Re: my comment in #71, I think we can next year remove the explicit availability toggle and just derive from count.
I think, not regarding #69, the logic was pretty good this year. A bit of adjustment should be made, but then we can use it for the actual state.
At the moment we have to count and then decide the availability. We have to count anyway. Counting when there are lots of items is not much harder, and if we're above the 20% threshold for the "green" state then it doesn't really matter if we miscount by a few.
Having to decide for 200+ items whether a low count (say, 2-15 items) means yellow or green was hard. Not for those items that had >100 initial stock, but especially for those that had around 3-30 initial items. I think if we automate that, then it will be more consistent (no two people decide differently for an item that has e.g. 3/15 items left). It also saves crew head-space at the event 🚀
We might want to make the percentage and minimum count parameters adjustable in "Store Settings". And if we manually want to force a different state for a single item then we can still post a fake count :P
I think during operations, counts are rare, and counting small numbers accuratly ist hard as people swag sizes and give stuff back, esp. with separate stock for backorders. That's why we had the colors independent.
@felixrindt This was your comment on #61 where we first discussed this, and I think it is not really accurate, at least from my expericence this year. The numbers are good enough, if they are around the thresholds it doesn't really matter and if they are not then a bit of error doesn't matter.
Of course, a manual override would have been (and still would be) useful in order to account for retuned items and human error.
@Doralitze you said that. I think we can still fix human error if we notice it by just recounting. If the availability is erroneously red, we can enter the (very small) number of items we find. If it is erroneously yellow, we can enter a big number by estimation, or 0 if we don't have the item anymore.
Angels this year understood very quickly how important accuracy is based on number of items left. They all knew that high-throughput items are hard to count to the single item as one or two are sold while they are counting anyway, and that low-throughput and low-count items need to be tracked more precisely.
Re: my comment in #71, I think we can next year remove the explicit availability toggle and just derive from count.
I think, not regarding #69, the logic was pretty good this year. A bit of adjustment should be made, but then we can use it for the actual state.
At the moment we have to count and then decide the availability. We have to count anyway. Counting when there are lots of items is not much harder, and if we're above the 20% threshold for the "green" state then it doesn't really matter if we miscount by a few.
Having to decide for 200+ items whether a low count (say, 2-15 items) means yellow or green was hard. Not for those items that had >100 initial stock, but especially for those that had around 3-30 initial items. I think if we automate that, then it will be more consistent (no two people decide differently for an item that has e.g. 3/15 items left). It also saves crew head-space at the event 🚀
We might want to make the percentage and minimum count parameters adjustable in "Store Settings". And if we manually want to force a different state for a single item then we can still post a fake count :P
@felixrindt This was your comment on #61 where we first discussed this, and I think it is not really accurate, at least from my expericence this year. The numbers are good enough, if they are around the thresholds it doesn't really matter and if they are not then a bit of error doesn't matter.
@Doralitze you said that. I think we can still fix human error if we notice it by just recounting. If the availability is erroneously red, we can enter the (very small) number of items we find. If it is erroneously yellow, we can enter a big number by estimation, or 0 if we don't have the item anymore.
Angels this year understood very quickly how important accuracy is based on number of items left. They all knew that high-throughput items are hard to count to the single item as one or two are sold while they are counting anyway, and that low-throughput and low-count items need to be tracked more precisely.