14

I am curious if there is there any good way to guarantee that my articles (posted on personal sites and other blogs) will last? Is there a reliable repository for articles, post and essays? Or are there any good practices of self-archiving?

Even when it comes to my websites, there is a dozen of things that can happen, e.g.:

  • forgetting to pay for the domain or to the provider,
  • the provider bankrupts,
  • university will likely close it after being no longer affiliated (and if not, it policy may change later),
  • having no to time to maintain it,
  • and the proverbial "Getting hit by a bus" (not to say that everyone dies, sooner or later).

When it comes to guest posts on other blogs - I can guarantee even less. And from practice, many good articles or essays I am linking to do disappear.

Of course, there is wonderful Wayback Machine at the Internet Archive, and (up to now) all my articles I care about are there. But it might be not a silver bullet, as it is hard to search inside.

Just in case, all of them are on CC BY or can be made so.

EDIT: I mean archiving in a public repository - for others, not me.

6

Great question. I agree that permanent archiving of blog posts and other digital content is an important challenge in open data. It might be helpful to break this down into parts:

Persistent identifiers

Having a persistent address at which potential users/machines can reference your content is crucial to good archiving, and most of the issues you list refer to this (losing your domain name, provider going bankrupt, etc.) One promsing way to address this is by registering a PURL or Persistent uniform resource locator. PURLs are widely used for major web ontologies like Dublin Core: http://purl.org/dc, to make sure these resources have consistent links. Users must register and they are relatively easy to set up (for instance, I have registered http://purl.org/cboettig as a partial redirect to each of the pages in my online research notebook. If you are familiar with DOIs for published literature these are largely analogous technology.

Archiving content

Making sure your link always resolves is not the same as guaranteeing your content continues to exist. Robust, geo-politically distributed archiving services like CLOCKSS or LOCKSS are probably the gold standard here, but not accessible to individual authors. Having your own distributed backup copies on public repositories is still a good idea. Depositing copies in an appropriate repository, such as fig*share* for scientific research content, is one way to achieve this level of archiving.

Tools like Git/Github can help archive the version history of your content, not just the most recent version.

Good metadata

Having good machine-readable metadata on your site will help search engines index it accurately and can help users/machines actually make use of it. Consider identifying author, titles, dates, tags, and other such data using RDFa, though even vanilla HTML5 has quite a few semantics available. Some examples here

You've already mentioned good licensing, which is key in making your posts useful as open data. See that the license information is properly embedded in machine-readable metadata as well.

  • Really nice answer. Would +>1 if I could. – Laurent Jun 3 '13 at 19:48
1

I use a mix of backups and version control.

I have a few sites built with a static site generator, which allows me to put all source code and content files in a version control system, like git. In addition to archiving your posts you also maintain a history of changes.

For dynamic sites I do backups of the source code and the database.

For external posts you could also use version control to archive them, which obviously does not prevent them from being unpublished by the external site owner.

  • 1
    I don't think if it answers the question. I can use version control for anything. Here as well, but the content is unlikely to change. Moreover, the question is not about archiving of my drive (which is trivial) but about making in publicly accessible and usable for as long as possible. – Piotr Migdal May 31 '13 at 20:51
  • I didn't realize that you wanted to make it publicly accessible. – ramiro May 31 '13 at 22:50
1

Just write topics in all possible places and create system which will sync them all.

Use blogs

http://blogger.com - google solution. I think it will live until google lives. Create blog - create posts

http://wordpress.com/ - open source blog platform from Automattic on microsoft clouds

Different wiki projects. SO is wiki project too.

Harder solution for non web-developer is create blog on http://heroku.com or http://openshift.redhat.com

Use http://pages.github.com/ as combine cvs and post info.

Create torrent from posts and post it on http://thepiratebay.sx/

In case of saving any fixes in topics you must have all texts in any CVS

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.