Getting pyinstaller 3.4 to work with Python 3.7

Python is an excellent language, but given that it’s free, it also comes with a lot of conspicuous loose-ends that you will not expect in commercially supported platforms like MATLAB.

Don’t expect everything to work right out of the box in Python. Everything is like 98% there, with the last 2% frustrate the heck out of you when you are rushing to get from point A to point B and you have to iron out a few dozen kinks before you can really start working.

When I tried use pyinstaller (v3.4) to compile my Python (v3.7) program into an executable, I ended up having to jump through a bunch of hoops:

  • pip install pyinstaller gives:
    ModuleNotFoundError: No module named 'cffi'
  • Then I looked up and installed cffi
    pip install cffi
  • After the dependency was addressed manually (it shouldn’t )  pip install pyinstaller worked
  • Then I tried to compile my first Python executable with pyinstaller, and I got this exception:
    File "C:\Python37\lib\site-packages\win32ctypes\core\cffi\", line 198
        SyntaxError: invalid syntax
  • I searched the exact string and learned that pyinstaller (v3.4) is not ready for Python 3.7 yet! How come pip installer didn’t check for it? I opened up the offending file and looked for line 198 and saw this:
    c_creds.CredentialBlobSize = \
        ffi.sizeof(blob_data) - ffi.sizeof('wchar_t')

    It’s a freaking line continuation character \ (actually the extraneous CR before CRLF) that rooster-blocked it.

  • I just deleted the line continuation and merged the two lines, and saved, then I was able to compile my Python v3.7 code (using pyinstaller 3.4) with no issues.

This is not something you’ll experience as a MATLAB user. The same company, TMW, wrote the MATLAB compiler as well as the rest. The toolbox/packages are released together in one piece so breaking changes that causes failure for the most obvious use case are caught before they get out of the door.

Another example of breaking changes that I ran into: ipdb does not allow you to move cursor backward.

Again, this is the cost associated with free software and access to the latest updates and new features without waiting for April/October (it’s the MATLAB regular release cycle). If hassle and the extra engineering time far exceed licensing MATLAB licensing costs, MATLAB is a better choice, especially if software is just a chore to get your company from point A to point B, and you are willing to pay big bucks to get there quickly and reliably.

Even with free software on the table, your platform choice is always determined by:

  • how much your time is worth wrestling problems
  • how much flexibility do you need (for customizing to your needs)
  • how much you are willing to pay for the licenses and support

In any case, Python community did good work. Please consider sponsoring PyInstaller and PSF if you profit immensely from their work. But if your company already paid for MATLAB and has real MATLAB experts on it*, don’t switch to Python just for the sake of chasing the trendy thing. You will loose time. Lots of time! Precious engineering hours! Over really stupid things like the ones above that is totally not your fault!  It’s better to dedicate a few days dealing with MATLAB-Python interface than wasting months over the clumsiness of Python if you have a long, complex workflow

Python’s largest value over MATLAB is that there’s a function developed for nearly every less than common situations (like all different variations of path and file management tools) so you don’t need MATLAB File Exchange to fill in the gap, but the downside is that Python, just like the rest of FOSS world (or Tektronix), has absolutely no sense of user experience studies: common operations are tucked under clumsy maneuvers such as making user jump through hoops to run a Python script in the interpreter (import do not execute the script after its cached, execfile was lile eval in MATLAB so they made it hell, or you need to import a freaking library like runpy to do something basic like this. I understand they cannot Why can’t they just add a root-level function like run() and call it a day? Python was just too eager to protect the namespace, and this comes at the expense of adding a lot of stupid maneuvers for really basic bread-and-butter stuff). Python is productive in the sense that there are many sophisticated features that are readily available that you don’t have to write your own library for the advanced cases, but the edge is like 1% of the use cases which I invest the time when I run into it: it’s not worth making 99% of my workflow miserable!

If you are into test instruments, MATLAB is like HP/Agilent/Keysight’s user interface while Python is like Tektronix’s UI where you’ll need to get to 4 levels of context menus to measure a freaking peak-to-peak voltage! Spending your smart people’s time to figure these stupid shit that came from poor UX (user experience) design is a poor investment! At least Python is free and has more advanced features; Tek charges similar to HP and offer nearly the same features yet it has absolutely no respect for the users’ time: they think all people have Stockholm syndrome after going through their steep learning curve for absolutely no benefits.

Why pay a junior engineer $50/hr for a whole week to go through the learning curve, plus months of maintenance work because they’ve used clumsy tools and clumsy methods, when I can code the same thing up in MATLAB in one hour for $500 and give you one page of easy to read code that you likely don’t have to debug through it down the line? If I designed a moderately complex system, most often I can get to the root cause of any bugs in around 15 minutes: the reason is that I spent my time thinking through the abstractions instead of diving into writing for-loops to hack through any common scenarios without first researching for the neatest mechanism. This way I don’t spend an afternoon writing for loops extracting a strings in the object properties of a categorical data type (a efficient way of marking complex repeating data by indexing unique items) and check if it’s really correct when I could have done some research and realize I can just cast it to a cell/monad/array of strings. The value of my experiences comes in knowing the right words and common lingo in many forms to describe many problems, like spotting a low level database operation implementation is just a RLE (run length encoding) in disguise. The hardest part of using search engines to find what you want quickly is knowing the right keywords. If you don’t know the concepts, you’d have jump into reinventing the wheel poorly when you could have used off-the-shelf proven code had you been able to frame the problem succinctly.

* Don’t be mislead by how many years of MATLAB one has! Majority of people those who claimed decades of MATLAB under the belt didn’t really learn the advanced concepts through studying documentation, MATLAB blogs or get training to take it to the next level of super-productivity! You can tell if they have a lot of for-loops for situations that doesn’t absolutely need it (memory saving, injecting elements to LHS assignment, parallel computing, not loading too many files at the same time, drawing graphics (order matters), big tasks that needs to be resumed if it throws an error) and using cells everywhere when they should have used structs or more modern types such as table(). Most of the modern, super-productive MATLAB language constructs happens since 2013, so I can safely say that MATLAB experiences accumulated before that aren’t too valuable for developing business logic part of the code (might be useful if you are writing tools and lower level MATLAB code like those in TMW).


Notify of
Inline Feedbacks
View all comments