I tried it out on a few comments from the long google memo articles, and the response was good. Specially sentiment, and reading difficulty level.

One thing that was annoying was marking the word "very" as a weasel word -- in the instances it seemed correctly used. "Controversial" vs "very controversial".

Also, modern convention of using straight quotes ("") has become the norm. Marking these for change to curly quotes is just fighting against a change that's already occurred in written english -- alteast for phrases and not spoken sentences.

Good work! I'd especially like for this to be an offline-checker that the browser provides for all textboxes (the way spellcheck is today provided).

For professional typesetters, straight quotes are not the norm. Proper quotes are still expected. However, if you are not a typesetter, you don't really need to worry about it.

> A a sensitive, unobtrusive and reliable tool for any writer, newspaper editor, teacher, blogger or student

> A a

I want with my whole heart to believe this was intentional.

Oh no, it was not. I just uploaded that page last night. Good catch, thank you.

more:

> Typley is more than just a proofreading tool. It's a complete writing environment.

I can sort of buy that proper grammar-checking requires "human-level intelligence", but how is it possible to detect illogic, for instance, without the same? Typely would have to understand an argument's structure and actual meaning to prove logical soundness. Doesn't that require human-level intelligence, or at least as much as with a grammar-checking task?

Typely is actually a huge list of regexes carefully targeted at exact bits. We don't do any text analysis beyond that and we try to stay away from being "too clever" which is a fine line that we don't want to cross.

www.grammarly.com does quite a good job.

No EULA or privacy statement on the website. Usually with these free services there is something in about using your data for profit.

Honestly, it's a "one man show" and I'm trying to do what I can with my time. The backend of this thing took so long that I pushed everything else too far and too early but I'm recovering.

I will add an eula and privacy. There is absolutely no data being stored or sent anywhere else. You will have to trust my words until I manage to upload those pages. I don't even have a marketing strategy yet, let alone selling "your data for profit". This tool was created for me and a short list of friends. Non of us are native English speakers/writers and I made typely to aid in writing.

Thank you for taking the time and I hope Typely can prove itself useful to you.

Not to say you don't need the EULA but the fact that it's a one man show has won my trust already.

>Honestly,...

Strangely enough, your post, copied and pasted "verbatim" in Typely gave (besides "curly quotes"):

EASY TO READ Reading ease

FOURTH GRADE Grade level

RICH Vocabulary

NEGATIVE Sentiment analysis

I would personally rate it as "very positive" in sentiment ...

Heck!, I added the above sentence and also my "very" was tagged as "weasel word"!

Since we don't have neutral sentiment baked in, it probably couldn't rate it as being positive because my post is, somewhat, apologizing.

It is easy to read and it is understandable by a fourth grader. The vocabulary was being reported as "Rich" due to the fact it's just a comment and does not repeat many "same" words. The story changes once you keep writing and the vocabulary will go down another level.

I have "in the works" some code to make the engine skip analysis for words inside quotes because of the clear indication of quoting someone else.

Does this have an api / plans to have an api / way to access it programmatically, or perhaps some of the tools are open source?

I've been building a spam/scam filter as a hobby, and some of these metrics might be useful to rank messages by spam probability. The spam bots aren't very sophisticated so I've been able to make very specific targeted metrics to make a spam rating without having to think about advanced spam detection, but I'd love to be able to have access to "reading level" and "vocabulary" metrics and stuff that I'm not inclined to make from the ground up on my own.

After visiting the page, with a typo, its still unclear to me what it does.

It's a tool that analyzes your text against a huge list of common mistakes writers usually do. The typos are fixed and pushed, CI is baking them now.

They call it a 'proofreader' but it doesn't actually do proofreading. It's more like a copyeditor, if the copyeditor's job was to be like a high-school English teacher.

From the sample loaded text:

> "Me and Brian got scared" > "Brian and me is the correct form"

This isn't true. "Brian and I got scared" is the correct form.

Point taken. My goal was to stop people from starting their sentence using 'Me and ...'. I guess I didn't give it enough thought and simply swapped the words.