Change Details

**Steps to replicate the issue** (include links if applicable): * go to https://stats.wikimedia.org/#/en.wikisource.org/reading/unique-devices/normal|line|2-year|~total|monthly or the same page for any other project **What happens?**: * it says there were 0 unique devices for this month, which is plain false. **What should have happened instead?**: * If the data exists, it should show the data. If it doesn't exist yet (it's the 31th), don't show a 0. **Other information** (browser name/version, screenshots, etc.): {F60980419} Also noting, that the "values under 1000 are not reported to preserve accuracy" message that appears is inaccurate, because, well, 0 < 1000. I suppose perhaps there is code in place that //should// prevent the 0 from being shown. This may be a RESTbase metrics problem.

**Steps to replicate the issue** (include links if applicable): * go to https://stats.wikimedia.org/#/en.wikisource.org/reading/unique-devices/normal|line|2-year|~total|monthly or the same page for any other project **What happens?**: * We see a sharp spike for the month of May 2025 and is continuing. **What should have happened instead?**: * The chart should reflect an accurate unique device count with slight increase or decrease. The fix for this issue is happening in T395934 **Other information** (browser name/version, screenshots, etc.): {F60980419} Also noting, that the "values under 1000 are not reported to preserve accuracy" message that appears is inaccurate, because, well, 0 < 1000. I suppose perhaps there is code in place that //should// prevent the 0 from being shown. This may be a RESTbase metrics problem.

**Steps to replicate the issue** (include links if applicable): * go to https://stats.wikimedia.org/#/en.wikisource.org/reading/unique-devices/normal|line|2-year|~total|monthly or the same page for any other project **What happens?**: * it says there were 0 unique devices for this month,* We see a sharp spike for the month of May 2025 and is continuing. which is plain false. **What should have happened instead?**: * If the data exists, it should show the data. If it doesn't exist yet (it's the 31th),* The chart should reflect an accurate unique device count with slight increase or decrease. don't show a 0.The fix for this issue is happening in T395934 **Other information** (browser name/version, screenshots, etc.): {F60980419} Also noting, that the "values under 1000 are not reported to preserve accuracy" message that appears is inaccurate, because, well, 0 < 1000. I suppose perhaps there is code in place that //should// prevent the 0 from being shown. This may be a RESTbase metrics problem.