Computers are so fucking weird.

[swelljoe@mas.to]

Computers are so fucking weird. Every day there's some new thing. Chromium on Linux (Fedora 43) is in some kind of sandbox, so I can't open local files, but only some of them? All of the directories that were created by the desktop (Documents, Downloads, Music, Videos, Pictures, etc.) are accessible, as well as a few hidden directories (.local, .nv, .pki, .var), but nothing else is. I can't even type it in explicitly and fully, it's a 404.

Firefox _can_ reach all files my user has access to.

[swelljoe@mas.to]

Also, I keep killing my editors (Zed and VS Code) trying to open a 32MB file with exactly one really long line of JSON. This doesn't seem like it ought to be "big" to any modern editor on a modern computer with 64GB of RAM, but here we are. jq can pretty print it without any trouble, so I could make a copy that's pre-parsed into individual lines, but it doesn't seem like I ought to have to massage text to even be able to load it.

Back to old reliable Neovim, with `:%!jq .` to format it.

[jumile@mas.to]

@swelljoe That file access problem sounds like the classic sandboxed package issue. Can't recall if it's Snap or AppImage that does it, though.*

Whichever one it is, there's an arcane command - or a GUI/TUI - that can change the access settings for individual packages.

*Sorry, it's been so long since I've used either package type that my memory is rusty. Both make me nope out. 😅

[swelljoe@mas.to]

@jumile yeah, I avoid Snaps and Flatpaks for this reason, among others (though Snap is worse than Flatpak, in my experience...maybe because the packagers are sloppier or maybe because Flatpak has better tools for making sure things are sane).

And, because I avoid Snaps and Flatpaks (and AppImages), Chromium is installed from an RPM. So, that is a very good guess, but not the case here.

[tedchoward@mastodon.social]

@swelljoe I keep a copy of BBedit on my Mac for the purpose of loading “large”text files.

Even vim will choke on a large file sometimes.

[swelljoe@mas.to]

@tedchoward in this case, it's not really that it's a big file (though it is big), it's that it's a (really) big single line.

Every editor is able to load it without too much trouble when it's been pretty printed and spread across a half million lines. Though only Neovim (of those I tested) can parse it and do syntax highlighting and such. VS Code disables parsing and pops a warning, but to its credit search and editing does work.

But, when on a single line loading it crashes VS Code.

[swelljoe@mas.to]

@tedchoward I historically very rarely ever need to look at giant files, but I guess I'm committing to doing this periodically, so I'll need to come up with a workflow that makes it easy...probably will have it fully automated at some point and not need to look at the actual data anymore, but for now, I need to check the shape of the data every now and then.

(If it's of any interest the files are from the Recreation.gov data set of all facilities in their database. https://ridb.recreation.gov/download )