Some of the bigger changes are documented in the Release Notes (see "What's New" in the Help system), but there are also many small usability improvements, particularly to the management of input data sets.
Allocations and the new Application Mode have come to Ventity. For more information on these in-development items check the Release Notes in the Help. We're eager to hear your feedback on these elements in particular so we can make Ventity the best software it can be.
In the beta, core language features and the basic user interface are now working well, but it's far from feature-complete. Our current goal is to get Ventity up to speed for the kind of modeling we (Ventana) do on commercial projects. We think we're close, but we still need time series data support, optimization, and some support for user interface building (sliders, buttons, etc.). We're very interested to hear what you would like to see next - please drop us a note at firstname.lastname@example.org or this address.
You can download a copy of Ventity from
Feel free to share the link with your colleagues. Download approval is now automatic and fast. Note that you don't have to fill out all the fields in the form.
If you're new to Ventity:
The Help system contains a lot of good "getting started" material, including a series of videos, which are also on Vimeo, starting with
The Ventity support forum is up and running at
http://www.ventanasystems.co.uk/forum/v ... c.php?f=48
What’s unique about Ventity? The really big change from most simulation languages is that we've replaced arrays with lists of entities, identified by attributes. This makes for a more natural and data-friendly problem description. Detail becomes a property of entities, not equations, which makes it a lot easier to change. Entities are modular, so they're easier to reuse and share among teams. Entities can be created and deleted dynamically, so there's no need to simulate a lot of zeroes for things that haven't happened yet. Similarly, a list can represent a sparse matrix, so there's no need to clutter your model with zeroes for project phases that aren't prerequisites or relationships that don't exist among people. All this happens in a development environment with modern graphics and windowing.
At the last beta release, we were just embarking on a project that really needs Ventity:
We’re now close to solving this problem. We can already build the model in Ventity, and its entity lists are much easier to work with than arrays. What we still lack is time series data input and optimization for model calibration. We expect the next major beta release to include these capabilities, at which point we’ll be able to do industrial-strength work.Here's an example that's just come up. A consumer products firm would like to model a portfolio of products across many regional markets, distribution channels, brands and package sizes. Not all of the brands are present in each market, and package options differ for each brand. We've done this many times in Vensim, but it takes a lot of data work to standardize things to work with arrays. With entities, it'll be easy - we'll create a Product entity identified by Market, Channel, Brand and Package attributes, and we simply won't bother to create combinations that don't exist. Later, when new products launch, we can add them to the model on the fly.
We think we're just scratching the surface of what's possible. That said, there are still some things that Vensim is really good at, like extreme speed, advanced algorithms and data connectivity, so we're continuing its development in parallel. We've rebuilt the infrastructure for 64bit and Mac builds lately, which puts us on a solid footing for future upgrades.One set of changes in the works will improve data-friendliness, in the spirit of what we’re doing with Ventity. For example, there are new formats that are friendly to relational data import and export to business intelligence visualization tools like Tableau.