Hi

I worked on a pretty interesting problem for the past few days: I had to process millions of database rows and wanted to do it in minutes instead of hours or days. With some help, I went from processing 30 to 50,000 rows per second and dropping the total runtime from around 50 hours to an hour or so. It was a really fun exercise, and I documented the process here.

Now I have to say one thing: some people reached out to me saying that I could have solved this purely on the database level or by using other languages and tools. Someone said I was "bragging my bicycle goes 40 km/h downhill after removing the training wheels, while ClickHouse is doing mach 3 on the same data with one query planner pass".

To me, that's comparing apples to oranges. There's a place and time for any tool; for my use-case, on my blog, PHP is definitely the choice that makes most sense. My goal wasn't to get the absolute best speed possible in the universe, it was to see how much I could optimize a problem within some fixed boundaries; and I'm happy I was able to do that!

Let me know your thoughts, you can always reply to this email if you want to!

From the feed

• A simple trick to filter out interview applicants — This made me chuckle. It's a good strategy, to be honest.
• Open source strategies — I wrote about how I manage open source work.
• A case study about Laravel and FrankenPHP
• Using AI is no longer optional — Even when using AI extensively, you still need your core skills, your ideas, your creativity.
• Newsletters feel different — I can only agree!
• Multithreading in PHP — I'm not sure if this is the direction PHP should take, but it's an interesting read nevertheless.

That's all I wanted to share, have a great rest of the week!

Brent