Ryan Simons 🇯🇲 Data Analytics | Big Data

CEO & Founder DenArthur Analytics of all things Data and Tech 💻📊
Motorsports fanatic 🏍️

Amateur cyclist 🚲

Lifelong Student

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

Have to celebrate every win that comes your way in business. We've made it this far, the work continues. 🥳🥳🥳🥳🥳🥳🥳🎉🎉🎉🎉

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@frectonz @neogoose_btw Perl is always the right answer 🙌🏿

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

Framework fatigue is real. Decided on Vanilla JS 3/4 years ago and I was instantly more productive.

https://t.co/7yyBZsfGH7

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@jnunemaker Been using them for endpoints on a static site for a few years. Pretty easy to use, loved how easy it was to get setup. Also did some POCs using them for AI LLM use. Documentation is top-tier.

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

For kicks tried to see how much of C I remember. Forgot just how simple it was to get an idea out. Neural network up in no time and piped the epoch data to Claude on the command line to provide feedback on the rate of learning and the network generally. Pretty cool 🚀

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@davidmullings It truly is frightening the content that gets recommended our way on these platforms even if you are careful

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@endingwithali I had to write out a backup and retention SQL procedure in an impromptu job interview exam. Got the job and and about 3 months later recognised the code in the midst of my first monthly backup activity with the team 😄

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

The majority of your company's data and know-how isn't where you think it is.
https://t.co/HB1vP9rnIV

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@arvidkahl This is 🥇

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@kamilaturap Did this recently for an old https://t.co/Px3HhvrJF1 4.5 project into my current stack of choice NodeJS / Express / Vanilla JS 🤯

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@dorkitude This is one of my toy machines, wouldn't necessarily recommend on a work machine though.

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@dorkitude I used it to try to fix Bluetooth on my Ubuntu Mac book Air. Gave it free reign of the system, I paid attention to make sure it wasn't nuking anything, and I did sudo where necessary. Was plenty impressed with what it tried, went as far as importing kernel modules.

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

Hey Junior Devs, just know that even with decades of experience, a 60 minute dev trip caused by "user_prod" instead of "user-prod" is enough to trigger an identity crisis 😁

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@wieslawsoltes @davidfowl Whoa, just got added to to Todo list

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@jeffdafo @gwenshap Huh? That's the conclusion you came to based on ALL of my responses above? That is curious in of itself. Good day to you sir.

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@sysxplore Agreed. Hopped all about the ecosystem for a while, Mandrake, Cent OS, Fedora plus others but Ubuntu is the go to. Now I veer outside only for experiment.

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@jeffdafo @gwenshap Each implementation or vendor will identify their niche and optimise for a different set of needs. Same reason we have so many car brands, same reason you can self host or use a VPS. To conclude the discussion ... 12 joins is nothing to be making architectural decisions about.

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@jeffdafo @gwenshap The DB can do significantly more than we give it credit for, joins are the least. If you can't get good response on the joins the first thing is to pick apart the explain plan to see that your SQL code is optimal. Table partitioning etc, a whole load is possible before app layer

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@jeffdafo @gwenshap If your dataset has grown enough that you can actively consider the investment of replicas and sharding them you are already in the position to manage replication. You're not taking this step on a whim, this is a solved problem.

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@jeffdafo @gwenshap more data up the stack on every single call where you made this decision, that you don't need to display in the UI.

[Ryan Simons 🇯🇲 Data Analytics | Big Data @ryansimonsjm]

@jeffdafo @gwenshap That's optimising for a different problem, one that may never come. Changing entire data layer implementations do happen but not as frequent as one would think, that gets debated a lot. In the meantime you would have resigned yourself to the performance penalty of passing...