Post AzO9SZDcC7UsWRuJcG by wndxlori@ruby.social
(DIR) More posts by wndxlori@ruby.social
(DIR) Post #AzMVt6OFIObPbf1Abw by foone@digipres.club
2025-10-19T09:01:22Z
0 likes, 0 repeats
"You may want to test changes made to the DText gem on a local instance. You are recommended to reconsider and rethink your life choices."I appreciate a README.md that knows to try to warn you away
(DIR) Post #AzMXNaLKazJV95ko9A by ppxl@social.tchncs.de
2025-10-19T09:18:03Z
0 likes, 0 repeats
@foone not a single day goes by where don't think if it wasn't more a peaceful life if I instead of IT turned to a gardening job
(DIR) Post #AzMb1BO765MzC7d1rk by foone@digipres.club
2025-10-19T09:58:48Z
0 likes, 0 repeats
THEY IMPLEMENTED THEIR COMMENT MARKUP LANGUAGE PARSER AS GENERATED C++ CODE?
(DIR) Post #AzMb5CJ5ceIpCeC26a by foone@digipres.club
2025-10-19T09:59:18Z
0 likes, 0 repeats
I'm sure there's a reason that makes sense. I will trust.
(DIR) Post #AzMbkpA83zihaGqoyG by foone@digipres.club
2025-10-19T10:07:03Z
0 likes, 0 repeats
@dysfun they warned me about testing my changes, not just by taking a look. and I had to look
(DIR) Post #AzMbol5z7rNy84mxO4 by foone@digipres.club
2025-10-19T10:07:29Z
0 likes, 0 repeats
@SRAZKVT I am relentlessly a glass-half-full kind of gal!
(DIR) Post #AzMe1dHPhAlBYvOuLw by foone@digipres.club
2025-10-19T10:32:34Z
0 likes, 0 repeats
found a fun bit in the database:so users who have made their account in the last 2 weeks aren't allowed to upload new posts, so when the database is initialized, it sets the account-created date for the admin user to two weeks ago.
(DIR) Post #AzMe4tj30UtDR7spTE by foone@digipres.club
2025-10-19T10:33:11Z
0 likes, 0 repeats
so this freshly created site has a user who says they registered on October 5th
(DIR) Post #AzMgVaExlDGujOfEDA by waldi@chaos.social
2025-10-19T11:00:16Z
0 likes, 0 repeats
@foone Learning from Microsoft is good learning. Microsoft Entra provides JWT that are always issued (the iat claim) five minutes before the request.
(DIR) Post #AzNpDvkI9vMoxgw1rs by foone@digipres.club
2025-10-20T00:12:21Z
0 likes, 0 repeats
@annarcana yeah I think that's the answer. god damn it rq
(DIR) Post #AzO9SZDcC7UsWRuJcG by wndxlori@ruby.social
2025-10-20T03:59:26Z
0 likes, 0 repeats
@foone I was going to say I worked with stuff like that over 20 (25)? years ago. But this is like… recent? 🫣
(DIR) Post #AzOB6zateNB9rVdZVQ by foone@digipres.club
2025-10-20T04:17:55Z
0 likes, 0 repeats
@wndxlori apparently ruby is eternal
(DIR) Post #AzOBCsccUAQxyYFbMm by foone@digipres.club
2025-10-20T04:18:40Z
0 likes, 0 repeats
@annarcana ahh, makes sense. I'm more used to python where our interpreter is just as slow, but the extension API is such a pain that we just deal with it :)
(DIR) Post #AzOEkLetqeIbRyy3Xs by foone@digipres.club
2025-10-20T04:58:47Z
0 likes, 0 repeats
@annarcana yeah exactly! No one writes cython or swig and walks away unchanged. it's telling something when the nicest api between C and python is the C fixed function interfaces. You know how badly you have to fuck up to make "calling a C ABI" the EASIEST solution?
(DIR) Post #AzOEpnylyr1sjUrhRo by foone@digipres.club
2025-10-20T04:59:46Z
0 likes, 0 repeats
@annarcana my latest C/Python project took one of the easy ways out: The python code talked to shared memory in a separate process implemented in C, for performance reasons.