I am a big fan of Mathematica, but see this response from my latest rejection:
Indeed the use of Mathematica is unfortunate. A user can indeed read
files in this expensive proprietary software, but they cant actually
use your developed formulae without buying the software. An important
remit of [Journal name - the journal did not want to be cited] is the
uptake of methodology by the community and this barrier significantly
reduces the potential impact of your work. I believe this was the
critical point the previous reviewer was making and success of your
resubmission hinged upon reformatting your work into a type of
user-friendly application.
One thing I can point out is that, with certain inputs, the CDF Player will be perfectly adequate for testing your code. If the purpose of the article is to present a scientific result, then that should be sufficient for replication. If the purpose is to introduce a package you created as a tool for others in your field, then it becomes a thornier issue. At which point you might want to see if other similar articles published therein make use of commercial software. If not, you might try to configure your package so that it can always be run in the CDF Player with arbitrary inputs, and moreover does not violate any licensing for Player usage (e.g. it may need to be noncommercial; check the license on this).
I’ll add that I also on occasion get hit with a certain software bias, I’ll call it, from anonymous referees.
Christian, I would prefer to say great opportunities ahead for Mathematica!
If you look at a site such as arXiv.org I don’t think you’ll see any CDF files available. In the future I believe you’ll see a lot of them and eventually they will replace PDF as the dominant form of technical document. Why? Because, as you probably well know, with their active and interactive capabilities they are far superior to other documents. But it will take time and at first you may have to do an end around, such as make them available from your own web site. People have to see them but once they see them, what’s not to like about them? Of course it will help if the initial examples are well written, perhaps mirroring a paper in interactive form.
Any serious paper is likely to have associated customized calculation and display routines. It is possible to transmit applications with packages to readers and they can install these in the Wolfram Research/Wolfram CDF Player/10.0/AddOns/Applications folder. Then the CDF document can use your packages. I’m not certain, however if this is approved by WRI. You can’t just unzip into that location. You have to copy the folder and give Administrator permission. It would be helpful if the free player created a $UserBaseDirectory where readers could place any additional packages outside of the Wolfram files.
Another problem is, if you have Mathematica installed, to get the player to open a CDF file in the free player, without access to Mathematica applications and packages, to test what the experience would be for a reader without Mathematica.
I suspect there are a number of tune-ups needed to make this a smooth experience for everybody.
Mathematica Cloud is another approach but I just don’t know enough about it yet. WRI really doesn’t give enough specific information, especially as regards developer packages.