This is the new JP ⇔ EN search engine for The searching concept has been drastically simplified, and the database has been redesigned to make searching a lot faster, so hopefully things feel more responsive, as well as more intuitive.

While you are typing your search term, an autoconversion suggestion box will let you pick alternative scripts for your search term: if you're typing western letters, anything that counts as romanised Japanese will show hiragana and/or katakana equivalent phrases to search with instead, and typing hiragana or katakana will give you the equivalent romanised terms.

Input conversion follows the same conventions as most IMEs, although for romanisation a system based on mirroring pronunciation is used. This means that you should be able to read the phonetic guide text for found entries (based on English pronunciation rules) and get it mostly right. For instance, eventhough "ti" is converted to ち when performing a search, it will be romanised as "chi", because that's its pronunciation based on the rules for written English.

The only wildcard currently available is the asterisk, *, which counts as "zero or more letters go here" wildcard. There are presets for start/substring/end searches after you searched for a term to widen your search, but if you want to search for word starts, substrings or word ends without having to first search normally, just use term*, *term* or *term respectively, and it effects the same thing.

Note that for English word searches there is currently no "this, and only this" search. If you search for "tiger", for instance, you will find anything that has the word "tiger" as bounded word, such as "There's a tiger in the market" or "tiger prawn". This seems more userfriendly than ruling all those kinds of hits out.

Virtually all entries are fully crosslinked, so clicking on any word in an entry that changes color when you mouseover will cause a search for that term. Clicking on the entry background, or on a word that does not change color, will load the detailed information about it in the right panel. Currently that means detailed explanations for kanji, and if the entry is a verb, all common conjugations for it.

Information blurp

This is a work-in-progress website redesign for and associated satelite sites. For the moment the focus will be on the base site, and making searching as easy as possible.

The new search box is an omnisearch, which means that it will search all the databases for results whenever it makes sense to do so. It will also search in all possible languages when it makes sense to do so, but to help you rule out undesired results the omnisearch offers on-the-fly input conversion so that you can pick a Japanese phrase to rule out English searches.

The landing page is kept as minimal as possible, to keep searches clean.

This box will show related information while browsing through the search results, such as information on kanji used in words, related terms and example phrases.

The styling is not fixed yet, and it's still very early days. I'll probably use different fonts for different aspects of the site, the futuristic looking font is mostly for presentation purposes, not for running text, for instance.

- Pomax 

New features
  • Everything is cross-linked. Simply hover your mouse over something, and if it lights up you can click on it to search the dictionaries with it.
  • Clicking an entry shows the details for all kanji used in that entry, so there is no need to separately look up kanji in a word anymore.
  • Input language auto-detection has been replaced with "search everything that makes sense" logic instead. Searching for "envelope" will search as both English and kana, for instance, because it can be turned into エンヴェォペ.
  • Convert-as-you-type romaji to kana in the search box. You can use your cursor keys to traverse the suggested kana conversions, or click on them with the mouse, or if you want to get rid of the conversion box, hit ESC.
  • Romanisation "furigana" for entry readings.
  • Much better database result caching. The first search may take a few seconds, any time after that, it'll take roughly 0.008s to fetch the results from cache.
  • new content management, with better support for noties, announcements, on-the-side blog entries, and comments
  • Login-based preferences. Simply sign up, log in, and customise the way you want things to look and feel. Being logged in means you'll also be able to post comments simply by typing your comments and hitting "post".

Post a comment