While December is a short month, nevertheless we have quite some progress to report.

GIN now has an inbuilt JSON viewer/editor based on the amazing JSON Editor Online from Jos the Jong. Just click on a .json file and you'll see it coming :-).

But what if we had not provided a link to a JSON file above? How could you have found one? That's where the real news is.

We are happy to announce that we have finally launched the GIN Data Search Facilities, which enable you, among other things, to search for files matching some criterion. For example, you could have searched for all JSON files with a query search and the query: "Path:*.json". But that's just the beginning. Basically, you can search for everything now, including the content of files. Try it out with a search for Spike Sorting or Electric Fish. The data search can be found under the explore tab and is equipped with the ability for wildcard and fuzzy searching. Complex search queries are also supported out of the Box!

The technology behind the search we call gindex. Gindex is based on modern search technologies like elasticsearch and it basically analyses repositories for content types it knows (Text, XML, PDF, JSON, odML, NEV to name a few ). From there, it builds a representation that is good for (full-text) searching. You can find more details and documentation here.

As a special Christmas gift, we have also added a Commit search which might make your data organization a bit more flexible :-).

With that, we wish everyone a Merry Christmas and the proverbial happy New Year!!

November 2017

Busy autumn for GIN!

Many things are moving forward for GIN this month and there is more news coming soon, but let's move to today's topic.

GIN @SFN2017
Of course, we have been present at the SFN meeting and presented our service (see our poster ). We are grateful to the "Neuroscience in Germany" for hosting us and our presentation!

Specifically for this event, we introduced our brand new ARM docker image! We did not want to bring a big server machine to the conference (well you know they are not exactly hand luggage) and therefore, we looked for a somewhat smaller and more mobile solution! Guess what? GIN performs nicely on a Raspberry Pi which, besides being a little bit of a toy deployment for us, is actually nicely illustrating GINs low resource footprint!

There have also been a lot of changes on the server side and quite a bit of our day to day housekeeping has been streamlined. We have moved our weblogs analysis from Webalizer to live-logging with GoAccess. We also migrated our monitoring (well we duplicated it, at least for the time being...) from Nagios to a combination of telegraf, influxDB and Grafana. So at "mission control", which is where we keep an eye (and a notification service) on the GIN-Server metrics, everything is new and shiny!

Finally, Louis has joined the GIN team and will soon help us improving our service with new tutorials and better usage instructions. Welcome to the team!

October 2017

October has been a busy month for GIN and there have been a number of updates:

First, long-awaited features have been finalized and added to the gin command line client. Foremost of these is that an automated criterion for files to be git-annexed has been introduced, dependent on their size (the default threshold is at 10MB), smaller files are added to git directly. This enhances diffing, patching, direct downloads etc. Also, a lot of smaller things have been added. As always the newest version can be found here

GIN-Web has been updated with small help messages. Also, we have changed the name of the DOI request file to datacite.yml to increase transparency. We took this opportunity to redesign the DOI UI a bit. For modern browsers code highlighting has been moved to web workers and finally, a lot of tutorial text and imagery has been added.

This brings us to the next point. Kateryna has joined the GIN team and will be working on Tutorials, user instructions, community building, and UI improvements. Welcome to the team!

Finally, we are excited to announce that G-Node is now listed as a recommended data store on the Scientific Data repositories list and on Springer Nature. We hope this will further our mission - to bring easier data sharing and reproducibility to scientist from neuroscience, and the general scientific community!

Christian Garbers edited this page 4 weeks ago