Merge PDF files without uploading them

Merging PDFs is the most ordinary office task there is, and almost every well known way of doing it involves sending your document to a company you have never heard of. That is a strange arrangement to have normalised, and it is the reason this applet exists in the form it does.

It is worth being precise about why it matters, because the objection is not paranoia. Uploading a file containing personal data to a third party service is processing, and it is processing by a processor you have not assessed, under terms you have not read, in a jurisdiction you probably have not checked. If the document is a signed contract, an invoice with bank details, an HR letter or anything clinical, that is a decision with your organisation's name on it, taken by whoever had the file open.

Most data protection policies already say something about this, usually in a sentence about not using unapproved third party services with personal data. The sentence is almost universally ignored, not out of defiance but because the alternative has been inconvenient. Removing the inconvenience is more effective than repeating the policy.

There are practical reasons too. Nothing is queued behind an upload and a download, so it is faster on a large file than any hosted service can be. There is no page limit, no watermark, no account, and no daily cap designed to convert you. And it works with no connection once the page has loaded, which is occasionally the difference between finishing something on a train and not.

One thing worth checking after any merge, whatever tool you use: bookmarks and internal links. Merging renumbers pages, and a table of contents that pointed at page four of one document now points somewhere arbitrary in the combined one. It is the most common way a merged document is quietly wrong, and looking at the first entry takes two seconds.

Where this stops being the right answer is repetition. If the same set of documents is combined the same way every month, that is a process rather than a task, and it should not depend on somebody remembering the order. That is the point at which it belongs in something that runs by itself.