-
Notifications
You must be signed in to change notification settings - Fork 44
R believes that X11 isn't install, even when it is? #149
Comments
CRAN required me to disable functions that use X11 library to retain imager on CRAN when autotools can't find X11 library. An idea to avoid the disablement is to modify the following line of configure.ac file Line 59 in 9173851
like this DONT_HAVE_X11=''
The idea is a temporary measure. |
I have now tested on several Mac machines, and they all have this error. Autotools does find X11, but Imager does not detect X11. This causes Imager to disable some functions even when it shouldn't. How does Imager detect when the X11 library isn't installed? Could there be a bug there? |
PS. Thanks for the quick and help replies! |
imager uses Lines 51 to 52 in 9173851
The disablement means R failed to call XOpenDisplay function of X11. |
I have also just seen this error on my colleagues Mac after they upgraded their R version to 4.0.5. It had been working previously with R version 3.6 edit: and it appears that downgrading back to R 3.6 does not fix the issue. |
Same for me as for @ajfedorec haven't been able to use the library ever since upgrading to R 4.0.0 or R 4.0.5. Downgrading to 3.6.1 didn't help either. |
The problem is |
We don't have a solution to the problem at the moment. |
Thanks for making your code for imager free for us all to use.
I am using the imager package specifically for the grabPoint() function in my own package.
Everything works fine when I run the package on a windows machine. But, on Mac, the function fails, with a message that X11 needs to be installed: "Error in select(im) : Please install X11 library to use this function."
This is where it gets weird. If we run, capabilities(), X11=TRUE. Similarly, we can run data.entry() which pulls up a working GUI in R. So, X11 is working fine, but imager, or really, just select() from cimg, doesn't believe X11 is there.
I was able to try on a second Mac, with an older imager version, and it worked.
This lead me to think that maybe, in your last commit, some code was introduced that makes R believe that X11 isn't install, even when it is?
Any idea if this is right? Or how to figure out where?
The text was updated successfully, but these errors were encountered: