How AltF Atlas is maintained

An entry has to be the best answer, or a meaningfully different answer, to a question someone actually asks. Three near-identical PDF mergers is worse for a reader than one, so near-duplicates are cut rather than collected — the value of a curated list is what it leaves out.

It also has to work in a browser tab with no installation and no server of your own. That is the boundary of the whole product; a desktop application, however good, belongs in the software directory instead.

A directory where everything is described as excellent carries no information. The `Where it stops` sentence on every entry is the reason to believe the rest of it: a size cap, a watermark, a missing export format, an account requirement.

Where a tool genuinely has no meaningful limit, the field says what it does not do instead. It is never left empty, and it is never filled with "no limits".

Access is recorded as what the site costs you before it does its main job — open, free account, or freemium — rather than what the top tier costs. That is the fact that determines whether a tool is useful to you in the next five minutes.

Processing is recorded as on-device only when the work genuinely happens in the browser and the file is not uploaded. This is a privacy claim, so where there is any doubt the entry says hosted. A wrong on-device claim is worse than a missing one.

Shutting down does not remove an entry — it moves it to the archive with a named successor, because "what replaced this?" is a question that outlives the site. Entries are removed outright only when they become actively harmful: aggressive dark patterns, a paywall in front of what was the free feature, or a change of ownership that turns a tool into an ad vector.

Nothing in the Atlas is a paid placement, and no entry can be bought. Where AltFTool has its own tool for the same job, the entry links to it alongside the external site rather than in place of it.