Launched in August 2020, Cultural Japan. More than a million items on Japanese Culture, in a user-friendly interface.

On searching it becomes apparently that it’s an aggregator. A search for cat for instance, brings in results by Japanese artists and photographers from all over the Anglosphere and Europe, whereas one might have expected it to draw on Japanese digitised collections. Only one result came from Japan, from the University of Tokyo General Library. However, it appears this is simply an artefact of my search. If I search for cat using the Japanese Kanji 猫, then the search results are indeed a mix.

It seems no large versions of pictures are on-site, and the searcher must follow an off-site link and hope for the best, negotiating various links and viewer types to try to get a large version.

In the case of a test picture, “Scouts near Niu-chang on a snowy night”, after about four pass-throughs the largest version at the final landing site was 800px. However, a quick Google Search for the same picture’s title “as a phrase” found a large version via Google Arts & Culture. This was not picked up by Cultural Japan. This is worth remembering if you get to the end of the click-chain and still only have a small picture.

It appears that the DOAJ is now accepting journals that are only ‘open’ in the sense that they’re available to institutions and members via JSTOR. e.g. four or five Berghahn Books titles that were listed today.

The Berghahn Books blog has no announcement I can find that these titles are about to go open access, and Google appears to know nothing either. Their only fully OA journal title appears to be Anthropology in Action, added to the DOAJ in May 2018.

New from Microsoft, a “Transcribe in Word” service. Upload an audio file of less than 200Mb, press play for transcription to begin. The words appear as if being typed into the Office 365 online version of Word (only). What makes it interesting is that it apparently identifies different speakers, and can also work in real-time when at conferences, lectures etc. Users can transcribe five hours per month from their uploaded files.

Microsoft is rather playing “catch up” here, and the market leader is StreamText offering real-time no-lag transcription at 9 cents per minute. Such rapidly advancing tech likely means it’s not a moment in time to be training as a sign-language interpreter for classrooms and large events. Although I guess heavy accents and mumbling professors may still require it.

GreasyFork is now accepting “Stylus format user CSS” as well as UserScripts.