katafrakt.bsky.social
⚗️ Just another #ElixirLang dev
👉 he/him
🔗 https://katafrakt.me
🥌 curling in my free time
440 posts
499 followers
151 following
Regular Contributor
Active Commenter
comment in response to
post
Annual reviews were always bullshit anyway (YMMV), so no surprise that now when we have bullshit machines at hand they are heavily used.
comment in response to
post
But my thinking is similar:
If there is a context in which certain syntax feature make the code more readable or more communicative, it's probably worth having it, even if it adds this unnecessary flexibility.
comment in response to
post
So do endless methods in certain contexts ;)
comment in response to
post
Just playing devil's advocate:
Would you say the pipe operator is an unneeded possible syntax? After all you can
elem = Map.get(map, :value)
elem = String.downcase(elem)
elem = String.replace("<id>", id)
...
comment in response to
post
I don't think it was fixed TBH. The "optimization" is still there and still is useless. Just the code around it is complicated further.
(proof: it already had to be even more fine-tuned: github.com/rails/rails/...)
But it's good to know at least something was done in that direction.
comment in response to
post
I described one of my favourites, "syntax errors based on data", here: katafrakt.me/2022/07/28/a...
comment in response to
post
Great! Code is even better than the video ;)
comment in response to
post
Is this some elaborate joke? Especially calling it a "search engine" part.
comment in response to
post
Wiem
comment in response to
post
ja pewnie też nie potrafię 🙈
comment in response to
post
I'm not sure what's going on but I want this.
Waiting for the video ;)
comment in response to
post
I hope that czyWybranoMiejsceDlaPsa function will already be in the standard library.
comment in response to
post
TBH what's the incentive? Python opens up the world of ML and probably more (Elixir has decent ML capabilities, but Python is still the king).
This would be possible with mruby for sure, but that means no access to most of the gems.
comment in response to
post
nevermind, solved
comment in response to
post
Not sure what this question means
comment in response to
post
"maximize code coverage" 🤨
comment in response to
post
These are a great guidelines. I wish more of such were publicly available and discussed. Fortunately I think things are changing for a bit better now, people are tired of jumping around the codebase to understand a simple feature.
comment in response to
post
[✔️] Yes, but the management disagrees
comment in response to
post
I'm not going to use the server anyway, which - I think - kind of diminishes its advantage over fzf. But I still think it might be a bit more solid than my current setup.
comment in response to
post
I agree it's not bad per se, but combined with other points.
If this PR had been reviewed, I wouldn't be so worried about the size or no content.
comment in response to
post
github.com/cachix/deven...
My favourite kind of PRs:
* vague title
* no description
* 20+ commits
* 2000+ lines changed
* merged without review
comment in response to
post
That's something I'm not encountering very often (or, TBH, at all) but I believe you ;)
comment in response to
post
I was comparing it to VibeVer, not BreakVer ;)
comment in response to
post
Sounds similar to PrideVer
pridever.org
comment in response to
post
One of my first blog posts ever was about overriding "new" 🙃
katafrakt.me/2015/05/15/w...
It does indeed feel a bit weird to return nil from that. I would expect some kind of an object always. But in principle I don't think the practice is bad.
comment in response to
post
strangely accurate
comment in response to
post
At least it returns JSON, not an HTML soup like Facebook Graph API.
comment in response to
post
One of the things that Event Sourcing brings to this is the change in responsibility for "maintaining" state transfers from the writer(s) to the reader(s) which changes the problem significantly - it is quite possible for independent readers to have their own projections of state without conflicts.
comment in response to
post
Wow, I never thought about it like this, but it's spot on.
comment in response to
post
Yes, I get it. I see this approach a lot, been guilty of it too (looking for dependency injection in Elixir for example).
But OTOH I hear a lot of people saying that document DBs simply don't have joins, which is not true as we can see.
comment in response to
post
What's the problem?
rethinkdb.com/docs/table-j...
comment in response to
post
I use Giscus. Used to have Disqus, but was also burned by ads.
giscus.app