It's a package for solving symbolic integrals that outperforms many systems. It has an extensive test-suite of over 72,000 integration problems and I encourage you to try how many of them can be solved by SymPy or Julia. Spoiler: I already helped porting it to Symja and if the python community can solve the problems with MatchPy, the port to SymPy can be finished as well.
No, you're not. It's tiresome, especially when people comment who obviously never had to solve hard integrals or pde's to even come up with a model which can then be implemented in an algorithm. I'm absolutely stunned how short-sighted many of the commenters are. I use Mathematica daily and, still, I appreciate Julia, Python, R.
All this talk about data-scientists only using open-source is ridiculous. At my university, departments like medicine, biology, chemistry, and life-science make a huge part. The most commonly used programs there are Excel and SPSS. Other departments use Matlab extensively. In finance, SAP is a big player. All of them are closed source and cost money.
Why can't people just appreciate that it is a huge step for Wolfram to release a free version of its kernel and think that this is an opportunity?
If no one ever criticized restrictions, we wouldn't have the freedom we have today.
Most people here are complaining about the license, not the software. They think the software is good but the license is holding it down. To me this looks like valuable feedback.
Hmm, I see what you mean. I'm under the impression that this isn't the main issue people are having here but what do I know.
One other reason why people might be so upset here is the naming of this product. "Free" is commonly used to describe libre software. This is at best an oversight by the marketing team and at worst deliberately set up to deceive developers.
Your last sentence reminded me of an interview with the Ton Rosendaal, the creator of Blender. He talks extensively about how Blender became OSS and not abandonware. I highly encourage to watch the whole interview (you need at least 5 mins to get used to his accent), but the important part starts around minute 14
I'm not even sure, why this blog post was necessary in the first place. No one argues that a company needs revenue to pay employees and WRI is trying to maximize its profit. That is absolutely fine. However, claiming that OSS cannot achieve great and well-designed software, just to cover the true reasons is simply wrong. Maybe someone should remind WRI that they use Qt, Cairo, Pango, OpenCV and many more OSS libraries under the hood. Qt, in particular, is almost as old as Mathematica itself when the information on Wikipedia is correct.
He acknowledges that: "As I said at the start, the open-source model can work very well in smaller, self-contained subsets of computation where small teams can focus on local design issues. Indeed, the Wolfram Technology stack makes use of and contributes to a number of excellent open-source libraries for specialized tasks, such as..." and lists some
I can only second this. Gonzalez & Woods presents a very good overview of the material in a highly practical and understandable way. Each section contains tons of references into the more specific literature that you can follow as soon as you reached the books limits. Finally, it is a pleasure to read if you are interested in image processing and it keeps you hooked.
For signals (as opposed to images), you should have a look at "Understanding Digital Signal Processing" by Richard G. Lyons (ISBN-13: 978-0131089891). I enjoyed this very much and if you grasped the contents from it, you should be able to understand audio-specific books easily.
I second Lyons. In addition, "Digital Signal Processing: A Practical Guide" by Steven Smith is also quite approachable -- immersion in calculus not required.
"A Digital Signal Processing Primer" by Ken Steiglitz is a nice but rigorous intro to the subject. Written by an EE academic, it's more mathematically rigorous than Smith or Lyons.
Allen Downey's "Think DSP" is also worth a look, though its focus is more conceptual than practical, IMO.
https://rulebasedintegration.org/
It's a package for solving symbolic integrals that outperforms many systems. It has an extensive test-suite of over 72,000 integration problems and I encourage you to try how many of them can be solved by SymPy or Julia. Spoiler: I already helped porting it to Symja and if the python community can solve the problems with MatchPy, the port to SymPy can be finished as well.
Disclaimer: I'm one of its developers