5 ms·

I can't seem to understand what the problem with "pile up" is? The whole idea of logging issues is to see how long they're open for! In fact good projects take

by ak39 8y ago

I can't seem to understand what the problem with "pile up" is? The whole idea of logging issues is to see how long they're open for! In fact good projects take the painstaking exercise to coalesce tickets that are in fact the same with an original item IN ORDER TO track the correct age of the ticket. This sounds totally backwards.

_pmf_ 8y ago
> This sounds totally backwards. It is. It's the result of low level managers using issue counts as KPI: now it's a metric that must be gamed.
- danabramov 8y ago
  We don't use GitHub issue counts as "KPI" at Facebook. The explanation is more prosaic: https://news.ycombinator.com/item?id=17316150 https://news.ycombinator.com/item?id=17316150
workinthehead 8y ago
Nope, wrong on all counts. You need to delete stale issues, it's no good having an endless backlog of someone's nice-to-haves that they no longer even care about. And, the "whole idea" of logging issues is not to see how long they are open for... it's to, you know, log issues, because they can't be fixed instantaneously. Measuring how long they are open for is just a pointy-haired boss side effect, because those are the kind of idiots that would think that kind of metric means anything.
- allthenews 8y ago
  >Measuring how long they are open for is just a pointy-haired boss side effect...kind of metric means anything. I don't think such cynicism is warranted. Open issue time is a useful metric, at the very least over time, of how quickly developers are fixing problems and how complex they are becoming. Certainly not something for a pointy haired boss to optimize exclusively for, but a useful bit of data to be taken into consideration with other measurements of performance.
  - bonesss 8y ago
    Of all the KPIs for development teams, how-f-ing-long-will-it-take is a pretty 'key' one, IMO. It's not the only factor, it's highly product dependent, but it is very relevant to most outward facing activity...
- valuearb 8y ago
  I agree that how long they are open is meaningless, but closing them before they are fixed throws the baby out with it's dirty bathwater.

by ak39 8y ago

_pmf_ 8y ago

> This sounds totally backwards. It is. It's the result of low level managers using issue counts as KPI: now it's a metric that must be gamed.

danabramov 8y ago
We don't use GitHub issue counts as "KPI" at Facebook. The explanation is more prosaic: https://news.ycombinator.com/item?id=17316150 https://news.ycombinator.com/item?id=17316150

workinthehead 8y ago

Nope, wrong on all counts. You need to delete stale issues, it's no good having an endless backlog of someone's nice-to-haves that they no longer even care about. And, the "whole idea" of logging issues is not to see how long they are open for... it's to, you know, log issues, because they can't be fixed instantaneously. Measuring how long they are open for is just a pointy-haired boss side effect, because those are the kind of idiots that would think that kind of metric means anything.

allthenews 8y ago
>Measuring how long they are open for is just a pointy-haired boss side effect...kind of metric means anything. I don't think such cynicism is warranted. Open issue time is a useful metric, at the very least over time, of how quickly developers are fixing problems and how complex they are becoming. Certainly not something for a pointy haired boss to optimize exclusively for, but a useful bit of data to be taken into consideration with other measurements of performance.
- bonesss 8y ago
  Of all the KPIs for development teams, how-f-ing-long-will-it-take is a pretty 'key' one, IMO. It's not the only factor, it's highly product dependent, but it is very relevant to most outward facing activity...
valuearb 8y ago
I agree that how long they are open is meaningless, but closing them before they are fixed throws the baby out with it's dirty bathwater.