I get that, I mean in terms of publicly visible user info - understandably most users don't want all their machine names visible.

This thread has sparked some interesting thoughts ...

As to the original suggestion, my take is that the total number of "installations", whatever that means, is useful information. However, so too is number of "active" installations -- again, whatever that means -- and a result returned in the last two weeks seems as good a measure as any other. So, personally, it would be useful to see "x active installations of y total", or something similar. But this isn't a big deal for me.

What about the issue of removing historical data about device installations?

This is much more interesting to me, and my first thought was to consider the GDPR implications. There is a need to remove data which is no longer of "use", but I think this only applies to personally identifying informations (PII). It makes sense to keep historical data about the processing a person has done for the grid, but do we need to keep device names, and are they PII?

At first I thought that device names are not PII, and therefore it doesn't matter if that data is kept. But on investigation I think that they might be construed as PII -- they do uniquely identify a machine, which is someone's property, and which is trackable. In some instances they are the machine serial number and therefore it is quite likely that a sales record could exist which links the machine to an individual by name and address (and likely more PII).

I now think that WCG should remove records for machines which have not returned results within two years, and should accumulate them into a general "other historical installations" record for each user. If users feel that their machine ids could identify them, they should either change the machine names or request WCG to hide their user data altogether, but I do think it would be prudent for WCG to do something for the users who are not aware of this potential issue.

Hi Erika,

Thank-you for the information. That sounds like WCG agrees with me and is taking appropriate action. My comment that "it would be prudent for WCG to do something for the users who are not aware of this potential issue" and my suggestion that "WCG should remove records for machines which have not returned results within two years, and should accumulate them into a general 'other historical installations' record for each user" still stand, but that's probably a 'nice to have' and may not be strictly necessary. However, it might be considered appropriate on the basis of deleting unnecessary historical information -- that is, applying an appropriate data retention policy.

It's good to know that you're listening, and keep up the good work!

Hi Erika,

You nearly had me there ...



Your data retention policy does NOT say that data for a particular device is removed after a period of THAT DEVICE being inactive. It only refers to deleting ALL the data associated with an inactive USER, where the definition of an inactive user includes all of their devices being inactive.

I still think it would be prudent to delete all the data for individual devices once that device has been inactive for a defined period. But, like I said, you could argue that this is not necessary and I'm not going to push the point.

As always, thank-you for listening and for your feedback.