Timeline for I accidentally came across one of my co-workers salaries
Current License: CC BY-SA 3.0
7 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 1, 2016 at 15:16 | comment | added | PeteCon | @Stephan: In an agile setting, the management/users will be able to see changes soon enough - especially if continuous deployment is in place. I still maintain that developers don't need real data in order to test programs - and indeed, real data often gets in the way of good testing. Great discussion, but I'm out, as we're digressing from the original question. Thank! | |
| Aug 1, 2016 at 1:27 | vote | accept | Luke | ||
| Jul 31, 2016 at 11:52 | comment | added | Dan Is Fiddling By Firelight | @Pedro while having access to real data isn't inherently wrong; if needed proper confidentiality controls (eg the same NDA that anyone else with access to salary data has) should have been put in place beforehand. However in this case I agree with Pete that there's no reason why it couldn't've been anonymized first. And from the description of the question it appears that there wasn't an intent do so do; but in this case it was inadvertently leaked due to the person who provided the data not realizing one bucket only held a single persons data. | |
| Jul 31, 2016 at 5:15 | comment | added | Stephan Branczyk | @Pete, In an agile setting, you'd want the ongoing prototype to be useful to management/users as soon as possible, and that usually means using real data and real names. Now, I am not saying that one shouldn't obfuscate such things, but ultimately it's a tradeoff decision. | |
| Jul 31, 2016 at 4:14 | comment | added | PeteCon | As a security analyst, I'd disagree here; it's pretty simple to change names to 'Fred Bloggs1' and add a random number to any figure that may be used. During UAT it's acceptable for the clients to work with real data, but development is (or should be) a no-no. | |
| Jul 31, 2016 at 3:38 | comment | added | Pedro | +1 for distinguishing between the two issues. But depending on their roles and the project plan, it might be perfectly acceptable for developers to have real data. OP mentions extract functionality; loading real data, even for tests, makes it much easier for the business users to validate. | |
| Jul 31, 2016 at 3:02 | history | answered | PeteCon | CC BY-SA 3.0 |