Some points in that page don't seem like valid criticism. Like, complaining that the compiler is not "zero dependency" on Linux, because it depends on glibc?

Also, allocating a fixed 1000 byte region at the start of a program is not a "memory leak". In fact even glibc has tons of such "initial fixed cost" allocations - IIRC valgrind has built-in "suppression file" to ignore these allocations. (It would be a memory leak if the compiler allocated new 1000 byte chunks every time println is called, but the article doesn't demonstrate it so I guess I'll never know.)

IIRC one of the original claims was to skip libc and do syscalls manually, like Go used to do on Linux.

So listing it makes sense in that context.

That sounds more like a bugfix. Doing syscalls manually is a rather perverse exercise, which go thoroughly proved.

No, this was never claimed.

Hardly an unbiased article:

> V can be bootstrapped in under a second by compiling its code translated to C with a simple cc v.c No libraries or dependencies needed.

Git is a dependency, which means perl is a dependency, which means a shell is a dependency, which means glibc is a dependency, which means that a lot of other things (including posix threads) are also dependencies. Pedantically, you could even go as far as saying that you could count the Linux kernel, the processor being used and the like as dependencies, but that's a bit out of scope for this.

==============

Ridiculous and unnecessary nitpicking.

Also benchmarking the language with a debug build, using a slower backend, and while running vfmt.

Here's a better benchmark: V compiling itself in 0.3 seconds:

https://www.youtube.com/watch?v=pvP6wmcl_Sc

So if it was a release build from the time, there wouldn't have been massive memory leaks, she would have been able to compile 1.2 million LoC per core per second, V would have only used 400kb of space to compile itself, and the curl binary wouldn't have been used for the download_file function? I agree that there's a cutoff where Linux dependencies don't really count as dependencies in this case.

This is great. I try to have an objective discussion about whether or not Christine's critique is fair, get no response, and am downvoted by who I assume are V lang fanboys. Not improving my impression of the language.

It's important to sort out these lies (and apologize for them if they were, in fact, lies) because no company will use a product they can't trust to do what it says. Though it's all at V lang's own peril I suppose.

The rest of her blog posts on V

https://christine.website/blog/series/v

I'm the author of that post if anyone has any questions.

Yeah let’s bash on V again instead of celebrating what this post is about: a really great web app.

Maybe it is time to move on to more productive discussions. People are obviously building pretty interesting things with V.

Calling this "a really great"web app is a significant stretch. It's an extremely bare bones clone of GitHub. Another case of over promising and under delivering from the V team.

I would like to believe Gitly is a really great web app. The problem is, how can I trust the claims that have been made about Gitly, when I have reason to distrust the group making the claims?

Christine's blog posts indicate V lang is blatantly over promising and under delivering. You can't claim a language is memory safe, when it has trivial memory safety issues. You can't claim a language compiles 1.2m LoC in a second, when it takes about 3 seconds to compile 50k LoC. If they said these are the goals for the language in the future, that's a different story but they did not say that. This was 2 years ago, so maybe things have changed and V lang is no longer over promising, but I don't have the stats on hand to back it up and so I remain skeptical. I like the idea of Gitly but I can't trust it does all they claim right now, given V lang's past.

If you have any stats on V lang fixing these issues, I'd honestly love to see them. Being able to deliver on all of these promises, would make V lang an extraordinary achievement.

I read that article, maybe .. maybe they wised up during the last 2 years.

It's a very large amount for the most talented person to learn, much less build multiple production-grade systems with.

We all should get some grace ofc, but does he also not need to show that he's learned not to make exaggerated claims? Otherwise there is no accountability and someone else gets fucked because they run across a bug because the dev didn't know use valgrind while claiming no memory leaks. I can't like that.

How many people would love to get their service on HN for a few hours?