Add something to the list, then reload the page - the data will still be there! I've tested it a lot and for me it is more consistent than the original. If it doesn't work for you, please also test the original project. The data stored by this system is synced across different projects (using the same storage uuid - see init block) No cloud variables used! The data is saved locally* and can be accessed from all devices in your network. It will be auto-cleared after a while (can range from a few minutes to days / weeks) (Scratch team, if you're not okay with me using the translate block like this, I won't do it again) I've added error correction (Hamming Codes: scratch.mit.edu/projects/1060987866 ). This means that every character consists of 11 bits! -- PRACTICAL USES -- Before using this storage system in your project, please read the warnings below! ⚡ Detecting if the Scratcher who uses your project has recently opened it before ⚡ Auto progress-saving features in games (so in case the user accidentally closes the tab, the progress won't be lost) ⚡ Sharing data (like highscores or user input) locally* across different Scratch projects ⚡ Storing data locally for a short amount of time ⚡ The 'allowed chars' variable has a size limit of 2048. If you want to add more capital letters, you also have to update the costume name. Before using this storage system in your project, please read the warnings below! Keep in mind that this is NOT A PERMANENT storage and doesn't work consistently for all users. -- WARNINGS -- TL;DR: I do not recommend actually using this because it's so inconsistent. It's just an experimental beta. ⚠️ The system is pretty inconsistent. If you switch tabs during encoding / decoding process, all saved data might be erased. ⚠️ This is not a permanent storage, it will be reset every now and then (can range from a few minutes to days / months depending on the device) ⚠️ For all of the use cases listed above, a cloud backend (with a Python library like scratchattach) is the faster and more stable option - however it requires knowledge in text-based coding languages. ⚠️ This system can NOT be used for cloud games / cloud highscores because the data isn't stored online - it's stored on your device and only visible to you. If you're looking for a cloud engine, check this: https://scratch.mit.edu/projects/440516876/ ⚠️ Encoding data is almost instant, decoding is slow. The more often the saved data is encoded / decoded, the longer 'get slot' will take. ⚠️ Only works on Scratch. Doesn't work on TurboWarp at all because the turbowarp translation api gives a 429 Error - To Many Requests. ⚠️ During the encoding process, around 30 clones are created, so beware of the clone limit (300). -- HOW DOES THIS WORK? -- TL;DR: It's based on Scratch's translate extension and the fact translations requested before are cached server-side for the network that requested them. Full explanation coming soon lol -- CREDITS -- Credit to by and by and by for inspiration and for inspiration and for some code (original project no longer shared) They used a quite similar concept, however my implementation of it is significantly faster and also more reliable. These people wrote that caching is happening locally - this doesn't seem to be the case, at least for me. The data stored by this system is synced across all devices that are in my network no matter what browser I use, therefore I believe the caching is happening server-side (I believe the Scratch API caches the response it sent to a specific network and will provide the same response again if the same network sends the same request again). -- APPENDIX -- *The caching of Scratch's translate API that allows this project to work does not actually seem to happen locally (read the paragraph above for more information).
