@PompeySteph@jk_rowling apparently you need to get degree and a certification in activism before you can be referred as such... @jk_rowling is as precisely right as every time on this particular topic!
@elonmusk Elon, come visit Ukraine, meet the president and ask him directly about all those questions you have bubbling here via Tucker channel and other lunatics. Don't limit yourself by just visiting Israel...
The @t3dotgg Problem
Teaching a generation of developers to race for quick dunks, ignore context when evaluating decisions, outsource thinking to who has the most internet points, and then crybully if anyone disagrees
RE: Exiting the cloud
I'm not sure if Theo has actually read the articles.
The first paragraph has the conclusion spelled out: "Renting computers is (mostly) a bad deal for medium-sized companies like ours with stable growth"
Theo begins his teardown by talking about:
- Why you can't just consider server costs (37signals posted a full line-item cost breakdown, including caveats around production/non-production servers, power, rackspace, bandwidth)
- The costs of employees to manage (37s has a ~10 person ops team and has said they have not needed to increase)
- Losing the agility to try things out (DHH explicitly mentions the opposite: high costs to try a new experiment around disk based cache; vendor lock-in means you need to wait for providers to support new offerings)
- You have to consider your growth curve (the articles say the same including examples like Black Friday traffic spikes and the launch of HEY with 300k signups in three weeks)
- "Majority of companies have variable workloads" (I don't think this is true and probably a consequence of being in the early stage start up and/or VC hypergrowth bubble...B2B SaaS companies generally have a slow ramp of growth)
- The only time to leave the cloud is if "your traffic has peaked or is slowly declining and you want to squeeze the last cent from every customer" (Or...your traffic is predictably growing and you can forecast capacity? What?)
And the absence of other useful discussion points: is it good to consolidate all servers to a few platforms? what has changed in the past 10 years when it comes to managing data centers? could you actually get better performance? how should I personally evaluate if a cloud exit make sense for my company?
Does Theo approach the situation with an open-mind, looking to see what lessons he could learn for his own projects and share with his community? No. He gives a few surface-level takes that are all covered in the articles he is lambasting.
The articles are all here (Theo doesn't link or encourage viewers to read for themselves -- that is "scary" and "concerning" and "might be used by your coworkers to make bad decisions").
You may have a knee-jerk reaction to the title or hate the author, but these are not long posts and they do provide the context and caveats to help you understand if your situation applies.
https://t.co/it1ufw928h
https://t.co/FxPFMlcThb
https://t.co/YR0yDr0YIM
https://t.co/mGO8ZFQTUe
https://t.co/4RcbwGccSO
https://t.co/E0eyz0JsKu
I'm sure I will get plenty of "lol u mad?" and "DHH bootlicker" comments. That is the kind of tribal, hostile to other opinion community you are building.
Not even going to cover the TypeScript Turbo part: that will be left as an exercise to readers to find the context of how Turbo is used and how Basecamp and Rails have historically managed open source libraries and their obligations.
It's a clear case of Gell-Mann Amnesia: when you see someone talking about a topic or situation you know about and are clearly wrong or misrepresenting, but then the next topic comes and we trust the same people to be giving accurate takes.
Challenge the people creating content to "Do better". It's hurting you and your abilities, not them.