How this site is checked

There are two kinds of page here and they are checked in two different ways. New material is reported and checked by the desk. Archive material is reproduced from the original Picdar site, and the only thing established about it is that the copy is faithful to what was published. Every page states which of the two it is, and the distinction is never blurred.

New material is written against primary sources: the published standard, the statute, the specification or the vendor’s own documentation. A field name, a version number, a duration, a term of copyright or a file-format limit appears on the page only if it can be pointed at in one of those, and the text says which one. Where a figure is disputed between sources, both are given. Where there is no source, there is no figure and no field: a gap is more useful than a number nobody can check.

Nothing on this site is a product ranking, and no score, star or verdict is attached to a supplier. That is deliberate. The beat is procedure and rights, both of which depend on a publication’s own circumstances, so a league table would be a worse answer than an explanation of the trade-off.

Archive pages are copied byte for byte from the company’s own site, including prices, telephone numbers, product names and claims that have not been true for twenty years. Broken links inside them are repointed to the surviving address where one exists, and nothing else is altered. The original addresses still answer, so an old citation or bookmark continues to work.

Dates are used for one thing only: the date shown on a page is the date the material on it was last written or checked. Pages are not restamped in bulk to look recent, and an archive page keeps the date the company published it.

The scale of the site is worth stating plainly, because it explains what this desk can and cannot check. As this page was last built it held 22 articles of new material across 5 sections, and 58 pages reproduced from the original Picdar site. 8 of those archive pages carry a publication date of their own, and those dates run from 2003 to 2006; the rest were published undated and are left that way. Those counts come from the site’s own index rather than from an estimate, which is why they are not round numbers.