I had the opposite experience. I had rented a house that had an empty patch of dirt and asked the landlord if we could plant some vegetables there (which I didn't have a lot of experience in). My older neighbor saw me and said that we had great soil in our neighborhood but I didn't really think about it. The vegetables grew amazingly well, we had way too many tomatoes and tons of broccoli that kept growing back and more. I barely took any care with it after planting, other than very sporadic watering and harvesting.
The next year we moved from that house to a new place, where we couldn't plant directly in the ground but the landlord was happy for us to use pots and planters. I eagerly planted my broccoli again thinking we'd have the same endless supply... but this time it barely produced anything and looked nothing like the last year. I bought some kind of soil bags from the gardening store after asking an employee which would be good for vegetable pots and planters. Something about the pots or the soil or otherwise made a gigantic difference even though we had moved probably less than a mile distance wise.
I'm a very amateur gardener so I may have made some other mistakes, but I think I treated the plants very similarly both years.
This makes sense as pots require very high precision of soil composition and water management. You need good drainage but a water reservoir under the pots with some kind of wick to keep it watering from the bottom. Also the size of the pot you need is huge compared to your intuition. A single tomato plant requires more than 100L of soil, 30% of which should be perlite, you need some pebbles for better drainage mixed in, and more than 50% should be compost. This often isn't enough for a tomato, so you need to add slow release fertilizers at different levels of soil. In a warm day (> 30C) you need to give 5L of water in the morning and in the evening. The water should stay in the pot but not flood it. Ah, also, they cannot handle cold as good as in soil either. So the pots should be moved indoors when it's too cold, and require shielding from a lot of wind.
If you prepare the pot, soil etc properly, you can get good results. It's very repeatable as it's a very precise recipe. If you put potting soil in a pot randomly it won't work at all.
I've had tomatoes grown successfully, with decent yield in pots that take maybe 30L of soil, and without any perlite. It isn't that hard if we're talking about hobbyist stuff, where results don't have to be ideal but just "good enough". Picking a variety that is hardy and forgiving, and a place with enough warmth and sunlight (at least in the North) is the most important part.
True. What I went for was extremely good yield, like 10-15kg per plant. That requires a lot of foilsge growth early on and proper pruning + a lot of water later on. Just good enough isn't that hard, though still the depth of yhe pot should be enough for root development.
A properly selected variety adapted to your climate needs both less soil, less water and and a lot less maintenance (still daily care but nothing crazy).
Yes, variety matters a lot. I've managed to grow tomatoes quite well in pots located outside in Northern European climate, simply by picking a traditional variety that is known to be easy to grow. (In Germany it's known as "Bonner Beste" if anyone is interested). It doesn't even require that much upkeep.
I saw significant difference using the Lite version, enough that I switched to Firefox with Origin instead. I expected it to be good enough and was surprised to see the difference.
Same. I use the element picker tool all the time to rid the web of crap I don't care about. Example:
whatever_class:has-text(/YouTube Shorts/)
On Android you can even do this on your phone in Firefox. The UI is a bit tricky on such a small device, but it's so worth it. I went so far as to uninstall Chrome (well, disable it) on my Android.
For whatever reason, the UBOL creator chose not to include zapper/picker in order to make it as "lite" as possible. It wasn't a Manifest v3 thing, as they've explained.
I have no problem with using a separate extension for zapping.
What do security minded people do about passwords? It seems like you either use the same password for everything, or you need some kind of password manager, but then I'm always worried about having all my passwords in one place meaning they all get compromised instead of just one.
It also feels like there's a convenience tradeoff with a lot of solutions. I could keep a physical binder full of passwords in my home office but that would be a pain to look up and enter things every time (and a big risk for anyone with physical access to my place).
Bitwarden for usability. Vaultwarden if you can and prefer to self host. Being on the internet you'll have to trust someone at some point. Can reduce risk by combining strong 2FA (not SMS/Email) alongside backing up your vault.
Ensure all your passwords get reset at some point after vaulting, long randomly generated from Bitwarden extension/app is easy enough. Ensure you enable strong 2FA at each service you have an account at too.
Passkeys tied to actual hardware, like the TPM-based solution in Windows Hello, whenever possible, Keepass where not.
Keepass DB cloud synced, but the passkey file I use in conjunction with a p/w to open it never leaves the machine(s) it's on. Also, key file needs Admin rights to read, so KP is run privileged, which also protects its process memory space from user-land snooping.
Even better than the TPM in Windows is a hardway FIDO2 or OTP key, I'd imagine. Those cannot be comprimised by a virus on your PC in the same way, assuming you don't leave the key in at all times and you only tap the button when explicitly logging into something that would require it.
I use a simple algorithm. So you don't actually remember the password, put the algorithm to produce the password for the site or service. Not perfect, but each passwords turns out to be unique (mostly). I don't know what experts think about that, but it has worked fine for me.
* If 1 to N password(s) leak the pattern may be obvious leading to your other accounts being compromised
* Not all sites have the same password “rules” so there is no algorithm that works for all passwords without you being aware of the rules of the given site. Rules that only you only (may) have access to at signup time.
* Typing passwords out manually sucks (slow and error prone)
1) only matters if you're a very high value target who is being manually target. Doesn't apply to 99.999% of people, who only need to worry about credential stuffing and brute force.
2) Similarly, it's not hard to come up with an algorithm that satisfies 99.9% of websites.
3) To a lot of people, managing a password manager sucks.
I personally do use a password manager and automatically generated passwords, but also understand that for many people it's the better option.
Yes! I'm totally aware, but, for the first point, attacks are generally automated. If someone tries to find the pattern, you are being personally targeted and you have bigger problems.
As per number 2, it is true and it sucks big time.
As per number 3, I don't really mind much. You don't generally have to use your password every time.
I agree that especially with modern LLMs, I would avoid following patterns like this.
Dedicated 2FA on a hardware device seems pretty resilient, I hope more banks incorporate it instead of SMS 2FA. Hosting vaultwarden also seems pretty good because it’s unlikely for you to be targeted, but requires selfhost maintenance.
It's absolutely not that- it's an early implementation of a drafted WHATWG standard in the "iteration" phase, put behind a development build and a development flag so people know that it isn't stable and might change and nobody who doesn't explicitly opt in will be affected. This is literally made to allow "people who try to set some standards" to test their proposed standards and iterate on them before finalizing the proposal.
As far as accessibility, the native browser select is almost always going to be more accessible than someone making a custom input using JavaScript so they can add some styling control. Having the native version support basic styling is a big accessibility win IMO, because it disincentives developers from making a less accessible alternative for the sake of matching some design file.
I'm lucky enough to live about 5 minutes from multiple different tool libraries. They've always had any tools I need in surprisingly good condition.
Makes a lot of sense for people like me that wouldn't need a hole saw more than once or twice a year at most to share one with the community instead of buying it myself.
If I was using something all the time though I would definitely buy my own. I have always been a little envious of friends with a great workshop of tools and supplies in their garage. I don't have the space for it or do enough tool related work to personally justify it for myself though.
I long for a space like this that I could spend some drop-in time at now (in my thirties and not looking to attend another full college degree program). If I could do one thing to reorganize society it would be to create a bunch of places like this with rich cultures of learning and collaboration that anyone could drop in on.
There was a decent technology meetup culture in the bay area pre-covid, but so many of the groups I enjoyed stopped around that time and never started back up again.
The end of a basketball game feels almost completely unwatchable for me when they devolve into constant fouling, stopping the game every couple of seconds.
There are possessions where the defense has the sole objective of ignoring everything other than fouling as fast as possible, which feels boring and can stretch 30 seconds of in game clock time to 5+ minutes of back to back stoppage. I get that it might be the mathematical best play because it forces the winning team to take free throws and then turn the ball over without taking any clock time, but they could architect the rules to avoid it.
I totally agree with the block call being a coin flip.... I'd extend it to almost every other call. NBA reffing seems like it absolutely sucks for such a large scale professional league. Basketball is a fast paced game so I know they can't catch everything, but when you're watching on TV you see so many things that are so inconsistently called. Those calls end up changing the outcome of the game when one team has 20 more ft chances than the other.
I tried a little Japanese (the second language I know best). Not sure where the content comes from but the translations were wrong, the vocabulary was strange, and otherwise the cards did not work out for me. Seems like they were auto generated via AI or google translate or something, in any case they were not accurate or natural sounding. Maybe other languages are better but I wouldn't know enough to tell.
Anki is one of my favorite pieces of software ever so I was definitely willing to try yours. Seems like the feature that defines your app is auto-generation of a card / set of cards. Didn't see any way to create your own card, all you can do is put in an English word and get the automatically generated card added to your deck. So if that's not working well the entire thing crumbles.
Interesting idea that might have a future if it finds significant improvement. I think if it was ironed out I would still prefer it to spit out an anki compatible deck that I can use in Anki itself instead of in the ricotta web app.
If you have iOS/macOS, I've made a native app for mining Japanese web/ebook content into Anki (or Manabi Flashcards which is getting FSRS soon too): https://reader.manabi.io
I'm currently working on adding first-class Mokuro integration for manga mode, bilingual user-provided or Whisper-generated captions for video, an HDMI-input mode with realtime lookups, and local/BYOK AI features. And once I add Yomichan dictionary support, I will be able to go multilingual too. I also have a beta available via the Discord that resolves some bugs that I'm trying to release very shortly.
I've added Anki integration and am working on WaniKani + JPDB sync. Hopefully more service integrations soon too. I like to have nice defaults but let users keep their flashcards wherever they like. Ricotta too if there were an API.
All fair points of feedback. I use it for latin languages so getting a take on the quality of other languages is something I'll work on. Could you give me an example that it gave you that missed the mark?
I could have it output the deck into Anki I like that idea. I wrote a script that did this from CSV but it didn't have audio so I moved to this making the first gen of this app - its open sourced here if you would like to repurpose it https://github.com/Amber-Williams/language-flashcard-app-gpt.... I could try and find my code I make to generate the Anki deck if anyones interested as well.
Custom cards is something I agree is a nice to have. If you have preferences please share I'll try to add them to future feature updates.
wow, this looks nice. I will download it this week to give it a go. At a first glance this looks like notion for learning languages.
I have been using Babbel, and while it's nice to have mini lesson to take each day... they offer no option to write out notes, or to look at the grammar and get some context, no extra examples and flashcard system is just awful.
Okay I had a closer look now. Tbh I don't download anything that isn't a major brand product or open-source. I'm guessing the amount of work you put into this you're planning to monetize the app right? If so my first point of feedback would be to support either a web app or mobile app.
I really enjoy the concept behind the app focusing on subjects rather than structured lessons. I enjoy going on learning rabbit holes so this would fit me better than structured lessons.
Lastly a video or demo of the app would help as the UI is quite complex. It's not clear how I would go from importing a web page to using the cards for example.
Feel free to ping me if you have updates - happy to help support another indie dev!
Just dropped in my email to try out the Japanese. I'll let you know.
I'm not really sure what to do with this -- I thought you wanted someone to take a look at some Japanese content, but I can't find any in the app/tutorial.
thanks! right, there is no content. you build your library along the way, the feedback request is about autogenerated flashcards (basically you create and import content and then it autogenerates flashcards for you)
The next year we moved from that house to a new place, where we couldn't plant directly in the ground but the landlord was happy for us to use pots and planters. I eagerly planted my broccoli again thinking we'd have the same endless supply... but this time it barely produced anything and looked nothing like the last year. I bought some kind of soil bags from the gardening store after asking an employee which would be good for vegetable pots and planters. Something about the pots or the soil or otherwise made a gigantic difference even though we had moved probably less than a mile distance wise.
I'm a very amateur gardener so I may have made some other mistakes, but I think I treated the plants very similarly both years.