Roman

All day migraine is not how I wanted to spend Saturday

GOOOOOD EVENING

We balling with this list

Don't get me wrong, I agree completely, and I'd love to have something akin to Rust's stronger enum system to underpin errors, options, and pattern matching.

It does feel like you have to architect around the languages semantics

Love the green

This might be a consequence of trying to fit a square in a round hole or whatever the phrase is. Go very much lends itself to microservice architecture, within which your error mapping on your service looks more like checking the HTTP response code of another service, if that makes sense?

Best remake

I'm back home now, so no more doggo for a while. Gonna miss him

You know, seeing stored procs with their earliest change log at 2003 doesn't seem so bad now

Enforcing boundaries when you haven't done that before can come off as confrontational, but it isn't. You just have to be firm and make it clear you don't want them to be part of your life anymore. There's no need to discuss it at length, just state intentions and cut ties and communications

If this is true, this is the funniest thing I've heard about VB. I need to laugh to cope with the pain of the legacy projects

I swear this is a consequence of those tiktoks people were making a few years ago romanticising it for some reason. It's really not much to ask to be met with an open mind and no expectations of some type to fill

Even if they make you write VB .NET or god help you, VB6?

You could potentially still do this, but you might just have to hardwave details on the error and swap for an alternate at some point. This isn't something I've seen done in Go, and I'm not sure what use case it'd come up in? Sounds very OOP

I think the intention really is to focus on then as more generic values that you keep a same error path for rather than more granular control where you might perform different control depending on the outcome. It's closer to an A then B, if it fails here then do this. Then proceed to next

You also don't- there's no exhaustive checks required for matching errors. Onus is on the programmer here to do their homework properly

You'd have to snoop the service package code and couple the handler to its internal error representations if any custom ones are made. It's not nearly as elegant as what rust can do with enums and error mapping, no. I wish

Can't do it, I don't have object permanence so if I don't see people they don't exist until I'm reminded about them

Always be an eternal learner if you can, the basics never end, and those foundational skills are always useful

This kind of granular error handling isn't usually implemented. Usually you just return an error of some variation and you can potentially check it in the calling function to match it and do different things, but typically it'll just be going down error path or bubbling up from there

Was alarmed that you've just started programming or something

This aged like wine www.cve.org/CVERecord?id...

Corposlop straight from the source today

TS is a glorified vscode extension

Ok, but what if we didn't sit in the quiet coach on the train and talk loudly with a screaming toddler

Happy new year, hoping 2025 treats you better

Happy NY moots ~~

He's a complete teddy, so he'd love that. Highly recommend getting one someday