
AWS DevTools Hero | Builder by heart | Building @nexus_share | https://xitter.cf/t.co/NtbD4wGU1N | Enterprise Architect @ FICO | Opinions are my own
Bensheim
Joined October 2007
- Tweets14.5K
- Following2.5K
- Followers2.1K
- Likes18.6K
Is the scary part of a deploy the code, or the fact that only one person really understands the rollback?
We keep asking if AI will replace developers. Isn't the quieter question what it's turning the job into?
My MCP server for @nexus_share now has 45 tools - is that too much? One of them can be used to upload images!
Is the hardest part of AWS the engineering, or deciding which three of the forty services you actually need?
One thing I did not expect while prepping Pegasus Galaxy for real players: I built AI agents that can play it, and they immediately burned tokens ;).
Agents as adversarial testers for your own game is a weirdly effective way to find the holes.
Would you trust AI players to balance-test your game?
One thing I did not expect while prepping Pegasus Galaxy for real players: I built AI agents that can play it, and they immediately burned tokens ;).
Agents as adversarial testers for your own game is a weirdly effective way to find the holes.
Would you trust AI players to balance-test your game?
Does your developer portal save time, or add one more place to look before you give up and ask in Slack?
Infra as code promised repeatability. Why does every environment still feel a little bit handcrafted?
Build-in-public reality for Pegasus Galaxy this week: I've been deep in the systems around the game, not the game itself.
The unglamorous truth of launching a game: the game logic was the easy part. Setting up things around it and creating a community is difficult.
Anyone who's shipped multiplayer: what broke first when real people showed up?
So true. Thats why #serverless is still a thing.
@edjgeek @julian_wood @focusotter and so many more know what I am talking about
That’s the best of course — what doesn’t run cannot be attacked. I meant that a managed runtime also takes care of patching, so for the Nodejs thingie for example, we didn’t have to do anything. AWS already did it.
And that’s the other thing: major providers usually get advance notification, so it’s often patched before it goes public. Priceless.
Golden paths are great until your use case isn't on one. Then is the platform helping, or in the way?
luckily I'm building on @awscloud with #Serverless so the issues didnt bother me much, but still.... TGIF!



