Wiser Now

No, seriously, what are the odds I got dumber?

2026-01-28, by DrFriendless technologymetaAngularMySQLGraphQL

I’ve just finished the gargantuan, nay, Brobdingnagian, task, of porting all the old blog posts from the WordPress blog to the new Astro blog. That is all except a couple for which I lost the images - I suspect the images were being stored in my Google Images and one day when I was cleaning that out I deleted a couple. But anyway I read all the old blog posts. The things that impressed me were:

I spent so much time mucking around with Auth0 that I never had time to make it useful for anything. The new login stuff is:

It does help that in the time between now and then I worked on a system with its own authentication, so it was a topic I learnt things about. That applies to many other facets of cloud / web / serverless development - I either learnt them at work or they became more mainstream and I became accustomed to the ideas they were showing me.

One thing I am missing a little is my Trello board. Some time ago I got email from whoever runs it now saying they thought I wasn’t using so I should do something… so I deleted it. I don’t miss the data, but I miss the functionality. Maybe one day I’ll write my own, and hook it into the site for general gaming shenanigans.

Another technology I spent a long time thinking about was NoSQL databases. The NoSQL concept has its place - I spend a lot of time doing SQL queries and then turning them into JSON to return to the GraphQL clients. Wouldn’t it be good to just get that JSON straight from the database?

Sometimes it would be, and that’s probably why MySQL added JSON-valued columns a few years ago. At work we stored random key-value data in XML encodings which couldn’t be queried unless they were decoded. Storing the same data in JSON columns meant they could be queried from the command line. So MySQL solved that problem and negated one of the great advantages of NoSQL databases. So for the foreseeable future I’ll be sticking with MySQL.

I also used to be interested in Google Analytics. The list if technologies killed by Google is infamous. The ones that affected me personally were Google Reader and Google Play Music, but the transition we had to do at work from Universal Analytics 3 to Universal Analytics 4 was brutal even if it didn’t count as a death. I now no longer trust Google to provide anything I use - no Captcha, no GCP, not one single service. Google owns Angular, which distresses me, but I was using Angular before they were involved. I hope it will survive for a while after they dump it.

In any case, having seen the relentless manner in which analytics from various sources pursues users, I have turned against on principle. In an online discussion about this sort of tech the other day, I asked if it just technowankery, and nothing convinced me that it wasn’t. If you can find your audience you don’t need to spy on them to improve your business, you need to provide a quality service. So, although there’s basically no money at stake in this little venture of mine, I’m going to keep it ethical and see how that goes.

Finally (i.e. I’m sure there was more I wanted to say but I forgot it all) I think I’m way better at cloud architecture now that I know what all the bits do. The fatal bug which killed the site for 3 years or more was fixed with a pretty simple queue. I think I need to add another queue at the moment, as some of my code is interacting badly with some of BGG’s code, and a queue will probably solve that problem.

Well, it is all very well to be wise, and write a lot of words. It is a whole other thing to knuckle down and fix one’s bugs. Thus I bid you adieu.