5 ms·

Python vs. Ruby – The Workshape.io Smackdown

59 points by carsie 12y ago

lion_del 12y ago
Interesting. A good addition would be to see your geographic data (I remember saying where I wanted to work). I wonder if either are more popular in certain geographies or not.
nvk 12y ago
Fun article, but there is a lot of selection bias here. Small sample, product focussed at younger devs, etc...
- GordyMD 12y ago
  This is a fair criticism. As we build up a larger data set we hope that more value can be attached to the it and the insights be more indicative of the community. Until then we are just interested to share what we are currently finding with the new primary data we have collected. We could definitely do with a statistician to help us out and minimise any selection bias, but i think that comes when we increase the amount in our data set by another order of magnitude or two. :)
  - ChikkaChiChi 12y ago
    The article does a good job of explaining your methods of data collection, and nothing in the content suggests otherwise. I wish more articles on the Internet had this kind of transparency.
teamhappy 12y ago
I would have expected the C/C++/Linux people to have higher seniority than the CSS/CoffeeScript/Node.js people. Can anybody explain this? Is it that they work in two different industries (i.e., they both build web apps, but the Python ones are heavier on the number crunching side)?
- GordyMD 12y ago
  The seniority level was gathered by asking each person the seniority they would like in their next job. This suggests that it is more common for people working with Ruby to think of picture themselves in a higher seniority role than their Python counterparts. It is not an actual reflection of their current level. Also it is completely subjective - Your definition of C/C++/Linux people being more senior to CSS/CoffeeScript/Node.js people is your point of view - other people's views may be different.
  - teamhappy 12y ago
    > It is not an actual reflection of their current level. I know. I meant to say "have a higher seniority in the graph." I assume the sample size was big enough to handle shy Pythonistas and/or overly-confident Ruby devs. :P > Your definition of C/C++/Linux people being more senior [...] I said "I would have expected [...]". > [...] other people's views may be different. Clearly: http://uk.businessinsider.com/best-tech-skills-resume-ranked-salary-2014-11?r=US http://uk.businessinsider.com/best-tech-skills-resume-ranked... My personal experience is that somebody who doesn't know C/C++ and some UNIX basics is going to write worse code on average. (Settle down.) My personal experience probably isn't representative of anything (I'm a graphics programmer), but I think it shows in the web dev world as well. Puma for example (a Ruby web server I know virtually nothing about) is crazy fast because the guy who wrote it knows about fork and signals. Another example is Bluebird (JS promises lib), which only became popular because the guy who wrote it knows how compilers work.
- dagw 12y ago
  Seniority and titles are also majorly influenced by company size. Years ago I worked as systems administrator at a larger company. I had a friend of mine who did very similar work at a very small company. We earned basically the same, were in basically the same industry, did basically the same work and for all intents and purposes had the same responsibility. My title was Systems Administrator and his was Chief Technical Officer. Being the lead developer at a 2 month old 3 man web shop is very different from being lead developer on Microsoft Windows.
kyllo 12y ago
Ruby devs build more web apps and Python devs do more systems dev and data crunching. Not exactly a groundbreaking revelation knowing that Python has much more in the way of bindings for C/C++ and Fortran libraries for statistical / math routines, but still interesting to see this reconfirmed.
- ChikkaChiChi 12y ago
  Source? My understanding was that before Rails, Ruby was considered primarily a language used for mathematics. Anecdotally, it does seem that I see a lot of stories about using python for statistics, but I haven't seen any hard numbers one way or the other. Python seems to also have a lot of web implementations. I'm not sure where you are getting your numbers from.
  - darrhiggs 12y ago
    I believe they are referring to the article > The skills and time load sections both re-inforce that Python is more heavily linked to Data Science than Ruby. Ruby is more of a web technology with strong affinity to front-end technologies and iOS.
  - kyllo 12y ago
    SciPy, matplotlib, pandas et al have existed and seen heavy use in the mathematical/scientific programming field for many years. SciRuby is a great project started last year to bring the same functionality to Ruby.
  - MrBra 12y ago
    what planet do you come from? :)
davman 12y ago
Unrelated to the article itself, but I have an ongoing issue with Workshape that I'm wondering if anyone else has. I've currently got a 97 "match" with a position there, but I still get "There are currently no matches for your desired role". Does anyone else have this? Am I doing something wrong?
- GordyMD 12y ago
  Hey. Apologies for the bad UX here. You are currently seeing the unmatched screen. The reason for this is most likely that your desired location does not match that of the position. I'll have a look into it though to confirm whether this is the case. If you want to get back to me so we can talk about it more then email me at gordon[at]workshape.io. In the future we will be updating the experience for users who do not get matches according to their profile. We currently show the closest matches, which I can imagine being frustrating. Instead we are thinking of surfacing insights from our data set based on your input. So, similar to this article, we'd give insights relevant to you about opportunities in the area.
rshaban 12y ago
The Secondary skills distribution chart seems the most interesting to me -- it seems to indicate that Python users are more interested in other programming languages and general Computer Science (C++/Java, Unix, machine learning), while Ruby users are more likely to be focused on app/web dev (CSS, Coffeescript, Objective-C). This makes sense given Rails' popularity and libraries like scikit-learn, numpy, etc.
ChikkaChiChi 12y ago
Are Rubyists significantly less likely to have database skills than those who have made a profession out of Python programming? The secondary skills chart seems to indicate this (at least for Workshape sample set). If so, is this the result of 10 years of working with an abstraction layer in Rails? Is it simply because Rails developers have a different focus? The Average Workshops vis seems to indicate this slightly, but not nearly as much as the premise would indicate.

59 points by carsie 12y ago

Interesting. A good addition would be to see your geographic data (I remember saying where I wanted to work). I wonder if either are more popular in certain geographies or not.

nvk 12y ago

Fun article, but there is a lot of selection bias here. Small sample, product focussed at younger devs, etc...

GordyMD 12y ago
This is a fair criticism. As we build up a larger data set we hope that more value can be attached to the it and the insights be more indicative of the community. Until then we are just interested to share what we are currently finding with the new primary data we have collected. We could definitely do with a statistician to help us out and minimise any selection bias, but i think that comes when we increase the amount in our data set by another order of magnitude or two. :)
- ChikkaChiChi 12y ago
  The article does a good job of explaining your methods of data collection, and nothing in the content suggests otherwise. I wish more articles on the Internet had this kind of transparency.

teamhappy 12y ago

I would have expected the C/C++/Linux people to have higher seniority than the CSS/CoffeeScript/Node.js people. Can anybody explain this? Is it that they work in two different industries (i.e., they both build web apps, but the Python ones are heavier on the number crunching side)?

GordyMD 12y ago
The seniority level was gathered by asking each person the seniority they would like in their next job. This suggests that it is more common for people working with Ruby to think of picture themselves in a higher seniority role than their Python counterparts. It is not an actual reflection of their current level. Also it is completely subjective - Your definition of C/C++/Linux people being more senior to CSS/CoffeeScript/Node.js people is your point of view - other people's views may be different.
- teamhappy 12y ago
  > It is not an actual reflection of their current level. I know. I meant to say "have a higher seniority in the graph." I assume the sample size was big enough to handle shy Pythonistas and/or overly-confident Ruby devs. :P > Your definition of C/C++/Linux people being more senior [...] I said "I would have expected [...]". > [...] other people's views may be different. Clearly: http://uk.businessinsider.com/best-tech-skills-resume-ranked-salary-2014-11?r=US http://uk.businessinsider.com/best-tech-skills-resume-ranked... My personal experience is that somebody who doesn't know C/C++ and some UNIX basics is going to write worse code on average. (Settle down.) My personal experience probably isn't representative of anything (I'm a graphics programmer), but I think it shows in the web dev world as well. Puma for example (a Ruby web server I know virtually nothing about) is crazy fast because the guy who wrote it knows about fork and signals. Another example is Bluebird (JS promises lib), which only became popular because the guy who wrote it knows how compilers work.
dagw 12y ago
Seniority and titles are also majorly influenced by company size. Years ago I worked as systems administrator at a larger company. I had a friend of mine who did very similar work at a very small company. We earned basically the same, were in basically the same industry, did basically the same work and for all intents and purposes had the same responsibility. My title was Systems Administrator and his was Chief Technical Officer. Being the lead developer at a 2 month old 3 man web shop is very different from being lead developer on Microsoft Windows.

kyllo 12y ago

Ruby devs build more web apps and Python devs do more systems dev and data crunching. Not exactly a groundbreaking revelation knowing that Python has much more in the way of bindings for C/C++ and Fortran libraries for statistical / math routines, but still interesting to see this reconfirmed.

ChikkaChiChi 12y ago
Source? My understanding was that before Rails, Ruby was considered primarily a language used for mathematics. Anecdotally, it does seem that I see a lot of stories about using python for statistics, but I haven't seen any hard numbers one way or the other. Python seems to also have a lot of web implementations. I'm not sure where you are getting your numbers from.
- darrhiggs 12y ago
  I believe they are referring to the article > The skills and time load sections both re-inforce that Python is more heavily linked to Data Science than Ruby. Ruby is more of a web technology with strong affinity to front-end technologies and iOS.
- kyllo 12y ago
  SciPy, matplotlib, pandas et al have existed and seen heavy use in the mathematical/scientific programming field for many years. SciRuby is a great project started last year to bring the same functionality to Ruby.
- MrBra 12y ago
  what planet do you come from? :)

davman 12y ago

Unrelated to the article itself, but I have an ongoing issue with Workshape that I'm wondering if anyone else has. I've currently got a 97 "match" with a position there, but I still get "There are currently no matches for your desired role". Does anyone else have this? Am I doing something wrong?

GordyMD 12y ago
Hey. Apologies for the bad UX here. You are currently seeing the unmatched screen. The reason for this is most likely that your desired location does not match that of the position. I'll have a look into it though to confirm whether this is the case. If you want to get back to me so we can talk about it more then email me at gordon[at]workshape.io. In the future we will be updating the experience for users who do not get matches according to their profile. We currently show the closest matches, which I can imagine being frustrating. Instead we are thinking of surfacing insights from our data set based on your input. So, similar to this article, we'd give insights relevant to you about opportunities in the area.

rshaban 12y ago

The Secondary skills distribution chart seems the most interesting to me -- it seems to indicate that Python users are more interested in other programming languages and general Computer Science (C++/Java, Unix, machine learning), while Ruby users are more likely to be focused on app/web dev (CSS, Coffeescript, Objective-C). This makes sense given Rails' popularity and libraries like scikit-learn, numpy, etc.

ChikkaChiChi 12y ago

Are Rubyists significantly less likely to have database skills than those who have made a profession out of Python programming? The secondary skills chart seems to indicate this (at least for Workshape sample set). If so, is this the result of 10 years of working with an abstraction layer in Rails? Is it simply because Rails developers have a different focus? The Average Workshops vis seems to indicate this slightly, but not nearly as much as the premise would indicate.