003 12:41 PM | TrackBack