Vadim Kantorov

Same problem. I tried to download GitHub zipballs with my deployment of this CORS proxy: `https://github.com/vadimkantorov/busyidetest/archive/refs/heads/master.zip`. This redirects to `https://codeload.github.com/vadimkantorov/busyidetest/zip/refs/heads/master`, and the browser runs this second request without CORS proxy....

`{redirect : 'manual'}` is useless apparently. At least proxy-side redirect would be useful...

If I go the latter path following Dropbox backend example, is there a way to access/extend [BaseFileSystem](https://github.com/jvilk/BrowserFS/blob/master/src/core/file_system.ts#L385) (I am consuming BrowserFS from my own TypeScript code) which seems not exported...

I've clone this repo, ran `npm install` and got an error: ``` src/generic/key_value_filesystem.ts(936,34): error TS2345: Argument of type '{ [name: string]: string; } | undefined' is not assignable to parameter...

`yarn install` has worked, then I changed directory to my project dir and ran `npm install ../browserfs` which worked as well. But then TypeScript fails to build my code with:...

Thanks for the suggestion. ES2017 helped for the SharedArrayBuffer! `npm install dropbox --save-dev` unfortunately didn't help for the Dropbox error (I've also removed old package from node_modules manually). Btw the...

I checked the source of [dropbox sdk](https://github.com/dropbox/dropbox-sdk-js/blob/master/src/index.d.ts#L3) (npm package dropbox), it contains `export = DropboxTypes.Dropbox` (if I understand well, underlying DropboxTypes isn't available for import at all). The backend, dropbox_bridge...

It also seems that there's overlap of VideoReader with torchaudio's StreamReader which was added in the latest release: https://pytorch.org/audio/0.12.0/io.html#streamreader, this StreamReader boasts even GPU-based video decoding. IMO it's quite needed...

If it's again a problem with Python 3.10 released and conda trying to use it, can we delete torchvision 0.2.2 from conda altogether? Better have an error than a crippled...

Currently it seems that torchvision is not compatible with OpenCV conda package: maybe because of libpng / libjpeg others