4 ms·

The article you linked I agree with almost fully, nice work. The only part I have issue with is when the teacher says “we fix things by "switching them off and

The article you linked I agree with almost fully, nice work. The only part I have issue with is when the teacher says “we fix things by "switching them off and on again"” as if that’s a bad thing. I see and agree with your actual point, but a power cycle is generally step one for almost all debugging.

nonrandomstring 4y ago
> the teacher says “we fix things by "switching them off and on again"” as if that’s a bad thing TBH, as an electronics engineer and computer scientist it's the first thing I do too :) I think my point in that article was that we shouldn't teach that it's the _only_ recourse, and after if it still doesn't work then just replace it. It's still one step better than the wisdom of my Grandmother's generation: "If it doesn't work bang it once on the top. If it still doesn't work bang it hard on the side twice. If it still doesn't work, call a man."
- starkd 4y ago
  The banging fix was also useful for getting better reception on the tv set. Surprising how often it actually worked!
  - Damogran6 4y ago
    Chip creep was the behavior for chips to walk out of their packages due to temperature cycling, percussive maintenance helped reseat things, sometimes enough to make them work again. (The Apple III was a beneficiary https://www.techjunkie.com/apple-iii-drop/ https://www.techjunkie.com/apple-iii-drop/ )
    - QuercusMax 4y ago
      My dad always talked about lifting the front edge of a component up 2 inches and then letting it drop - he called it the "2-inch test". He said that either it would reseat whatever was flaky, or make it super obvious that it needed to be reseated.
corobo 4y ago
> but a power cycle is generally step one for almost all debugging If you're not the sysadmin and the debugging relates to a production server, stop power cycling it and causing it to fsck on the way up while I'm trying to actually fix it tho lmao. Bonus points: you're frying all the information I need to actually see what is going wrong, now I'm digging in logs and monitoring graphs hoping to catch a clue Definitely not something that happened repeatedly this morning, haha

nonrandomstring 4y ago

> the teacher says “we fix things by "switching them off and on again"” as if that’s a bad thing TBH, as an electronics engineer and computer scientist it's the first thing I do too :) I think my point in that article was that we shouldn't teach that it's the _only_ recourse, and after if it still doesn't work then just replace it. It's still one step better than the wisdom of my Grandmother's generation: "If it doesn't work bang it once on the top. If it still doesn't work bang it hard on the side twice. If it still doesn't work, call a man."

starkd 4y ago
The banging fix was also useful for getting better reception on the tv set. Surprising how often it actually worked!
- Damogran6 4y ago
  Chip creep was the behavior for chips to walk out of their packages due to temperature cycling, percussive maintenance helped reseat things, sometimes enough to make them work again. (The Apple III was a beneficiary https://www.techjunkie.com/apple-iii-drop/ https://www.techjunkie.com/apple-iii-drop/ )
  - QuercusMax 4y ago
    My dad always talked about lifting the front edge of a component up 2 inches and then letting it drop - he called it the "2-inch test". He said that either it would reseat whatever was flaky, or make it super obvious that it needed to be reseated.

corobo 4y ago

> but a power cycle is generally step one for almost all debugging If you're not the sysadmin and the debugging relates to a production server, stop power cycling it and causing it to fsck on the way up while I'm trying to actually fix it tho lmao. Bonus points: you're frying all the information I need to actually see what is going wrong, now I'm digging in logs and monitoring graphs hoping to catch a clue Definitely not something that happened repeatedly this morning, haha