Yesterday I proposed the idea of a “minimum interesting service worker” that could provide a link (or links) to archives or mirrors when your site was unavailable as one possible solution to the desire to make personal #indieweb sites more reliable by providing at least a user path to “soft repair” links to your site that may otherwise seem broken. Minimum because it only requires two files and one line of script in site footer template, and interesting because it provides both a novel user benefit and personal site publisher benefits. The idea occurred to me during an informal coffee chat over Zoom with a couple of other Indieweb community folks yesterday, and afterwards I braindumped a bit into the IndieWeb Developers Chat channel^1. Figured it was worth writing up rather than waiting to implement it. Basic idea: You have a service worker (and “offline” HTML page) on your personal site, installed from any page on your site, that all it does is cache the offline page, and on future requests to your site checks to see if the requested page is available, and if so serves it, otherwise it displays your offline page with a “site appears to be unreachable” message that a lot of service workers provide, AND provides an algorithmically constructed link to the page on an archive (e.g. Internet Archive) or static mirror of your site (typically at another domain). This is minimal because it requires only two files: your service worker (a JS file) and your offline page (a minimal self-contained static HTML file with inline CSS). Doable in <1k bytes of code, with no additional local caching or storage requirements, thus a negligible impact on site visitors (likely less than the cookies that major sites store). User benefit: If someone has ever visited your personal site, then in the future whenever they click a link to your pages or posts, if your site/domain is unavailable for any reason, then the reader would see a notice (from your offline page) and a link to view an archive/mirror copy instead, thus providing a one-click ability for the reader to “soft-repair” any otherwise apparently broken links to your site. Personal site publisher benefits: Having such a service worker that automatically provides your readers links to where they can view your content on an archive or mirror means you can go on vacation or otherwise step away from your personal site, knowing that if it does go down, (at least prior) site visitors will still have a way to click-through and view your published content. Additional enhancements: Ideally any archive or mirror copies would use rel=canonical to link back to the page on your domain, so any crawlers or search engines could automatically prefer your original page, or browsers could offer the user a choice to “View original”. You can do that by including a rel=canonical link in all your original pages, so when they are archived or mirrored, those copies automatically include a rel=canonical link back to your original page or post. The simplest implementation would be to ping the Internet Archive to save^2 your page or post upon publishing it. You could also add code to your site to explicitly generate a static mirror of your pages, perhaps with an SSG or crawler like Spiderpig, to a GitHub repo, which is then auto-served as GitHub static pages, perhaps on its own domain yet at the same paths as your original pages (to make it trivial to generate such mirror links automatically). If you’re using links to the Internet Archive, you can generate them automatically by prefixing your page URL with https://web.archive.org/web/*/ e.g. this post: https://web.archive.org/web/*/https://tantek.com/2024/151/t1/minimum-interesting-service-worker Possible generic library: It may be possible to write this minimum interesting service worker (e.g. misv.js) as a generic (rather than site-specific) service worker that literally anyone with a personal site could “install” as is (a JS file, an HTML file, and a one-line script tag in their site-wide footer) and it would figure everything out from the context it is running in, unchanged (zero configuration necessary). This is post 14 of #100PostsOfIndieWeb. #100Posts ← https://tantek.com/2024/072/t1/created-at-indiewebcamp-brighton → https://tantek.com/2024/173/t1/years-posse-microformats-adoption Post glossary: GitHub static pages https://indieweb.org/GitHub_Pages HTML https://indieweb.org/HTML JS https://indieweb.org/js rel-canonical https://indieweb.org/rel-canonical service worker https://indieweb.org/service_worker Spiderpig https://indieweb.org/Spiderpig SSG https://indieweb.org/SSG References: ^1 https://chat.indieweb.org/dev/2024-05-29#t1717006352142600 ^2 https://indieweb.org/Internet_Archive#Trigger_an_Archive

Yesterday I proposed the idea of a “minimum interesting service worker” that could provide a link (or links) to archives or mirrors when your site was unavailable as one possible solution to the desire to make personal #indieweb sites more reliable by providing at least a user path to “soft repair” links to your site that may otherwise seem broken.

Minimum because it only requires two files and one line of script in site footer template, and interesting because it provides both a novel user benefit and personal site publisher benefits.

The idea occurred to me during an informal coffee chat over Zoom with a couple of other Indieweb community folks yesterday, and afterwards I braindumped a bit into the IndieWeb Developers Chat channel¹. Figured it was worth writing up rather than waiting to implement it.

Basic idea:

You have a service worker (and “offline” HTML page) on your personal site, installed from any page on your site, that all it does is cache the offline page, and on future requests to your site checks to see if the requested page is available, and if so serves it, otherwise it displays your offline page with a “site appears to be unreachable” message that a lot of service workers provide, AND provides an algorithmically constructed link to the page on an archive (e.g. Internet Archive) or static mirror of your site (typically at another domain).

This is minimal because it requires only two files: your service worker (a JS file) and your offline page (a minimal self-contained static HTML file with inline CSS). Doable in <1k bytes of code, with no additional local caching or storage requirements, thus a negligible impact on site visitors (likely less than the cookies that major sites store).

User benefit:

If someone has ever visited your personal site, then in the future whenever they click a link to your pages or posts, if your site/domain is unavailable for any reason, then the reader would see a notice (from your offline page) and a link to view an archive/mirror copy instead, thus providing a one-click ability for the reader to “soft-repair” any otherwise apparently broken links to your site.

Personal site publisher benefits:

Having such a service worker that automatically provides your readers links to where they can view your content on an archive or mirror means you can go on vacation or otherwise step away from your personal site, knowing that if it does go down, (at least prior) site visitors will still have a way to click-through and view your published content.

Additional enhancements:

Ideally any archive or mirror copies would use rel=canonical to link back to the page on your domain, so any crawlers or search engines could automatically prefer your original page, or browsers could offer the user a choice to “View original”. You can do that by including a rel=canonical link in all your original pages, so when they are archived or mirrored, those copies automatically include a rel=canonical link back to your original page or post.

The simplest implementation would be to ping the Internet Archive to save² your page or post upon publishing it. You could also add code to your site to explicitly generate a static mirror of your pages, perhaps with an SSG or crawler like Spiderpig, to a GitHub repo, which is then auto-served as GitHub static pages, perhaps on its own domain yet at the same paths as your original pages (to make it trivial to generate such mirror links automatically).

If you’re using links to the Internet Archive, you can generate them automatically by prefixing your page URL with https://web.archive.org/web/*/ e.g. this post:

https://web.archive.org/web/*/https://tantek.com/2024/151/t1/minimum-interesting-service-worker

Possible generic library:

It may be possible to write this minimum interesting service worker (e.g. misv.js) as a generic (rather than site-specific) service worker that literally anyone with a personal site could “install” as is (a JS file, an HTML file, and a one-line script tag in their site-wide footer) and it would figure everything out from the context it is running in, unchanged (zero configuration necessary).

This is post 14 of #100PostsOfIndieWeb. #100Posts

← https://tantek.com/2024/072/t1/created-at-indiewebcamp-brighton
→ https://tantek.com/2024/173/t1/years-posse-microformats-adoption

Post glossary:

GitHub static pages
https://indieweb.org/GitHub_Pages
HTML
https://indieweb.org/HTML
JS
https://indieweb.org/js
rel-canonical
https://indieweb.org/rel-canonical
service worker
https://indieweb.org/service_worker
Spiderpig
https://indieweb.org/Spiderpig
SSG
https://indieweb.org/SSG

References:

¹ https://chat.indieweb.org/dev/2024-05-29#t1717006352142600
² https://indieweb.org/Internet_Archive#Trigger_an_Archive