Tor instance's captcha doesn't seem to work #30
Loading…
Reference in New Issue
No description provided.
Delete Branch "%!s(<nil>)"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
System: Fedora 40
Browser: Tor Browser Alpha (14.0a1)
Instance: http://4getwebfrq5zr4sxugk6htxvawqehxtdgjrbcn2oslllcol2vepa23yd.onion
Steps to reproduce:
The default instance works correctly on the browser.
I tried messing around with NoScript's settings (enabling all toggles, overriding Tor's security preset) but it was to no avail.
Well... I'm unsure of what to say here. Was able to solve the captcha, and I can do subsequent searches too.
Maybe try resetting your browser's settings? Do images load on other websites? The Captcha image is quite big in size (~150kb), maybe I can use a better image format than JPEG2000. What do you think about using WEBP if your HTTP headers ask for it?
Images load on the default instance, and other websites too (even ones on Tor).¹
I didn't mention on my first post, but I'm using Wayland.
Regarding your question, I'm not that much of a technical user (yes, despite using Tor), so unfortunately I can't contribute to this part of the discussion...
I tried reinstalling to reset my config but it didn't work.²
I tought of using the stable version but it doesn't run my machine for some reason.
I'll test the website on my phone.
Search works on my phone (even images display). So there is either an issue with Tor's alpha or my machine.
Can you hit CTRL+SHIFT+I for me, head over to the network tab, refresh the page, and show me what you see?
If you can also take a screenshot of the /captcha request, that would be good for me. Right now it just seems that it just takes a long time to load because of the network.
This is what the network tab looks like.¹
Sorry if I misinterpreted it, but this is what the network tab on "http://4getwebfrq5zr4sxugk6htxvawqehxtdgjrbcn2oslllcol2vepa23yd.onion/captcha" looks like.²
The glowies seems to be blocking the connection. Right click the captcha image for me and select "Reload image", see if that affects anything. This might be due to the captcha's filesize, I'll work on making it lighter in the next commit or 2.
Besides, can you click the top right button to "get a new identity"? This might be a problem related to your tor relay.
Reloaded the image five times, no results.
Even tried "Open Image in New Tab", got "Unable to connect".
Clicked on "Tor Circuit" -> "New Tor Circuit for this site" five times, no results.
I can't reproduce the issue. I don't know what to tell you. One last thing you can try, can you load up
/proxy?i=https%3A%2F%2Ftechcrunch.com%2Fwp-content%2Fuploads%2F2016%2F05%2Fp1040495.jpg%3Fresize%3D50
on the hidden service and tell me if anything loads?
Huh, by clicking on the "https" icon -> "Connection secure" -> "More information", the CAPTCHA loads after some time, on this media section (not on the actual page).
Seems to work fine, I got the image.
No problem, you are already helping me enough by going through this issue and maintaining the search engine.
Today I also got the captcha to load by changing the header from "https" "http" (again, it doesn't show up in the actual webpage, but on a separate tab).
This seems to be a bug with the tor browser. Requests should not be upgraded to
httpswhen using tor.For the last screenshot you sent, you are getting back an html page because I am telling the browser to use the cached version of the image in case the client attempts to reload the image. It's also a security feature where it prevents 4get from generating a different captcha with the same answer.
Indeed, I checked it's issue tracker and that does seem to be the case:
By toggling the configuration option listed above the captcha loads correctly. I think this issue can be closed now. Thank you for guiding me through this process.
Sick, glad you could fix the issue. Lmk if you find anything else.