| 00:31.30 | *** join/#riftuidev zz_Cidan (~cidan@cpe-23-242-200-64.socal.res.rr.com) |
| 01:01.51 | *** join/#riftuidev ZorbaTHut (~tocohl@maximilian.pavlovian.net) |
| 01:01.51 | *** mode/#riftuidev [+o ZorbaTHut] by ChanServ |
| 01:09.01 | ZorbaTHut | downside to having a large home network: replacing your desk takes five hours of rearranging cables |
| 01:10.50 | Ivnedar | probably quicker to build a new room extension to the house and start over there |
| 01:16.37 | *** join/#riftuidev zz_Cidan (~cidan@cpe-23-242-200-64.socal.res.rr.com) |
| 01:17.43 | ZorbaTHut | in retrospect that might not have been a bad idea |
| 01:25.14 | Baanano | ZorbaTHut: any idea, or could you ask somebody, how much it takes for the server to generate sales statistics after GMT midnight? it's been 1h23' and there are a lot of items that don't have sales statistics yet |
| 01:27.14 | ZorbaTHut | baanano, I'd be willing to bet it's midnight PST |
| 01:27.31 | ZorbaTHut | 95% of Trion is west-coast so we kinda standardized on PST :V |
| 01:29.29 | Baanano | <PROTECTED> |
| 01:29.43 | ZorbaTHut | ah, in that case it's probably GMT, just takes a while to load |
| 01:29.55 | ZorbaTHut | I imagine "how long it takes" will depend on how many items were on the AH |
| 01:30.21 | ZorbaTHut | note that the analyze command *just* pulls from a precomputed database - if the data hasn't been inserted in the database yet, it won't generate the data ondemand |
| 01:31.07 | Baanano | I bet it's just doing something like time() % (24 * 60 * 60) == 0, which coincidentally is midnight GMT, probably not intentional :P |
| 01:31.16 | ZorbaTHut | this would not surprise me :) |
| 01:37.56 | Baanano | btw, after GMT midnight some analyze commands result on server disconnection - I believe it's not related to the amount of calls as i've done it at full speed earlier without any issues |
| 01:38.55 | ZorbaTHut | huh, odd |
| 01:43.56 | Baanano | maybe it's requesting the exact item whose statistics are being calculated in that moment |
| 01:45.33 | seebs | I've at least occasionally done "move everything to another desk" as a step in the way to redoing a desk. |
| 01:46.19 | Ivnedar | ZorbaTHut: any idea if that analyse data might one day end up on the ftp site with the discovery data? :D |
| 01:48.15 | ZorbaTHut | Huh. Interesting idea, I don't think anyone's brought it up. I'll ping Daglar. Note that it probably won't be updated daily |
| 01:48.38 | ZorbaTHut | seebs, problem is I don't have a spare desk in the area :/ |
| 01:48.57 | Ivnedar | I was expecting it to show up with the "when servers reboot" type periodic thing, not daily :D |
| 01:52.10 | ZorbaTHut | realistically it'd be "when the discovery data is updated", which isn't entirely regular |
| 01:52.25 | Ivnedar | that seems to be whenever servers are rebooted :D |
| 01:52.37 | Ivnedar | well, intentionally rebooted :P |
| 01:53.21 | Ivnedar | It may be too much data to realistically store there and make available, I just thought it would be cool to have available, and see what sites do with it :D |
| 01:54.36 | Ivnedar | It might be worth having someone ask on the forums if anyone would be interested in having the data made available |
| 01:54.44 | Ivnedar | might prompt more of a discussion as to whether it would be used |
| 01:56.26 | Ivnedar | bbl, lunch time :) |
| 01:56.28 | seebs | Zorba: You need a migger house and more desks. |
| 01:56.36 | seebs | bigger |
| 02:01.38 | Noshei | gah, can't wait for poe's update |
| 02:01.45 | Noshei | so many days to wait still! |
| 02:44.21 | Ivnedar | far out, our implementation of servicenow is rubbish. |
| 02:45.01 | Ivnedar | 7 minutes after an incident is assigned to me, it emails me to let me know that the initial response SLA was breached. I resolve it 20 mins later after fixing the problem. 30 mins after that it emails me to say that the resolution SLA is breached ... |
| 02:48.16 | Baanano | :s is it broken or do you really have those SLA? |
| 02:51.57 | Ivnedar | Ahh, the SLA was based on when it was initially raised with another unit 2 hours earlier, that they ignored and then bounced to me. Fun |
| 02:59.22 | Ivnedar | And apparently they've set the SLAs up so that the resolution SLA continues ticking until the requestor accepts that the incident has been resolved |
| 02:59.36 | Ivnedar | which could be days later |
| 03:02.50 | Ivnedar | oh wow, Microsoft's hyperlapse looks awesome |
| 03:05.31 | Ivnedar | I wonder if that would work for a video game flyby |
| 03:05.33 | Ivnedar | It should |
| 03:12.12 | Baanano | what is it? |
| 03:12.35 | Ivnedar | http://research.microsoft.com/en-us/um/redmond/projects/hyperlapse/ |
| 03:12.59 | Ivnedar | doing timelapses of headmounted cameras gets a lot of motion in it that make the timelapses look terrible |
| 03:13.08 | Ivnedar | since it's just taking every 10th frame |
| 03:13.55 | Ivnedar | hyperlapse constructs the path that the camera took, works out which direction each frame faces, then works out a smooth path through that with a stable camera facing, then reconstructs the new timelapsed frames from the frames it has along the path |
| 03:14.40 | Ivnedar | so you get a much smoother looking timelapse video |
| 03:58.55 | Noshei | thats pretty cool |
| 03:59.41 | Ivnedar | yeah, I was very impressed |
| 04:54.28 | *** join/#riftuidev Noshei (~quassel@c-24-9-171-99.hsd1.co.comcast.net) |
| 05:16.41 | *** join/#riftuidev Noshei (~quassel@c-24-9-171-99.hsd1.co.comcast.net) |
| 07:24.00 | *** join/#riftuidev borbra (~quassel@grm-128-39-146-62.studby.uia.no) |
| 13:39.22 | *** join/#riftuidev jeffreyhamby (~quassel@216.136.24.180) |
| 13:53.27 | *** join/#riftuidev jeffreyhamby (~quassel@216.136.24.180) |
| 13:54.43 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 13:54.43 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 13:56.11 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 13:56.11 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 14:15.32 | *** join/#riftuidev Baanano (~chatzilla@212.122.123.164.dyn.user.ono.com) |
| 14:23.59 | *** join/#riftuidev Baanano (~chatzilla@212.122.123.164.dyn.user.ono.com) |
| 14:31.25 | *** join/#riftuidev Baanano (~chatzilla@212.122.123.164.dyn.user.ono.com) |
| 14:33.42 | *** join/#riftuidev Baanano (~chatzilla@212.122.123.164.dyn.user.ono.com) |
| 16:16.59 | Baanano | ZorbaTHut: about items not having data shortly after GMT I believe it was a misunderstanding on my part - there doesn't seem to be a big process that collects statistics every night, but they are collected incrementally each time an auction sells -> most items didn't have data because it was too early and no auction had been sold yet -> that's consistent with NA shards getting that data... |
| 16:17.01 | Baanano | ...filled at a faster rate than EU ones, as it was NA prime time while most EU was sleeping |
| 16:20.03 | Baanano | about server disconnects on Command.Auction.Analyze, i fixed it introducing heavier throttling on my part -> maybe the queue limit is too high |
| 16:26.40 | Baanano | about it not happening during EU prime time, *might* be related to server load -> eg. being loaded, the server receives/responds request at a slower pace that doesn't cause issues; while later, with less people playing, the rate at which traffic is produced is higher and some client or server guard causes disconnection |
| 16:33.37 | Baanano | anyway, that's just speculation, no way to prove or refute it on my part |
| 18:08.04 | *** join/#riftuidev jeffreyhamby (~quassel@216.136.24.180) |
| 20:23.47 | Baanano | wow, this is crazy => Bloodiron EU, so far today -> Plat spent: 19.452 / Plat earned: 18.479 / Plat sunk: >=5.803 / Plat sunk (old model): >=975 |
| 22:17.57 | seebs | What's "plat sunk" measuring? |
| 22:25.25 | Baanano | plat lost to deposits & fees |
| 22:29.20 | seebs | So what's the 975 measure? How much you would have spent on deposits using the old model? |
| 22:29.29 | seebs | That's ... a lot. |
| 22:30.04 | Baanano | yep, fees mostly, deposits under the old model were negligible |
| 22:31.44 | Baanano | btw, it's not me, it's the full traffic going through the AH |
| 22:33.02 | Baanano | in fact, plat sunk figures must be higher, as there are listings i'm not seeing because they're cancelled / sold too fast |
| 22:34.14 | Baanano | plat spent / earned should be more accurate as they come from server provided sales data |