Green Guide - Software Engineer

[russelltrow]

Context

Develop a post Green Software Practitioner course resource for Software Engineers.

Resources

• https://github.com/Green-Software-Foundation/learn/blob/green-guides-dev/docs/green-guides/software-engineer.md

Actions

• Define the persona
• Find SMEs to work on the guide
• Create first draft of guide
• Refer to the LFC131 feedback on "More Software Developer Content"

[russelltrow]

In connection to Analyse feedback collected from the Linux Foundation course there are lots of requests for "More Software Developer Content"

I've reproduced a selection of the more useful comments below:

• I expected some advice for reduce carbon in every day sw development
• I thought it would have been more focused on software practices to reduce emissions.
• I wish it focused more on what could be done from a design standpoint
• I was expecting coding practices that developers can do to help the software greener
• I was expecting more content on how to develop green software (more related to the programming part).
• Was more on power consumption and less on what software engineers can do to promote green applications
• Thought this would cover actual techniques for improving software, but it just goes into basics of energy usage.
  it did not give me enough practical details as to how to limit carbon emissions from developing software
• Would have liked some more practical advice on how to directly impact carbon emission in the workplace.
• Kind of high level approach, I wonder if the title (software) is a bit off topic. However the course itself was fine
  I expected more direct hints/tips/solutions in stead of a long list of definitions and theories, altough some of the theory was interesting
• I would have loved more specific information about implementing green software practices. An additional future course specific to front-end practices would be great!
• I would like to see more details on the way software could be programmedmore energy effecient. e.g. database syncs/backups on specific time or using cache in regions (follow the sun).
  I liked the course, and treat it as an introduction. What I am missing though is more practical information on implementing the SCI framework, as this course is directed "at software practicioners".
• It's good to know the terminology and electropolitical stuff. I was hoping for some practical advice like "here's how to use a carbon-awareness API in my cronjob to decide when to run it.""
• Very interesting content, teaching concepts and vocabulary that are essential to the ongoing debate about climate. I was expecting more technical content about how to design applications such that they are less carbon intensive.
• Other than hardware I was expecting there to be more examples of how software can be more efficient. This would include ways of writing more performant code, the design of user interfaces and user experiences that require less energy directly or indirectly.
• I am missing two things: (1) the sobriety in the design of software application, which is a bit different from the efficiency, and (2) regarding the SCI I am missing examples with numbers, and maybe more information about concrete ways to measure E.
• It is not really focusing a lot on software. All aspects related to it can be summed up in few pages, still I enjoyed the overall view on the topic. Probably an advanced course on green software could be of help (or maybe already exists)
• I think I was expecting a bit more detail rather than such a high level overview. I would find a course that goes into more detail really useful, especially the process of making an application as green as possible to get a real sense of what that might look like.
• The focus was on the science, business, and politics providing context to green software. This was valuable, but not the first time I've reviewed similar materials. I was hoping to have a more guided introduction on how to apply these in application development and infrastructure management. This might be met with wrapping the course with links to patterns and case studies.
• From the course title, I was expecting strategies or methodologies to produce software that consumes less energy (for instance, in the last Kubecon someone shared a graph stating that Python is the most inefficient language in terms of energy consumption). In contrast, although the title being "green software", the content is focused mainly in hardware, electrical installations and theoretical concepts, nothing directly related to "software".
• This course doesn't cover code optimization or how useful some software are compared to others. It concentrates on the carbon footprint, which is the least concern towards the future of our planet. Moreover, if find very ironic that Google and Microsoft are shown as good "green" examples, as they obviously work on planned obsolescence on their softwares.
• Course name is "Green Software" and it didn't include much information about how to change ones approach to creating software to make in more "greener". How to adjust app to perform computing during renewable energy is supplied, how to make sure framework I want to use has good utilisation. That's what I was expecting from training with such a title. Nevertheless topic is not aligning with the content, it was interesting.
• I think there is room for more content. In the measurements chapter, there can be mentions as to why still considering the total carbon emissions of a software project is important (avoiding Jevon's paradox). In the hardware efficiency chapter, there can be mentions of effects beyond just carbon emission; for instance, data centers consume large amounts of water for cooling. In the section about demand shifting, I would also include specific examples of types of programs that could be demand shifted (can a front-end application be demand shifted? what about a machine learning program? etc.)
• Create video tutorials or demonstrations that illustrate practical aspects of green software development, showcasing tools, techniques, and best practices.
• As mentioned above... I was missing a chapter regarding software efficiency. Therefore, I am thinking about "Why is C as a programming language more sustainable compared to Python?" Solution: compiled languages are more efficient compared to interpreted languages and a little bit more elaboration on that. Or "What is the impact of recursive functions when you can write a for-loop instead" Solution: memory consumption due to the overhead of allocating memory for each function execution. Another one would be "How to make program conditions more sustainable?" Solution: prefer switch-case over if-else statements and maximum can be achieved when sorting the conditions descending based on the probability of occurrence. I would be more than happy to assist with that chapter!