The Unix Haters' Handbook

Richard wrote: I don't want to debate this (it's not my space), but I think, like much of the unix-haters-handbook, there is more to this than has been discussed here. The info system is a lot like many page setting systems which came before it (and there is a long history, littered with things like scribe, bolio, runoff and many others) One nice thing about it is that you can "compile" source documents into many different formats. This tends to be really useful for some applications. It can also be done (somewhat) with man pages but it's not easy and there are fewer options for inserting various types of 'metadata' for referencing other information. There's also methods for structuring the information in ways that go beyond the format of a single man page. I've personally used info/tex-info to generate system documentation and then went on to make html pages, pdf's, printed pages and single text files (which can be searched with any text editor or grep). I found it easy and nice to use (but I still miss scribe). On topic here, the MIT lisp machine manual was written in bolio, which is somewhat close to tex-info and one industrious guy was able to take the source files I unearthed and (30 year later) compile them into really nice html, automagically. I'm sure he could turn them into even nicer pdf file with a small amount of work. That says something about the 'language' and how it encodes the information. http://common-lisp.net/project/bknr/static/lmman I'm not sure you could do that with man pages. Maybe you could. So I'd offer that info pages are more than just cumbersome poorly hypertext'd documents. They are part of a larger system which has goals beyond just simple man pages. -brad