Updates

Until 31 October, the first credits you buy are matched. Add fifty dollars' worth and a hundred dollars' worth arrives; add two hundred and four hundred does. There is no ceiling and no code to enter — it happens on its own the first time you buy, once per account, and the button tells you the doubled figure before you pay.

Nothing else about credits has changed: what a review costs, the estimate shown before every run, and the ten credits every new account starts with are all as they were.

Almost everything here came from what you told us in the first days. Thank you for writing in — several of you described the same two frustrations, and this round is mostly those.

You can now talk to the reviewer. Press Ask on any comment and a conversation opens beside the review. Ask what the comment actually means, how you would fix it, whether it is fair, or what you would say to an editor who raised it. The reviewer answers with your whole manuscript and its own whole review in front of it, so you can take the conversation anywhere in the paper.

It can also go and look things up while it answers: search the literature, check a citation against your own bibliography, read a page on the web, and look at one of your figures. Where it rests a claim on a source, the source is listed under the answer. Where it is giving you its own judgement, there is no source, and you should read it as judgement.

Each comment keeps its own conversation, and so does each figure, so a question about your identification strategy does not end up in the thread about your abstract. Every thread stays with the manuscript, and you can choose how carefully each answer is worked out — the same three depths the review uses. What an answer costs is shown underneath it.

A concise review, for when the full one is too much. Several of you said the review said too much to act on. Next to the depth setting there is now a choice between Extensive and Concise. Concise reads your paper exactly as thoroughly — the same passes, the same tools — and then does one more thing: it goes back over everything the readers found, merges the points that are really the same point arrived at by different routes, and keeps the ten that bear most on your central claims. Nothing is thrown away; a smaller point survives inside the comment it belongs to. It is fewer comments, not thinner ones.

Figures. The reviewer looks at every figure in the manuscript now, not only the ones it could place from the section headings, and you can open a conversation about any of them.

A handful of smaller repairs came from your reports too: mathematics in a quoted passage is set properly rather than printed as code, the conversation stays where you left it when you reload the page, and the review and the conversation now call a figure by the same name.

We are live! Give your manuscript a try before going to journals. May you be blessed in the battle of minds!