Streets shifted a several meters

Hi guys

Lots of streets in my city are not correctly placed on the satellite pics.
How do I know which is wrong: satellite pics or mapped streets?

Currently I am ignoring the satellite pics and I am drawing areas based only on the streets, but things are getting messy.

What should I do? :open_mouth:

Here are some answers to a similar question asked recently on the help page.

You can compare the mapped streets with recorded GPS tracks. You seem to use iD: Go the right side menu, open the background settings (symbol with stacked sheets) and activate the check mark to display OpenStreetMap GPS Tracks. They will display as thin colored lines

There are not many GPS tracks in that area but the few that there are suggest to me the mapped streets are more accurate then the imagery.

In the same menu you could chose a different imagery source but in your area they all seem to have the same images. On the bottom of that menu you can also shift the imagery to match the existing map. That might make it easier for you to apply changes consistent with the rest of the map.

the area:

Use a professional differential GPS receiver to survey the true location.

Only a few meters is well within the error bounds of consumer products.

You might be able to get better with a consumer product by averaging over many hours.

Unless the error is extremely intrusive, it is normal to use the currently mapped position as the correct position and adjust other sources to that when using them to add new detail. Some care may be needed where the offset is changing.

There are lots of Strava tracks in the area to help with alignment of satellite images.
As viewed in JOSM:

Using JOSM I find that switching off the visibility of the data layer and changing the opacity (or dragging the slider from side to side) of the Strava layer is a good method to assist with aligning the satellite imagery with the Strava and gps traces.

Plus about 10 for this one! I find the Strava layer with reduced opacity very, very useful in both aligning imagery and for determining the location of trails and roads hidden by tree cover, etc.