| 21:33.25 | *** join/#uphpu jbot (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 21:33.25 | *** topic/#uphpu is Utah PHP Users Group | uphpu.org |
| 21:34.09 | mindjuju | ~rlart |
| 21:42.41 | mindjuju | the dr chipper looks cool, just don't have a yard to use it for |
| 22:03.17 | mindjuju | ugh, i've got to go through about 30k documents replacing a stupid phrase!!! i thought there was a switch for Grep that allowed for find and replace, not just find?? |
| 22:13.05 | mindjuju | any grep users in the house? |
| 22:13.22 | mindjuju | ~lart macnewbold |
| 22:15.04 | macnewbold | ~bonk mindjuju |
| 22:15.13 | jbot | ACTION bonks mindjuju over the head |
| 22:15.13 | macnewbold | yes, grep rocks da house |
| 22:15.46 | macnewbold | find/replace would be a switch for perl |
| 22:15.50 | macnewbold | I'm pretty sure grep is read-only |
| 22:16.06 | macnewbold | there's a perl oneliner for it in 'man perlrun' |
| 22:16.10 | macnewbold | look for replace in there |
| 22:16.38 | mindjuju | yeah, if i were using a *nix box, this would be easy |
| 22:16.44 | mindjuju | as it is, i'm using a port of grep |
| 22:16.51 | mindjuju | but i just found a port of gsar |
| 22:17.42 | macnewbold | you could grab activestate perl |
| 22:17.45 | macnewbold | it's for winders |
| 22:17.49 | macnewbold | and then you oculd do it right :) |
| 22:20.01 | jsmith | mindjuju: Use "sed", not "grep" |
| 22:20.16 | macnewbold | ah, yes, sed is another good one. |
| 22:20.25 | macnewbold | one of perl's "ancestors" |
| 22:20.45 | macnewbold | me too |
| 22:21.00 | macnewbold | I haven't gotten too close to awk yet, but it can be handy too |
| 22:21.59 | jsmith | Sure, you can find sed for Windows -- I've used it dozens of times |
| 22:37.20 | mindjuju | hey, any of you guys ever build custom web tracking |
| 22:38.58 | jsmith | Like stats tracking? |
| 22:40.35 | mindjuju | yup, what client hit what page |
| 22:41.22 | mindjuju | the problem i'm seeing is that clients hit page a, then go to page c, then shimmy over to page q, then back to a then back to c then back to a |
| 22:42.07 | mindjuju | this is my first web tracking so still a little rough, but you know what i was just thinking while i was typing this in |
| 22:42.34 | mindjuju | i should cookie the pages they hit, record on myside, and query the cookie, if marked in cookie, then don't record, else record |
| 22:42.40 | mindjuju | and make the cookie expire after a day |
| 22:43.23 | mindjuju | cause we do want to know that they went to pages a, c, q, t today and a,c,d, f tomorrow, we just don't want to know how many times in a day that they hit |
| 22:43.29 | mindjuju | which is what i'm currently tracking |
| 22:43.50 | mindjuju | anybody agree or want to offer an alternative? |
| 22:46.42 | jsmith | Well, I worked for Omniture for a couple of years, so I'm obviously biased |
| 22:47.11 | jsmith | But yes, pathing is a tough problem to solve -- but if done right, it can be very insightful |
| 22:47.38 | mindjuju | way cool! i didn't know you worked there |
| 22:47.58 | mindjuju | are you still with them? |
| 22:49.19 | macnewbold | one thing to probably take into account is page loads vs image loads and things embedded in the page |
| 22:49.39 | macnewbold | if you were to analyze my paths, on many sites there's a lot of forks in it |
| 22:49.48 | macnewbold | cause I open more than one link in new tabs at one time |
| 22:52.36 | mindjuju | we're actually looking at data for 2 groups of people, we are tracking the effectiveness of some of our web based products by recording when a client hits an offering and recording if they download the offering so on that level, i have to track page hits and name |
| 22:53.09 | mindjuju | on the second side, i need to know what pages are bringing in potential clients and monitor them |
| 22:53.35 | mindjuju | no personal info, just what pages did they seem to like |
| 22:53.44 | mindjuju | that way we make more of the good stuff |
| 22:54.49 | mindjuju | i'm not really using the cookie right now, so might as well use it for something |