Glenn Maynard said: >We can interpret DFSG#2 to mean "the form closest to source that still >exists" if we want, but it's extremely questionable to try to interpret >"preferred form for modification" as "preferred form for modification, >or any form, no matter how unreasonable it is to edit, if the preferred >form for modification has been lost".
I interpret it as "prefered form for modification, out of those forms which exist". (I think it may actually be specifically the preferred form for modification of the licensor, or of the distributor. This would be relevant in cases where there is some argument about which of several existing forms is the preferred form.) I would reject GPL'ed items where the authors deliberately destroyed all copies of their original source files not because "preferred form for modifcation" was not available -- but rather because the authors would clearly be trying to evade the intent of the GPL using a technicality, and that indicates unreliable authors acting in bad faith, who I wouldn't want to touch with a ten-foot pole. The actual case where machine-code-as-source has come is in ROMs for old games where the source code was lost many years ago, but where it's still useful and feasible to edit the ROM images. But this discussion is really not relevant to the issue at hand, because.... >In any case, I don't think anyone has actually claimed that IBM has >lost the source. Asking them for it is probably the best thing to >do next. Absolutely. --Nathanael
