tag:blogger.com,1999:blog-2631429776292642259.post4263716939487039101..comments2024-02-14T04:37:42.427+02:00Comments on What's up on aprs.fi: Some database cleanupsHessuhttp://www.blogger.com/profile/12550787353137508264noreply@blogger.comBlogger8125tag:blogger.com,1999:blog-2631429776292642259.post-85895784748249727602009-04-02T22:35:00.000+03:002009-04-02T22:35:00.000+03:00Maybe I'm being draconian, but I suggest you remov...Maybe I'm being draconian, but I suggest you remove the corrupted callsigns from the database, since they're not technically "valid data" anyway.<BR/><BR/>...dtwW6DTWhttps://www.blogger.com/profile/08296054359177080557noreply@blogger.comtag:blogger.com,1999:blog-2631429776292642259.post-53381916018782958322009-03-30T19:19:00.000+03:002009-03-30T19:19:00.000+03:00Maybe a bit on the side, but please enlight me:I h...Maybe a bit on the side, but please enlight me:<BR/>I have a couple of objects I want to get rid of, but how do I do it?<BR/>test5oea, TEST5OEA2, Halten2 - do they time out in some way? The objects are deleted in my Xastir map, but still show up in aprs.fi.<BR/><BR/><BR/>Btw. do you have a way of reporting bugs?<BR/>I think I found a minor bug in the Status view: the status for my electric car (http://aprs.fi/?c=status&call=LA5OEA-5) shows the time twice instead of date and time (I guess) e.g.:<BR/>15:54:56 - 15:54:56 UTC: LA5OEA-5: 3595Km; 58%; 41.2DAh;12.50...Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2631429776292642259.post-76028959872612696902009-03-30T09:37:00.000+03:002009-03-30T09:37:00.000+03:00LA3QMA: Strange destination callsigns (like a digi...LA3QMA: Strange destination callsigns (like a digipeater path in the place of the destination call) are not much of a problem for aprs.fi, it only causes problems for the individual's transmitted packets if the digipeater path is wrong.<BR/><BR/>Corrupted source callsigns are more of a problem, since they clutter up the prefix browsing view, and generate a lot of rows in the database, rows which are never really referenced. And it's a bit hard to differentiate them from valid short-lived objects.Hessuhttps://www.blogger.com/profile/12550787353137508264noreply@blogger.comtag:blogger.com,1999:blog-2631429776292642259.post-11176734556626462622009-03-29T11:38:00.000+03:002009-03-29T11:38:00.000+03:00Strange callsign is one problem.But i also see cal...Strange callsign is one problem.<BR/><BR/>But i also see callsigns like "WIDE" etc<BR/><BR/>And people using UI-View using the unproto wrong. <BR/>UI-View has to use the word APRS before the path. So if you see things like WIDE1-1 via WIDE2-2 instead of APU25N via WIDE1-1,WIDE2-2 then this is most likely a ui-view user.<BR/><BR/>So instead of dropping those packets is it possible to make a link with a list of non valid packets? Or do you just want them out of the sql database?LA3QMAnoreply@blogger.comtag:blogger.com,1999:blog-2631429776292642259.post-86605137845823029972009-03-26T15:47:00.000+02:002009-03-26T15:47:00.000+02:0010 days might have been a limitation of some other...10 days might have been a limitation of some other implementation, or maybe a specification by Bob. But I think it's a better feature to store for a longer time and let the user decide which time range to view.<BR/><BR/>(M series.)Hessuhttps://www.blogger.com/profile/12550787353137508264noreply@blogger.comtag:blogger.com,1999:blog-2631429776292642259.post-51008142478039623132009-03-26T13:50:00.000+02:002009-03-26T13:50:00.000+02:00-oh didn't think of that :$-Wasn't that the standa...-oh didn't think of that :$<BR/><BR/>-Wasn't that the standard? on the aprs network? or am i so wrong? i remember reading that somewhere a couple of years ago.. when i first started experimenting on aprs<BR/><BR/>-great to hear :)<BR/>I'm also struggling to make caching better coordinates based.. not easy..<BR/>(not aprs.. google maps and some media data..)<BR/><BR/>further i see you have a cisco/juniper network.. that's the best ;)<BR/>J-series?<BR/><BR/>73's<BR/>ON3GPSMaescoolhttps://www.blogger.com/profile/06959086293469846740noreply@blogger.comtag:blogger.com,1999:blog-2631429776292642259.post-23246907817959769892009-03-26T13:25:00.000+02:002009-03-26T13:25:00.000+02:00Maescool: There are a lot of APRS items and object...Maescool: There are a lot of APRS items and objects, and tactical callsigns (like airplane/ship callsigns), in completely valid use, so a regexp to only match amateur callsigns is out of the question.<BR/><BR/>10 days? What standard? :) aprs.fi is currently storing a year's worth of history, which is a nice feature.<BR/><BR/>I'm already using memcache for some tasks, of course, but it's a bit tricky to use for finding a set of targets in a specific area. It's very quick because it doesn't provide things like range searches. Most of my load is currently caused by writes (inserts, updates, deletes), the reads are mostly cached already, one way or another.Hessuhttps://www.blogger.com/profile/12550787353137508264noreply@blogger.comtag:blogger.com,1999:blog-2631429776292642259.post-12173868383088357472009-03-26T12:58:00.000+02:002009-03-26T12:58:00.000+02:00aren't there regexes to check the calls.. or wasn'...aren't there regexes to check the calls.. or wasn't there a list to check the validity.. like the programs we use on contests :)<BR/>+ wasn't the standard norm to just store for 10 days?<BR/>PS. check out memcache for optimalisation ;) i use it on a high traffic site, brings the load way down!Maescoolhttps://www.blogger.com/profile/06959086293469846740noreply@blogger.com