Shulou(Shulou.com)06/02 Report--

This article mainly explains "what are the shortcomings of Python". The content of the explanation is simple and clear, and it is easy to learn and understand. Please follow the editor's train of thought to study and learn what are the shortcomings of Python.

Python is a general programming language released in the 1990s.

Although the syntax of Python is concise, it was not known in the first decade, nor did it really enter the work of developers. Perl is the preferred scripting language and Java has become the mainstream in the field of object-oriented programming. Of course, any language takes time to mature and will be adopted only when it is more suitable for development tasks than existing tools.

For Python, this time first appeared in the early 21st century, when people began to realize that it is easier to learn than Perl and provides interoperability with other languages.

This discovery allows many developers to integrate Python into their applications. The emergence of Django finally led to the demise of Perl, and Python began to get more development opportunities. Although the popularity is not as popular as the newer Java and JavaScript than Python.

Fast forward to now, according to the 2019 StackOverflow developer survey, Python has overtaken Java to become the second most popular programming language.

Python is the fastest growing programming language in the last decade. The popularity of Python is closely related to the emergence of big data in the 1910's and the development of machine learning and artificial intelligence. Enterprises urgently need a low-threshold rapid development language that can help manage large-scale data and scientific computing tasks. Python is well suited for challenging tasks in these areas.

In addition to these advantages, Python is an interpretation language that supports dynamic typing. More importantly, it is supported by Google, and Google has invested in the development of Python's deep learning framework, Tensorflow, which makes Python the preferred language for data analysis, visualization and machine learning.

However, while the demand for machine learning and artificial intelligence is still growing, Python will not always be at the top, and it will grow old one day. Like every programming language, Python has its own weaknesses. These weaknesses make it easy for Python to be replaced by programming languages that better suit the needs of the enterprise. Although there is already an R language, the emergence of newer languages such as Swift, Julia, and Rust actually poses a greater threat to the current king of data science.

Rust is still trying to catch up with the major machine learning communities, so I believe that Swift and Julia will be the languages that will overthrow Python and eventually dominate data science.

Python lacks type safety and is very slow

All good things come at a price, and the dynamic typing feature of Python is no exception. Dynamic typing can make it easier to write code quickly without defining types, but it increases the risk of BUG when the code is running, especially when the size of the contemporary code base increases, making it harder for developers to run code in a production environment. BUG, which is easily discovered by the compiler, is not recognized in Python, which creates development obstacles and eventually slows down the development process of large-scale applications.

To make matters worse, unlike compiled code, Python's interpreter analyzes each line of code when it is executed. Compared with other languages, it not only leads to more overhead, but also greatly reduces the performance of the program.

Julia allows you to avoid these problems. Although it is dynamically typed, it has a real-time compiler. The JIT compiler either generates machine code before execution or compiles using a previously stored cache, which makes Julia perform as well as statically typed languages. More importantly, it has a key feature called multi-dispatch, which is similar to OOPs's function overloading, although it can only be useful when running the program. The power of multi-dispatch is that it can handle different parameter types without creating separate function names or nested if statements. This feature not only helps developers write more compact code, but it is also a victory for numerical computation, because, unlike Python, we can easily extend the solution to handle all types of parameters.

Even better, Swift, as a statically typed language, is highly optimized for its LLVM (low-level virtual machine) compiler. LLVM makes it possible to quickly compile into assembly code, which makes Swift run super efficiently, almost as fast as C language. In addition, Swift has better memory security and management tools, namely automatic reference counting. Unlike the garbage collector, ARC is more deterministic because it reclaims memory as long as the reference count returns to zero.

As compiled languages that provide type annotations, Swift and Julia are faster and more powerful than Python. This alone may be enough for developers to recommend them instead of using older languages and considering many other factors.

Python has limitations in parallelism

If slowness is not an obvious disadvantage of Python, then the limitations of this language in parallel computing are definitely the most obvious.

In short, Python uses GIL (Global interpreter Lock), which improves the performance of a single thread by preventing multiple threads from executing simultaneously. There is a big obstacle to this process: developers cannot use multiple CPU cores for intensive computing.

Here, I must agree that we have done a good job of taking advantage of the interoperability between Python and C / C + + libraries such as Tensorflow and PyTorch. But the Python wrapper does not solve all debugging problems. When examining the underlying code, we still rely on C and C +. In essence, we can't take advantage of Python at a low level, which is what makes Python different.

This factor will soon determine the decline of Python and the rise of Julia and Swift. Julia is a language specifically designed to solve the shortcomings of Python. It mainly provides three features: collaborative programs (asynchronous tasks), multithreading and distributed computing, all of which show that concurrent and parallel programming is infinitely possible. This structure enables Julia to perform scientific calculations and solve big data's problems faster than Python.

On the other hand, Swift has all the tools needed to develop mobile applications and has no problem with parallel computing.

Both Swift and Julia have the interoperability and strong support of Python

Although Python has shortcomings in terms of speed, multithreading, and type safety, it has a huge ecosystem with a large number of libraries and packages. In the face of Python, Swift and Julia are still babies in the field of machine learning, and the number of libraries they have is very limited. However, their interoperability with Python makes up for the lack of library support in Julia and Swift.

Julia not only allows programmers to use Python code (and vice versa), but also supports interoperability with C, R, Java, and almost all major programming languages. This versatility will certainly give the language a good boost and increase its chances of rapid adoption among data scientists.

On the other hand, Swift provides interoperability with Python through the PythonKit library. The big selling point of Swift (which originated from Apple) is that it has strong support from Google, which fully supported Python decades ago.

In addition, the fact that chrislattner, the founder of Swift, is now working on Google's AI brain team shows that Swift is being trained as a replacement for Python in the field of machine learning. The Tensorflow team invested in Swift through the S4TF project, further proof that the language is more than just a wrapper for Python. On the contrary, with its differentiated programming support and low-level working capabilities like C, Swift is likely to be used to replace the underlying deep learning tools.

Thank you for your reading, the above is the content of "what are the shortcomings of Python". After the study of this article, I believe you have a deeper understanding of the shortcomings of Python, and the specific use needs to be verified in practice. Here is, the editor will push for you more related knowledge points of the article, welcome to follow!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.