man(1) oddity - was: HEADS UP: bzip2(1) compression for manpages...

[Harti Brandt]

On Mon, 19 May 2003, Ruslan Ermilov wrote:

RE>On Mon, May 19, 2003 at 11:53:49AM -0300, Daniel C. Sobral wrote:
RE>> Andy Farkas wrote:
RE>> >
RE>> >Arrghh. Why you not believe me when I have proved the user experience is
RE>> >different between 4.x and 5.x ?????
RE>>
RE>> He isn't denying that. He's just claiming these 10 seconds are not a
RE>> result of the man page being catpaged.
RE>>
RE>Right.
RE>
RE>> I do have a question... on these examples, does the cat page exist or
RE>> not, for each?
RE>>
RE>It does not.  If it exists and is up-to-date, it's just uncompressed
RE>(if it's compressed) and displayed.
RE>
RE>> And what happens in the other case (ie, if the cat page
RE>> does not exist in these examples, what happens when it does exist)?
RE>>
RE>If the catpage does not exist, it's either created (if the user
RE>has the write permission to the cat* directory) or the raw
RE>manpage gets formatted, and the output is piped to the PAGER.
RE>Piped, not ";"ed, hence no message.
RE>
RE>If we are to add the message, it should be "Formatting and
RE>displaying the page, please wait..." which is silly (IMO).

You should try 'man sh' on an 166MHz Pentium, then you see that this
is not silly... groff is a C++ monster and on slow machines takes quite a
while even to grok the macros, not talking about spitting out the first
page.

harti
-- 
harti brandt,
http://www.fokus.fraunhofer.de/research/cc/cats/employees/hartmut.brandt/private
brandt at fokus.fraunhofer.de, harti at freebsd.org