Author: way0utwest

  • Daily Coping 3 Jan 2023

    Today’s coping tip is to use one of your strengths in a new or creative way.

    I have a lot of strengths. They’ve helped me to be successful. This tip is a hard one for me, as I need to get out of my comfort zone and try something new.

    One of the strengths that has helped me is explaining concepts, ideas, etc. to others. It’s a lot of what has helped grow SQL Server Central over the year.

    I’m going to take some time to explain how I approached some problems in Power BI and then take feedback. I’m sure there will be lots of suggestions and advice that helps me improve, but my goal is to provide a model that even though I’m not a guru or an expert, I’m going to display what I know, and then improve based on what I learn from others.

    Hopefully I get more of you to blog and showcase your knowledge in the same way.

    I’m starting with two projects. A Flyway one (first post tomorrow) and a Power BI one next week.

    I started to add a daily coping tip to the SQL Server Central newsletter and to the Community Circle, which is helping me deal with the issues in the world. I’m adding my responses for each day here. All my coping tips are under this tag.

  • Daily Coping 2 Jan 2023

    Today’s coping tip is to start the new year off with something new – listen, watch, read something completely different.

    I started listening to more country music last year. It’s an interesting genre, especially from the guitar standpoint. I’ve learned a few songs and want to work on a few more.

    I’m starting today out with a few playlists that are outside of the normal things I listen to. No work (maybe), but I’ll put in the top songs from

    • Kelsea Ballerini
    • Luke Combs
    • Thomas Rhett

    It will be an interesting day.

    I started to add a daily coping tip to the SQL Server Central newsletter and to the Community Circle, which is helping me deal with the issues in the world. I’m adding my responses for each day here. All my coping tips are under this tag.

  • How Would Level 5 Self Driving Work?

    We had an incredibly cold front move through Denver last week. It was 46F one early afternoon, which dropped to 0F by the evening and down to around –15F overnight. Crazy cold, with snow.

    Driving the Tesla was interesting.

    This is part of a series that covers my experience with a Tesla Model Y.

    Driving to Town

    I went to the gym the next day, and saw this as I headed out of the neighborhood:

    20221222_164301

    It’s not a great photo, but what I noticed was that the 43mph had the D above it, for Drive, but not steering wheel for Autopilot. I use it on this road at times when traffic is light. You can somewhat see the road is a bit visible, and you can see some of the yellow through the snow in the middle, but no right line.

    The cameras can’t tell where the road is, so they won’t let the car drive. I’m not sure LIDAR or RADAR, or anything would work here.

    Here’s another shot in town:

    20221222_164902

    If you look past the steering wheel, you can see the road looks better. Not great, but better. There are curbs here, so radar/lidar might work. Certainly humans can see them, but the car again doesn’t have the steering wheel showing Autopilot is available.

    Update 1:

    One other interesting note, a few days later, I was driving this same road, but more snow had fallen against the center curb. In essence, the middle lane was narrower because of snow piled up on the left side (in the image above).

    I got repeated “lane departure” warnings, well 2-3 in a 1/2 mile, as the car thought I was crossing the lane boundary on the left. Strange, and disconcerting.

    Update 2:

    The next day as snow was melting, but the roads were very dirty, the car struggled to find the middle dividing line. I could see it as a human, but it was dim and barely visible. The car swerved a few times as it couldn’t always see where there was a turn lane on either side (L or R).

    Can Level 5 Work?

    This is a great example, to me, of why Level 5 is not likely to occur for cars. Maybe Level 4 within certain conditions and in well mapped/known locations. Or perhaps with sensors in roads that cars read. However, a little snow (this was 2-3 inches) has blinded the car.

    Humans drive on these roads all the time in snow. To be fair, many drive poorly, and often they don’t know where the lines are. However, we negotiate. We slide our car slightly left or right when there is other traffic. We aren’t perfectly in the lanes, and depending on the plows, we might end up “losing” a lane and just working with one path that isn’t really in either lane.

    Can computing learn to do this? Maybe, but it’s feels like an even more complex problem, especially when you now are asking the computer to react to how other drivers are moving on the road. Negotiating a route with them. Can a computer see another driver wave them on or ask them to stop?

    Tesla isn’t Level 5 or anywhere close. It’s Level 2, and perhaps Level 3 with the beta self-driving software (which I haven’t seen). I don’t know that I think anyone is closer, though I’ve heard the Google/Waymo taxis in Phoenix are pretty good. I’d like to try one and see, but those aren’t Level 5.

    I’m not sure I would trust Level 5 if someone claimed it, at least not unless every other vehicle was also a L5 controlled car.

    I do enjoy Autopilot and it was worked well for me. The FSD version was OK, but not enough better for me to subscribe to it. I look forward to cars becoming better and safer, but I’m not expecting to be able to go to sleep in the car and have it take me to my destination anywhere, anytime. I just don’t think L5 is anywhere within reach.

  • Looking Back at 2022

    This is the last workday of 2022. Next week starts a new year, and as I’ve often done, I wanted to look back at the year. This time I decided to look back month by month, at some of the headlines and memorable data-related topics. I’m tackling things month-by-month.

    In January there was a set of “tech experts” who shared their thoughts on the best database management systems. This one is worth a read for the humor involved. I wouldn’t really consider many of these to be DBMSes. I thought about including this one as an April 1 joke, but it was a real story. I get asked for my opinion at times by writers researching a topic they don’t understand. I hope I don’t come across like a few of these people.

    February was another sad data breach story. In this case, from the state of Washington where many tech people live and have startups. It wasn’t clear initially what happened, but later articles noted this was from a stolen device. To me, this was a great reminder why dev machines (and databases) should NOT have PII. Mask/obfuscate/anonymize that data please. Or at least delete my name from your dev systems.

    March had another humorous story (to me): Oracle is going to lure people away from AWS and SAP with their new offering. I could believe the latter, but not the former. Oracle hasn’t ever been good about pricing and it seems more people are leaving Oracle than coming to it.

    April is the month of April Fools, but this isn’t a joke. Another data breach, again from a dev system. This one from Fox News, which included information of not only employees, but guests and celebrities. The claim that this was a dev system and not production doesn’t matter if the data is real. Please people, keep prod PII out of dev.

    May had a funny post from Hacker News. Someone put their whole life in a database. The comment that caught my eye on HackerNews: “Men will literally devote hundreds of hours to building a bespoke database tracking every moment of their lives instead of going to therapy.”

    I worked a lot on my weight and diet in 2022. June had me finding a public database to help me choose better food. A public database on processed foods. Great idea, but everyone has an agenda. I hope this has some crowdsourcing and reasoning and isn’t just one person’s opinion.

    In July, another data breach. This time in China with information for 1 billion people. Wow. I dislike large databases for this reason. It’s also a good reminder why you ought to remove information from your databases over time, at least the PII part. Again, delete my name, if nothing else.

    There are so many types of database platforms. Have you heard of a vector database? Apparently, it’s for managing vector embeddings, whatever those are. An August article on the strange growth of the database market.

    September started this crazy AI art craze. There was a call to remove living artists from the database of works that an AI uses. Makes sense to me. I think artists deserve support and while I like AI doing new things, maybe wait until the artist isn’t producing work.

    October showed a good reason why we need ongoing patches or open-sourcing of code for retired systems. There was a 22-year-old vulnerability reported in SQLite.

    In November, what other news could there be than Lego Steve? It was a tiring week.

    December is just ending, but I’m ending on a reason why databases without auditing are a problem. Men behaving badly in this one. Gathering public information (or even semi-public) at scale can be problematic, and the information gets abused. Better controls, but also more auditing and triggering of some actions to prevent this (and other) sort of abuse.

    Let me know if you remember these events, or perhaps if there’s a favorite memory of the 2022 data world that you wish I’d included.

    Steve Jones

    Listen to the podcast at Libsyn, Stitcher, Spotify, or iTunes.