Methodology

How pairs are chosen

Each comparison sets an older technology beside a newer one that either belongs to the same lineage or solves the same underlying problem. Postal mail and email both carry written messages; sextants and satellite positioning both answer the question of where you are. We do not pair things at random or for novelty, and we avoid comparisons that only make sense as a contest between unrelated products.

The eight criteria

Every article looks at the same eight criteria, so that different comparisons can be read in the same way:

• Speed: how quickly the job gets done, from start to finish.
• Portability: how easily it moves with a person.
• Capacity: how much it can hold, carry, or handle.
• Energy: what power or fuel it needs, if any.
• Durability: how well it lasts under ordinary use and over time.
• Maintenance: how much upkeep it asks for.
• Repairability: how realistic it is to fix when something fails.
• Privacy and user control: who else can see or shape what you do with it, and how much say the owner has.

Why there is no overall winner

A tool is good or poor only for a particular purpose. Where the evidence supports it, a criterion carries a contextual label such as "Better for portability" or "Better for privacy and user control." Where it depends on circumstances, we say so and leave the label empty. We do not add the criteria up into a score, because doing so would hide the tradeoffs that make the comparison worth reading. Conclusions offer picks for particular needs instead of a single verdict.

Sources

We prioritize museums, libraries, archives, universities, standards bodies, and primary documentation such as patents, manuals, and original announcements. Encyclopedias are used sparingly as secondary references. Each article lists the references that support it, with a short note on what each one backs up.

Handling dates

A technology rarely has one date. We separate invention, patent, prototype, commercial launch, and widespread adoption, and we say which one we mean. When sources disagree, or several people worked on the same idea at once, the article says so rather than choosing a favorite.

Handling numbers

We do not publish numbers we cannot tie to a source. Where a figure would be unsupported, we use qualitative wording such as "far larger" or "much slower." We do not present approximations as exact values.

Reference models and periods

Technologies changed over time, and a claim about "the telephone" may be true of one decade and false of another. Each article states the model or period it uses as its reference point, so readers know what is being compared.

Revisions

When an article changes in a meaningful way, the revision is logged with its date on the article. Substantive corrections are described in plain terms; the process is explained on the corrections page, and our broader standards are set out in the editorial policy.