I actually started writing this tutorial using Jessica Garson’s recent Glitch starter project, but after installing a few of the most common Python math/dataviz packages, I quickly run out of space
This past while, I was working on an in-browser tool to explore an archive file’s contents. At the very end of the workflow, I have a file’s contents taken from the archive, and there’s no file type information attached. So I’ve been using new Blob(...) with no type setting and letting the browser figure it out.
Well on Wednesday I found out that the browser that everyone else uses doesn’t bother guessing the file type in that situation.
Okay that’s not the end of the world, I just have to learn how that browser prefers to be instructed to guess the file type. It’s pretty standards-compliant, after all. I’ve even heard people say that standards are written in order to describe what that browser does
I found this standard MIME Sniffing Standard , which I gather from it that using unknown/unknown is more explicitly “please do guess” than not setting it at all (not sure tho).
And I went on to try setting that, and the browser that everyone else uses still doesn’t bother to guess. https://scrawny-palm-cornucopia.glitch.me/ (edit: you are warned that some test links will save a randomly named file into your downloads. thanks again, browser that everyone else uses )
So now I’m thinking about ripping out the whole previewing feature, because I have no confidence that Firefox will bother maintaining this behavioral difference from the browser that everyone else uses.
A little late to this thread since I was computerless this weekend!
I have been thinking a lot about community and in particular this question came up in my Twitter timeline yesterday and I haven’t been able to stop rolling it around in my head. If you have thoughts about the health of communities, I would love to hear them!
Random things on the internet I have been enjoying lately:
My friend Lynn runs this project called A Single Div, and she’s been sharing her Divtober artistry there
I’ve also been playing around with a website project that’s not quite ready to share, but I will make sure to share it here first when it’s "finally ready.”
This week, I’m working on putting together a Halloween movie pub quiz to cap off a mini film festival I host every year, and putting the finishing touches on my lightning talk for this year’s Roguelike Celebration. It’s been nice to have a couple projects that scratch some of the same itches as programming, but don’t have annoying error messages or syntaxes to get used to.
I’m also really tickled by Samperson’s DOOMpad project, I’m going to try and make some time to play around with it this weekend.
second update this week: as a result of looking too desperately at specs, I now have a commit in the w3c File API spec GitHub - w3c/FileAPI: File API
anyway, no revelations from reading the spec. for now I’ve marked all extracted files as application/octet-stream.
replies too:
if this is something you run manually and you don’t mind the wait, maybe you could install dependencies into /tmp and symlink them into your home dir?
what communities are you thinking about when you read this question? neighborhood? online? public/private?
also what does “transactional” mean here? the parenthesized bit sounds like “quantitative”
dang, great graphics regardless of whether it’s done in css! looks like a lot is done with effects that css allows multiple of, e.g. background gradients and shadows. clip-path also allows high complexity, but I see she is opposed to using that in this project
and here’s your open thread!
)
