There are two buried leads in the Segment/Twilio news:
1) This is the biggest YC acquisition by a large margin. Cruise was next at $1B.
2) The company was setup so that everyone on the team shared in this win, including alumni. Let's expand on this...
[1/11]
Chia has grown almost 1 exabyte (974PiB) in the last 24 hours. The scale of this is hard to comprehend.
Creating plots at this rate burns up the equivalent of one 1TB Samsung Evo SSD every 3 seconds.
An exabyte is 125,000 8TB drives or $25,000,000 in Seagate USB drives.
A shitty part about getting older is no longer being able to hold strong convictions. This makes it hard to be a Twitter thought leader. Like I'd love to be a "huge monorepo advocate" or a "major serverless shill" but all I can think of are the downsides.
Be very skeptical when you read statements like "40-70% of the US population will be infected over the next 12-18 months" because that's just not the way reputable scientists talk about forecasts.
I've always had a hard time taking the whole "agile" movement seriously given that, in my experience, velocity and agility have SO MUCH MORE to do with the software's architecture than how work is planned. [1/7]
"she is worshipped as a deity. her followers only know the symbol of the kubernetes logo, but have no concept of kubernetes as a concrete concept. they do know they would be relegated to a life of hard labor, manually entering kubectl commands without her."
$559 million worth of Chia's currency (XCH) has been farmed, so at this point Chia likely is a net economic negative, just based purely on a few surface-level metrics.
Queues are bad, but software developers love them. You'd think they would magically fix any overload or failure problem. But they don't, and bring with them a bunch of their own problems.
"this kubernetes administrator works 14 hour days and hasn't taken a shower in a month, hasn't left the office in days. at this point they are just typing kubectl with random strings!!!"
10/
@holman
has compiled a dope list of startups with extended exercise windows. You should work for one. It says a lot about the character of its leadership.
5/ This speaks to the incredible character and long-term thinking of the founding team.
Major key: they now have a whole roster of talented folks who they can call on for the rest of their careers. These folks can trust that they'll be taken care of. Very powerful.
100% compatible thoughts:
1) JavaScript is a garbage PL
2) It is worth being proficient at JavaScript
3) Frontend is its own highly respectable field, not a stepping stone to more “hardcore” pursuits
the quarterly planning model where everyone stops and pretends to plans an entire quarter of work all at the same time is one of the greatest business scams
"she further ascends. her consciousness is now fully integrated into the machine. her corporeal being has long been discarded. society no longer even thinks of kubernetes, but it is there. she is there. administering the systems."
Over the year I've been at
@planetscaledata
, developers have consistently asked us the question: "when are you going to support PostgreSQL?"
The answer: We aren't. That's it. But why? 🧵 1/
2/ Point
#2
was long-term thinking that started many years ago when they eliminated the post-employment exercise time limit (typically 90 days) from their employee option agreements. This is *KEY*
@mscccc
i know we're just having fun here but a lot of this is because ppl often don't differentiate between magic values (meaning can't be extracted contextually) and context-explained constants (like an inline SQL query)
Junior engineers: focus on improving by learning documented best practices
Mid-level engineers: hones skills watching and learning from their colleagues
Senior engineers: god damnit what the fuck does this JIRA even mean??? who the fuck told the customer we would do this???????
people who have never been inside a datacenter before: yeah the cloud is overpriced, you can just buy servers from dell
me: ok so if i add a perc h810 to my 730xd to wire it to an md1200 using sff-8088 cables, can i add 12 SATA disks as JBOD passthrough?
After 4 and a half incredible years (and an exit!), I’m closing out my time at
@segment
soon.
Next month, I’m joining engineering at
@planetscaledata
. I quickly became enamored by the team and the vision, and couldn’t let this one pass.
🥂
engineering management is actually just spending your entire day pretending to understand things that you are embarrassed you didn't already know about.
8/ They can choose to leave bad bosses. They can choose to leave unethical companies. They can take unpaid leave to care for family members. They can choose to work on things that are fulfilling and empowering, not only for the paycheck.
The thing that nobody talks about with the whole sidecar pattern is how much CPU it burns. If you're moving a lot of data in/out through the proxies it can be non-trivial. Adding 10-15% to your compute budget is a serious ask.
6/ It's tempting to get cajoled into not doing this or ditching it by the rest of the cap table, some of whom see alumni as "deserters" or "unbelievers" undeserving of a windfall, but the reality is that these sorts of restrictions just amplify privilege.
4/ Instead of only those who were already well-off, who could put the cash up to exercise becoming *even richer*, this changed the lives of a substantial number of households. Perhaps 100s have dramatically increased financial freedom.
7/ I am THRILLED that so many of my hard-working colleagues and former colleagues can now live a life freer of the burden of financial pressure, one of the key ways capitalism terrorizes labor.
like most caches, Docker's image cache comes with a capacity limit, at which point it will start to free up space by pruning LRU images. oh wait, no it doesn't. you copy-paste some random commands in from the internet every time you run out of disk space. yeah yeah. cool cool.
3/ It means that exiting employees who couldn't afford to exercise their options after they left the company didn't need to, and still very much benefited from this exit. Unfortunately, most startup employees don't exercise when they leave.
Things I once held as sincere beliefs:
* Databases are bad queue backends
* Static typing is a waste of time
* Monoliths are always worse
* Strong consistency isn't worth fighting for
* Using the "right" tool always trumps using the one you know
* The JVM sucks
“Working from home is great!” I nervously explain to Wilson, the volleyball that has become a permanent fixture on my dining table. “At least he doesn’t try to tell me about the superiority of GCP,” I tell myself, as a kind of consolation.
Why is their no relatively easy-to-use, decomposed library for distributed fault-tolerant consensus (i.e. "libpaxos")? Why do we have to use etcd or ZooKeeper or comply with a very needy Raft library? It's Time For Some Game Theory... 🧵