[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Re: Please review changed man-file of w3m



Hello Justin,

Justin B Rye schrieb am 11. Nov 2014 um 11:37

> Markus Hiereth wrote:
> > Justin B Rye schrieb am 10. Nov 2014 um 00:46
> [...]
> >>       Sheets. W3M can also serve as a general purpose browser and
> >>       pager (like "more" or "less") for text files named as
> >>       arguments, passed on standard input, or accessed via the net.
> > 
> > Maybe we can renounce the parenthesis with more and less and shorten
> > the last phrase a bit:
> > 
> >        W3M can also serve as a general purpose browser and pager for
> >        text passed in files, on standard input, or accessed via the
> >        net.
> 
> I'd prefer to avoid talking about "passing" text in a file (which also
> upsets the "Xed via Y" parallelism).  Come to think of it you can also
> invoke "w3m ." and navigate to a local file - but maybe we don't need
> to mention that.  We could just drop the parenthesis:
> 
>          W3M can also serve as a general purpose browser and pager for
>          text files named as arguments, passed on standard input, or
> 	 accessed via the net.
> 
> Or reorganising slightly (we're not really contrasting "browser mode"
> and "pager mode" here - we've done "web browser" already):
> 
>          W3M can also serve as a general purpose directory browser and
> 	 as a text pager for files named as arguments or passed on
> 	 standard input.
> 
> Or even:
> 
>          W3M can also serve as a pager for text files named as arguments
> 	 or passed on standard input, and as a general purpose directory
> 	 browser.

Each of the three versions is an improvement. The latter is sorted the
best.


  
> >>       When given one or more command line arguments, W3M will
> >>       attempt to work out the target's MIME type before deciding how
> >>       to handle it. Web URLs provide W3M with content-type headers;
> >>       for relative or absolute file system paths, W3M relies on
> >>       filenames.
> > 
> > instead of "work out" more concrete: 
> > 
> >        When given one or more command line arguments, W3M will handle
> >        targets according to content type. For web URLs, this
> >        information is delivered as HTTP header value and for relative
> >        or absolute file system paths, W3M relies on filenames.
> 
> Let's straighten out both phrases to have active verbs:
> 
>          When given one or more command line arguments, W3M will handle
>          targets according to content type. For web URLs, W3M gets
>          this information from HTTP headers; for relative or absolute file
>          system paths, it relies on filenames.

That's fine.


  
> >>       With no argument, W3M expects data from STDIN and assumes
> >>       "text/plain" unless another MIME type is given by the user.
> > 
> > Maybe we should use "standard input" instead of "STDIN" in the
> > descriptive first paragraphs of the manpage. We used "standard input"
> > in the first paragraph of section DESCRIPTION.
> 
> Yes, fair enough.  And although I habitually write STDIN/STDOUT/STDERR
> all-caps it might be more standard to use stdin/stdout/stderr
> throughout.

It's up to you. I just would prefer complete words "standard input" in
the descriptive parts of the manual page.

  
> >>       -cookie, -no-cookie
> >>              turn cookie-handling on or off
> > 
> >        -cookie, -no-cookie
> >               turn cookie-handling on or off. Corresponds the pair of
> >               boolean variables accept_cookie and use_cookie in
> >               configuration files
> 
> That should be "Corresponds ^to^ the pair...".  We have no particular
> reason to go into this much detail when we don't mention other
> variables like show_lnum or bg_color.  And saying it like this still
> fails to hint at the difference between "using" and "accepting"
> cookies.  The point of putting it this way:
> 
>          -cookie, -no-cookie
>             use cookies and accept new ones, or do neither
> 
> was that it explains that, so if I look at the config file I'll see
> what they do.  Maybe even
> 
>             use stored cookies and accept new ones, or do neither

That's fine


 
> >>       -post file
> >>              use POST method to upload data defined in file. The syntax to be
> >>              used is var1=value1[&var2=value2]...
> > 
> > Maybe we should use the typewriter font for the file's content 
> > 
> >                used is \fCvar1=value1[&var2=value2]...\fP
 
> Oh, okay.  That formatting has no effect in normal manpagers, but I
> assume it affects other kinds of output.  It would be easy to end up
> marking a lot of this text as typewriter font...

All the font variations (bold, italics, monospace) appear with
  groff -Tps -man file.1 > file.1.ps

The content of this file for data to be posted is the only occurence
where I miss the monospace font. Everywhere else it is used.

  
> >>       -4     IPv4 only (equivalent to -o dns_order=4)
> >>
> >>       -6     IPv6 only (equivalent to -o dns_order=6)
> > 
> > I think there is more sense in a hint to the items in the
> > configuration files than in providing a second way of setting the
> > value of the configuration variable
> > 
> >       -4     IPv4 only. Corresponds to dns_order=4 in
> >               configuration files
> > 
> >       -6     IPv6 only. Corresponds to dns_order=6 in
> >               configuration files
> 
> Good idea,
> 
> >>   Options for data output (with immediate exit):
> > 
> > I personally dislike appended information in parenthesis. Can we use
> > "followed by"
> > 
> >    Options for data output, followed by immediate exit:
> >    Options for data output, immediate exit after operation:
> >    Options to request data, immediate exit after delivery
> >    data request options, immediate exit after delivery:
> 
> Definitely the first one.

OK, let's use it.


> >>              Convert a web page to text with appended links
> >>               $ w3m -o display_link_number=1 w3m.sourceforge.net > out.txt
> > 
> > With retry_http=TRUE, the target is accepted. Otherwise, w3m will
> > complain "Can't load w3m.sourceforge.net"
 
> Oh, sorry, when did that get set?  Of course I was trying to ensure it
> would fit on one line and not trigger justification - there's probably
> a more appropriate nroff markup solution for that.  If necessary:

Silly thing, that these seven characters more of http:// cause line
breaking.

Merging this example with the "Start with no imput" would solve the
problem. With -B for bookmarks instead of -v, the content of out.txt
was less boring.



> Sorry about the lack of revised version, I've run out of time; I'll
> get back to it this afternoon.

I just searched for information about the layout commands for manpages
and this page ...

  http://www2.research.att.com/sw/download/man/man5U/groff_man.html

contains some valuable hints. I need to go through our manpage file
and check how .SS, .SH, and .TP are used. In addition, .TP accepts a
number as argument that affects indentation.

Therefore I would suggest that you introduce the last language
improvements and I remove still exiting outcommented lines and try make a better use of the layout macros.

Best regards
Markus


Reply to: