crbelaus.com
I write software (mainly #ElixirLang) and I do martial arts. Sometimes I blog at https://crbelaus.com/
61 posts
711 followers
128 following
Regular Contributor
Active Commenter
comment in response to
post
Other shows I enjoyed recently are: Yellowstone, The Americans and Ozark.
comment in response to
post
Silo is amazing. Best sci-fi show I've watched lately along with Severance. Others (Fallout and The Expanse) are great but not at the same level IMHO.
comment in response to
post
For example "upcoming_events" may be such a scope that is extracted. Each LiveView or controller builds its own query based on that scope and then applies the required ordering, filtering, etc.
comment in response to
post
This is what we do at work and so far it has worked great. We try to keep an eye on common "scopes" (as Ruby on Rails calls them) that show up in multiple queries and extract them out.
comment in response to
post
I started using it in my day job (about 12 people in the team) and for my personal projects as well.
You can use it on any git repo without forcing others to switch (the rest of the team is using git).
comment in response to
post
This is the way.
comment in response to
post
Por el momento nada, pero le tengo muchas ganas. Todo lo que he visto es brutal, especialmente esto:
bsky.app/profile/giac...
comment in response to
post
Elixir 1.18 improves the type inference and checking and is now able to detect dead code, clauses that will never match, etc.
hexdocs.pm/elixir/1.18....
This will only get better over time. For example upcoming Elixir versions will be also able to do type inference on guards.
comment in response to
post
This allows the compiler to understand which data structures you are using an helpfully warn you if you try to access non-existing fields.
This makes it so much easier to move fields around and rename them confidently.
comment in response to
post
Just added the book to my reading list. I’ve also found the exercises and will start doing them tomorrow morning.
Thanks for the recommendation @thebrianb.bsky.social !
comment in response to
post
Thanks! I will take a deeper look. I've been also recommended this book: www.goodreads.com/book/show/14...
comment in response to
post
Good luck! (Or break a leg 😂)
We in the Elixir community are thankful for your work!
comment in response to
post
Looks like I have to learn other things than playing the guard!
comment in response to
post
Very interesting. Thanks! 🙇♂️
comment in response to
post
He doesn’t recommend me using a standing desk, but I have a strong feeling that switching between sitting and standing while working would benefit me.
Anyone with a similar problem has tried the standing desk? Thoughts?
comment in response to
post
I was going to suggest this same approach: triggers that convert the updated/deleted row to a JSON object that can be logged or archived.
@brandur.bsky.social provides a great example in brandur.org/fragments/de...
comment in response to
post
Wait. I have to check this out.
comment in response to
post
That's true. We switched from GitHub to GitLab and were able to import everything (issues, pull requests, etc.) but that depends on how good the integration is between providers.
Some other providers may not have this integration and thus you would lose the archive.
comment in response to
post
You may have just convinced me LOL 😂
comment in response to
post
My main gripe is bisecting but I just found out that git bisect supports `--fist-parent` (since Q4 2020) so that makes merges as easy to bisect as squashes. You could easily find the merge that introduced the problem and then, if you need it, go deeper and find the particular commit within.
comment in response to
post
That makes sense. I guess it is mostly a function of what the team prefers and is more familiar or comfortable with.
I have my preference but I am happy with any merging strategy as long as the team agrees to do it that way and does it consistently.
comment in response to
post
Your argument makes sense, but IMHO that is not a fair comparison. I'd say that quashing is more similar to log archiving than deletion.
The original commits that made the MR are still available in GitLab/GitHub/etc. along with the feedback from the reviewers, links and other relevant information.
comment in response to
post
Also, in my experience MR squashing only works well if MRs are small and focused on single features. If a MR does multiple unrelated things, squashing is a no-go. But I would say that the MR shouldn't do that in the first place.
comment in response to
post
So, to summarise: I believe that squashing helps to ensure a relative good commit quality on larger/diverse teams. It also ensures that every commit passes the checks and is deployable, which helps with bisects, cherry-picks and reverts.
comment in response to
post
In my experience same thing happens with MR commits. As the team grows it harder and harder to force everyone to use the same commit style (avoid "Fix error", "Fix typo", "Handle feedback", etc) so squashing ensures a minimum unit of work without those unwanted commits.
comment in response to
post
I think about squash merges as the code formatter for MRs. Theoretically we could all agree on a particular code style so we wouldn't need a formatter. In practice when the team grows it is impossible to keep everyone using the same style. So we require automatic formatting.
comment in response to
post
Only Murders in the Building
comment in response to
post
Congratulations! This looks great
comment in response to
post
Pull request sent!
github.com/Wilfred/diff...
comment in response to
post
While preloading with joins results in less DB queries it is not always the most performant option, particularly when used with has_many or many_to_many relationships.
I would recommend using preloads without joins by default (parallel preloads are great). Then measure and optimise if necessary.
comment in response to
post
I've noticed the same.
Somehow my posts get much more interactions and replies than on X despite being the same content. Also, I got more followers here than on X, despite my X account being much older
comment in response to
post
📌