They article shows the “alarming” first message of the Twitter threads then if you look at the further analysis (in later replies) it’s from a function which is rage shake which is a bug reporting tool which then the researchers say oh ok no problem…
As an ex-consultant I would never hire a consultancy to do technical work or want to work there again now I’ve seen behind the curtain.
The main issue is resourcing is pretty much ALWAYS awful as their goals (maximum resource billing) don’t align with yours…people go on about it, but the term body shop is pretty apt.
I was in a leadership position and its pretty standard operating procedure for you to ask for a senior Java developer and be told you’re getting a junior Python test engineer just because it’s whoever is sitting around twiddling their thumbs (…and that’s from an internal perspective, if you’re the client you’ll just get lied to about their skills)
This basically leads to teams being comprised mostly of people who have absolutely no clue what they are doing - Is no wonder that a lot of the projects either fail, go over budget or have severe performance/security issues…and as a bonus you’ll get charged per day for a person as much as a permanent FAANG employee costs.
Worst offender I saw in my time doing it was Sapient, they seem to just bring anyone off the street, pure incompetence.
Currently using TypeORM in a production system, has caused us so many issues to be honest highly recommend everyone avoiding.
IMO they’ve added every feature under the sun but not bothered on focusing on quality, some things just don’t work at all, or you get stuck with horrific performance issues. If we weren’t so far down the line with it I’d rip the whole thing out although it’s causing so much pain I might just bite the bullet and do it anyways
I did see MikroORM the other day and looked promising so cheers for the quick review, if anyone else has any recommendations would be good to hear!
This made my day, absolutely wonderful achievement! Especially the area that you’re now in...it’s really not an easy task to retrain yourself for ML.
I’m 12 years in professionally and still loving every second of it! Currently been slugging out Leetcode problems all hours of the day to try and get myself to an org with a proper engineering track. So maybe further down the line I still might be able to solve computer problems (in some way or form) during my day.
I really hope to someday to be able to come onto HN and do the same thing as you :)
Do hate this about javascript. one of the main reasons on the web side (apart from js not having a good STL) however is enterprises are still locked in to IE11 which requires all these tranpilers/bundlers to use modern syntax which usually are the heaviest of all of these dependencies.
I would assume it’s due to a lot of the banks running windows and having large AD rollouts means it’s a bit of a gateway to the rest of their cloud services.
You have to use Azure ADFS for things like office 365/teams, so would make sense for people to keep all their eggs in one basket.
Came into the same issues. I hate example projects like these because they are misleading and read like silver bullets.
The split stack resources thing is extremely difficult to manage due to serverless not supporting CF params. so you have to use CF cross stack references which are horrible or hack in support like we did.
I’ve seen issues where you can’t put the value of a variable as an arn in an event due to the poor abstraction across the top of it as they’ve tried to make it cross provider, completely pointless. Just breaks.
Stick with cf/terraform if you’re starting a big project.
Been using it recently to try and run a large serverless app offline for dev purposes, it kind of works but the experience is ok...not great.
They seem to be piling all their energy into creating mocks for new (paid) services when it might be worth consolidating as the original mocks have a lot of issues.
- Documentation is non-existent, expect to trawl through github issues to work out how something works as it’s quite opaque (api gateway invocation)
- CloudFormation implementation is completely broken (no intrinsic functions) so unless your stack is simple you’re pretty much required to use any AWS api based devops tools (e.g. Terraform)
- API’s are not fully complete, means terraform either breaks on redeploy when it tries to get a resources status or at best case triggers redeploy of certain resources each time (the most exotic thing we’re using is SNS)
- Test suite is...light, seen a few things go through their CI and break it
- There’s a bit of non-consistent behavior - you’ll set an env var and find it’s not implemented for a certain case and be left scratching your head
- Expect to have to make pull requests yourself to fix things
This isn’t a winge, I understand it’s partially open source and you can just fix the issues yourself when they come up like we are doing. But just a heads up for who may naively look at it and think it’s a silver bullet...you’ll have to go through all these steps.
Does anyone know of a real beginners level guide to Bazel that isn’t the docs. Got a typescript monorepo with a few backend services/react frontends I wanted to build and I got a bit lost in the complexity of it due to package.json handling
Testing is valuable but my opinion of it has changed over the years.
On a new/fast evolving product I prefer to have a solid suite of integration/e2e tests and a lighter unit test suite near the edge of a server with no mocking of deps (e.g. if it uses a db, spin up a local instance). I would also test something that is non-trivial to understand or a critical dependency in the system e.g. a rules engine.
The reason being is - the code is in so much flux that the internal interfaces change constantly. In agile new requirements come along and you end up chucking a lot of the old code out of the window and wasting time.
As you move towards completion of the project and the internal interfaces shore up then increase the tests. So when it’s in maintenance mode someone else can easily make modifications and has documentation on maybe why something works in a certain way
Just to pick your brain (I’m an ML noob and it triggered an idea in my head). I was wondering if GPT-2 could be used to generate small paraphrases for an input sentence for search suggestions? e.g “I’m going to the mall today” -> “I will be going to mall today”.
They article shows the “alarming” first message of the Twitter threads then if you look at the further analysis (in later replies) it’s from a function which is rage shake which is a bug reporting tool which then the researchers say oh ok no problem…