On Thu, September 14, 2006 1:06 pm, Joe Shaw wrote: > In general I am rather nervous about indexing remote resources for a > number of reasons:
Yes, this is understandable. However, I assume that the QueryDomain was ment to adress this. > > * Doing so can be extremely taxing on the network. I believe that a queryable of this sort would be able to play very nice with resources as it does not have to constantly index as fast a possible all the information available. how many times have you gone to a link on Google that lands you on a dynamic page and the Google indexer has not updated the information. You are looking for a pb&j recipe and the recipe of the week has tuna sald. It seems for me to hapen alot which causes me to goto the cache several times a week. Ignoring, the fact that the site should properly setup a robot.txt file, users will accept slightly outdated indexes of external information. During DoQuery shouldn't the queryable hit the lucene index to match results? The only network traffic after play nice with the network indexing is when the user opens the bookmark. Or am I missing something? > > * For random resources (like Google search and bookmarks, but > unlike something that you have an account for, say Gmail or > Flickr) there are privacy and security concerns because Beagle > is essentially phoning home. What? Beagle phones home? GTFO... I am not sure if Google/bookmarks are ok or gmail/flickr are ok in your opinion? Assuming, beagle does not phone home all that is needed is some form of UI for setting the QueryDomain. The user should decide what they want in search results. I like the domain concept as it makes life easier to deal with 4 categories as opposed to my 21 queryables that I have working in development. I am thinking of external sources that are not controled directly by the user (gmail/flickr) not as random but external indexes ot query. BTW bookmarks are controled by the user just like the photos in flickr. However, integrating external index into results is exactly I as user want out of beagle: "Beagle is a search tool that ransacks your personal information space to find whatever you're looking for. Beagle can search in many different domains." For me "personal information space" includes on a daily basis Google, Yahoo!, eBay, Gmane, Bookmark wildcard lookup, ... Now I can understand the situation: I am a historian and I have lots of nazi research documents on my harddrive and I have John Doe in my contact list and the query John Doe and nazi showing up on Zeitgeist would not make my contact too happy, but that is what the query domains are for right? I am perfectly willing to admit that I missed the argument completely but isn't the concept of personal information space expanding? And doesn't beagle provide a better service than say Google (non-desktop search) by allowing you personal information space to be integration to the world on the user's terms? Adam _______________________________________________ Dashboard-hackers mailing list [email protected] http://mail.gnome.org/mailman/listinfo/dashboard-hackers
