Search location by barcode

[melconrad]

Could we have an option to locate samples by barcode to eliminate the need to search by uid and date?

[m-murphy]

So this was an explicit design decision, but I'm open to discussion. My rationale was if you have the barcode of a tube, that means you have the tube in hand since you scanned it. If you didn't have the tube in hand and were looking to search by barcode, that suggests to me that you probably have moved the tubes around without updating the db, in which case the search isn't going to do you much good.

@jbriggs7 also voiced some support for this feature, would you both mind explaining the exact use case this would fill? I'm a bit wary of it since it's a little too easy to move samples around and trick yourself in the process. Is the problem that the UID and date are too tricky to get right the way things are currently setup?

[melconrad]

The rationale is really just to be able to use the barcode as a uid. Currently, Jenny finds samples for a number of individuals in the lab and there seems to be some difficulty in having the dates in the correct format. . .especially when the format is dd/mm/yy. We figured an alternative would be to just have people merge the barcode into their databases and use that to search instead of the id/date combo. Searching by date will probably be easier once the bug that makes the date be one day off is fixed, so maybe it won't matter so much then.

…

On Mon, Sep 30, 2019 at 2:22 PM Maxwell Murphy ***@***.***> wrote: So this was an explicit design decision, but I'm open to discussion. My rationale was if you have the barcode of a tube, that means you have the tube in hand since you scanned it. If you didn't have the tube in hand and were looking to search by barcode, that suggests to me that you probably have moved the tubes around without updating the db, in which case the search isn't going to do you much good. @jbriggs7 <https://github.com/jbriggs7> also voiced some support for this feature, would you both mind explaining the exact use case this would fill? I'm a bit wary of it since it's a little too easy to move samples around and trick yourself in the process. Is the problem that the UID and date are too tricky to get right the way things are currently setup? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#11>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACBP4LH64Z2A55K3B7U24L3QMJU3DANCNFSM4I36SSXA> .

[bgreenhouse]

Seems like the top priority then is to fix the date issue - then we can see if necessary to do anything else. Max is this something you can delegate to Noam?