The short answer: use a capture extension, not the print dialog. Print paginates — the same article came out as 26 pages with 9 breaks cutting through a sentence. A capture writes one continuous sheet and can stamp the page with where it came from and when.
Firefox, desktop and Android Chrome 116+, Edge, Brave, Vivaldi
Free, MIT licensed, runs on the device. Edge asks once to allow extensions
from other stores; Opera needs its Install Chrome Extensions add-on
first. Then: open the page, press Alt+Shift+Y or click the icon.
| Browser print | Full-page capture | |
|---|---|---|
| Same article comes out as | 26 pages | 1 sheet |
| Page breaks cutting a sentence | 9 | 0 |
| Text you can select and search | 94.8 % recall | from OCR: 92.6 % |
| Costs anything | no, built in | no |
Print wins on text. If all you need is a readable, searchable copy and the pagination does not bother you, the function already in your browser is enough — and this page says so rather than pretending otherwise. A capture is worth it when the layout matters, when a break would fall through a table, or when the source details have to travel with the file. Method and raw data
A capture extension reads what your own session already shows. A licensed journal article, a course page, an order confirmation — saved as you see it. No server-side converter can do this, because it does not have your session: measured across 20 mixed sources, a server-side reader was refused by 5 of them outright. Capturing a page you may read is a copy for your own use; it is not a route to content you do not have access to.
Only Firefox. Chrome for Android installs no extensions at all, so the question only arises there. Of 248 page-saving extensions checked against the add-ons API, 60 declare Android support and none had been tested on a device before we did. The Android measurement
A URL in a bibliography is a promise about a page you do not control. Checked against 150 sources taken from real reference lists: 19.3 % were gone, 8.7 % had no archived copy anywhere, and where a snapshot existed it was a median of 603 days old. For a page that declares no publication date, the time of retrieval is the only date a reference can carry — and it exists only at the moment you look. What happens to a source after you cite it
A capture can write that in: authors, title, DOI, licence and the exact time, inside the PDF and in an RIS record beside it — a format that imports into Zotero, Citavi, EndNote or Mendeley.
For a bibliography rather than a single page, most of the work needs no browser. Of 20 mixed sources, 10 became complete records with RIS and BibTeX in 8.1 seconds over plain HTTP — no account, no key. The other ten are handed back with the reason, so you know which addresses need you. The recipes · the measurement