If you have any questions about Seeker that aren't answered here, send me a message at mailto:stefan@antioch-college.edu. How Does Seeker Work? How does Seeker work? It's been known for some time that Netscape Navigator can be configured to tell other applications to open URLs and local path names (the system clipboard's path name, for example, would be "Macintosh HD:System Folder:Clipboard"). Here's an example script. When this script is run in an editor such as Apple's Script Editor, it will register Fetch's creator code and the file transfer protocol with Netscape Navigator, which results in Fetch being launched whenever Navigator encounters an "ftp://" URL. register protocol "FTCh" for protocol "ftp" To put it simply, Seeker greatly expands upon this registering technique. Scripts that register certain protocols for other programs like Anarchie, NetFinder, Claris Emailer, and Eudora are widely available at the usual software archives, including the Info-Mac HyperArchive at http://hyperarchive.lcs.mit.edu/HyperArchive.html. Seeker Not Working Seeker just doesn't seem to work. Why is this? Seeker won't work well if, when it launched, finds numerous copies of Netscape Navigator, or if Netscape Navigator finds numerous copies of Seeker when performing a search. You'll also need to be sure that your copy of Netscape Navigator (or Netscape Communicator) is named "Netscape Navigator". This is because there seems to be an unfortunate limitation of AppleScript that I'm not yet sure how to work around. If you know how to use AppleScript to tell an application exclusively by its creator code, please let me know. My email address is mailto:stefan@antioch-college.edu. And thanks Dave, but no, I don't plan to port Seeker to Frontier! Netscape Navigator 3.0.x Compatibility I use Netscape Navigator 3.0.x and it seems that Navigator sometimes forgets Seeker's protocols when I quit Navigator. Why is this? This a bug in Netscape Navigator - not with Seeker. As fas I know, only versions 3.0 and 3.0.1 are affected by this bug, so if you use Navigator 2.x or 4.x you should be fine. Fortunately there are two fairly simple work-arounds for this problem. The first is that you can launch Seeker each time you want to perform a search instead of allowing Navigator to automatically launch Seeker for you. The second work-around is to install the Netscape Navigator preferences file included with Seeker (found in the accompanying "For Navigator 3.0.x Users" folder). You can do this by placing a copy of the included Netscape Preferences file in the "Netscape ƒ" folder that's found in the Preferences folder of the System Folder of your startup disk. The only drawback to this work-around is that you'll need to reconfigure your favorite settings (be sure to quit Navigator before installing the new Netscape Preferences). Unsupported Search Engines Seeker doesn't support my favorite search engine. Could support be added for it in a future release of Seeker? I may be adding a plug-in-type architecture to Seeker in the future. Registered users that request support for a particular search engine can receive a plug-in by email that they can place into a Seeker Plug-ins folder. Another good reason to register! If you're a webmaster and you would like Seeker to support your website, read the below section, How Seeker Uses the GET Method. Strange Behavior I've tried to go to XYZ's website by entering "xyz" into the Navigator Location bar but instead of Netscape Navigator going to "www.xyz.com", Seeker launches and then doesn't do anything. Why does this happen? As most users of Netscape Navigator know, entering "xyz" into the Location bar will resolve to "http://www.xyz.com/". While this is a nice time-saver, when using Seeker, you must precede a website's address with the usual "http://" protocol (http://xyz). You can also enter the full web address without the "http://" protocol (e.g. www.xyz.com). Otherwise, entering "xyz" into the Location bar will only confuse Seeker into thinking that you want to search the web using XYZ's search engine. This is because of a bug in Netscape Navigator's protocol handling. I hope to include a work-around in a future release of Seeker. Hopefully you'll find that typing the full www.xyz.com really isn't so bad, especially when new top-level domains are introduced (.rec, .store, .web, etc.). Note that this is only a problem when "xyz:" is a protocol that Seeker uses. Unrecognized Protocols I've tried to perform a search using XYZ's search engine but Netscape Navigator says that it doesn't recognize the "xyz:" protocol. Why is this? The first thing to do is to make sure that Seeker has been launched at least once before the search was attempted. Also, it wouldn't hurt to check your spelling of the protocol with its listing in the Supported Protocols chapter. It is important to note that is a site has ".com" in its name (ABCNEWS.com, SHAREWARE.COM, Quote.com, etc.), you must include the ".com" in the protocol. If that's not the case, try quitting Seeker and relaunching it. Another possibility is that Netscape Navigator's preferences file was either damaged or created from scratch. Seeker relies on Netscape Navigator's preferences file to function, as that is where Navigator lists the protocols Seeker registers each time it is launched. What to do if this happens is to delete the Navigator's preferences file from its preferences folder (the "Netscape ƒ" folder that is found in the Preferences folder of your startup disk) and start a new one from scratch. Remember, Seeker will have to be relaunched in order to reregister all of its protocols. Supported Protocols What are the protocols that Seeker 1.0 supports? Seeker's supported protocols are fairly intuitive and can be easily guessed. For example, AltaVista's USENET search is "altavistausenet", Apple Software Updates is "appleupdates", and Excite NewsTracker is both "excitenewstracker" (and "excitent"). Check the chapter on Supported Protocols for a complete listing of all 107 protocols. Protocols are listed in alphabetical order with the name of search engine and website's URL. Internet Explorer Support Is there a version of Seeker for Microsoft Internet Explorer? No. While it is possible to adapt Seeker to work with Internet Explorer, there aren't any plans to do this any time soon. To be completely honest, Microsoft gives me the creeps and I have no desire to support their products. However, if enough people show an interest it then Seeker for Internet Explorer is a possibility. Cyberdog Support Is there a version of Seeker for Apple's Cyberdog? Do you want to use Seeker with Cyberdog? Let me know by sending me a message at mailto:stefan@antioch-college.edu. How Seeker Uses the GET Method How Seeker Uses the GET Method For now Seeker can only take advantage of Netscape Navigator's use of the "GET" method of submitting form data. If you want Seeker to support your website's search engine, it would be easiest to include in your search engine the ability to accept GET queries. However, some search engines, including Excite's personal search engine (e.g. the one used at HotWired), normally use the POST method but Seeker is able to "mimic" the GET method by ending the URL with a query string. When performing a search at HotWired, the generated URL appears like this. "http://search.wired.com/cgi-bin/architext/AT-HotWiredsearch.cgi" Because the HotWired search engine supports it, placing "?search=" at the end of the CGI's physical URL works just fine. "http://search.wired.com/cgi-bin/architext/AT-HotWiredsearch.cgi?search=knitting%20patterns" Unfortunately, not all search engines that use POST work when trying this trick, and even some engines that do use strange terms (in place of HotWired's use of "search") that can't be easily guessed, such as AltaVista's use of "aq" for its Advanced Query search. It would be appreciated if you could either support the GET method, or if your engine uses POST, that the engine supports the above-mentioned trick. Now that we know what is involved, let me know if you're interested in Seeker supporting your search engine. And if you know any more about this (my knowledge of CGIs is rather limited), I would very much appreciate any input you might have. My email address is mailto:stefan@antioch-college.edu.