> > No, he's not kidding. This is pretty much the same question > I asked last week. > > Just because there's no manpage doesn't mean that it's impossible > to use - or don't people read source code any more? >
Of course one reads source code. It's called Intellectual Property Security through Obscurity (IPSO, for short, without any FACTO appended). I asked a very pertinent question last week while I was in Broomfield: "What constitutes DDI Compliance?". My take on this is a shell script I wrote in 1991 which looks for external symbols in a module and tries to find a man page. If it doesn't have a man page, it's not in the DDI. You could stretch that a bit and say "well, search the ARC materials" and if you find an adequate definition there, maybe you could say some interface was documented and an exportable "spec". The point here is that if we use things like the ARC and have an interface taxonomy, we should adequately document those interfaces. We don't have to write a developer's guide for everything, but something more than what is generally there. The libscsi case material and lack of documentation hardly constitutes a substitute. If it were backed up by 10 years of mailing list discussions like the linux "sg" driver is, that wouldn't be a problem, But it isn't. There's some cryptic blog entries, and some source code and > libses makes use of sgen(7d), but libscsi doesn't mandate use > of libses. > It's less than clear how you would implement it. > Pluggable fwflash(1m) uses both libscsi and libses, and the code > for almost of it is open. > > What's your underlying objection to using libscsi and/or libses? > > None, per se. They were referred to as a "now we have them, so all [developer/user usability] problerms are solved" reference. I beg to differ.
