Newsletter Archive / Snapshots, Image Caching


Date: Fri, 19 Dec 2025 08:04:55 +0100 (CET)
From: support@blip.town
Subject: Snapshots, Image Caching

Thanks for testing Bliptown! There have been a lot of updates over the last week or so that I wanted to tell you about.

New layout: gallery-columns

Just briefly, there’s a new layout class: gallery-columns. The gallery-grid layout is horizontally oriented, and so isn’t suited to pages you want vertically oriented (i.e. in a fixed number of columns). The gallery-columns layout arranges images (or rather, elements) into three columns both on desktop and mobile.

Snapshots

You can now save, restore and download snapshots of your site. I haven’t written any documentation on snapshots, so here goes...

A snapshot is a point-in-time copy of all the files in your site. Taking a snapshot saves a copy, then subsequent snapshots only save what has changed, creating a full history of your site.

You can browse snapshots from the Files page, under “+ more”. Each snapshot taken on Bliptown will include a brief commit message listing which files have been added (A), modified (M) or deleted (D).

Restoring a snapshot will restore all the files to /snapshots/<HASH>, e.g.

/snapshots/06aef65/index.md

Also, no longer are Bliptown sites held prisoner of blip.town! You can download a tarball (like a zip) of any snapshot. You can also clone a snapshot respository with git, if that’s something you’re familiar with.

Restored snapshots are not included in new snapshots.

You can’t delete snapshots. (Well, you can, if you SSH into the server and rewrite the git history, but if that doesn’t make any sense to you then no, you can’t delete snapshots.)

Image Caching

Bliptown now has a simple file-based web-optimised image caching system, allowing your site to serve smaller images while also linking to larger originals.

You can create a cache from the Settings page, which will generate copies of all the supported images on your site, resize them to fit within 2048 pixels on the longest side, and save them at 75% quality. If the image is already smaller than this it will not get resized.

The image files are saved in a /.cache directory and the filenames are SHA1 sums of the URL path. A request to the original image path will be redirected to the cache path if the cached file is the most recent.

You can bypass bypass the cache by appending a cache=0 query parameter to the URL:

![image of doggy](/images/doggy.jpg?cache=0)

Cached files are not included in snapshots. The cache is only generated manually, so if you don’t need a cache, don’t worry about it!

Please let me know if you have any feedback on these features! As with everything on Bliptown, they’re still in alpha state, so I appreciate your testing and reporting any bugs.

Significant Bugfixes

  • Support for UTF-8 filenames.
  • Eliminated empty <p> elements around file inclusion directives.
  • Added CSS overrides for #drawer contents to avoid accidental (or perhaps intentional?) user styling.
  • Made #viewport/main CSS accommodate expanding content.
  • Added workaround for iOS bug displaying emoji instead of disclosure triangles.
  • Display login message instead of 404 error for protected pages.
  • Fixed creating new page attempting to recreate current page.
  • Made bliptown_helper fork for concurrent commands, avoiding request deadlocks.

As always, Bliptown is completely free and open source software. The source code can be explored here: https://blip.town/code

Sincerely,

–
Paul W. Rankin
https://blip.town

← back