3 ms·

> Justify the dev time to save micro-pennies worth of electricity to me instead. KEY (user_id) I mean, it's a dozen characters. Do you need to know how fast I

by pkulak 3y ago

> Justify the dev time to save micro-pennies worth of electricity to me instead. KEY (user_id) I mean, it's a dozen characters. Do you need to know how fast I type before you run the calculation?

shpx 3y ago
First tell me the minimum amount of time typing this would have to take for you to agree it's not worth it and I will try to keep adding things like the time it takes for someone to ask you to do this, for you to start VS Code, find the file, press ctrl-s, deploy the changes, and possibly some estimation of how long it takes a new developer to read and understand this system (please tell me how fast you speak and an agreeable value for how fast the average developer reads as well for this part) vs a simpler one until it goes over that limit.
MaulingMonkey 3y ago
> Do you need to know how fast I type before you run the calculation? I'll assume 100WPM, call that two words, billed at $200/hour and call that $0.06, falling under "too cheap to be worth arguing against", which falls under the aforementioned: >> If it'a a 5 second "this is probably the right choice" kneejerk reaction, maybe it's fine. That said, there's a decent chance those 6 cents won't pay for themselves if this is a login on a single user system, with any theoretical benefits of O(...) scaling being drowned out by extra compile times, extra code to load - and I'd be plenty willing to NAK code reviews that merely attempt to replace /etc/passwd and /etc/shadow with this, as the extra time code reviewing the replacement still has negative expected ROI, and it'll be a lot more involved than a mere dozen characters. Now, maybe we're attempting to centralize login management with Kerberos or something, perhaps with good reasons, perhaps which does something similar under the hood, and we could talk about that and possible positive ROI, despite some actual downtime and teething concerns?
- TylerE 3y ago
  Now document the two words, run the test suite to verify no-breakagem commit to source control, and push to production. Suddenly those two words cost a lot more than $0.06, and that's IF everything goes smoothly.
  - Aeolun 3y ago
    This falls under the “if we need to test that indexes still work on our mysql/postgres we have bigger problems” header.
    - TylerE 3y ago
      Depending on your IO pattern, adding an index can make writes quite a bit slower.
jefftk 3y ago
(author) If you're already using a relational database you should almost certainly set up indexes on your table ids and foreign keys. But that's pretty different from the examples in the post! I'm not anti-index, I'm anti-"if you ever have a full table scan in production you're doing it wrong".
- Aeolun 3y ago
  ‘If you ever have a non-deliberate full table scan in production you are doing it wrong’ then?
  - grog454 3y ago
    If you ever have non-deliberate anything in production there's a non zero chance you're doing it wrong.

by pkulak 3y ago

> Justify the dev time to save micro-pennies worth of electricity to me instead. KEY (user_id) I mean, it's a dozen characters. Do you need to know how fast I type before you run the calculation?

shpx 3y ago

First tell me the minimum amount of time typing this would have to take for you to agree it's not worth it and I will try to keep adding things like the time it takes for someone to ask you to do this, for you to start VS Code, find the file, press ctrl-s, deploy the changes, and possibly some estimation of how long it takes a new developer to read and understand this system (please tell me how fast you speak and an agreeable value for how fast the average developer reads as well for this part) vs a simpler one until it goes over that limit.

MaulingMonkey 3y ago

> Do you need to know how fast I type before you run the calculation? I'll assume 100WPM, call that two words, billed at $200/hour and call that $0.06, falling under "too cheap to be worth arguing against", which falls under the aforementioned: >> If it'a a 5 second "this is probably the right choice" kneejerk reaction, maybe it's fine. That said, there's a decent chance those 6 cents won't pay for themselves if this is a login on a single user system, with any theoretical benefits of O(...) scaling being drowned out by extra compile times, extra code to load - and I'd be plenty willing to NAK code reviews that merely attempt to replace /etc/passwd and /etc/shadow with this, as the extra time code reviewing the replacement still has negative expected ROI, and it'll be a lot more involved than a mere dozen characters. Now, maybe we're attempting to centralize login management with Kerberos or something, perhaps with good reasons, perhaps which does something similar under the hood, and we could talk about that and possible positive ROI, despite some actual downtime and teething concerns?

TylerE 3y ago
Now document the two words, run the test suite to verify no-breakagem commit to source control, and push to production. Suddenly those two words cost a lot more than $0.06, and that's IF everything goes smoothly.
- Aeolun 3y ago
  This falls under the “if we need to test that indexes still work on our mysql/postgres we have bigger problems” header.
  - TylerE 3y ago
    Depending on your IO pattern, adding an index can make writes quite a bit slower.

jefftk 3y ago

(author) If you're already using a relational database you should almost certainly set up indexes on your table ids and foreign keys. But that's pretty different from the examples in the post! I'm not anti-index, I'm anti-"if you ever have a full table scan in production you're doing it wrong".

Aeolun 3y ago
‘If you ever have a non-deliberate full table scan in production you are doing it wrong’ then?
- grog454 3y ago
  If you ever have non-deliberate anything in production there's a non zero chance you're doing it wrong.