https://www.youtube.com/watch?v=TSlUFJKqQCA&feature=youtu.be

This is an interesting interview with the Michelson developer for Tezos. Tezos is one of the main competitors against Ethereum and is in my opinion a serious competitor. Michelson has the potential to be a far superior programming language over Solidity which in my honest opinion is extremely challenging even for experienced programmers. I don't work much with Solidity particularly because it's difficult to write secure code in such a new language with obvious design flaws. 

That being said, it is not necessarily true that Michaelson will be more secure than Solidity out of the gate. It all depends on the decisions made by the language designer and the methodology of the developers who create the initial development kit, libraries, etc. 

My thoughts on the design of secure programming environments
---

I use particularly the phrase "secure programming environment" because it's not enough to simply provide a language. Even if for sake of argument Solidity were secure, it is not true that the code written under Solidity or the development environment is of enough quality to help developers produce secure code. The environment typically for functional programming languages is more security oriented in general but once again any new language is likely to have gaps, holes, mistakes, which can take time to fix.

The most important aspect I can see as a developer is the need for secure libraries. EOS to it's credit seems focused on providing an operating system for developers which is actually what developers need. In my opinion I would be happy with a nice development kit and a secure set of libraries to handle the most common algorithms, cryptography, etc. It should not be left up to each programmer to have to reinvent the wheel in order to write a DHT, or a blockchain, or a multisig, or a public key cryptography implementation for secure messaging. These very basic and often used "recipes" should be written securely once by the most skilled developers in the space and then the best practice should be to simply use the libraries which have been tested over a long period of time.

In a language like C++ we have libraries like this which is why it's possible to develop something like Graphene. In Solidity we do not have something like this which is why it's hard even for people with decades of programming experience to create a DAC, or DAO, or secure messaging, because in essence it's trial and error and mistakes happen. To reduce the rate of error requires relying on best practices (coding practices), and on standardized secure libraries which even newbie developers can make use of and all developers can trust. I hope Tezos provides tight, high quality, easy to use, secure and standardized libraries to encourage these best practices.