Of mice and software

Found myself a few weeks ago in a situation where I quickly had to buy a new mouse because I had misplaced my quite-good-premium-brand one. A quick visit to a major online shopping site (I don't need to add a name, they are all more or less the same) turned up about a zillion mice from 5 to 150 euro. 85% of them from unknow, dubious or copy-cat brands. This seems to be the norm nowadays for online shopping. This has become quite a profitable business model (thanks to fantastic advances in manufacturing, designing products and a global economy) but not always the best for the consumer. Many of these "mice" are not up to snuff (no pun intended), but then again, some of them are almost the same quality as the 150 euro premium brands.

Hmm ok.. but Dirk, how are you going to connect this slight annoyance you seem to have with online shopping and dubious cheapo-products to software architecture?

Good question reader... let's dive in!

About a month ago I was talking with a fellow architect about what software components to use and where. There are so many choices now... choice is good and choice means we have a healthy industry but choice also means we have these 5 euro copycat products hovering close to you architectural shopping basket. Especially with the whole AI hype, it seems quite easy to design your own cool, must-have component and sell it or slap it on GitHub or a Nuget repository (this has also become a rather lucrative business model). I am not going to name any products or companies here here but as an architect you are often asked or forced to pick a variety of components, tools and software middleware solution to populate your landscape. Do you pick the 5 euro mouse or the 150 euro one? Of course, if this could be answered by a random guy on LinkedIn like me, world peace would be next. The answer is, like always, quite a bit more nuanced.

Yeah, the age old rule is still the first one you have to apply: Choose Boring Technology (thanks Dan McKinley - https://mcfunley.com/choose-boring-technology). Always start there and then work your way down to exciting crazy experimental stuff IF your requirements match. 95% of the time the boring stuff will do a perfect job and you avoid the risk of "Oh dear, component <any name you want> is not supported any more / became commercial / has a serious security risk / has bad performance or just doesn't work" questions that will come after you a few months into production. Avoid stuff that is Hype-Driven or, yeah I know, based on personal preference. You might be a big fan, but your company architecture might not be.

Well, I guess I will buy my slightly boring mouse now made by a boring company I have used in the past.

Next time we will talk physics!