 This is just a weekend project. I downloaded the average weather data for weather stations around the world. I think there was, yeah, 72,000. And then I averaged out each month. And I thought it would be cool with Meteor just to be able to query on the front end with Minimongo. So it's about 1.5 megs of weather data. So it takes a while to come down. But once it's there, then you have your months and your ideal temperature range. And then you can kind of see which stations match the criteria through the year. Right now I'm limiting it to large cities. You can kind of change it. But I limit the rendering to 250 points because when you try to draw all of them, it takes forever. So there's all of them. As you can see, every time I click it takes a while. But these are all the weather stations that met that kind of request. I also note the bounding box of the map box. So it's only querying within this range as you zoom in. But it was just a fun little project. So the next step I want to do with this is using aviation data. So this is monthly average temperature. So if it's really hot in the daytime and really cold at night, it's not really the same if you're looking for what's the high temperature in Singapore in June. So I want to get average high, average low, and then also precipitation data, which I believe the aviation data sets will also include. But the biggest thing with this is clean data for all around the world. I could find nice US data sets. I could find nice German data sets. But stuff that's all around the world that will easily mesh together was kind of difficult. But I think airports will have that. But yeah, just a fun little project. Do you re-query depending on what's in the geo and the map? Yeah, so I think I ended up using the Mongo Geo feature. So Mongo supports geo points in geo querying. You could search by a circle or a square or a polygon, things like that. But the meteor support, like mini-Mongo querying for geo is kind of strange or it wasn't working or something. So you could actually just query less than, greater than, lat and long. And that worked. Yes, I had to... It was all CSV files and tab files. So I used just a simple script to clean it all up and then plug it into the Mongo database on the back end. So that's what gets pulled down every time. But it's up, it's on wengo.io. It's live if you want to play around with it. But yeah, I had to kind of clean out the junk. There's also... Oh, I guess I forgot about this part. But I have the yearly average when you click each one. But the station names are sometimes very odd. So I think one's called Batman and things like that. I don't know exactly who's running the show. But it's an official data set from one of the GOV websites. But yeah, the temperatures are just off. I mean, they're accurate. But being an average of the whole month, it's not really a good view. But yeah, any other questions? It's really bad. Meteor kind of chokes when you're trying to force a mega and a half of weather data into it. But it works. It's not failing on me. Once I restrict... I think the biggest thing was also drawing on the map. Drawing 5,000 points on a map. As you scroll, the scroller was kind of hard. So that limiting it helped a lot. But doing... So I have a batch script that goes through a raw JSON file and stuffs it into Mongo. So every time I would clean up the data a little bit, change the base dataset file, I'd clear the database and put it back in. And that took quite a while. Like multiple minutes. So you're not using the up-log? As far as querying or... So when I insert, I only do it on the server side. Because it's just at those command scripts you can create. Yeah, that's usually how secondaries work, up-logs. But reading, I don't think is the... What do you use the technique for? I mean, reading... I think there's a lot of data changing, maybe. Like if there's constantly stuff being pulled. The one thing, like as you notice the loading bar across as it's going, I think server-side rendering might speed it up a lot. Because the overhead to having Mongo or Meteor handle all of that syncing might be a lot higher than if I let the server return a mega-and-a-half of blob data in the source. But I just have to make sure it doesn't try to then sync it back to the back end. But yeah, there's a few little tweaks. Maybe. Yeah, with the new Wireshark. Wired Tiger, yeah. Is that what people are using in that production yet? I think it's safe. It's not. Yeah. I think it's supported in Meteor, but I'm not sure if it's rolled out to like the modules and those guys that are there, so it's fine. I'm friends with the Compose guys. I think they're going to keep away from it for a little while. Just to make sure, you know, battle-tested. But cool. Thanks.