For social reasons, communities engage in ForgiveAndForget. We want to forgive and forget past mistakes, in order to rebuild constructive relationships with people. Software such as some wikis explicitly implement ForgiveAndForgetInSoftware; as pages get edited, older revisions are forgotten by the software (see KeptPages).
There are two problems with this approach, however:
As more and more documents only exist electronically, future historians will depend on the availability of software to read these documents and on the continued archiving of these documents.
The FreeSoftware proponents believe for example that open standards and free software are our best bet when it comes to finding the right software to read electronic documents in the future.
Clearly, ForgiveAndForgetInSoftware makes meaningful archiving impossible. While we might decide on a wiki to summarize the lessons we learn from past conflicts, future historians will be unable to reinterpret history for their own generation.
However, when catering to future historians, you don’t need to make the past highly accessible. For example, future historians may be interested in a heated online debate between two politicians, one of which would later become the president of the EU. However, it is not necessary to RememberThePast by linking to that debate on the politician’s front page and allowing it to be indexed by google and freely copied. Instead, it may be more appropriate to move it to a seperate “archive” section, or even store it offline, provided it is not deleted.
Does this mean that we would want to keep all page versions, for example? – AlexSchroeder
To the extent practical, yes, that’s my view. There need to be ways of excising some content--the example I think of is someone posting a third-party’s personal information, like an address or a credit card number by way of an attack on that person, their identity, their financial integrity--but other than that, the metrics should be technical (storage space, bandwidth, functional latency), rather that enforcing social goals through technical means. KeptPages, in my view, is a HardSecurity-like approach to ForgiveAndForget. People forget when they forget, not when the software drops a diff. And when they remember, its better that they have an opportunity to reinterpret the past from the original sources, and to be reminded of the ways the history went down counter to their selective, imperfect, and perhaps self-serving, memory-formation processes. --JoeAnderson
I agree. Also, you want to delete massive pr0n or warez uploads. I do worry about claims of financial integrity. A credit card is one thing. But you can say things that affect people’s financial integrity. – LionKimbro
The benefit of KeptPages in this context is that it is automatic. Without this mechanism, how would we determine what revisions should be deleted and what revisions should be kept? Voting? Something like DeletedPage? We’d have to invent a lot more process, which I am unwilling to do. – AlexSchroeder
It’s a good question, who decides, how, and how removal--especially removal preserving OpenProcess--is implemented. I think its answer depends on the laws of the venue in which the site is hosted, and who is hosting it. Current mechanisms as they stand aren’t sufficient for warez, illegal pr0n, or card info--that stuff should be deleted on the front end, immediately or at least as soon as detected, and should never enter the archive in the first place, however obscured access to it might be by being buried down in the list of previous versions.
I agree that the archive needn’t be kept immediate to hand, and perhaps would be better set aside and clearly marked as an archive. Continue to use KeptPages for automatic pruning of the current corpus. A series of periodic, downloadable, tarball archives taken to coincide with KeptPages expiry might suffice. In any case, this needn’t necessarily be taken as metadiscussion that needs to be immediately applied to this wiki. I certainly have no sense of urgency about this. I do think, all things considered, it is a better way to go, though. Whatever path you take, though, I guess there will always either be someone along to game the system, or someone along to accuse someone of gaming the system. --JoeAnderson
The problem with copies made by RecordKeepers or archiving software is well illustrated by the famous Dr. Fun cartoon about the Ghost of Usenet Past coming back to haunt you . Social pressures may be successful in restraining this practice, as similar pressure already exists in the judicial system: The necessity for warrants and due process before admitting evidence. Currently, there are lots (hundreds of thousands) of examples of people searching Google for friends, girl/boyfriends. Employers often do Google background checks. There are the ClaireSwires of the world. --anon.
It seems to me that the current community expectation in all sorts of media is that digging things said out of archives and using them against the author is ridiculous. When somebody digs up old pictures of presidential candidates smoking pot or foreign ministers throwing stones, then that is just a non-issue. At least it should be. – AlexSchroeder
These days, historians place great value in individual material, however. Private letters, diaries, all these little individual pieces of information help us provide a new interpretation of the past. And yet, digging out all these letters, reading about the attrocities of the second world war doesn’t stirr up the conflict again. Why? Because closure has been found. Conversely, in Yugoslavia, closure had not been found, and the wars of the 90s are partially a result of this. – AlexSchroeder
It’s important that archives are marked as such. So, on a blog, you have a date above your old blog entries, so people can easily understand “ahh, he felt that then, but that was a couple years back”. On a wiki, you want to make sure that the stuff you’ve decided to ForgiveAndForget is in the VersionHistory, but it’s not in the current version, in other words, it’s not in the WikiNow. That’s one advantage of a wiki over a blog - because you can edit the current version, the software actively supports ForgiveAndForgetInWetware, even if you don’t go as far as implementing KeptPages. – anon
The advantage of a blog is that old material is automatically marked old, whereas on Wiki it is a manual, labour-intensive process. – anon
Freenet. The community decides what to keep and what to cancel. What is requested is copied and remains, what is not requested gradually vanishes. History is a question of capacity. The more capacity the more history the better. TerraWiki?. – MattisManzel