little-brisk:

has anyone found a good way to back up their tumblr that isn’t a complete fucking mess? all i want is an archive of my original posts, a function which every blogging platform since the invention of blogs has made extremely easy but we all know the hellsite wouldn’t be our home if it was a functional blogging platform, so

{{maryellencarter replied to little-brisk: I think @brin-bellway knows things about tumblr backups… }}

@maryellencarter​ summoned me here, since I have experience in this topic and am pretty much always up for talking about archiving methods.

So, I know of a few different options, depending on how exactly you define “good” and whether you specifically want *only* OPs or would be good with a full-blog archive too.

—

1. tumblr-utils. On Linux, you just download their zip, put the file labelled “tumblr_backup.py” in your home directory (I keep the rest of the folder elsewhere in case I ever need it for anything), and then open a command-line terminal and run “python tumblr_backup.py little-brisk”. I’ve never done it on other operating systems, but I expect it’s fairly similar; since this is a relatively popular method, we might be able to find someone with firsthand experience doing it on your operating system to give you more details.

By default, this will give you a local folder (a sub-folder of the folder your tumblr_backup.py file is in) containing both reblogs and OPs, with images included but not video or audio, and posts indexed by month but not by tag. There are various options you can add to the command [link]: for example, “python tumblr_backup.py –save-video –save-audio little-brisk” will include the audio and video as well as the images. Warning: if you use the “–tag-index” option and you have any tags (whether organisational or commentary) with a slash in them, the archiver will crash. I have not tested the “–no-reblog” option, but it does exist.

Depending on how many posts you keep and how many images were in them, the folder may be several GB in size and contain tens of thousands of files (most of them tucked away under the image and individual-post sub-folders). IME one *needs* to zip a large blog’s folder in order to move it around: attempting to copy 30k individual files to a USB drive or suchlike tends to result in stuff like “Estimated time remaining: 4 months”. Zipping a large blog only takes an hour or three, though, and then you can copy it to places in just a few more minutes.

—

2. WordPress. This is mostly for if you really want your backup to be public-facing: the formatting *is* a bit of a mess (especially on long reblog chains and tag rambles), and I personally have been chipping away at fixing my WordPress’s formatting for an entire year and am still in 2017 (admittedly, I’ve been doing some mental crop rotation lately).

I’m not sure if they offer an OP-only option.

If you’re feeling particularly paranoid and are willing to fiddle around with a few software guts, you can self-host this, or keep a kitted-out WordPress server stashed away on your computer in case you want to import into a self-hosted platform later [here is a post I wrote on how I did this].

—

3. I know it must be possible to use wget because I’ve seen people do it [link], but I’m not sure exactly which options you need to add to the terminal command to make it work properly. I just now made a few promising-looking variants of my Dreamwidth-archiving command [link], and none of them could get past the front page of any of the blogs I tested them on.

—

TumblThree is Windows-only and Soup is OPs-only, so I haven’t used them, but I hear some people like them.

---

Tags:

#reply via reblog #101 Uses for Infrastructureless Computers #Tumblr: a User’s Guide #the more you know #WordPress