Tried out the feed finder on my blog again and I have another bug to report - it seems the URLs on the page can cause a crash within the web app! my blog (at matthew.science) uses Zola SSG, and it seems the URLs are formatted with a preceding //: '<a href="//matthew.science/posts/riscv/">Basics of the RISC-V ISA</a>'
This causes the following error: TypeError: URL constructor: //matthew.science/posts/riscv/ is not a valid URL.
Theoretically, I think that should work. It (At least it used to be) specified that //site/some/path should assume the uri protocol of the current context. So if it was a link on an http page, it should assume http, same with ftp and https etc. It should work sorta like how a leading slash assumes the current site context.
This was back before the Web became the one true way and is the reason it uses 2 slashes, to distinguish protocol local from site local.
Comments
Tried out the feed finder on my blog again and I have another bug to report - it seems the URLs on the page can cause a crash within the web app! my blog (at matthew.science) uses Zola SSG, and it seems the URLs are formatted with a preceding //: '<a href="//matthew.science/posts/riscv/">Basics of the RISC-V ISA</a>'
This causes the following error: TypeError: URL constructor: //matthew.science/posts/riscv/ is not a valid URL.
Theoretically, I think that should work. It (At least it used to be) specified that //site/some/path should assume the uri protocol of the current context. So if it was a link on an http page, it should assume http, same with ftp and https etc. It should work sorta like how a leading slash assumes the current site context.
This was back before the Web became the one true way and is the reason it uses 2 slashes, to distinguish protocol local from site local.
Thank you! Didn't see that case yet but I'm glad you commented, now I can fix it
Update: it's fixed