Textile SchoolPublishing since 2009Fibre to finished garment
The working reference for the textile industry.
For the student learning to spin, the designer choosing a finish, the merchandiser costing an order and the mill running the machine. Look something up here and you are told where the answer came from, when it was last checked against that source, and what it does not cover.
A reference, not a magazine. The larger part of this site is now cited records rather than articles, and every claim on this page is counted from the catalogue as it renders.
It began in 2009 because finding a straight answer on fibre properties, weaving or garment production meant sifting through scattered sources. That has not changed, and it is still the whole job — the difference is that an answer here now arrives with its evidence attached. The scope runs the full spectrum, fibre to fashion, and past clothing into medical textiles, automotive upholstery and home furnishings.
What is here
Six things, and the question each one answers
3,038 test methods
Standards
The buyer named an ISO method, the mill’s laboratory reported an Indian one, the third-party lab used the American one. Are they the same test — and if not, what changed?
516 tools
Calculators
Not a record but a result. Put your own figures in and read the answer, with the formula, a worked example and the standard it rests on printed underneath.
1,459 models
Machinery
What a machine actually runs at, taken off the builder’s own published datasheet and cited back to the page it was read from.
306 series
Trade events
When the show runs and where. The series is the page and the edition is a property of it, so the page does not die the day the hall empties.
91 instruments
Compliance
Which obligation lands on which date, on whom, under which law — and whether that date is enacted or still only indicated.
806 in the archive
Articles
17 years of explanation — spinning, weaving, knitting, dyeing, garmenting — written before the registers existed and kept because a register cannot introduce a subject.
How it works
What a record has to carry, and what happens to it afterwards
An article is finished the day it is published. A record never is — and the four steps below are the whole difference between them.
Somebody reads the source and enters the facts
There is no ingestion pipeline behind the registers, because there is no uniform source to ingest: one machinery builder publishes clean HTML tables, the next puts its figures in prose, a third serves a consent wall to anything automated. So a person reads a published datasheet, a standards catalogue or an official journal, and the record cites the page and the date it was read.
source URL required
It publishes marked provisional until the editorial desk signs it
Unsigned is not the same claim as unchecked-in-a-while, and conflating the two cost this site its search presence for as long as the registers existed. A record nobody has vouched for says so at the top of its own page, at full size, in plain language.
says so on the page
The signature is a date and a source, not a status
Signing records the document the record was checked against, when, and when it must be checked again — a review interval set by how fast that kind of fact moves. A trade-fair date and a test-method number do not rot at the same speed.
verification date is the byline
Past its review date it removes itself from the index
A record that has gone stale is withdrawn from search rather than left sitting there looking current. That is the difference between this and an article: an article is finished the day it goes up, and nothing about it obliges anybody to look again.
de-indexes itself
What the site refuses to publish
Every site says it is committed to accuracy. These are the commitments written into the software, so an editor in a hurry cannot get past them and neither can we.
Any standard’s text
A record carries the number, title, scope, edition and equivalences — never the procedure, the apparatus or the rating scale. That text is copyrighted and you buy it from the body that wrote it. We will tell you which body that is.
A modified equivalence that does not say what was modified
ISO/IEC Guide 21 gives three codes and they decide whether a laboratory report is acceptable. A MOD record with an empty deviation note will not publish: accepting a modified-equivalent report without knowing what changed is where the money is lost.
An obligation with no official source behind it
No compliance record goes live without a URL that resolves to the instrument itself, and prose that reads as advice is refused. We record what a law says and date the check. We will not tell you whether it applies to you.
A date presented as settled when it is not
Half of what matters in compliance is indicated rather than enacted, and delegated acts slip. A date without a confidence marker beside it cannot be saved at all.
A figure that cannot be pointed back at a document
The price series stores what a source published, as published, with the source and the observation date on every row. Nothing derived is stored: a cross rate is computed on read from the two rates actually quoted.
An answer to “does this apply to me”
Not a rule the software enforces — a line the site does not cross. Interpretation is what your counsel, your certifier and your laboratory are for. What we owe you is the fact, the source and the date.
Where it came from
Four dates, and where each one is known from
Three of these are read out of the catalogue as this page renders. The first cannot be — the site is older than the oldest article still standing on it — so it is dated from an outside record and the record is linked.
2009
The site goes online
Started because finding a straight answer on fibre properties, weaving or garment production meant sifting through scattered sources — under the line it carried then, “where the textile industry assembles to learn”. This is the one date here not counted from the catalogue: the archive of the site does not reach back to its own beginning, so it is dated from an outside record instead, and cited.
Oct 2010
The oldest article still standing
Everything before it has been rewritten or retired, which is why the catalogue cannot date the founding. For the decade and a half after it the site was what that first line made it: things written to be read once.
Aug 2026
The first calculator
The first page here that produced an answer instead of a paragraph, and the first that had to be right rather than readable.
Aug 2026
The registers open
Every register the site keeps opened within 9 days of the first record. By the end of them the reference held more cited records than every year of publishing before it had produced articles.
Where it stands
Most of the catalogue has not yet been signed by a human
4,164 of 4,934 signed
The other 770 say provisional on their own page, at the top, at full size. They were drafted from published catalogue metadata and nobody has yet checked them against the source.
A signature is a claim that a record was checked against a specific document on a specific date, and on its own it is not enough. A signed record must also hold something worth reading: a scope, an equivalent in another standard, a calculator that implements it, or real specifications. One that does not stays out of the search index until it is filled in. So the backlog is real, it is visible on every page it affects, and it comes down at the speed a person can review.
Two more things we would rather you heard here than discovered. A source URL that returns a page is not proof that the page is the source, so that is tested separately instead of trusting a status code. And a machinery record naming a model but holding none of its figures stays out of the search index, because a page with nothing on it should not be in a result.
What it is for next
Four objectives, each with a number that says whether it worked
Sign the backlog
Every provisional record reviewed against its source by a person. This is the slowest thing on the list and the only one that cannot be automated, which is exactly why it is first.
measured on this page
Keep logging the series that cannot be back-filled
Prices are the one thing here nobody can catch up on. A competitor starting in 2028 can never hold a figure observed in 2026, so the collector runs every day whether or not anybody is reading it — and the derived numbers, like a spinning margin, exist only because both legs are kept.
daily, permanently
Wire the archive into the registers
A standard should cite the article that explains it; an article should point at the calculator that does its arithmetic. Sixteen years of explanation and four registers of fact are more useful joined than stacked, and joining them adds no pages.
no new URLs
Open a register only where the facts are public and stable
A laboratory directory was specified, assessed and declined: thousands of entries that change constantly, with no reliable public source. Machinery survived the same test because a specification does not change after the model ships. The next register will be argued the same way, in public, before it is built.
declined is a result
Tell us we are wrong
A register is only as good as the corrections it attracts
The person best placed to catch an error is the one who came here already knowing the answer. There is a route for each kind.
A machine of yours is wrong, or missing
Builders and owners can send the published figures for a model. A record that gains its specification gains the right to be indexed.
A show is listed with the wrong dates
Organisers can correct an edition or add one. We will not assert a date nobody published.
A standard, an obligation or a calculation is wrong
Name the record and say what the source actually says. A correction with a citation is acted on; one without becomes research.
You want to write something
Still open, with a clearer brief than before: explain a thing the registers can only record.