The move to TLS-secured websites is great for privacy and a good step forward for the Internet. That being said, as someone who collects and uses retro computing technology, this poses a problem; TLS-secured website will not work in older browsers. Further, heavy use of Javascript, larger image sizes, etc. make loading even none TLS-secured websites problematic. —When you’ve got megahertz and megabytes to work with rather than giga or tera, large payloads take forever!—
Over the last few days I’ve been ruminating a solution to this problem. Here’s what I’m thinking. I’d like some sort of proxy that will:
- Take an HTTP or HTTPS request from an older computer,
- Make the modern request on the older computer’s behalf
- Accept the response,
- Compress/convert all images to smaller sizes and older formats,
- Strip all Javascript,
- Minify HTML & CSS,
- & return a IE3/Netscape Navigator 3 valid HTTPS payload with a self-signed certificate the older browser is capable of and will always trust.
Ideally this proxy could be served from a RaspberryPi or a similarly low-powered device. For an application such as using in a Newton or other portable device, it might even be possible to run this from a RaspberryPi Zero inside the device.
I’ve researched forward and reverse proxies and man in the middle type proxies, but honestly I’m not a networking guy and I don’t completely understand what they do. My feeling is that what I want to do can be accomplished with one of these type proxies, but I’m not entirely sure.
So, if you have guidance on if this can be accomplished with existing proxies, how it might be accomplished with existing proxies, or a good starting place for if I end up needing to write this from scratch, please contact me via Twitter. I think a proxy such as the one I have proposed would be extremely useful for the retro computing community at large and I would love to be a part of making it happen.
