> Only at small scales are full stack engineers valuable.
Seems like the ideal career is to start somewhere big and successful enough that allows you to specialize (after some poor generalists sabotaged their own careers by making it a thriving environment for specialists). Because even small scale businesses think they should start by hiring specialists.
I have recently migrated ~8y of Apple Numbers spreadsheets (an annoyingly non-portable format) to plaintext accounting.
It took me many hours and a few backtracks to get to a point where I am satisfied with it, and where errors are caught early. I would just suggest anyone starting now to enable --strict --pedantic on ledger-cli from the day 1, and writing asserts for your accounts as well e.g. to check that closed accounts don’t get new entries.
I really miss data entry being easier and not as prone to free-form text editing errors (most common are typos on the amount or copying the wrong source/dest account), but I am confident it matches reality much better than my spreadsheets did.
I'm not sure that's what you're implying, but DNS-level blocking is not more powerful than filtering in the browser, at least in the sense of what it is capable of.
Content-filtering can and most definitely does block domains entirely, but it can also filter page elements served from the same domain which match a known ad "signature".
Though, if you can't run an adblocker, e.g. your Smart TV's browser then sure, DNS-level blocking is your best bet.
I can attest that CSS is very effective for obfuscating e-mail.
I displayed my academic e-mail on my webpage for over half a decade using CSS to flip the text direction[1] without getting significant spam.
Even if I know my current branch, having my prompt show me untracked/uncommitted/unpushed changes helps to identify if something didn’t work because I’m in a dirty state, or if something I ran (unexpectedly) caused a dirty state.
For example, I don’t expect running scripts/build.sh to modify tracked files in the repo. Seeing part of the prompt go from “” to “?2!3” (two untracked, three changed files) makes that glaringly obvious.
If your pyproject.toml does not list all your dependencies (including dependencies of your dependencies) and a fixed version for each, you may get different versions of the dependencies in future installs.
A lock file ensures all installations resolve the same versions, and the environment doesn’t differ simply because installations were made on different dates. Which is usually what you want for an application running in production.
I have a very involved `gq` alias that helps me find and finish pending work. It works either in the current repository or a folder containing multiple repositories.
Basically, my assumption is that `gq` should return empty, which means I have a clean slate, and can start taking on new work. Otherwise, there is ongoing work that needs attention.
It just lists:
* modified/untracked files
* stashed changes
* local-only branches (not tracking a remote branch)
* branches out-of-sync with their upstream (either ahead or behind)
* branches that aren't the main branch (even if tracking and in-sync with a remote upstream)
Getting this command to return empty is a surprisingly effective way to stay productive, especially when losing focus due to too much work.
It's basically inbox-zero for git.
But it only works if you like working with a clean worktree.
Seems like the ideal career is to start somewhere big and successful enough that allows you to specialize (after some poor generalists sabotaged their own careers by making it a thriving environment for specialists). Because even small scale businesses think they should start by hiring specialists.
reply