Just to clarify for everyone… arXiv is a public preprint repository that is independent from the ISIC challenge. However it has become he de facto place to place pre-prints for this sort of domain, and we used this mechanism successfully last year.
There is a process of getting articles approved (i.e. an editorial process) prior to them being made public, and unfortunately a number of the submissions were of such poor quality they were rejected. Given the number of submissions that we received (which is a good thing!), since I believe the arXiv editors are volunteers they decided that given the large number of poor quality submissions, they were going to block further submissions to the ISIC challenge.
As some of the other moderators stated, we are looking for a solution. However the purpose of this challenge is to advance knowledge in the data science community, particularly for our domain of interest ( machine learning, computer vision and dermatology). Thus actually having a tangible manuscript that can serve as the basis for future work is an important part of the challenge.