  <h1 align="center"><b>Robozilla</b> </h1>
  <ul>
    
  <li><a href="#explanation">Explanation of the Process</a></li>
  <li><a href="#not_checked">Categories Not Checked by Robozilla</a></li>
  <li><a href="#correcting">Correcting Errors</a> 
    <ul>
      <li><a href="#general">General Methods</a></li>
      <li><a href="#deeplinks">Correcting Erroneous Deeplinks</a> </li>
    </ul>
  </li>
  <li><a href="/ddp/06015/">List of Error Codes</a></li>
  <li><a href="#reds">Using Robozilla/Reds</a></li>
  </ul>
  <p>Robozilla is an automated link checker which is periodically run by <a href="/ddp/03011/">metas</a>. 
    Each link is checked twice, and if it exhibits an error both times it is flagged 
    red.</p>
  <h3><b><a name="explanation"></a></b>Explanation of the Process</h3>
  <p>Every link residing in the main directory is tested to ascertain whether 
    it is accessible. The HTTP status code returned by each server is evaluated. 
    If the code is an error code, then the link is flagged. A number indicating 
    the type of error is placed in brackets after the erroneous site's title. 
    URLs are only marked red on the 'edit side', i.e. the public do not see the 
    error markings.</p>
  <h3><b><a name="not_checked"></a></b>Categories Not Checked by Robozilla</h3>
  <p><a href="http://dmoz.org:8080/Bookmarks/">Bookmarks</a> and <a href="http://dmoz.org:8080/Test/">Test</a> 
    are not checked for erroneous links. However, if a link is marked red in the 
    main directory, it will be marked red in every category it is listed in, including 
    Bookmarks and Test.</p>
  <h3><b><a name="correcting"></a></b>Correcting Errors</h3>
  <p>The first step is to attempt to visit the site yourself. You may find that 
    the site is now functioning, in which case you can click the 'This URL works 
    for me' button.</p>
  <p> There are two main types of error. The first is where Robozilla has been 
    unable to access the server: connection errors. The second is where Robozilla 
    has managed to access the server, but an error code has been returned. A <a href="/ddp/06015/">full 
    list of error codes</a> is available. If you do not wish to correct the error 
    yourself please do not blindly unreview the site; leave it for somebody else. 
  </p>
          <h4><b><a name="general"></a></b>General Methods</h4>
        
<p> 1 - Ensure that the URL is syntactically correct. For example check 
  that it:<br>
  a) does not start with a doubled htttp:// prefix -- if it does replace 'http://http://' 
  with 'http://' <br>
  b) does not contain spaces -- they are illegal characters. In domain names either 
  remove them or replace with a hyphen, '-', and in file or directory names, subsitute 
  either '%20' (the encoded form of the space character), a hyphen, an underscore, 
  or remove it alltogether.<br>
  c) is not doubled, e.g. "http://www.example.com/http://www.example.com/". </p>
        <p>2 - If a URL has been found to exhibit a Robozilla error, the edit 
          screen includes two additional options to aid you in investigation of 
          the error. One of these is a link to the version of the page that has 
          been cached by <a href="http://www.google.com/">Google</a>, and the 
          other to <a href="http://www.archive.org/">archive.org</a> --both of 
          which store archived copies of web sites. If such a version exists, 
          then by visiting it you may be able to determine: the new URL of the 
          site--if provided on the previous site, or whether the site has closed 
          down for good--for example, the front page may contain words to the 
          effect of: "I no longer have sufficient time to maintain this site, 
          so I'm shutting it down." Alternatively, locate some unique text on 
          the page, such as the owner's name for example, and search for that 
          via a search engine -- if the site has been moved its new loaction may 
          be returned. </p>
        <p>3 - Using a whois service such as the web-based <a href="http://www.allwhois.com/">allwhois.com</a>, 
          you can query the domain name. The domain name's record will inform 
          you whether or not the registration has lapsed. If it has then it's 
          likely that the domain name is defunct.</p>
        <p> 4 - Use a search engine to query the site's title -- if it's reasonably 
          distinctive you may be in luck. </p>
        <p><b><a name="deeplinks"></a>Correcting Erroneous Deeplinks</b> </p>
        <p>If the URL is a deeplink, i.e. it specifically includes a filename 
          or directory name, then try removing the file/directory information, 
          one level at a time. For instance, consider: http://www.somedomain.com/main_directory/sub_directory/main.html 
          which exhibits a <a href="/ddp/06015/#404">404 
          error</a>. You could attempt to fix this URL by following the steps 
          below: </p>
        <p>1 - Remove the filename, 'main.html', and try accessing http://www.somedomain.com/main_directory/sub_directory/ 
          <br>
          2 - If the modified URL still does not point to the correct page, continue 
          removing directory names until you find an accessible page.</p>
        <p>If a filename is specified and the above method yields no results -- 
          perhaps the file is at the root level of the server -- you may wish 
          to try modifying the file extension. Common file extentions include: 
          '.htm', '.html', '.asp', '.cgi', and '.pdf'. For example, if the URL 
          http://example.com/main.htm exhibited a 404, you could try http://example.com/main.html</p>
      <p align="left">If you find the correct URL then ensure you have updated 
          the listing accordingly. You may find that using this method you are 
          able to access a page on the site, but not the page that the old URL 
          pointed to. In this case, you may use the site's internal search function, 
          its site map, or just browse around to try to locate the page's new 
          address.</p>

        <p>If you are unable to correct the URL then either unreview or delete 
          the site -- use your judgement. Also see the next paragraph:<br>
        </p>
     
      <h3><b><a name="reds"></a></b>Using Robozilla/Reds </h3>
    
    <p>If you are unable to find the new URL of a site which you regard as particulalry 
    noteworthy, you may want to move it to <a href="http://dmoz.org:8080/Test/Robozilla/Reds/">Test/Robozilla/Reds</a> 
    (a.k.a T/R/R) where editors who specialise in correcting erroneous URLs will 
    investigate the URL, and if successful, will return the URL to its original 
    category. You should send URLs to the closest appropriate sub-category of 
    <a href="http://dmoz.org:8080/Test/Robozilla/Reds/">Test/Robozilla/Reds</a>, so 
    for example if you are editing in <a href="http://dmoz.org:8080/Arts/Movies/Titles/">Arts/Movies/Titles</a>, 
    erroneous URLs should be sent to <a href="http://dmoz.org:8080/Test/Robozilla/Reds/Arts/">Test/Robozilla/Reds/Arts</a>. 
  </p>
  <hr>
  <b><img src="/ddp/i/159.gif" width="16" height="16">Note</b>: Please only send 
    unqiue, one-of-a-kind URLs to <a href="http://dmoz.org:8080/Test/Robozilla/Reds/">Test/Robozilla/Reds</a>; 
    there is not the manpower there to investigate all of the directory's errors. 
  <hr>
     <div align="left">
  <p>To aid you in this process, a bookmarklet has been created. To use the bookmarklet 
    add this '<a href="javascript:with(document.forms%5B0%5D)%7Btypecat.value='Test/Robozilla/Reds';void(newnote.value=cat.value+'%20=%3E%20Test/Robozilla/Reds');void(omit%5B2%5D.click());void(submit%5B1%5D.click());%7D">bookmarklet</a>' 
    link to your browser's bookmarks. </p>
</div>
  
    <div align="left">If you'd like to help in <a href="http://dmoz.org:8080/Test/Robozilla/Reds/">Test/Robozilla/Reds</a> 
      consider <a href="http://dmoz.org:8080/editors/editcat.cgi?cat=Test/Robozilla/Reds">applying 
      for the category</a>; they need the help.</div>
