On Friday, January 06, 2006 3:44 PM [EDT], Jason Linhart <[EMAIL PROTECTED]> wrote:
> Aengus wrote: >> >> I don't think there has even been any significant debate about new >> features, at least on this list, in the last couple of years - >> Analog is a fairly "mature" appliction at this point. > > There have been a couple of significant new feature suggestions over > the last few years, but interest was short lived when it was > discovered that Stephen wasn't interested in them. There have been occassional suggestions, but no serious debate - as you said, nobody was all that pushed about doing anything about it. > My personal favorite was a database of which character sets are in use > by each search engine so that search terms could be mapped into a > uniform character set. > > Some form of visit counting, despite the contradictions that entails, > is a frequently suggested feature. AWStats has gained a great deal of > popularity because it does visit counting and Analog doesn't. > > Other suggestions that I can remember right now include support for > GeoIP, visit duration estimation, pre-made configuration settings for > pulling more information out of the agent string, creating more than > one copy of a single report but with different configurations in a > single run of the program. Plus, there were others I can't remember > right now. > > Working on any of these projects would require both some programmers > with significant time to devote to Analog and someone to coordinate/be > the librarian. Some of these are good candidates for the external Helper App model - GeoIP is an obvious example, as DNS resolution is already done this way. (You can get a free GeoIP country database at http://www.maxmind.com/app/geoip_country. Using it to modify a DNS cache file to include Country information would be fairly simple. I don't think there are any free GeoIP data sources with higher resolution than this). The Character set issue for search terms might be another good candidate for an external helper app that preprocesses the logfile, though there might be some advantages in having Analog do it. Visit counting can be done now if you're using session cookies, but session duration would need some additional code (though you can get it easily enough if you post-process the report). In other words, if you actually log this information, you can get Analog to report it. If you don't log the information, Analog won't make the assumptions for you. Again, it would be theoretically possible to preprocess your logs to add a Session code in the User field. (Actually, that might be fairly easy - sort your logfile by IP address, assign a "sessionID" on each line, and increment the sessionID whenever the gap between to accesses from a single IP address is greater than your "visit timeout", or whenever the IP address changes. I think even my rudimentary coding skills might be up to that :-). One way or another, if the people with the skills to do any of these things wanted them badly enough, there isn't anything stopping them doing it. I'm not saying that they wouldn't be valuable additions, or that they shouldn't be developed, just that the reason they haven't been developed isn't because Stephen has prevented it, but because nobody else actually wrote the code to do it. Aengus +------------------------------------------------------------------------ | TO UNSUBSCRIBE from this list: | http://lists.meer.net/mailman/listinfo/analog-help | | Analog Documentation: http://analog.cx/docs/Readme.html | List archives: http://www.analog.cx/docs/mailing.html#listarchives | Usenet version: news://news.gmane.org/gmane.comp.web.analog.general +------------------------------------------------------------------------
