bradwood

8 ms·

61.
by bradwood 4y ago
What is so wrong with DNS over UDP? It's worked like that since it was invented. Surely making sure your cluster has a UDP-accessible DNS service can't be that hard. I think this is not an unreasonable decision for the musl mainta
62.
by bradwood 4y ago
So true. The main use case to solve is unpacking the damn thing - a cause of a lot of friction in my household. A great solution to this was created by a firm called Fisher and Paykel who made a 2 drawer dishwasher designed to permanently
63.
by bradwood 4y ago
And when it's not crashed it's slow as cold shit.
64.
by bradwood 4y ago
1Password teams
65.
by bradwood 4y ago
score == 20GBP for 50 quid, I've heard "reddies" quite a bit too, but my personal favourite -- an "Ayrton" for 10GBP (rhyming slang: Ayrton Senna -> tenner)
66.
by bradwood 4y ago
We use AWS Fargate with a small ec2 box orchestrating the jobs. There is a howto on Gitlab's docs that walks you through it all
67.
by bradwood 4y ago
Yes. I'm a premium user and I got so pissed off with their runners that we ended up building a runner farm in AWS. This is so much faster and better but as a premium user I cannot refuse to take the bundled CI minutes on their shitty S
68.
by bradwood 4y ago
And before anyone says "monorepo", GL's cicd support for monorepos is piss-poor too.
69.
by bradwood 4y ago
Same reason there is no specification for GUI interfaces. There are trends, and patterns, and conventions in both GUI and CLI design but it's always going to be about the UX, the use case, the target audience, etc.
70.
by bradwood 4y ago
Alex Honold and Free Solo should be on this list too.
71.
by bradwood 4y ago
Functional != Perfect Go live in Russia, North Korea, Belarus, Somalia, Nigeria, China, etc to see how it can not function.
72.
Woman wins case against British Airways over flight vouchers(bbc.co.uk)
45 points by bradwood 4y ago|22 comments
73.
by bradwood 4y ago
This is at worst a polemic and at best a half-baked argument that doesn't account for half the benefits of serverless. With serverless I have no IP networks to manage and secure, no servers to manage, patch, secure and a whole failure
74.
by bradwood 4y ago
Home-brewing. It's not too physical, but there is a lot of carrying around of malt and carboys and gear and of course, the cleaning.
75.
by bradwood 4y ago
I considered moving our B2B stack over from AWS Cognito to Auth0 around 6 months ago. My experience with their pre-sales & bus-dev personnel was so dire and their approach so entitled that I vowed to never give them our business. I&#x27
76.
by bradwood 4y ago
ramda - JS
77.
by bradwood 4y ago
I have a Dell Precision 5540 and it's a nice bit of kit. Fast, fully Linux compatible, good for gaming and dev with a decent onboard GPU. It's a bit pricy though.
78.
by bradwood 4y ago
Heroes of Might and Magic III
79.
by bradwood 4y ago
Now sure how this in any way trumps Kleppmanns's authoritative work Designing Data-intensive Applications .
80.
by bradwood 4y ago
Manjaro's installer is a lot easier to use.
81.
by bradwood 4y ago
Which universe?
82.
by bradwood 4y ago
It's not cool living under an apartheid regime. Even if you are the beneficiary of it. Source: white South African born in the 70s (with strong Jewish lineage).
83.
by bradwood 4y ago
So be it, if that's the case. At least it showed that a proper trade-off between features and bugs was made at the time. Remember that the triage can and should be rerun from time to time. Already open bugs could then be down- or upgra
84.
by bradwood 4y ago
FFS. We're not living in the 90s anymore. Spent a few hours configuring LetEncypt and move the fuck on. If they are incapable or unwilling to do this in this day and age I'd have to wonder about how they can be trusted to engineer
85.
by bradwood 4y ago
Ask yourself: given the system in its current state what change will make my customers most happy? Is it fixing this bug, or is it shipping this new feature? Do the thing that will produce the most customer happiness in the shortest amount
86.
by bradwood 4y ago
This is a blunt and unworkable approach. Stopping the shipping of any feature because some bug exists doesn't take what is important into account: the user experience. A more refined approach is to triage the bugs into 4 or 5 levels of
87.
by bradwood 4y ago
That saying predated k8s.
88.
by bradwood 4y ago
I agree with most of this except the part at the end that advocates "normal hosting". Don't do that either. Just go all in on a hyperscale cloud's serverless offering. No patching, no servers, no upgrades, no networking,
89.
by bradwood 4y ago
I think you'll find "I think you'll find" trumps "well actually". ;)
90.
by bradwood 4y ago
Well, if you're doing immutable infrastructure, it's probably perfectly reasonable. Not sure about rolling releases on production gear though.

More ›