HN user

moorow

143 karma
Posts0
Comments23
View on HN
No posts found.
An Update on Heroku 6 months ago

We saw this coming (like most people) a while ago when Heroku started flaking without status updates, and moved part of our workload to Fly. We ended up moving off Fly as well (significant unreliability and just some very strange network load balancer issues that would cause us downtime) and went to Railway, and that's been fantastic so far. We've moved our whole workload onto it.

RemoveWindowsAI 7 months ago

As someone in this exact position, I swapped to CachyOS this week and have been extremely impressed. I've yet to come across a game that didn't function extremely well (I have not tested bf6 yet, which I suspect the anti cheat may fail on).

Watson was a marketing exercise designed to sell a bunch of disconnected text and image processing libraries pulled together by consulting services. It did not function as advertised.

At one point we worked with a large energy company that was basically sold something LLM-like (large-scale indexing and searching/querying of documents) in 2016 or so. IBM had a team of 90 people doing full-time data ingestion for something like 26,000 documents. We got asked to do a counter-product in two weeks, which was literally just a TF-IDF search and some smarts around ingesting different types of documents. Both solutions performed approximately equally, except one cost something in the order of $185m and one cost $40k. Watson continued running for about a year until an external data science contractor realised they could query Watson for highly confidential board meeting notes, and it would provide full previews into the documents. The project was shuttered shortly after.

Alas, nobody gets fired for hiring IBM.

I mean I understand that's the goal, but in this specific CVE it looks like the issue was introduced in Bitnami's own scripts sitting on top of everything, so a ideally-zero-CVE underlying OS isn't going to solve that problem at all.

It also seems like this set of changes was made in this specific way to forcibly disrupt anyone using the existing images, many of which were made off the backs of previously existing non-bitnami open source projects, so I assume you can understand why people are annoyed.

But again, anyone with any knowledge or experience of Broadcom saw this coming, so...

As someone who's currently experiencing significant and crippling gastroparesis due to GLP-1's, being reductive about the side effects is not particularly helpful.

On the other hand, if 20 hours of leetcode practice is all that stands between you and half a million dollars a year, isn't that a pretty good indicator that the interview process isn't hiring based on your skills, talent and education, and instead on something you basically won't encounter in the workplace?

Dunno about you guys, but we're not charging doing payroll, management or regulatory compliance to our clients unless it's specifically requested/required by the client.

You could say "oh but that cost is bundled into your rates", but that cost is also bundled into your product fees for a product start-up, so..?

We don't have to do r&d documentation but we also don't get r&d reimbursements. Not a lot else different from a back-office perspective.

I own a tech consulting company in Australia. What rules are you talking about? The only change that's come in recently that's affected us is the changes to fixed-term contracts, and that's an entirely fair change to stop employees from getting dicked over by bad employers. Likewise, moving super through one-touch payroll is a great change that literally only affects dodgy employers.

The reason we're a country that digs shit out of the ground in lieu of doing anything else is the same reason why virtually all investment in the country is in real estate: it's not taxed highly enough to encourage people to diversify, and it's a sector that's too big to fail. Why would you invest in your mate's new tech company and potentially lose it all when you can throw it into a property with almost literally zero risk and far better returns?

Probably the most entertaining thing about the article (aside from the article itself) is seeing a comment section full of Americans who have precisely zero understanding of Australian humour or writing style.

As a former-PhD-Data-Scientist-who-quit-the-industry-because-it-was-full-of-fraud-and-went-back-into-software-engineering-and-is-now-an-Australian-consulting-to-Americans, this is even more hurty. Someone sent this to me and I thought I'd dreamposted it.

Great article, the anecdotes physically pain me in the same way that watching Utopia does.

GPT-4o 2 years ago

Not sure if you think training LLMs is carbon neutral, but if so I have some news about the barrier reef that you're not going to be that pleased to hear

Australians celebrate Christmas by going to the beach, and July 4th with sledding and patriotic snowball fights.

It may surprise you to learn that Australians don't celebrate the American day of independence lol

As someone that lives in Perth, our experience as a while through COVID was actually a net positive. Not to take away from anyone who lost loved ones, but Perth basically gained the benefits of COVID (WFH shift, better health funding, better remote technologies) and very few of the negatives.

It is an extremely rare situation though, and would not have been possible if it were not the most isolated capital city in the world.

Not OP, but as someone that's now rebuilt 4 serverless systems into monoliths (across 4 orgs) and achieved some pretty stunning results, serverless would be useful if all of these things are true:

1) Your workload is fairly simple 2) Your workload is unpredictably spiky from low lows to high highs 3) Latency is not a concern

If any of these aren't true, you're probably better off using a monolith (but you should think logically about your individual use case).

You can define container cleanup policies within the registry, matching on age, tags, etc. https://docs.gitlab.com/ee/user/packages/container_registry/...

Yeah, I gotta say - this is not great. Particularly given the overarching guidance on GitLab CI is "break everything up into 30 jobs", having to store containers as either an artifact or in the registry in order to pass them to sast or test jobs or w/e means that most projects are going to hit this limit instantly. Our largest work project has registry usage sitting at 800g - admittedly, we could be a lot better at cleaning up tags. But staying under 50 for the whole namespace, not even just registry, is going to incredibly difficult.

A fair while ago, we used self-hosted and swapped to SaaS because it was the same price and we didn't have to deal with upgrades etc - and we used our own CI runners anyway. Is the unspoken intention to force users like us onto self-hosted? Should we instead be using a non-GitLab registry to store this stuff?

Very long term GitLab user, very disappointed by this.

I remember a project when I was consulting that was floated, a few years ago. It was a blockchain-backed ledger for tracking fuel charges across a truck fleet - effectively, when the truckies filled up with fuel, they'd record it in the ledger. It can't be modified or altered, nobody can defraud the company because it's an immutable record!, shouted the sales team.

I mean, the truckie could always put the wrong value in when he entered it.