Download websites that use DreamWeaver

About DreamWeaver

DreamWeaver never ran on any server we crawl. The application sat on a designer's desktop, and what survives on the web is the file it saved, still being served long after whoever saved it moved on.

Adobe sells the application as a desktop tool for authoring pages, and it has been on the market since 1997, well before the frameworks it shares this dataset with existed. A page carries its fingerprint because the editor wrote meta tags, template comments and generated JavaScript into the output, and those artefacts survive in the file for as long as nobody rewrites the page.

As of October 2026 that fingerprint is present on 132,324 sites, and the version picture is what makes the entry worth reading. We can separate 20 distinct versions in the wild, and the one encountered most is 8, a release from the era when the product still belonged to its original publisher rather than to its current owner. On a hosted service, version spread reflects a rollout schedule. Here it reflects the year each page was authored and then abandoned, which turns the field into a rough date stamp for the page rather than a statement about software anyone is running.

The JavaScript artefacts deserve a note of their own, because they are part of why the fingerprint survives. The editor generated its own behaviour scripts for rollovers, image swaps and form validation, and those functions were written into the page rather than loaded from a library that could later be swapped out. A page can be re-styled, re-hosted and moved behind a modern proxy while still carrying them, which is exactly what makes the marker durable and exactly what makes it easy to over-interpret.

Geography is flatter than on most entries. 43,455 of the sites sit in United States, a share well below what a single-market product would produce, and .com leads the suffixes. A tool sold worldwide through a design-software channel spreads the way its buyers spread, and the buyers were agencies and in-house designers everywhere.

Two groups want a list of these pages, and they want it for opposite reasons. Agencies and platform vendors read it as a redesign pipeline, because a page still carrying the fingerprint of a desktop authoring era is a page nobody has rebuilt, and the business behind it either has no web supplier or has one who never touched the markup. Security and risk teams read the same list as an exposure map, since abandoned pages tend to sit on abandoned stacks, with old form handlers, stale includes and hosting nobody logs into.

The honest caveat matters more here than on most entries. A fingerprint proves how a file was authored, not that anyone still holds a licence or still edits the site. Some of these pages are maintained in a modern editor and simply inherited old boilerplate, and some belong to businesses that closed years ago while the hosting kept renewing. The list is evidence about pages, not a customer roster for the vendor.

The distinction is also why an export beats a lookup. Checking one domain tells you almost nothing, because a single old page proves nothing about the organisation. The full set, filtered by version, country and suffix and joined against whatever else the site runs, is what separates the merely old from the genuinely abandoned, and that separation is the thing worth buying.