I have a macro that uses XMove/XClick commands that works in one machine (Windows 10) and are named “something_dpi_108.png”. The issue is that a exported version of the macro running in another machine (Mac) begins to have issues in all visual commands that use images and is needed to take a screen shot of every image in the other machine to make it to work.
The thing is in the other machine, when is selected the image, automatically is created with “something_dpi_120.png”.
Then is there like a universal way to make these images work in more than one machine without the need to take screenshot again for each one? Since the thing becomes complex when there are several images to replicate in other machines.
Can I edit each “something_dpi_108.png” image and edit in an image editor to change size or something and rename it to “something_dpi_120.png”?
Ui.Vision contains already some logic to adapt the images to changing screen resolutions, but this does not always work correctly. One way to make this work better is to lower the confidence level from the default 0.6 to e. g. add @0.45 to the image name. => e. g. Xclick | something_dpi_120.png@0.45
But even with this change, sometimes the better (or only) solution is to recreate the image on the second machine. Improving the image recognition further is on our todo list.
Another workaround is to use text recognition (OCR) instead. Text recognition is resolution independent! So instead of Xclick | image.png use XClickText | text... or XClickTextRelative | text…`
Thanks for your answer. The original image is dpi_108.png in machine1 and is already used like dpi_108.png@0.90 in machine1 since without the 0.9 matches several other points. Then when this image is used in machine2, the technique dpi_108.png@0.90 doesn’t work.
Additionally of OCR you mention, use XClick by coordinates is independent of the machine or coordinates will also change from machine to machine?