A simple fix that Digital Ocean could try is if you get one or more pull request marked as "spam" or "invalid", then the account won't be eligible for the prize. If this was highlighted right at the beginning of the fest, people wouldn't stoop to such cheap levels. Smh.
FWIW this year's spam PR storm has been heavily influenced by a popular Indian youtuber who posted a "how to get a free t-shirt" video showing the PR creation step-by-step. This has caused tens of thousands of identical PRs across GitHub adding the same few words to READMEs.
I gave up on contributing to open source projects when I got a dressing down by an overworked maintainer I wrongly pestered. He was right, but I never again mustered the will to help after that terrible first impression. I think Linux Open source has unfortunately fostered a toxic culture of "asshole geniuses", where you don't have to be nice to others when you are hyper-competent at what you do. No matter that this behaviour repels a lot of people and creates a massive barriere to entry which contributes to the chronic shortage of maintainers.
It's a hard problem to solve. I try to always be extra nice especially when closing any issues or PRs, because 99% of the time the contributor really believes it is the best thing they could do for the project, and I don't want to make them feel the contribution was not worth their time or effort. Some people are more diplomatic about saying 'no' or offering correction than others, though, and there is a lot of toxicity in many areas and companies.
Astra, I would suggest that fostering the personality type you mention is far from limited to the open source movement. Many companies either tolerate, or foster such people in a range of jobs. I am thinking of two particularly bad ones as I type. I could add that over the centuries we have seen more that we should in world politics as well, plus the ones that have just the asshole element, but wrongly think that they are geniuses. :-)
I didn't look up how exactly i works - but i think, it would be good if only accepted pull requests count, so you probably would be spammed once or twice, but you can just ignore it, and close them over time, and in 2-4 years at least the totally obvious, spam should be over. It would be a little sad for people who honestly tryed, but aren't that experienced - but maybe you can also flag these, to give them rewards - instead like now the other way around
Idea: Put up some open-source repositories that are nothing but comments, and accept any and all pull requests. Then everyone can win stickers and T-shirts!
If I wasn't drowning in school work right now and working part-time I would absolutely help you out on your open source stuff! This time next year I'll be free from school so maybe I can help out then :)
I feel like hacktoberfest is a great way to highlight the importance of choosing meaningful metrics. I'm curious if things would be different if a separate metric had been chosen to gauge "contribution".
That's the tough thing-most other metrics that don't create the perverse incentive would be subjective. I think the best outcome would be more "we will give these shirts away to the first X people who sign up, regardless of contribution" if they're just giving away the shirts. Then still ask people to contribute but don't require it. In a similar vein, I get a ton of low-effort PRs submitted to my projects (usually automated code linters that don't match my style and don't pass CI tests) that I know are pushed out just to give the person's GitHub profile more 'credibility'.
@@adrianteri Apparently so. I'm guessing it's because it's a _lot_ of work getting enough maintainers to open the floodgates on their own repos... rightly so :)
I'm not sure Digital Ocean can fix this issue long term, other than canceling Hacktoberfest altogether. It's the old intrinsic vs extrinsic motivation. When motivated by extrinsic rewards, we tend to do the minimum necessary to get the reward, like that guy who spammed the user.yml file on 6 repos. Alfie Kohn wrote a book about it called, Punished by Rewards, which is worth reading for a deeper dive on how extrinsic motivation backfires. Here's an excerpt: www.alfiekohn.org/article/reading-incentives/
This is nothing new on github, there are swarms of people who do nothing but submit spelling corrections to hundreds of repos, and then pretend they're a developer...
I contributed to an open source project (it had nothing to do with Hacktoberfest) and my pull request sat around for about a year until I decided to send a friendly reminder (the maintainer had expressed the intention to review and likely accept my contribution) every month until he got around to it.
I think the clear problem is that 1 month is not enough time for someone who has never contributed code to contribute meaningful code. If the goal is to encourage new users there has to be guidance and stages leading up to that first pull request.
First time contributing Something to an open source project is the hardest part imo, because you have no Idea what to do or expect. Glad I finally did tho, its a lot easier to just do it now.
It maybe did not help that my first contribution was in Ruby, a language I never used before outside of a vagrantfile, and was to the vagrant repository itself of all places, which seemed kinda "scary" at the time. 😄
My ancestry is also from The Netherlands and I have a Dutch last name. I've visited 3 times and it such a great country. I hope you can visit someday too. You will love it.
Just reads like gatekeeping by folks who never thought through their decisions. It's okay to not respond to trash open source commits. It's okay to not want the upstream people push back. You can filter and use bots and GitHub config
I wouldn't consider it gatekeeping-I think most maintainers would agree that we absolutely love getting 'first time contributors', and as I said in the video, I usually go out of my way to spend more time helping those contributors, especially if the PR has some merit, and even if it has a long way to go before merging. It's not gatekeeping to be angry at a company for creating an incentive for 5x more spam PRs to appear in the matter of a few days.
@@JeffGeerling I've read that an interruption can cause the loss of 15 minutes before you can get back into the flow of your work. If you have an hour to work on your project, those four spammed interruptions just wiped out any chance of making progress that day.
@Jeff Geerling You have great videos and the content is always interesting, even if for me sometimes it gets a little bit high level and out of my range of knowledge. In 2022, I watched this as I am Googling about what Hacktoberfest is (as this year's event started today), so I can get started on my journey of contributing to Open Source, and becoming part of the community. After watching, now I am aware and am trying to tread lightly and not be a hindrance to a person who worked hard on their project. I am not going to go full canned mystery meat😁 like the example in the video, but I would like to use this as a learning experience. How would you suggest going about as a beginner and trying to ACTUALLY contribute during this event, or in general? Thanks for all the content and I am looking forward to more!
A simple fix that Digital Ocean could try is if you get one or more pull request marked as "spam" or "invalid", then the account won't be eligible for the prize. If this was highlighted right at the beginning of the fest, people wouldn't stoop to such cheap levels. Smh.
Maybe a little tweaking is necessary, but yes. I think this would be interesting to look into
They’ve done exactly that this year.
@@skollskollskoll Yeah I know xD
FWIW this year's spam PR storm has been heavily influenced by a popular Indian youtuber who posted a "how to get a free t-shirt" video showing the PR creation step-by-step. This has caused tens of thousands of identical PRs across GitHub adding the same few words to READMEs.
Easy fix. Just pull the old "Shipping rates apply"
"This is why we can't have nice things..."
Purrfect
I gave up on contributing to open source projects when I got a dressing down by an overworked maintainer I wrongly pestered.
He was right, but I never again mustered the will to help after that terrible first impression.
I think Linux Open source has unfortunately fostered a toxic culture of "asshole geniuses", where you don't have to be nice to others when you are hyper-competent at what you do.
No matter that this behaviour repels a lot of people and creates a massive barriere to entry which contributes to the chronic shortage of maintainers.
It's a hard problem to solve. I try to always be extra nice especially when closing any issues or PRs, because 99% of the time the contributor really believes it is the best thing they could do for the project, and I don't want to make them feel the contribution was not worth their time or effort.
Some people are more diplomatic about saying 'no' or offering correction than others, though, and there is a lot of toxicity in many areas and companies.
Astra, I would suggest that fostering the personality type you mention is far from limited to the open source movement. Many companies either tolerate, or foster such people in a range of jobs. I am thinking of two particularly bad ones as I type. I could add that over the centuries we have seen more that we should in world politics as well, plus the ones that have just the asshole element, but wrongly think that they are geniuses. :-)
Not always hyper-competent but believing to be hyper-competent
Opt-in sounds like a good plan!
This happens. Do not despair. Push back.
I didn't look up how exactly i works - but i think, it would be good if only accepted pull requests count, so you probably would be spammed once or twice, but you can just ignore it, and close them over time, and in 2-4 years at least the totally obvious, spam should be over.
It would be a little sad for people who honestly tryed, but aren't that experienced - but maybe you can also flag these, to give them rewards - instead like now the other way around
Congratulations. Compute module 4 on the way. Great work on this channel.
#ohbaby! Hey, congratulations! So happy for you!
Thanks! 👶
Idea: Put up some open-source repositories that are nothing but comments, and accept any and all pull requests. Then everyone can win stickers and T-shirts!
Google Summer of Code does it a lot better.
I’m first Jeff
Second!
If I wasn't drowning in school work right now and working part-time I would absolutely help you out on your open source stuff! This time next year I'll be free from school so maybe I can help out then :)
Exactly what I was thinking. Full time work + cert studying + Masters + kids. That’s my time lol.
I feel like hacktoberfest is a great way to highlight the importance of choosing meaningful metrics. I'm curious if things would be different if a separate metric had been chosen to gauge "contribution".
That's the tough thing-most other metrics that don't create the perverse incentive would be subjective.
I think the best outcome would be more "we will give these shirts away to the first X people who sign up, regardless of contribution" if they're just giving away the shirts. Then still ask people to contribute but don't require it.
In a similar vein, I get a ton of low-effort PRs submitted to my projects (usually automated code linters that don't match my style and don't pass CI tests) that I know are pushed out just to give the person's GitHub profile more 'credibility'.
Sometime I think 2015 or 16 Repo's and maintainers had to opt-in for contributions to be eligible ...Has the system changed?
@@adrianteri Apparently so. I'm guessing it's because it's a _lot_ of work getting enough maintainers to open the floodgates on their own repos... rightly so :)
When a metric becomes a target, it fails to be a good metric
I'm not sure Digital Ocean can fix this issue long term, other than canceling Hacktoberfest altogether.
It's the old intrinsic vs extrinsic motivation.
When motivated by extrinsic rewards, we tend to do the minimum necessary to get the reward, like that guy who spammed the user.yml file on 6 repos.
Alfie Kohn wrote a book about it called, Punished by Rewards, which is worth reading for a deeper dive on how extrinsic motivation backfires.
Here's an excerpt:
www.alfiekohn.org/article/reading-incentives/
Thanks for sharing the article, it explains a lot
Wow, that's bad. Thanks for the video.
This is nothing new on github, there are swarms of people who do nothing but submit spelling corrections to hundreds of repos, and then pretend they're a developer...
0:46 A planted tree is like a free t-shirt for the Earth!
I contributed to an open source project (it had nothing to do with Hacktoberfest) and my pull request sat around for about a year until I decided to send a friendly reminder (the maintainer had expressed the intention to review and likely accept my contribution) every month until he got around to it.
Change it to where you have to be recommended by a friend. Who has been a proven contributor, they get 3 recommends.
Congratulations on your baby!
I think the clear problem is that 1 month is not enough time for someone who has never contributed code to contribute meaningful code.
If the goal is to encourage new users there has to be guidance and stages leading up to that first pull request.
First time contributing Something to an open source project is the hardest part imo, because you have no Idea what to do or expect. Glad I finally did tho, its a lot easier to just do it now.
It maybe did not help that my first contribution was in Ruby, a language I never used before outside of a vagrantfile, and was to the vagrant repository itself of all places, which seemed kinda "scary" at the time. 😄
Your shirt looks familiar. Oh it is because I have the same one!
Go Netherlands!
My ancestry is also from The Netherlands and I have a Dutch last name. I've visited 3 times and it such a great country. I hope you can visit someday too. You will love it.
cool
Just reads like gatekeeping by folks who never thought through their decisions.
It's okay to not respond to trash open source commits. It's okay to not want the upstream people push back. You can filter and use bots and GitHub config
I wouldn't consider it gatekeeping-I think most maintainers would agree that we absolutely love getting 'first time contributors', and as I said in the video, I usually go out of my way to spend more time helping those contributors, especially if the PR has some merit, and even if it has a long way to go before merging.
It's not gatekeeping to be angry at a company for creating an incentive for 5x more spam PRs to appear in the matter of a few days.
@@JeffGeerling I've read that an interruption can cause the loss of 15 minutes before you can get back into the flow of your work. If you have an hour to work on your project, those four spammed interruptions just wiped out any chance of making progress that day.
Congrats to the actual news in this video! Hope everything will be good and healthy for both your wife and and the to be newborn child :)
Thank you so much!
@Jeff Geerling You have great videos and the content is always interesting, even if for me sometimes it gets a little bit high level and out of my range of knowledge. In 2022, I watched this as I am Googling about what Hacktoberfest is (as this year's event started today), so I can get started on my journey of contributing to Open Source, and becoming part of the community. After watching, now I am aware and am trying to tread lightly and not be a hindrance to a person who worked hard on their project. I am not going to go full canned mystery meat😁 like the example in the video, but I would like to use this as a learning experience. How would you suggest going about as a beginner and trying to ACTUALLY contribute during this event, or in general? Thanks for all the content and I am looking forward to more!