Tag the 1.0.2 final release.

git-svn-id: http://viewvc.tigris.org/svn/viewvc/tags/1.0.2@1433 8cb11bc2-c004-0410-86c3-e597b4017df7

CHANGES

@@ -1,56 +1,70 @@
-Version 1.0 (1-May-2006)
+Version 1.0.2 (released 29-Sep-2006)
+
+  * minor documentation fixes
+  * fix Subversion annotate functionality on Windows (issue #18)
+  * fix annotate assertions on uncanonicalized #include paths (issue #208)
+  * make RSS URL method match the method used to generate it (issue #245)
+  * fix Subversion annotation to run non-interactively, preventing hangs
+  * fix bug in custom syntax highlighter fallback logic
+  * fix bug in PHP CGI hack to avoid force-cgi-redirect errors
+
+Version 1.0.1 (released 20-Jul-2006)
+
+  * fix exception on log page when use_pagesize is enabled
+  * fix an XHTML validation bug in the footer template (issue #239)
+  * fix handling of single-component CVS revision numbers (issue #237)
+  * fix bug in download-as-text URL link generation (issue #241)
+  * fix query.cgi bug, missing 'rss_href' template data item (issue #249)
+  * no longer omit empty Subversion directories from tarballs (issue #250)
+  * use actual modification time for Subversion directories in tarballs
+
+Version 1.0 (released 01-May-2006)
 
   * add support for viewing Subversion repositories
-  * add support for running on MS Windows (2003-Feb-09)
-  * generate strict XHTML output (2005-Sep-08)
+  * add support for running on MS Windows
+  * generate strict XHTML output
   * add support for caching by sending "Last-Modified", "Expires", 
-    "ETag", and "Cache-Control" headers (2004-Jun-03)
+    "ETag", and "Cache-Control" headers
   * add support for Mod_Python on Apache 2.x and ASP on IIS
   * Several changes to standalone.py:
     - -h commandline option to specify hostname for non local use.
     - -r commandline option may be repeated to use more than repository
       before actually installing ViewCVS.
     - New GUI field to test paging.
-  * add new, better-integrated query interface (2004-Jul-17)
-  * add integrated RSS feeds (2005-Dec-22)
+  * add new, better-integrated query interface
+  * add integrated RSS feeds
   * add new "root_as_url_component" option to embed root names as
     path components in ViewCVS URLs for a more natural URL scheme
     in ViewCVS configurations with multiple repositories.
-    (2002-Dec-12)
-  * add new "use_localtime" option to display local times instead of
-    UTC times (2002-May-06)
+  * add new "use_localtime" option to display local times instead of UTC times
   * add new "root_parents" option to make it possible to add and
     remove repositories without modifying the ViewCVS configuration
-    (2004-Jul-16)
-  * add new "template_dir" option to facilitate switching between
-    sets of templates (2005-Feb-08)
+  * add new "template_dir" option to facilitate switching between sets of 
+    templates
   * add new "sort_group_dirs" option to disable grouping of
-    directories in directory listings (2005-Mar-07)
-  * add new "port" option to connect to a MySQL database on a nonstandard
-    port (2005-Dec-22)
+    directories in directory listings
+  * add new "port" option to connect to a MySQL database on a nonstandard port
   * make "default_root" option optional. When no root is specified,
-    show a page listing all available repositories (2005-Feb-04)
+    show a page listing all available repositories
   * add "default_file_view" option to make it possible for relative
     links and image paths in checked out HTML files to work without
     the need for special /*checkout*/ prefixes in URLs. Deprecate
-    "checkout_magic" option and disable by default (2006-Apr-03)
-  * add "limit_changes" option to limit number of changed files shown
-    per commit by default in query results and in the Subversion revision
-    view (2005-Dec-23)
+    "checkout_magic" option and disable by default
+  * add "limit_changes" option to limit number of changed files shown per
+    commit by default in query results and in the Subversion revision view
   * hide CVS "Attic" directories and add simple toggle for showing
-    dead files in directory listings (2004-Jul-31)
+    dead files in directory listings
   * show Unified, Context and Side-by-side diffs in HTML instead of
-    in bare text pages (2004-Jun-22)
+    in bare text pages
   * make View/Download links work the same for all file types
-    (2004-Jan-21)
-  * add links to tip of selected branch on log page (2005-Oct-03)
-  * allow use of "Highlight" program for colorizing (2005-Dec-20)
+  * add links to tip of selected branch on log page
+  * allow use of "Highlight" program for colorizing
   * enable enscript colorizing for more file types
-  * add sorting arrows for directory views (2004-Jul-21)
-  * get rid of popup windows for checkout links (2004-Jan-21)
+  * add sorting arrows for directory views
+  * get rid of popup windows for checkout links
   * obfuscate email addresses in html output by encoding @ symbol
-    with an HTML character reference (2004-Jul-29)
-  * add paging capability (2001-Dec-31)
+    with an HTML character reference
+  * add paging capability
   * Improvements to templates
     - add new template authoring guide
     - increase coverage, use templates to produce HTML for diff pages,
@@ -59,43 +73,36 @@ Version 1.0 (1-May-2006)
     - add new template variables providing ViewCVS URLs for more 
       links between related pages and less URL generation inside
       templates
-  * add new [define] EZT directive for assigning variables within
-    templates (2004-Apr-21)
+  * add new [define] EZT directive for assigning variables within templates
   * add command line argument parsing to install script to allow 
-    non-interactive installs (2005-Jan-06)
-  * add stricter parameter validation to lower likelihood of CSS
-    vulnerabilities (2002-May-24)
+    non-interactive installs
+  * add stricter parameter validation to lower likelihood of cross-site
+    scripting vulnerabilities
   * add support for cvsweb's "mime_type=text/x-cvsweb-markup" URLs
-    (2002-Oct-10)
-  * fix incompatibility with enscript 1.6.3 (2002-Feb-05)
-  * fix bug in parsing FreeBSD rlog output (2003-Jul-24)
+  * fix incompatibility with enscript 1.6.3
+  * fix bug in parsing FreeBSD rlog output
   * work around rlog assumption all two digit years in RCS files are
-    relative to the year 1900. (2005-Sep-30)
+    relative to the year 1900.
   * change loginfo-handler to cope with spaces in filenames and
-    support a simpler command line invocation from CVS (2003-Feb-11)
+    support a simpler command line invocation from CVS
   * make cvsdbadmin work properly when invoked on CVS subdirectory
-    paths instead of top-level CVS root paths (2006-Mar-17)
-  * show diff error when comparing two binary files (2002-Jan-23)
-  * make regular expression search skip binary files (2002-Jan-17)
+    paths instead of top-level CVS root paths
+  * show diff error when comparing two binary files
+  * make regular expression search skip binary files
   * make regular expression search skip nonversioned files in CVS
-    directories instead of choking on them (2002-Sep-27)
+    directories instead of choking on them
   * fix tarball generator so it doesn't include forbidden modules
-    (2002-Feb-22)
   * output "404 Not Found" errors instead of "403 Forbidden" errors
-    to not reveal whether forbidden paths exist (2005-May-17)
-  * fix sorting bug in directory view (2002-Apr-18)
+    to not reveal whether forbidden paths exist
+  * fix sorting bug in directory view
   * reset log and directory page numbers when leaving those pages
-    (2005-Jan-29)
-  * reset sort direction in directory listing when clicking new
-    columns (2004-Jul-21)
-  * fix "Accept-Language" handling for Netscape 4.x browsers
-    (2002-May-23)
-  * fix file descriptor leak in standalone server (2004-Jul-17)
-  * clean up zombie processes from running enscript (2002-Jun-15)
-  * fix mysql "Too many connections" error in cvsdbadmin (2003-Jul-24)
-  * get rid of mxDateTime dependency for query database (2003-Feb-09)
+  * reset sort direction in directory listing when clicking new columns
+  * fix "Accept-Language" handling for Netscape 4.x browsers   
+  * fix file descriptor leak in standalone server
+  * clean up zombie processes from running enscript
+  * fix mysql "Too many connections" error in cvsdbadmin
+  * get rid of mxDateTime dependency for query database
   * store query database times in UTC instead of local time
-    (2003-Feb-09)
   * fix daylight saving time bugs in various parts of the code
 
 Version 0.9.4 (released 17-Aug-2005)

INSTALL

@@ -64,10 +64,10 @@ Congratulations on getting this far. :-)
     To start installing right away (on UNIX): type "./viewvc-install"
     in the current directory and answer the prompts.  When it
     finishes, edit the file viewvc.conf in the installation directory
-    to tell viewvc the paths to your CVS and Subversion repositories. Next,
-    configure your web server to run <INSTALL>/cgi/viewvc.cgi, as
-    appropriate for your web server.  The section `INSTALLING VIEWVC'
-    below is still recommended reading.
+    to tell viewvc the paths to your CVS and Subversion repositories. 
+    Next, configure your web server (in the way appropriate to that browser)
+    to run <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/viewvc.cgi.  The section
+    `INSTALLING VIEWVC' below is still recommended reading.
 
 
 INSTALLING VIEWVC
@@ -118,12 +118,12 @@ installation instructions.
    then edit the files in <VIEWVC_INSTALLATION_DIRECTORY>/templates.
    You need knowledge about HTML to edit the templates.
 
-4) The CGI programs are in <VIEWVC_INSTALLATION_DIRECTORY>/www/cgi/.  You can 
+4) The CGI programs are in <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/.  You can 
    symlink to this directory from somewhere in your published HTTP server 
    path if your webserver is configured to follow symbolic links.  You can 
-   also copy the installed <VIEWVC_INSTALLATION_DIRECTORY>/www/cgi/*.cgi 
+   also copy the installed <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/*.cgi 
    scripts after the install (unlike the other files in ViewVC, the scripts 
-   under www/ can be moved).
+   under bin/ can be moved).
 
    If you are using Apache, then see below at the section 
    titled APACHE CONFIGURATION.
@@ -149,26 +149,26 @@ Either METHOD A:
 2) The ScriptAlias directive is very useful for pointing 
    directly to the viewvc.cgi script.  Simply insert a line containing
 
-     ScriptAlias /viewvc <VIEWVC_INSTALLATION_DIRECTORY>/www/cgi/viewvc.cgi
+     ScriptAlias /viewvc <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/viewvc.cgi
 
    into your httpd.conf file.  Choose the location in httpd.conf where 
    also the other ScriptAlias lines reside.  Some examples:
 
-     ScriptAlias /viewvc /usr/local/viewvc-1.0/www/cgi/viewvc.cgi
-     ScriptAlias /query /usr/local/viewvc-1.0/www/cgi/query.cgi
+     ScriptAlias /viewvc /usr/local/viewvc-1.0/bin/cgi/viewvc.cgi
+     ScriptAlias /query /usr/local/viewvc-1.0/bin/cgi/query.cgi
 
    continue with step 3).
 
 or alternatively METHOD B: 
 2) Copy the CGI scripts from 
-   <VIEWVC_INSTALLATION_DIRECTORY>/www/cgi/*.cgi
+   <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/*.cgi
    to the /cgi-bin/ directory configured in your httpd.conf file.
 
    continue with step 3).
    
 and then there's METHOD C:
 2) Copy the CGI scripts from 
-   <VIEWVC_INSTALLATION_DIRECTORY>/www/cgi/*.cgi
+   <VIEWVC_INSTALLATION_DIRECTORY>/bin/cgi/*.cgi
    to the directory of your choosing in the Document Root adding the following
    apache directives for the directory in httpd.conf or an .htaccess file:
 
@@ -183,7 +183,7 @@ and then there's METHOD C:
 
 or if you've got Mod_Python installed you can use METHOD D:
 2) Copy the Python scripts and .htaccess file from 
-   <VIEWVC_INSTALLATION_DIRECTORY>/www/mod_python/
+   <VIEWVC_INSTALLATION_DIRECTORY>/bin/mod_python/
    to a directory being served by apache.
    
    In httpd.conf, make sure that "AllowOverride All" or at least 
@@ -246,7 +246,7 @@ there are some additional steps required to get the database working.
    Optionally, you can create a second user with read-only access to the
    database.
 
-3) Run the <VIEWVC_INSTALLATION_DIRECTORY>/make-database script.  It will 
+3) Run the <VIEWVC_INSTALLATION_DIRECTORY>/bin/make-database script.  It will 
    prompt you for your MySQL user, password, and the name of database you 
    want to create.  The database name defaults to "ViewVC".  This script 
    creates the database and sets up the empty tables.  If you run this on a 
@@ -285,7 +285,7 @@ there are some additional steps required to get the database working.
    To get real-time updates, you'll want to checkout the CVSROOT module
    from your CVS repository and edit CVSROOT/loginfo.  Add the line:
 
-     ALL <VIEWVC_INSTALLATION_DIRECTORY>/loginfo-handler %{sVv}
+     ALL <VIEWVC_INSTALLATION_DIRECTORY>/bin/loginfo-handler %{sVv}
 
    If you have other scripts invoked by CVSROOT/loginfo, you will want
    to make sure to change any running under the "DEFAULT" keyword to
@@ -295,7 +295,7 @@ there are some additional steps required to get the database working.
    If you are running the Unix port of CVS-NT, you'll need to use a 
    slightly different command line:
    
-     ALL <VIEWVC_INSTALLATION_DIRECTORY>/loginfo-handler %{sVv} cvsnt
+     ALL <VIEWVC_INSTALLATION_DIRECTORY>/bin/loginfo-handler %{sVv} cvsnt
      
    The extra 'cvsnt' parameter tells the handler script to parse the 
    commit information in a different way. 
@@ -316,7 +316,7 @@ there are some additional steps required to get the database working.
      #!/bin/sh
      REPOS="$1"
      REV="$2"
-     <VIEWVC_INSTALLATION_DIRECTORY>/svndbadmin rebuild "$REPOS" "$REV"
+     <VIEWVC_INSTALLATION_DIRECTORY>/bin/svndbadmin rebuild "$REPOS" "$REV"
 
    If you allow revision property changes in your repository, create a
    post-revprop-change hook script containing the same commands as the

lib/blame.py

@@ -42,25 +42,28 @@ def link_includes(text, repos, path_parts, include_url):
   match = re_includes.match(text)
   if match:
     incfile = match.group(3)
+    include_path_parts = path_parts[:-1]
+    for part in filter(None, string.split(incfile, '/')):
+      if part == "..":
+        if not include_path_parts:
+          # nothing left to pop; don't bother marking up this include.
+          return text
+        include_path_parts.pop()
+      elif part and part != ".":
+        include_path_parts.append(part)
 
-    # check current directory and parent directory for file
-    for depth in (-1, -2):
-      include_path = path_parts[:depth] + [incfile]
-      try:
-        # will throw if path doesn't exist
-        if repos.itemtype(include_path, None) == vclib.FILE:
-          break
-      except vclib.ItemNotFound:
-        pass
-    else:
-      include_path = None
+    include_path = None
+    try:
+      if repos.itemtype(include_path_parts, None) == vclib.FILE:
+        include_path = string.join(include_path_parts, '/')
+    except vclib.ItemNotFound:
+      pass
 
     if include_path:
-        url = string.replace(include_url, '/WHERE/', 
-                             string.join(include_path, '/'))
-        return '#%sinclude%s<a href="%s">"%s"</a>' % \
-               (match.group(1), match.group(2), url, incfile)
-       
+      return '#%sinclude%s<a href="%s">"%s"</a>' % \
+             (match.group(1), match.group(2),
+              string.replace(include_url, '/WHERE/', include_path), incfile)
+    
   return text
 
 

lib/query.py

@@ -403,6 +403,7 @@ def main(server, cfg, viewvc_link):
       'query' : query,
       'commits' : commits,
       'num_commits' : len(commits),
+      'rss_href' : None,
       }
 
     if form_data.hours:

lib/vclib/bincvs/__init__.py

@@ -474,7 +474,7 @@ def _tag_tuple(revision_string):
     t = map(int, string.split(revision_string, '.'))
     l = len(t)
     if l == 1:
-      raise ValueError
+      return ()
     if l > 2 and t[-2] == 0 and l % 2 == 0:
       del t[-2]
     return tuple(t)

lib/vclib/svn/__init__.py

@@ -483,10 +483,18 @@ class BlameSource:
     self.last = None
     self.first_rev = first_rev
     
+    # Do a little dance to get a URL that works in both Unix-y and
+    # Windows worlds.
     rootpath = os.path.abspath(rootpath)
+    if rootpath and rootpath[0] != '/':
+      rootpath = '/' + rootpath
+    if os.sep != '/':
+      rootpath = string.replace(rootpath, os.sep, '/')
+      
     url = 'file://' + string.join([rootpath, fs_path], "/")
     fp = popen.popen(svn_client_path,
-                     ('blame', "-r%d" % int(rev), "%s@%d" % (url, int(rev))),
+                     ('blame', "-r%d" % int(rev), "--non-interactive",
+                      "%s@%d" % (url, int(rev))),
                      'rb', 1)
     self.fp = fp
     

lib/viewvc.py

@@ -14,7 +14,7 @@
 #
 # -----------------------------------------------------------------------
 
-__version__ = '1.0.0'
+__version__ = '1.0.2'
 
 # this comes from our library; measure the startup time
 import debug
@@ -946,7 +946,7 @@ def get_file_view_info(request, where, rev=None, mime_type=None, pathrev=-1):
                                          pathtype=vclib.FILE,
                                          params={'content-type': 'text/plain',
                                                  'revision': rev,
-                                                 'pathrev': rev},
+                                                 'pathrev': pathrev},
                                          escape=1)
   if request.cfg.options.allow_annotate:
     annotate_href = request.get_url(view_func=view_annotate,
@@ -1313,18 +1313,15 @@ def markup_stream_php(fp, cfg):
   if not cfg.options.use_php:
     return None
 
-  sys.stdout.flush()
-
-  # clearing the following environment variables prevents a 
-  # "No input file specified" error from the php cgi executable
-  # when ViewVC is running under a cgi environment. when the
-  # php cli executable is used they can be left alone
-  #
-  #os.putenv("GATEWAY_INTERFACE", "")
-  #os.putenv("PATH_TRANSLATED", "")
-  #os.putenv("REQUEST_METHOD", "")
-  #os.putenv("SERVER_NAME", "")
-  #os.putenv("SERVER_SOFTWARE", "")
+  # The following HACK may be be used to allow a PHP CGI executable to be
+  # invoked instead of a CLI executable, on systems that do not have PHP's
+  # CLI (command line interface) installed. Just uncomment the following lines:
+  #os.unsetenv("SERVER_SOFTWARE")
+  #os.unsetenv("SERVER_NAME")
+  #os.unsetenv("GATEWAY_INTERFACE")
+  #os.unsetenv("REQUEST_METHOD")
+  #os.unsetenv("SCRIPT_FILENAME")
+  #os.unsetenv("PATH_TRANSLATED")
 
   return MarkupPHP(cfg.options.php_exe_path, fp)
 
@@ -1425,14 +1422,17 @@ def view_markup(request):
     streamer = markup_streamers.get(ext)
     if streamer:
       markup_fp = streamer(fp, cfg)
-    elif cfg.options.use_enscript:
-      markup_fp = MarkupEnscript(cfg, fp, request.path_parts[-1])
-    elif cfg.options.use_highlight:
-      markup_fp = MarkupHighlight(cfg, fp, request.path_parts[-1])
 
-  # If no one has a suitable markup handler, we'll use the default.
-  if not markup_fp:
-    markup_fp = MarkupPipeWrapper(fp)
+    # If there wasn't a custom streamer, or the streamer wasn't enabled, we'll
+    # try to use one of the configured syntax highlighting programs.
+    if not markup_fp:
+      if cfg.options.use_enscript:
+        markup_fp = MarkupEnscript(cfg, fp, request.path_parts[-1])
+      elif cfg.options.use_highlight:
+        markup_fp = MarkupHighlight(cfg, fp, request.path_parts[-1])
+      else:
+        # If no one has a suitable markup handler, we'll use the default.
+        markup_fp = MarkupPipeWrapper(fp)
     
   data['markup'] = markup_fp
   
@@ -2123,7 +2123,7 @@ def view_log(request):
   if cfg.options.use_pagesize:
     data['log_pagestart'] = int(request.query_dict.get('log_pagestart',0))
     data['entries'] = paging(data, 'entries', data['log_pagestart'],
-                             'revision', cfg.options.use_pagesize)
+                             'rev', cfg.options.use_pagesize)
 
   request.server.header()
   generate_page(request, "log", data)
@@ -2830,7 +2830,7 @@ def generate_tarball_header(out, name, size=0, mode=None, mtime=0,
 
   out.write(block)
 
-def generate_tarball(out, request, reldir, stack):
+def generate_tarball(out, request, reldir, stack, dir_mtime=None):
   # get directory info from repository
   rep_path = request.path_parts + reldir
   entries = request.repos.listdir(rep_path, request.pathrev, {})
@@ -2847,30 +2847,43 @@ def generate_tarball(out, request, reldir, stack):
   if reldir:
     tar_dir = tar_dir + _path_join(reldir) + '/'
 
-  # Subdirectory datestamps will be the youngest of the datestamps of
-  # version items (files for CVS, files or dirs for Subversion) in
-  # that subdirectory.
-  latest_date = 0
   cvs = request.roottype == 'cvs'
-  for file in entries:
-    # Skip dead or busted CVS files, and CVS subdirs.
-    if (cvs and (file.kind != vclib.FILE or (file.rev is None or file.dead))):
-      continue
-    if file.date > latest_date:
-      latest_date = file.date
+  
+  # If our caller doesn't dictate a datestamp to use for the current
+  # directory, its datestamps will be the youngest of the datestamps
+  # of versioned items in that subdirectory.  We'll be ignoring dead
+  # or busted items and, in CVS, subdirs.
+  if dir_mtime is None:
+    dir_mtime = 0
+    for file in entries:
+      if cvs and (file.kind != vclib.FILE or file.rev is None or file.dead):
+        continue
+      if file.date > dir_mtime:
+        dir_mtime = file.date
 
-  # push directory onto stack. it will only be included in the tarball if
-  # files are found underneath it
+  # Push current directory onto the stack.
   stack.append(tar_dir)
 
+  # If this is Subversion, we generate a header for this directory
+  # regardless of its contents.  For CVS it will only get into the
+  # tarball if it has files underneath it, which we determine later.
+  if not cvs:
+    generate_tarball_header(out, tar_dir, mtime=dir_mtime)
+
+  # Run through the files in this directory, skipping busted ones.
   for file in entries:
-    if (file.kind != vclib.FILE or
-        (cvs and (file.rev is None or file.dead))):
+    if file.kind != vclib.FILE:
+      continue
+    if cvs and (file.rev is None or file.dead):
       continue
 
-    for dir in stack:
-      generate_tarball_header(out, dir, mtime=latest_date)
-    del stack[:]
+    # If we get here, we've seen at least one valid file in the
+    # current directory.  For CVS, we need to make sure there are
+    # directory parents to contain it, so we flush the stack.
+    if cvs:
+      for dir in stack:
+        generate_tarball_header(out, dir, mtime=dir_mtime)
+      del stack[:]
 
     if cvs:
       info = os.stat(file.path)
@@ -2878,7 +2891,8 @@ def generate_tarball(out, request, reldir, stack):
     else:
       mode = 0644
 
-    ### read the whole file into memory? bad... better to do 2 passes
+    ### FIXME: Read the whole file into memory?  Bad... better to do
+    ### 2 passes.
     fp = request.repos.openfile(rep_path + [file.name], request.pathrev)[0]
     contents = fp.read()
     fp.close()
@@ -2888,21 +2902,22 @@ def generate_tarball(out, request, reldir, stack):
     out.write(contents)
     out.write('\0' * (511 - ((len(contents) + 511) % 512)))
 
-  # recurse into subdirectories
+  # Recurse into subdirectories, skipping busted ones.
   for file in entries:
     if file.errors or file.kind != vclib.DIR:
       continue
 
-    # skip forbidden/hidden directories (top-level only)
+    # Skip forbidden/hidden directories (top-level only).
     if not rep_path:
       if (request.cfg.is_forbidden(file.name)
           or (cvs and request.cfg.options.hide_cvsroot
               and file.name == 'CVSROOT')):
         continue
 
-    generate_tarball(out, request, reldir + [file.name], stack)
+    mtime = request.roottype == 'svn' and file.date or None
+    generate_tarball(out, request, reldir + [file.name], stack, mtime)
 
-  # pop directory (if it's being pruned. otherwise stack is already empty)
+  # Pop the current directory from the stack.
   del stack[-1:]
 
 def download_tarball(request):
@@ -2916,6 +2931,8 @@ def download_tarball(request):
   sys.stdout.flush()
   fp = popen.pipe_cmds([('gzip', '-c', '-n')])
 
+  ### FIXME: For Subversion repositories, we can get the real mtime of the
+  ### top-level directory here.
   generate_tarball(fp, request, [], [])
 
   fp.write('\0' * 1024)
@@ -3188,8 +3205,9 @@ def build_commit(request, files, limited_files, dir_strip):
   commit.rss_date = make_rss_time_string(files[0].GetTime(), request.cfg)
   if request.roottype == 'svn':
     commit.rev = files[0].GetRevision()
-    commit.rss_url = 'http://%s%s' % \
-      (request.server.getenv("HTTP_HOST"),
+    commit.rss_url = '%s://%s%s' % \
+      (request.server.getenv("HTTPS") == "on" and "https" or "http",
+       request.server.getenv("HTTP_HOST"),
        request.get_url(view_func=view_revision,
                        params={'revision': commit.rev},
                        escape=1))

templates/include/footer.ezt

@@ -5,7 +5,7 @@
 <table>
 <tr>
   <td><address>[cfg.general.address]</address></td>
-  <td style="text-align: right;"><strong><a href="[help_href]">ViewVC Help</strong></td>
+  <td style="text-align: right;"><strong><a href="[help_href]">ViewVC Help</a></strong></td>
 </tr>
 <tr>
   <td>Powered by <a href="http://viewvc.tigris.org/">ViewVC [vsn]</a></td>

viewvc.conf.dist

@@ -459,6 +459,10 @@ highlight_convert_tabs = 2
 use_php = 0
 
 # path to php executable
+# (This should be set to the path of a PHP CLI executable, not the path
+# to a CGI executable. If you use a CGI executable, you may see "no input file
+# specified" or "force-cgi-redirect" errors instead of colorized source. The
+# output of "php -v" tells you whether an given executable is CLI or CGI.)
 php_exe_path = php
 # php_exe_path = /usr/local/bin/php
 # php_exe_path = C:\Program Files\php\cli\php.exe