Publish on autopilot
After months of development and extensive testing, changing concepts and approaches, we’re finally getting ready for our private beta launch of flowli! The public beta launch will probably happen sometime early next year.
The last few months have been extremely interesting and challenging. Naturally, when we set off to develop flowli – we wanted our CPP (Content Publishing Platform) to be extremely easy to use, whether you’re a designer, programmer, or end user.
To achieve this, we’ve introduced and then rejected quite a few approaches: we started using MS SQL Database for storing the data, but have since replaced it with a document-oriented database (MongoDB). We started with a rather inflexible file system based structure that we’ve now canned in favour of categories and tags. We started with a traditional tree-view hierarchy that has now been replaced with templates, widgets, categories and tags.
Okay, you’ve probably heard of all these individual bits before, but their unique combination makes flowli not only into something that is truly different to all the content management systems crowding the market, but also something that is far more flexible. Throughout the development we stuck to our simple aim to create a platform that is amazingly easy to use, whether you’re building a small microsite or a huge, complex website with tons of related articles, categories, listings, media etc. (we’re actually about to power such a beast with flowli; more further down).
Some of the ideas and concepts we introduced for flowli have also been recently promoted and published by Microsoft (eg. Microsoft Razor, .NET 4 templating). Our own version of Razor and templating system will make it easy to write any type of widget, as long as one knows some HTML, CSS and English. Furthermore, populating widgets (and templates) with content couldn’t be more straightforward: tell the widget which category, tag, article or media file you want displayed and it ‘pulls in’ all matching content automatically. The same is true, for example, if you want list a number of articles: just tell the system which tag (or category) to ‘pull in’ and all relevant texts will be displayed, following the structure and design of your template (HTML and CSS).
Yesterday, I demonstrated the latest flowli backend to one of our clients, Project Syndicate, who will soon be using it to manage their website with 1000s of articles, comments, user accounts, blogs, videos, podcasts and about everything else you can imagine. Their first reaction was “where are all the tabs and pages, where is all the functionality?” This was a proof to me that we succeeded in gutting every ounce of unneeded UI, presenting the user with a (perhaps surprisingly) minimal control surface. Everything content-related is happening in the flowli library, where one doesn’t only see the raw texts, images and documents, but (if this is enabled) also an article’s comments, for example. All the templates are managed in the template editor, and the way the templates are displayed and populated with content in what we call ‘flow control’. One final tab for a few admin settings – that’s it. Everything else happens under the hood and is automatic (but obviously can be overwritten in templates and through the API).
In the coming days and weeks we’ll be sharing some more information about the system and, if all goes well, will publish a number of websites powered by flowli.
