MangaWire opens its doors: here is what you will read

News

MangaWire is a manga news outlet published in English, French and Spanish. Sections, editorial line and sources: here is how we work every day.

MangaWire is live. The idea fits in one sentence: follow manga news, check it, and publish it the same day in English, French and Spanish.

Four sections

Photo: Samuel Regan-Asante / Unsplash.

News covers the day’s announcements: new series, adaptations, series endings, sales figures. Releases lists upcoming volumes, publisher by publisher and month by month. The industry section looks at publishing houses, serialization magazines and platforms. Guides answer the questions you ask before opening a volume: where to start, which order to read in, which edition to pick.

Who writes on MangaWire?

One named author, whose byline appears on every article and links to a profile.

Photo: Alexander Zalan / Unsplash.

That is no small matter in a field where much of the information circulates under pseudonyms or without any byline at all. A signed article commits somebody: it lets you judge a source’s reliability over time, and go back to it when something does not add up.

It also means there is somebody to hold responsible, which is the point of a byline.

The trade-off is accepted. A single-byline site covers fewer subjects than a newsroom, and covers them more slowly. We would rather publish three verified articles a week than ten lifted from elsewhere.

A single byline also has an editorial consequence: the site takes a point of view, that of someone who follows the manga market and says so. We do not pretend to an agency neutrality one person cannot sustain.

The day other bylines are added, they will appear the same way, with a profile and a background you can consult. No article will ever run without an author.

Why cover the market rather than the stories?

Because it is the least occupied angle, and the one that can be checked.

Chapter summaries, series reviews and plot debates are covered abundantly, often very well, by enthusiasts who give them considerable time. We have nothing to add there.

What is missing is the tracking of what surrounds the works: who publishes what, at what pace, at what price, in what format, with what results. That information exists, it is public, and mostly it needs finding, cross-checking and putting into tables.

The angle has a decisive practical advantage: it deals in facts. A box-office figure can be checked, a publication date can be checked, a change of publisher can be checked. We can be wrong, and it shows; that is exactly what makes correction possible.

It has a limit too: not everything interesting is quantifiable. The quality of a translation, the judgement in an adaptation or the worth of a work do not go in a table, and we will not pretend otherwise by treating them as though they did.

How we sort what we publish

Every item gets one of three statuses, and that status appears in the article.

That distinction is the first thing we look at, before deciding whether a subject even interests us. It settles most of the questions that follow: a rumour does not get a declarative headline, reported information names who reported it, and confirmed information links to the original announcement.

Readers sometimes ask why we bother with the middle category at all. The answer is that most manga news lives there: an outlet has something, the publisher has not spoken, and pretending otherwise in either direction would be false.

The status can change, and often does. Something reported on Monday becomes official on Thursday; we then update the article, with the date of the change. The reverse happens too: a rumour that never checked out stays marked as one, and the article is not deleted for it.

One case deserves separate mention: information that is accurate but incomplete. An adaptation confirmed without a date, a studio or a platform is not an imminent release, and presenting it as one would mislead as surely as a rumour. We then set out what is settled and what is not.

What we do not do is worth stating just as plainly. We relay neither scans, nor chapters leaked before publication, nor plot details drawn from them, even when accurate and even when everyone is discussing them. It is a rule without exception, set out in our editorial line.

Those four sections answer different uses. News is read the same day and dates quickly; releases get consulted before buying; industry is read to understand; and guides are built to stay accurate three years out. That difference in shelf life is why we do not treat them alike.

What we hold ourselves to

Every article cites its sources at the bottom, with a direct link. When a piece of information is not confirmed by an official source, the article says so at the top, in a box. When we get something wrong, the correction is dated and stays visible in the article, never quietly erased.

These commitments are not intentions: they are built into how the site runs. An article citing no source cannot be published, the check preceding release rejects it. This is not an editorial policy somebody could set aside on a busy day, it is a technical constraint.

Why three languages rather than one?

Because manga sells across three large markets outside Japan, and an announcement made in Tokyo concerns all of them at the same instant.

Photo: Collin Wigger / Unsplash.

The English-speaking world carries most of the sector’s media coverage, and Japanese information often transits through it before reaching other languages. France is among the world’s leading manga markets relative to population. Spain and Spanish-speaking America form a comparable bloc, with publishers and calendars of their own.

That transit has a cost. An announcement moves from Japanese to English, then from English to French or Spanish, and each stage adds delay and an opportunity to get something wrong. Readers outside English often receive the information a day or two late, and sometimes distorted by loose translation.

The lag is not only about timing. Each market has its own publishers, prices and schedules, and something true in Japan may have no concrete meaning locally for years. Handling the three languages separately lets us say, in each, what an announcement actually changes for the person reading it.

English readers are, in a sense, the least penalised by the chain, since they sit nearest its source. The three-language approach is therefore less about speed here than about consistency: the same facts, checked once, stated three times.

Publishing all three on the same day, from the same sources, removes that chain. It is no technical convenience: it means writing every article three times, with the same facts and the same links, which triples the verification work.

What you will not find here

Better said now, to avoid disappointment.

Photo: Adem AY / Unsplash.

Four things, none of which we will change our minds about quietly.

No chapter reviews as they are published. That format means commenting on material often read before official release, and it occupies ground others already cover well.

No “ten greatest manga of all time” lists. They read well and teach nothing; they exist to be shared, not to be useful.

No headlines phrased as questions when the answer is one word. If we know, we say so in the headline; if we do not know, we do not publish.

No advertising dressed as an article. The day a piece is commissioned or funded by a third party, it will say so at the top, before the first paragraph.

Saying what we do not know also has a practical use. It tells a reader where to keep looking, instead of leaving them with the impression that the question has been settled.

What you will find are figures where they exist, the sources behind them, and an account of what we do not know. That last habit costs us the most readers and is the one we care about most: an article saying “the screen count was not disclosed” is less satisfying than one that invents it, and it is the only one of the two worth reading.

How often do we publish?

Several times a week, on no fixed schedule, and never to fill space.

A news publication is under permanent pressure towards frequency: search engines favour active sites, and readers settle into a rhythm. That pressure pushes you to fill quiet days with reposts, lists and articles that teach nothing.

We would rather accept the quiet days. When nothing verifiable is happening, we do not publish, or we publish a guide, which depends on no news cycle.

The pressure has a side effect that gets little discussion: it pushes you to publish fast and verify afterwards, which is precisely the order that produces mistakes. A publication that commits to a cadence ends up treating verification as the adjustable variable.

The position has a cost we can measure. The sitemaps feeding news aggregators only keep articles under forty-eight hours old, and a pause of a few days drops us out of them. It is a deliberate trade between visibility and usefulness.

How to read an article on this site

Every page follows the same structure, and knowing the markers saves time.

The first paragraph gives the answer. Not context-setting, not a rhetorical question: the fact, with its figure and its date where they exist. A reader in a hurry can stop there and leave with the essentials.

The headline states the fact, not the feeling. You will find no superlatives, no exclamation marks, no phrasing built to create a gap you need to close. If a headline reads flat, it is because it describes exactly what the article contains.

Subheadings are phrased as questions, because that is how questions occur. They let you jump straight to what interests you without reading the whole piece.

Tables gather what compares: dates, prices, formats, publication rhythms. When information fits in a table, we do not dilute it into paragraphs.

At the foot are the sources, with a direct link to each. They are not decorative: they exist so you can check what we assert, and we encourage you to open them when a figure surprises you.

Finally, the update date sits at the top. If it differs from the publication date, the article has been revised, and the nature of the revision is stated at the foot.

How the site is built

Static pages, no advertising and no trackers, and it shows in use.

Photo: Albert Stoynov / Unsplash.

Each page is computed once, at publication, then served as is. There is no database queried on every visit, no third-party script loaded from another domain, no banner to dismiss before reading. An article page weighs around thirty kilobytes and appears in a fraction of a second, including on a slow connection.

The choice has consequences we accept. The site does not know who reads it, which denies us fine-grained statistics and rules out personalising anything. We take that to be the right trade for a reading publication.

The flip side is that a static site can display nothing personalised: no computed recommendations, no server-side reading history, no feed tuned to your tastes. What you see is what everyone sees, and that uniformity is a guarantee of sorts.

Two automated checks run before every publication and block release if they fail. The first verifies the search elements: title lengths, image weights, reciprocity between language versions, validity of internal links. The second refuses any article published without a source.

Those guards do not replace proofreading, and they detect no error of fact. They only ensure that no lapse of form slips through on a day of haste. The Glénat calendar corrected on 20 September illustrates it well: no automated check could tell that a series had been credited to the wrong publisher. Only a human verification catches that, and in this case it came late.

How we choose a subject

Three criteria, applied in that order.

The first is verifiability. If the information rests on nothing consultable, it does not become an article, however interesting. That removes a good share of what circulates on busy days.

The second is usefulness. An announcement changing nothing for an English, French or Spanish-speaking reader does not warrant a piece: a series available only in Japan, with no prospect of translation, is a note rather than a subject.

Those first two criteria mostly discard what arrives through social media with no clear provenance, which is also what circulates most.

The third is angle. Much is already covered elsewhere, and repeating it identically adds nothing. We look for what others have not handled: a figure put in perspective, a calendar reconstructed, a publishing mechanism explained.

A subject clearing all three gets written. A subject clearing only two waits, and sometimes never comes.

That filter discards considerably more than it lets through, which is the intention. Publishing less allows better verification, and a site that publishes little but accurately becomes, over time, more useful than one that is exhaustive and approximate.

What to do with announcements you see elsewhere

Sort them, and it takes under a minute.

The rule that settles most cases fits in a question: where did this information come from? A post linking to nothing checkable is not an announcement, it is an assertion. A screenshot with no link proves nothing, and an account followed by hundreds of thousands is no more reliable than another if it never cites its sources.

We have set that method out in a guide devoted to reading announcements, with the phrasings that should raise a flag and the checks to make. It applies to any licence and any publisher.

The habit generalises well beyond manga. Anything arriving through a feed, on any subject, submits to the same question about where it came from.

It is probably the most useful thing on the site, and it is also what makes our own articles checkable: the habits it describes apply to what we publish as much as to anything else.

What now?

The site is live, the sections are open, and the rest depends on what the news brings.

Three efforts are under way. Release calendars, publisher by publisher, which require regular collation. Guides, which build slowly because they have to stay accurate for years. And tracking market figures, the most demanding of the three, since public data is scarce and scattered.

Calendars are the least glamorous of the three and probably the most used. Nobody shares a release calendar, and everybody checks one before buying.

Longer term, the aim is a consultable archive: calendars going back in time, guides kept current, runs of figures comparable year to year. A news publication is judged on its speed; a useful site is judged on what you can still find in it two years later.

You can follow publications through the RSS feed, which carries every article as it goes up, with no account and no sign-up.

For any remark, correction or subject suggestion, the contact page remains the best route. Error reports are handled first: it is the one category of message we answer without fail.

One last word on what this site is not: an exhaustive source. We will never cover everything that publishes, and that is not the goal. Thirty solid articles a year beat three hundred nobody stands behind.

What happens when we get something wrong?

The correction is made in the text, a dated note is added, and the article stays online.

The procedure is set out in our corrections policy, and it is worth explaining here because it runs against instinct.

A visible correction is also an argument in itself. A site that never shows errors is not a site that makes none: it is one that does not say so, and a reader has no way of telling the two apart.

The instinct is to amend quietly, or to delete. That is understandable and it is a fault: a reader who shared the wrong article never learns it was wrong, and nobody can check what we wrote yesterday.

So we do the opposite. The text is corrected, the update date changes, a note at the foot says what was altered and why, and all three language versions are revised together. The error stays legible.

The arrangement has already been used. A release calendar published in August carried wrong publisher attributions; it was rebuilt entirely on 20 September, with the corresponding note. We would rather that be visible than imply a site that never errs.

It has a limit worth admitting: it does not repair everything. A reader who saw the wrong article and will not return keeps the false information, and no correction note reaches them. Which is why verification before publication counts for more than correction after.

Three languages, one thread

Every article exists in all three languages, with the same publication date and the same sources. The language switcher at the top always leads to the translation of the article you are reading, not to the home page.