Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
jhugo
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
34 ms
·
451.
▲
by
jhugo
4y ago
> I agree that buses are cramped and that trains are way more comfortable, but I think we can also agree that whatever space you don't have in a bus, it will be even worse on a plane... Depends which class you travel, and with which
452.
▲
by
jhugo
4y ago
This is a fair point — I have never been involved with, nor really patronised, tied houses, so I know less about the economics of them. The alcohol is still far more profitable than the food regardless; it's just that in a tied house a
453.
▲
by
jhugo
4y ago
Because you could be (and many high-value targets likely are) using your own IP address space "internally", rather than RFC1918. (Plus, IPv6.)
454.
▲
by
jhugo
4y ago
You can rebind the exact same origin then.
455.
▲
by
jhugo
4y ago
Well, they don't need to word it quite as ridiculously, and it depends whether SomeSite and Felt require attribution.
456.
▲
by
jhugo
4y ago
Yeah, it was really hard to take it seriously after reading that.
457.
▲
by
jhugo
4y ago
Mmm, maybe. Average occupancy of buses is usually only around 40%, which probably pushes them above the per-passenger emissions of a modern aircraft, and being typically diesels, they have other emissions than CO2 to worry about. On a perso
458.
▲
by
jhugo
4y ago
That's not the case; the vast majority of a pub's margin still comes from alcohol. GP comment is just wrong.
459.
▲
by
jhugo
4y ago
It sounds strange because it's just false. The vast majority of the margin in UK pubs comes from alcohol sales. I've worked and invested in the industry and it would be a very very unusual pub where this was not the case. As an as
460.
▲
by
jhugo
4y ago
In case it's not clear, nothing in my comment suggests that k8s will magically solve all your problems. It just provides abstractions that make growth (in size and complexity) of systems easier to manage, and helps to avoid lock-in to
461.
▲
by
jhugo
4y ago
This seems impossible. Frankfurt is nowhere near the great circle route (shortest path) between Seattle and India, and Qatar flies directly from Seattle to Doha, which is near the route, and which itself has frequent direct flights to many
462.
▲
by
jhugo
4y ago
The overhead of containerisation is mostly in the learning curve for teams that are not already familiar with it (and the consequent risk of a poor implementation). A well designed build pipeline and deployment is at least as efficient to w
463.
▲
by
jhugo
4y ago
You seem to be describing very early stage companies, and if so I agree, host it on your laptop if you need to, it makes zero difference. But it's not binary with Netflix on one side and early stage on the other. There are a lot of com
464.
▲
by
jhugo
4y ago
I certainly wouldn't recommend that anyone "focus on multi-cloud" in an early-stage company (unless of course multi-cloud is a crucial part of their product in some way). Kubernetes is basically an industry standard at this p
465.
▲
by
jhugo
4y ago
Yes, completely agreed. Multi-cloud is really not that difficult nowadays, and it puts you in a better negotiating position (when you end up spending enough to be able to negotiate), as well as giving you more location flexibility and the a
466.
▲
by
jhugo
4y ago
It depends. There are many things that carry a cost early but pay for themselves many times over later. Whether that will be the case for your startup depends whether you end up needing to scale quickly or not. It's also worth consider
467.
▲
by
jhugo
4y ago
> I think the set of cases where "we need to scale up rapidly if it succeeds" and "Kubernetes solves all of our scaling needs and we aren't going to have problems with other components" is almost empty. I agree,
468.
▲
by
jhugo
4y ago
If using k8s doesn't reduce the overall complexity of your system, then you probably don't need it. But not all systems are the same as yours.
469.
▲
by
jhugo
4y ago
> At the same time. I can scale heroku to 500 servers, and still be under the cost of one ops person. I can make that change and leave it there. I can do that all in under 30 seconds. Oh. And CICD is built in as a github hook. Even with
470.
▲
by
jhugo
4y ago
Anything can be a red flag signalling premature optimisation, depending on the scope of the problem you're solving. Like any dogma, "you don't need k8s" will sometimes help you, sometimes harm you.
471.
▲
by
jhugo
4y ago
I did initially assume that, but then it doesn't make any sense at all. Lufthansa doesn't fly between the US West Coast and Asia except via Europe, and a large number of other airlines offer direct routes between the US West Coast
472.
▲
by
jhugo
4y ago
I used to fly between London and Berlin a lot, and absolutely loved Tegel then. LCY-TXL was a great route. From my apartment to sitting on the plane at LCY took 20 min; from my desk in Berlin to sitting on the plane at TXL took 40 min. Prob
473.
▲
by
jhugo
4y ago
> I like flying lufthansa. They have the shortest flight times to Asia from the west coast. Can you elaborate on this? Where do you mean by "west coast"?
474.
▲
by
jhugo
4y ago
If there's no train, though, the turboprop plane is going to be better for the environment than all those people driving.
475.
▲
by
jhugo
4y ago
So a dog configured your CI then.
476.
▲
by
jhugo
4y ago
Nothing stops you using site- or link-local addresses and NATing to a single (or several) public IPv6 addresses, just as you do with IPv4. There was no "removal of NAT", it's just not necessary any more. A proper firewall wit
477.
▲
by
jhugo
4y ago
I don't know Italy laws, but everywhere that I'm familiar with, debt doesn't automatically pass to heirs. The heirs can always choose to not take the house, and thus not take the debt either. The house will be sold to pay the
478.
▲
by
jhugo
4y ago
Being less convincing saves the scammers time, because you won't reply unless you're easily fooled.
479.
▲
by
jhugo
4y ago
A DNS or firewall problem that causes Gitea to intermittently spend 10+ seconds responding to requests, when running in a cluster alongside other containers that have no such problems? Can you describe a plausible mechanism for that? We did
480.
▲
by
jhugo
4y ago
We tried to migrate to Gitea at my company, ~100 repositories with a bunch of mirrors, lots of CI, renovate-bot using the API, etc. It was a disaster. It was unusably slow (10+ seconds to load the list of pull requests, API calls to open a
More ›