This repository contains the files from an attempted restoration of a Wayback-Archived Adobe Muse website about iPods.
I stumbled uppon the link http://classicipods.test/identify-your-pod.html while searching how to identify an iPod Classic 5th gen. from an iPod Classic "5.5th" gen. Needless to say, I hit a ERR_NAME_NOT_RESOLVED page. Like anyone would do, I search it up on Internet Archive's Wayback Machine just to be faced with a window alert with the following message:
Some files on the server may be missing or incorrect. Clear browser cache and try again. If the problem persists please contact website author.
A quick Google search says this is propably an error related to Adobe Muse files (never heard of it until now). Peeking at the page source-code and comparing it to the captured URLs list for the domain we could easily guess it is because Wayback crawler couldn't follow (and archive) some files generated by the Adobe Muse program, especifically the ones hidden/nested inside this tag:
<!-- (...) -->
<!-- RequireJS script -->
<script src="/web/20180807013753js_/http://classicipods.com/scripts/require.js?crc=4159430777" type="text/javascript" async data-main="scripts/museconfig.js?crc=4179431180" onload="if (requirejs) requirejs.onError = function(requireType, requireModule) { if (requireType && requireType.toString && requireType.toString().indexOf && 0 <= requireType.toString().indexOf('#scripterror')) window.Muse.assets.check(); }" onerror="window.Muse.assets.check();"></script>
</body>
</html>
<!-- (...) -->The important part there is the data-main="scripts/museconfig.js?crc=4179431180" attribute. Searching this file on GitHub filtering by JavaScript can give us a hint of what its content was like.
I used wayback_machine_downloader to download all latest files from the Archive:
wayback_machine_downloader http://classicipods.com -d website -c 20This downloaded around 130 files but some of them (mainly assets) had some messed up trailing characters because they were called with a crc query param on the pages. These might be refering to a Cyclic Redundancy Check, being used as a versioning mechanism and/or for cache-busting. Because I don't care about the versions for now, I remove these strings from the filenames running the one-liner for file in *; do mv "$file" "${file/\%3fcrc*/}"; done command inside each asset folder.
Although Adobe Muse is descontinued, we can find some websites online using it or at least hosting its generated files.
Analysing the internals of the generated Muse files, we can infer the Some files on the server may be missing or incorrect. error message is thrown because a verification on the page failed. This can be seen near the top of index.html (formatted bellow for readability):
<script type="text/javascript">
// Update the 'nojs'/'js' class on the html node
document.documentElement.className = document.documentElement.className.replace(/\bnojs\b/g, 'js');
// Check that all required assets are uploaded and up-to-date
if(typeof Muse == "undefined") window.Muse = {};
window.Muse.assets = {
"required":[
"museutils.js",
"museconfig.js",
"jquery.musemenu.js",
"webpro.js",
"musewpslideshow.js",
"jquery.museoverlay.js",
"touchswipe.js",
"jquery.watch.js",
"jquery.musepolyfill.bgsize.js",
"require.js",
"index.css"
],
"outOfDate":[]
};
</script>From this asset list, our downloaded files contained only ./scripts/require.js. Checking this same list on all the other HTML files show the list above share all the required scripts apart from each page CSS file (some of which were not archived thus sadly can't be restored).
Opening the ./scripts/require.js file we have, we get to know how it is check for the Muse-generated assets (for a specific version, at least):
for (var a = document.getElementsByTagName("meta"), b = 0, c = a.length; b < c; b++) {
var d = a[b];
if ("generator" == d.getAttribute("name")) {
"2017.0.1.363" != d.getAttribute("content") && Muse.assets.outOfDate.push("require.js");
break
}
} // beautified snippetChecking the head from each HTML file confirms its "generator version" to be <meta name="generator" content="2017.0.1.363"/>. Using a combination of searching from Google + GitHub we can find some projects that contains files that match our generator version (some examples #1, #2, #3, #4, #5, #6, and many others).
Matching these hosted files with our list, I used https://download-directory.github.io/ to get them.
Another thing we need to do is explained in the End of service for Adobe Muse FAQ:
What happens to my existing Adobe Muse-built website? Will the site disappear? Websites built with Adobe Muse will continue to be live on the internet if the sites are hosted on any third-party hosting platform such as GoDaddy or Bluehost (for example). Sites hosted on Business Catalyst will have to be republished to another hosting platform to remain live after March 26, 2021, when Business Catalyst hosting is no longer available. As of July 31, 2020 sites published using Adobe Muse will no longer be editable using the Adobe In-Browser editing solution. For more details, see the Business Catalyst announcement page (https://businesscatalyst.com/#announcement).
There is a "http://" CDN link from businesscatalyst.com in ./scripts/museconfig.js (http://musecdn2.businesscatalyst.com/scripts/4.0/jquery-1.8.3.min), I simply replaced it with a cdnjs link. I don't really care that much about "http://"-only targeting since we can test it localy or host it for free on GitHub Pages (https by default).
There are many options how to serve the website contents locally. You can either copy it to a webserver localhost folder, or simply run npx serve . inside the ./website folder.
I decided to comment out some scripts from the page contents before putting up a hosted demo. This is mostly because, after getting the website to run, it became clear this was a "storefront" for a man named Sean to provide maintenance/mods services and/or selling parts for iPods. In respect to Captain Sean (as he states in the ./website/home.html page), I decided to not leave it as is because we don't know the current state of the matter for his projects.
Just to be clear that there is no ill intention with this project. This was done only to document and shed light to a part of the web that may be gone unnoticied and, probably, has already been lost. Adobe Muse websites, although most of them unknown, are still part of internet history and the lack of archiavability on its part (before and after its sunset) is whorysome at least. As shown in this very project some files are missing (css file for ./website/ipod-mini.html for example).
Since it relies heavily on loading files through javascript, crawlers are not able to easily navigate through all needed files. One easy way to solve this issue is to gather all Muse generated files from its manifest file. This file contains all folders, files, and versions used to generated the website content.