Technically you might argue that I should show 1 AM if the user did run at 2 AM in summer time, but everyone I personally know keeps their schedule over DST transitions. That is, if they did something at X o'clock before the transition, they'll keep doing it at X o'clock after it (sleep be damned). So generally showing 2 AM would be the most correct solution.

You cannot get that information from just UTC if you don't know where the user was when they made those historical events. Thus you either have to keep a history of their location (complicated) or just store the local timestamps (or at least the offsets) at the time of event. Always being able to convert from UTC with no extra data assumes that the user will never move, which might be fine if your application is limited to users in a single country.

The naming of "timestamp with time zone" is one of my favorite pet peeves. It's one of those things that you can say "well technically it's true" about.

The article suggests that for past events, UTC and this timestamptz would be acceptable as a general rule, but even there it depends on what you will be doing with the data. If you intend to interpret it as a series of local occurrences and try to visualize/summarize that data later, you may be in for a surprise as your user has moved to another timezone and now all the past events are translated to the wrong local hours [1]. For example, your system might end up showing that the user's best time for jogging based on historical data is at 2 in the night.

[0] https://news.ycombinator.com/item?id=48558005

[1] https://blog.nytsoi.net/2022/03/13/utc/

The article suggests that for past events, UTC and this timestamptz would be acceptable as a general rule, but even there it depends on what you will be doing with the data. If you intend to interpret it as a series of local occurrences and try to visualize/summarize that data later, you may be in for a surprise as your user has moved to another timezone and now all the past events are translated to the wrong local hours [0]. For example, your system might end up showing that the user's best time for jogging based on historical data is at 2 in the night.

[0] https://blog.nytsoi.net/2022/03/13/utc/

But I also just followed it for the first time from beginning to end, not doing anything. Because I assumed someone had already done the math.

Comments URL: https://news.ycombinator.com/item?id=47121210

Points: 91

# Comments: 53

> Then drop any elements and attributes that are not allowed by the sanitizer configuration, and any that are considered XSS-unsafe (even if allowed by the configuration)

The XSS-unsafe functions are all named "unsafe". Although considering web programmers, maybe they should have been named "UnsafeDoNotUseOrYouWillBeFired".

[0] https://developer.mozilla.org/en-US/docs/Web/API/HTML_Saniti...

If you compile it to JS, then the guarantees change to JS's guarantees.

Personally I've felt that the JS target is a big plus and hasn't detracted from Gleam. Writing a full stack app with both sides being in Gleam and sharing common code is something I've enjoyed a lot. The most visible impact is that there's no target specific functions in the stdlib or the language itself, so Erlang related things are in gleam_erlang and gleam_otp, and e.g. filesystem access is a package instead of being in the stdlib. If you're just into Erlang, you don't need to interact with the JS target at all.

LiveView does provide the tools to simulate latency and move some interactions to be purely client side, but it's the developers' responsibility to take advantage of those and we know how that usually goes...

1800 kWh is very little. We use around 12000 kWh and our neighbours' new house uses around 8000 kWh annually and most of that is heating. I'm not sure how many houses can hit 1800.

Last I heard there was a feature branch for testing a software implementation of floating point that would fix these issues, but naturally it would be a lot slower. I haven't tried it myself.