It’s basically that in conjunction with Deft, a… mode? for searching across plain text files in a directory. You’d presumably need to add the backlink syntax which I believe is something like ] as well as any possible metadata like tags but just a #+ TITLE should be enough. I haven’t used it heavily myself but I have got it set up and it appears the focus is on backlinking rather than folder heirarchy. It probably takes less time to write one from scratch than to figure out how to use my 200 lines of poor-taste, uncomented Python. If it’s anything else, it just lets you download it.md file, it builds the sidebar links (as above) and renders that + whatever it gets by running the file through markdown html file, it just builds the sidebar links (as above) and renders that + the HTML And finally it concatenates this with the blob of HTML above and drops it out.It then scans the directory for files and folders with names other than index.md and index.html, and builds the sidebar.If it finds a file called index.html, it just takes that blob of HTML.If it finds a file called index.md, it runs it through markdown and gets out a blob of HTML.If it can’t find a file called index.md or index.html, it generates a “clean” directory listing.Translates the URL into a local path, makes sure said path is not /etc/passwd or something similarly harmful, and then:.That being said, it’s a completely trivial, 200-line CGI script that basically: It’s a little quirky and has a bunch of hardcoded things that probably make it useless for others, but I’ll see if I can find some time to clean it up. However, I do go through them periodically and sometimes throw away some of the stuff that I definitely suspect I won’t care about, not even for nostalgia. These are mostly on non-technical subjects, but there’s a bunch of tech stuff in there, too. A lot of this is obtained by distilling notes that I take on paper, but a lot of that doesn’t lend itself easily to wikifying, so I have a bunch of old-fashioned folders around. Editing that in a browser is hell.Įdit: ah, worth pointing out. Some of the materials I organize are really long – e.g.Without a “real” file structure behind it, this quickly becomes unmanageable. I have a lot of mixed media to organize – it’s not just text, it’s also thousands of PDFs (datasheet, reference manuals etc.), schematics, images, ebooks, whatever.I have 20 years of notes and reference material there, the only way I can find something is by grepping an otherwise well-organized directory structure. I want to be able to easily grep through text files, not use some in-browser search function.It’s an uncommon setup but I ended up with it because: I’ve been using the last thing for about 3 years now and I don’t see myself switching any time soon. Got tired of it and spent another afternoon writing some bare Python CGI that implements the subset of werc’s functionality that I care about. Spent an afternoon trying to fix werc but my rc foo wasn’t exactly top notch after two years of not using it. I wanted better Markdown support, and “real” directory listings, including files with various quirky names which for some reason broke werc. This one’s actually great but I eventually stoped using plan9port and various things started to be hard to fix Cool but I was lazy and build times ended up longer and longer because I just didn’t want to deal with it. Brief experiment with a homebrew static site generator.I’m just not smart enough for such a smart tool). (I admit that, to this day, I have no idea how people manage to use Docker productively. Throwing containers at it only made it worse – I had to debug both whatever broke inside the container and Docker. I revisited it a few years later when containers supposedly helped solve that. Not self-contained, I had to care after a pretty complex stack. There were a bunch of WYSIWYG editor add-ons – like all in-browser WYSIWYG editors, all of them were really terrible. Spent quite some time migrating data from gitit to it. As soon as one of them isn’t packaged by whatever distribution I’m using, it’s bad. I don’t use Haskell, and don’t really care about it. Unfortunately, it drags a ton of Haskell-related dependencies. I loved it because it’s self-contained and Git-managed – I can write everything from a real text editor, not some in-browser thing. DidiWiki, I think? I liked it because it was small and self-contained.Worked great until I had to add schematics & co. It looks like it’s a lot of it because I tried all this over a span of 15-20 years or so.
0 Comments
Leave a Reply. |