4 ms·
tiddlywiki to me is scary concept compared to plaintext, it always frightens me that data becomes corrupted or not recoverable from from that html/js blob. But
by diimdeep 5y ago
tiddlywiki to me is scary concept compared to plaintext, it always frightens me that data becomes corrupted or not recoverable from from that html/js blob.
But it is also very cool concept, it could be much powerful than it is, think smalltalk like VM inside HTML file.
- fsiefken 5y agoyou can use the node based tiddlywiki and it saves the tiddlers in seperate text files. with the markdown plugin it saves to markdown. With obsidian you can use the markdown as well, or with kiln you can export to gemini or the small web
- karencarits 5y agoWell, in a sense it is plaintext as your data are kept in simple divs within the html file - just open it in notepad and all your data is there!
- bogwog 5y agoThat's why I use it with git. Not only do I get the ability to revert a catastrophic fuckup, I also get the ability to securely sync across all of my computers. The only thing missing to make it perfect would be a mobile client, but that's not huge deal for how I use it. The thing I don't like though is that, because it's all in a single giant html file, it is prone to getting huge. Mine is pretty big, and every time I open it (in Firefox or TiddlyDesktop) there's a noticeable delay. I'll have to look into breaking the monolith up into smaller files to avoid that.
- grumblepeet 5y agoThere is a mobile client that works excellently called Quine 2. Ive been using it for a couple of years with a mobile accessed Tiddlywiki personal wiki and it works pretty good. I keep the file on iCloud and can open it from desktop and mobile.
- bogwog 5y agoThanks for the suggestion, that app works great! I managed to set it up with my repo using an app called Working Copy, so now I have the full workflow going on all my devices