Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 15 Feb 2004 14:10:26 +0200
From:      Ville =?ISO-8859-1?Q?Skytt=E4?= <scop@FreeBSD.org>
To:        wolf@ti.com
Cc:        freebsd-cvsweb@FreeBSD.org
Subject:   Re: Issues/patches for 2.9.2
Message-ID:  <1076847025.9498.262.camel@bobcat.mine.nu>
In-Reply-To: <402D62C4.3050300@ti.com>
References:  <402D62C4.3050300@ti.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 2004-02-14 at 01:50, Christopher Wolf wrote:
> I installed the Beta cvsweb.cgi a few days ago and found a few issues 
> I'd like
> to suggest improvements to.  Below are some context diffs and comments on
> each issue.

Thanks!  Comments below.

> *With GNU diffutils version 2.7 and an 80 character wide development
> environment(!), 164 characters is not wide enough for side-by-side
> diffs.  A change to 168 shows all 80 columns with room for the diff
> indicators:*

Applied.

> *When trying to download a tarball/zipfile, the cvs command checks the
> files out into one directory, and the zip/tar command tries to read
> another: the cvs command checks out the module tree path
> (i.e. master/rts/shared/) while the zip/tar command refers to the base
> module (i.e. shared/ in this example).  This change modifies the cvs
> checkout command to use the same name as the tar/zip, although a more
> complex fix is probably needed:
> *

Applied a version that restores the way it was handled in revisions
before 1.191 (ie. FreeBSD-CVSweb 2.9.1); was accidentally broken there. 
It's pretty much similar as the patch you sent, and even a bit simpler
(using $basename directly instead of File::Basename).

> When the configuration variable $shortLogLen is set to something
> longer (such as 500), the file and age columns in the directory view
> will wrap, which is hard to read.  This change makes those columns
> nowrap, so that the log message is what wraps.  This could be added to
> the userid columns too, but those seldom contain spaces.*

The patch you sent was not valid XHTML (we send XHTML 1.0 Transitional
nowadays), and since the nowrap attribute is deprecated already in HTML
4, I reimplemented it using CSS and committed a couple of other related
enhancements while at it.

> When a particular revision is chosen from a revision tree graph, most
> of the log links are turned off!  I could not see a reason why, and
> the developers here wanted to use the tree to find a particular
> revision based on tag names, and then see the differences that made
> that revision.  This change adds a config file variable that allows
> turning on those links:*

I don't see why they were turned off either, and committed a change that
accomplishes what you suggested, ie. just does The Right Thing without
adding a config param for it.

> When enscript's -x option is set to 'x', the resulting image map code
> is not compatible with the newer mozilla browsers, which are fairly
> particular about using 'name' vs 'id' for identifying the image map.
> This change calls out to use both name and id and the result is
> compatible with IE and Mozilla:*

Unfortunately changing the cvsgraph option to "-x 4" makes it output
HTML 4 map markup which not valid XHTML.  I put in a kludge which adds
the name attribute to the map element on the fly, and still preserves
the "-x x" option to cvsgraph and XHTML compatibility.

> The colors used to display diffs are hard to read, so I made the
> following change to force changes in diffs to appear with a grey
> background, making them much easier to pick out when scrolling large
> files:*

Applied, modifying it so that I added another enscript highlight file,
lang_cvsweb_diff.st which is used to colorize diffs only.  The
background color change did not look too good in non-diff code
colorization.

> I also found that enscript's regexps in diffs.st would often, when
> displaying side-by-side colors diffs, get confused by lines of code
> ending in | (think OR) or > (think #include <xxx>) if it had a trailing 
> tab. 

This file is not distributed with CVSweb, so I'd suggest that you send
those modifications to the Enscript author(s), see
http://people.ssh.com/mtr/genscript/

Thanks again for the patches, I'll release a new beta including these
changes today.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1076847025.9498.262.camel>