With the recient news saying that support of mems’ creation and browsing will soon be discontinued, and the only option of saving mems provided by MemRise itself is one-by-one clicking and downloading images while copypasting text into a separate text document, I wrote a script to mass-download mems from the site.
The script itself is available here, however, it is written in Wolfram language, which is not something widely used, so I don’t expect many people to be able to run it by themselves.
There is an option for me to run the script on my end, downloading the required mems and uploading them afterwards for everyone to use.
If you need help with saving mems, all you need to do is post a request in this thread indicating the name of your MemRise account (which may differ from the forum name) or the name of any user you want to download mems from. (if you need to download mems from several users, it would be nice of you to make a list with each of them put in quotes; example: “user1”, “user2”, “user3”).
Here GoogleDrive/MemDump you can find all the mems I backed up so far by personal requests of other people as well as the mems I wanted to archive for myself (most of which were made by users from Remembering the Kanji course). The names indicate authors of the mems with the total number of their mems put into brackets. So far there are about 20,000 mems in total. Archives that contain pictures are stored in separate folders, while the ones that have only text mems are saved as spreadsheet files. I plan to keep expanding this database as long as I have the storage space.
Please note, that downloads may take some time, and there might be some problems with certain types of mems, as I still continue to encounter new irregularities in the ways MemRise stores the data. But I will try to do my best to process as many requests as I can, and the script already seems to work for the majority of cases as it is.
Here you go: johnastsang (956)
There was a problem with mems, where MemRise stored the item they attach to as a picture instead of a special character for some reason (there are several of those on your page 45, for example), I should probably look into those later. The rest seem to be successfully stored.
Please tell me, if you find anything wrong with them.
Thanks a lot! This is brilliant. It’s okay that a few mems with special characters cannot be saved.
The special characters from page 45 are Shavian letters . They appear as pictures in the course (thus cannot be saved) but they are supported by unicode (e.g. 𐑐𐑑𐑒𐑓𐑔𐑕). They are used to spell English phonetically. Interesting to look at:
I’m glad I could help.
Yes, those as well as several mems for Khmer letters on pages 42-43 (in case you’ll want to save those manually), which have audio as a front of the cards.
Thanks so much, you’re the MVP!!! Could I ask you to save jackfrutz’s mems?
Also, until when will you be accepting requests? I have ~300 kanji left to finish so I’ll have more mems by the end of the month
I was just reviewing my RTK words today, and discovered a lot more people I used mems from but didn’t back up yet. Jackfrutz was literally the first one in the list So yes, I plan to add his mems to the disk right away.
I don’t plan to go anywhere, so as long as memrise allows people to access mems, I’ll be able to back those up. Just let me know (in this thread or via PM), when you would like me to store yours.
On a side note, it makes me happy yet sad that it’s easier to browse your mems in a spreadsheet than it is in your own mem profile page
Done. I finished adding jackfrutz’s mems to the folder, and also included a lot of other authors.
I’m glad to hear that. My goal right now is to just preserve as completely as possible the raw data, and I haven’t come with any ideas yet about how it can be utilized later. So it is good, that even in this plain form it is not completely unusable.
One idea that comes to mind is trying to integrate the mems into kanji koohi. I use that site in parallel with memrise anyways. There’s no support for pictures but maybe one day there will be or the creator can be talked into it haha
That’s just for the kanji mems though… It would be handy if a program could be written in a reusable way so anki or other sites could easily use it to import the data too.
can you save mine please its took me a lot of time to make it and i dont want to lost it
Please help archive the mems of @cos. He’s created some fantastic courses, with mems preseeded.
I can’t emphasize enough how enthusiastic he was in creating courses:
I’m not the only person who spent a huge amount of time creating courses on memrise back when they were actively nurturing the community and community-created courses. I literally spent thousands of hours on my courses, over several years. For the fish idenfitication course, I took thousands of photographs myself, identified them and cropped and edited and labelled them, sorted them, etc. For the Hebrew course, I actually paid someone else with my own money to record voice for many of the items, so I could have two different people’s voices for each word or phrase, to help people get a better sense of how to say or hear things.
This example should be able to show why mems are a vital part of his courses. This is also for your consideration, @MemriseSupport.
Destroying mems that people casually jot down is bad, but removing the notes that course-creators write for others’ benefits is terrible. The fish course, as an example, is only 50% usable without the mems. Consider for a minute the time and effort @cos invested to take the thousands of photographs, writing the mems, and putting them into a course. I hope you feel something, @MemriseSupport.
Sadly I’m not into Scripts yet they have been mentioned for many years for lots of uses.
Please could you save mine (although there aren’t many).
What concerns me is number I use (created by other people) that won’t be saved probably.
Memrise - DW7
This may be a repeat of the most popular.
Memrise - DW7
I’ve added your mems to the common folder.
It seems though, that in some cases there are issues caused by the changes in direction of writing (or maybe it’s just me not recognizing symbols in different fonts). It’s hard for me to detect these issues (I’m sorry, I don’t know any Arabic), but if you have any notes on which items appear to be improperly saved, please let me know and I’ll try to fix it.
Both cases have a lot of mems, attached to the items with pictures. Sadly, the current version of the script still cannot process those. I’ll be looking into this matter in the near future and post an update here when I have any progress.
Thanks for looking into this for me @Eltaurus
I’ve added your mems to the folder.
They seem to have been downloaded in their entirety and without any errors (at this point I’m surprised that this is also possible ), but if you notice any problems, please let me know.
Managed to download your mems as well. Tell me, if you find anything wrong with them.
(note, that all pictures from the front and back of the cards are now placed in the separate “additional media” directory)
All mems are now stored in the folder.
Also, while working on these I had to rewrite parts of the script processing items with images and sounds. Because it works much better now, I went ahead and updated your folder as well, so there are no more missing cards with Shavian characters or Khmer sounds. You can check it out.