I recently published a paper with IEEE. During the publishing stage, I had to provide them the figures named [MyName]FigureX_a, [MyName]FigureX_b etc...
By accident, I got two figures wrong and they were named FigureX_b and FigureX_a instead of FigureX_a and FigureX_b. During publishing, I spotted this error and told them to fix the problem. They did fix it at the time and I approved the changes.
However, when I received the final complete PDF, the error was present once again. In the e-mail, it stated that "Please note this is not the final version of the article and corrections will not be reflected in this PDF". I thought that the change would be reflected in the final version so did not bother to e-mail.
I have just seen the final version on the web-site and the error is still present. This error should be pretty evident to anybody reading the paper. The figure is part of many other subfigures so the inconsistency is apparent. The overall message of the figure is not lost and the results are still the same apart from the order being wrong.
I am unsure what to do. On one hand, I want to contact IEEE and complain. I unfortunately do not have the proofs of the pdfs that were provided for review but I have my correspondence saved. On the other hand, I think the error is minor enough that any reader should understand what happened.
The file that will be on PubMed is correct though. I could in theory submit it to ArXiv with the correct ordering, copyright notices and other information required by IEEE(DOI,PMID etc.).
Does this require any drastic actions?