The Messidor database, which contains 1200 eye fundus images for diabetic retinopathy screening and diagnosis, has been publicly available since 2008. The database was created by the Messidor project to evaluate automatic lesion segmentation methods. This paper analyzes the feedback on the Messidor database after more than six years of distribution, focusing on the number of citations, web site visitors, and download requests. The analysis shows a significant increase in these metrics over time, particularly from less developed countries. The study also highlights the importance of clear and complete database descriptions, user acknowledgment, and automatic validation procedures for managing public databases. Despite the images becoming outdated, the Messidor database remains a valuable resource for researchers in retinal image processing, with 47 papers citing it since 2008. The experience gained from managing the Messidor database provides recommendations for future database design and maintenance.The Messidor database, which contains 1200 eye fundus images for diabetic retinopathy screening and diagnosis, has been publicly available since 2008. The database was created by the Messidor project to evaluate automatic lesion segmentation methods. This paper analyzes the feedback on the Messidor database after more than six years of distribution, focusing on the number of citations, web site visitors, and download requests. The analysis shows a significant increase in these metrics over time, particularly from less developed countries. The study also highlights the importance of clear and complete database descriptions, user acknowledgment, and automatic validation procedures for managing public databases. Despite the images becoming outdated, the Messidor database remains a valuable resource for researchers in retinal image processing, with 47 papers citing it since 2008. The experience gained from managing the Messidor database provides recommendations for future database design and maintenance.