119
WeatherMaster, Open Source, and the Future of Android; An Interview with Pranshul
(gardinerbryant.com)
this post was submitted on 04 Jul 2026
119 points (96.9% liked)
Technology
86068 readers
4293 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Strongly agree. You can dress up the interface in all kinds of ways, but if the data being presented isn't good the rest of it doesn't matter.
For a long time the data from Weather Underground was the gold standard. Then IBM bought them, made the data an expensive subscription, and effectively killed it for non-corporate users.
Then Dark Sky broke new ground with hyper-local weather reporting. Which was great until Apple bought them, killed the Android app, and made it exclusively available to iPhone users.
The best source of free data that is still available in the US is still the National Weather Service. Unfortunately the quality has been degraded by massive budget cuts while whole sections of the API have been taken away.
Finding a high quality data source at a reasonable price has become almost impossible.
I still use WU to this day as it's the most accurate and informative weather app in my experience, and the subscription to remove ads is $2.18 CAD per year. Is that expensive or have I misunderstood? It has been worth it to me by comparison to other options, and it is only to remove ads (though I haven't experienced the free version in almost a decade).
IBM sold The Weather Company to a private equity firm in 2024 and it has been operating independently ever since.
My experience was actually with the APIs and not with the main app. They originally allowed free access to the API for individuals, with registration, and a lot of excellent third-party apps let you enter your API to use them. When IBM purchased it, they priced the use of the APIs so that there was no option that made sense for individuals. If you weren't using it in volume, and paying tens of thousands of dollars a year, you were just out of luck.
I hadn't heard that IBM sold it off. I will have to go back and look at the original app. I had thought it also went through a big price jump, but that may have settled out in the meantime. It's been a while now.
The fucked-up part is that there are various networks of croudsourced weather data, but I think most or all of them are proprietary, or at least centralized (which means enshittification could put consumer API access at risk).
We need a service that's peer-to-peer (or at least federated) and open-data-licensed. And also not affiliated with the National Weather Service because Trumpism puts even that one at risk.
I generally agree with you. I suspect the current weather station hardware is mostly designed to report in to its corporate owners, but that could probably be worked around. Building a decentralized network for collecting weather data should be possible.
The hard part is that turning the raw data into actual forecasts requires a lot of processing power. My first thought was that it could be done by a large distributed network, similar to what was done with the SETI at Home or the protein folding project. My second thought is that there might be enough processing power that way, but it would be too slow to be useful. There's a lot of lag time in that kind of decentralized arrangement. It would likely produce accurate forecast of what happened several hours ago. I'm not sure how to solve that part of the problem.
Hmm, I hadn't really thought about that; I was just thinking about stations reporting current conditions. But yeah, you're right that that's the important part. Is weather modeling software another one of those areas like CAD where the state-of-the-art is locked up in proprietary shit, or is it government/scientific enough that the software is public? If one were to start building a distributed weather prediction system, are we talking about refactoring existing software to be distributed or reading research papers and implementing algorithms from scratch?