The Twitter algo update prioritising verified accounts on the timeline and replies is live… it degraded my Twitter experience.
My feed used to be 95% tech: overnight it turned into ~40% tech, and mostly verified accounts I don’t follow posting dumb stuff I never asked for like:
I found myself busy blocking these random verified accounts, which have nothing to do with tech.
I said I’m holding off on Bluesky before because it has no DMs. That was before my feed got emptied from tech.
LinkedIn has more tech content in my feed today than Twitter 🤯
Rolling out massive, site-wide updates for all users, based on intuition and no data/testing was eventually going to lead here:
The irony is that to make Twitter experience pleasant: #BlockTheBlue is the answer.
Ironic because I subscribe to Twitter Blue since launch (I use bookmark folders, and have a huge collection for my research) and so tons of people will, obviously block me as well.
Twitter is now something I have not felt before: boring.
It's also increasingly full of irrelevant tweets, bad takes, and meaner than usual comments, coming from Blue accounts (not legacy blue: but paying Blue ones). Being a paying Blue account, I'm now forced to decide whether… twitter.com/i/web/status/1…
The irony: because I am Blue, this tweet is being inserted into people's timelines who never asked for it, didn't want to see it, and it's degrading their experience as well, while also proving the point very well.
Such a big difference between devs who are capable and willing to learn new things when they need to (eg new languages, frameworks) & have a "sure I will sit down and figure this out the next few days" attitude, versus people who are used to given formal training to do the same.
When I worked at JP Morgan, the co organized dedicated training for languages like Python or Java. It was how lots of my colleagues were used to learning.
But eg at Uber, Eng2 & above assumed they pick up a language in a few weeks, as a baseline expectation made clear.
Whenever a company can provide training, sure this is a net positive.
But as an individual, you go further and faster if you learn on the go. The more languages/frameworks you learn, the faster picking up the new one is.
It's so worth making it a habit. The earlier, the better.
On the second day it’s already clear that Bluesky has found early PMF with a good part of what used to be Tech Twitter here: more so than Mastodon has for me.
I find myself spending as much time there as I do Twitter. If you get an invite, check it out. It lives up to the hype.
I say this with sadness and as someone who has spent nearly a decade on Twitter and owes a lot to it:
This site is dying a slow death thanks to the senseless and continuous policy changes which slowly but surely erode niche communities such as what Tech Twitter was.
Bluesky offers hope because of all the parallels it has with early Twitter.
Back in the welt Twitter days, it was devs who were over-represented on Twitter: so is the case on Bluesky now. You could hack on top of Twitter API: so can you on top of the AT protocol.
Maybe this is strange, but earlier in my career, I was always "afraid" of my managers at first. When they called a meeting with me, I thought I could be in trouble for something. I never expected much good from my "boss."
When I became a manager, I realized how this was all odd:
As a manager, I was judged based on the performance of my team. My interest became for my team to do well: and, thus, for all team members to do well!
To have a great team, I needed to work *with* people: help them, call out great work, and not intimidate them or take credit.
So why did I barely feel that my managers were "on my side" from the first day? And why is this still rare to see?
Don't exactly know. Managers have authority: some focus more on this. And some managers don't realize that everyone doing well on their team makes them look great.
"Our company has a new Head of Engineering, who now wants to roll out daily standups for all teams. What is your take?"
That this approach is as effective as eg mandating all teams to do 2-week sprints. Curb a bit of autonomy and mandate the "how"... solving what, exactly?
If it is clear on "what the problem is that needs to be solved," and daily standups _really_ solve this problem, then this is top-down problem solving. If everyone agrees: good!
If people either don't agree, or don't know what the problem is: then more process to have process.
My observation is that a lot of this is the new exec coming in & feeling internal pressure to showcase how they "hit the ground running," and do this by taking a practice their old company did, and forcing it top-down.
It gives the sense of doing something. Even when it doesn't.
I have questions. How does water intrusion into *one* data center take a whole zone (which should be multiple, physically separate and redundant DCs) offline?
The point of availability zones is to avoid issues in one DC taking down the whole zone. What am I missing?
Oh, I just see: an issue in one DC took down a whole region! So all AZs within that region are down.
Wow, this is very bad: the point of AZs is exactly for this to not happen.
Google has explanation to do after they got the region and zone back up and running.
Spot on by @joshuaseattle. Out of AWS, GCP and Azure, only AWS guarantees that their zones are 3 (or more) physically separate DCs. Turns out for Google, even a region isn’t physically separate!
On paper both GCP and Azure have more zones & regions than AWS: but in practice…
A Sr EM friend in EU received great offers a year ago from 2 companies, valued over $1B each. They chose the less "risky" one, business-model-wise.
The company they rejected laid off 40% staff 6 months ago. Bullet dodged. Not quite:
Their company: the one they joined thanks to what seemed like a more sensible business model and space (pure software play, high growth etc) - just laid off 50% of staff, my friend included.
Which ever unicorn offer they would have taken, the end result would have been the same.
Friend corrected: they had offers from 4 scaleups, but only seriously considered 2.
#1 (unicorn): 40% layoffs Nov 2022
#2 (unicorn, their current company) 50% layoffs Apr 2023
#3 (publicly traded, ~$3B valuation): 19% layoffs in Feb