>Offpunk answers with this error message :
>
>ERROR while caching
>gemini://gemini.circumlunar.space/users/hundredrabbits/
>*****
>
><class 'OSError'> = [Errno 97] Address family not supported by protocol
>


>My instinct tells me it may have to do with IPv6.

Hello,

You may be right about IPv6, as I have blocked IPv6 on the OS level. Can this prevent offpunk from working ?
On the same laptop I can access the gemini capsule with Lagrange browser.


>Could you also try to do that while online to get the full stack trace?

ERROR4: <class 'OSError'> : [Errno 97] Address family not supported by protocol
Traceback (most recent call last):
   File "/home/moussa/.local/bin/offpunk/netcache.py", line 805, in fetch
     path,newurl=_fetch_gemini(url,**kwargs)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
   File "/home/moussa/.local/bin/offpunk/netcache.py", line 641, in _fetch_gemini
     s = socket.socket(address[0], address[1])
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
   File "/usr/lib/python3.11/socket.py", line 232, in __init__
     _socket.socket.__init__(self, family, type, proto, fileno)
OSError: [Errno 97] Address family not supported by protocol

                 hundredrabbits (1 links)   (last accessed on Mon Dec 11 23:32:29 2023)
                 2023-12-06 22:44:23.272796
                 ERROR while caching
                 gemini://gemini.circumlunar.space/users/hundredrabbits/
                 
                 *****
                 
                 <class 'OSError'> = [Errno 97] Address family not supported by protocol
                 *****
                 
                 If you believe this error was temporary, type reload.
                 The ressource will be tentatively fetched during next sync.

On 23/12/11 11:33, Y G wrote:
>>My instinct tells me it may have to do with IPv6.
>
>Hello,
>
>You may be right about IPv6, as I have blocked IPv6 on the OS level. Can this prevent offpunk from working ?
>On the same laptop I can access the gemini capsule with Lagrange browser.
>
>
>>Could you also try to do that while online to get the full stack trace?
>
>ERROR4: <class 'OSError'> : [Errno 97] Address family not supported by protocol
>Traceback (most recent call last):
>   File "/home/moussa/.local/bin/offpunk/netcache.py", line 805, in fetch
>     path,newurl=_fetch_gemini(url,**kwargs)
>                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
>   File "/home/moussa/.local/bin/offpunk/netcache.py", line 641, in _fetch_gemini
>     s = socket.socket(address[0], address[1])
>         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


If IPv6 is disabled, the code is expected to crash when connecting to 
the socket. But, as your trace shows, the crash happens at socket 
creation, not connection. And socket creation was outside of the 
try/catch.

I’ve pushed a tentative fix. If IPv6 connection fails, it should fall 
back to IPv4 now.

Could you test the latest trunk?

> I’ve pushed a tentative fix. If IPv6 connection fails, it should fall
> back to IPv4 now.
>
> Could you test the latest trunk?


Tested, checked and approved ! At least on the same 
gemini.circumlunar.space capsule.

Thanks for your speedy patch.

On 23/12/12 12:06, Y G wrote:
>> I’ve pushed a tentative fix. If IPv6 connection fails, it should fall
>> back to IPv4 now.
>>
>> Could you test the latest trunk?
>
>
>Tested, checked and approved ! At least on the same
>gemini.circumlunar.space capsule.
>
>Thanks for your speedy patch.

Great! Thanks for testing.

I must admit that, retrospectively, it sounds really impressive. 

I’ve always admired great hackers and wondered how they managed to do 
this. Then I realize that what was happening was this:


User: Hey, I got this cryptic error
Dev : Mmmmm… sounds like IPv6.
User: How did you know that I had IPv6 disabled? Here’s the full trace.
Dev: Here’s the patch.

That really sounds amazing. Exactly how I always saw great hackers, 
myself being on the User side.

Now that I’m on the Dev side, I realize the patch is only about 
indenting three lines so they enter the catching loop. It is a lot less 
impressive than it sounds. ;-)


I will nevertheless keep this as a proud and rare moment where I managed 
to look like someone who understands what he is doing.

Le 12/12/2023 à 00:18, Ploum a écrit :
> Great! Thanks for testing.
>
> I must admit that, retrospectively, it sounds really impressive.

I must admit that I found your answer impressive too

> That really sounds amazing. Exactly how I always saw great hackers,
> myself being on the User side.
>
> Now that I’m on the Dev side, I realize the patch is only about
> indenting three lines so they enter the catching loop. It is a lot less
> impressive than it sounds. ;-)

As expected when you have developped your own knowledge and practice, 
especially on a tool you made yourself ! but nobody but you could say 
that "it is less impressive than it sounds".

The lightness of Offpunk probably helps to find quickly what should be 
patched.

> I will nevertheless keep this as a proud and rare moment where I managed
> to look like someone who understands what he is doing.

You did :)

You are a hacker, since you developped a software called «Offpunk». Even 
as a user of such a software I am under the impression I am a hacker.

I wish you will benefit of some more «proud and rare moments» !