It is about time that meta gets its process act together. Meta administrators and stewards are largely unanswerable to anyone. The Wikipedia ArbComm ruled that the spam blacklist was not to be used for content control. (I.e., "fringe" is not an argument that should support a blacklisting.) But this is completely ignored at meta, it's outside of ArbComm's "remit," and I saw that even the most abusive blacklistings, where the links being added were actually positive contributions, were sustained on review. Basically, it's Moulton's maxim: bureaucracies don't correct their own errors. To do so would be to admit error.
(The large majority of blacklistings that I've seen were appropriate. But that makes it all the more important to have an independent review process, to deal with the exceptions.)
If anyone wants to look at this, there are two sites in particular that remain on the global blacklist. I'm sure there are many other examples.
lyrikline.org. This was locally whitelisted on en.wip as a result of my work. (more accurately, the english language interface was whitelisted.) I never finished the job of replacing all the links that had been summarily deleted by a blacklist volunteer. No evidence was presented of even a single truly inappropriate link. The blacklisting "discussion." Lyrikline.org is a library of readings from prominent poets, from all over the world. It's supported by the German government. It has won awards for its work, and is extraordinarily useful as a source or external link for articles on the many poets represented there.
A study of the lyrikline blacklisting is at User:Abd/Blacklist/lyrikline.org.
Denial of delisting request. Notice charges of copyvio, which are preposterous for this site." In other words, a person, seeing this as a very useful site to link to, worked hard (as covered in my page on the history of this) to do it, and was slapped down, blocked, and the site blacklisted. De.wikipedia went on to whitelist the site, and since it is hosted in Germany, one would think they would know! It was also effectively whitelisted on en.wp.
Denial of delisting request. Notice that "excessive link placement" is the only evidence presented of "spamming.
lenr-canr.org. This was originally blacklisted on en.wp JzG, in a manner that was covered in RfAr/Abd and JzG. This was the decision where ArbComm ruled that the blacklist was not to be used for content control. However, when JzG was challenged on en.wp over his blacklisting, he rushed to meta and requested global blacklisting, which was granted. He made a series of false, deceptive, highly misleading arguments, such as calling the signature of the site owner Jed Rothwell, a long-time critic of JzG, "spam," implying that these were links. They weren't links at all. In other words, the primary evidence of "spamming" was not links at all, but just Jed identifying himself as "librarian, lenr-canr.org," editing as IP, and these edits, for the future, would not be interdicted by the blacklist at all.
What was interdicted was the legitimate usage of lenr-canr.org as a source of convenience links to published papers, hosted there with permission. If, out of the thousand peer-reviewed papers there, there is one true copyvio, it's possible, but not terribly likely. (Jed Rothwell normally obtains permission from both publishers and authors before hosting, but there are some prominent authors in the field where they have given permission, Jed has requested confirmation from the publisher, who did not respond. Jed is nonprofit and the most he would suffer, if the publisher truly didn't want him to host the document, is a take-down notice. Which hasn't happened. But that condition isn't the norm, affecting at most a couple of pages, not nearly enough to justify blacklisting the site.)
I was able to get many pages on the site whitelisted on en.wikipedia. Whitelisting was denied, as I recall, for only one page, and this was because of some unclarity like that mentioned above. That page looked like it might be copyvio. Technically, though, to legally prohibit linking to a page, it must be *known* that it is copyvio. There would be no risk, but I did not push the issue of this one page, I was much more concerned about the bulk.
A huge amount of evidence could be presented on this one. For another day, another place, perhaps.
