Bounties Harm FOSS | Prime Reacts

Recorded live on twitch, GET IN
/ theprimeagen
Article: ziglang.org/news/bounties-dam....
Authors: Andrew Kelley - andrewkelley.me/ | Loris Cro - x.com/croloris?s=20
MY MAIN YT CHANNEL: Has well edited engineering videos
/ theprimeagen
Discord
/ discord
Have something for me to read or react to?: / theprimeagenreact
Hey I am sponsored by Turso, an edge database. I think they are pretty neet. Give them a try for free and if you want you can get a decent amount off (the free tier is the best (better than planetscale or any other))
turso.tech/deeznuts

[@Guergeiro]

People already get mad when their PR is not merged in favour of another that solves the same problem, can you imagine how bad it would become when money is involved?

[@ThePrimeTimeagen]

yeah

[@Jiftoo]

This kind of competition hurts quality since people would be incentivised to push _something_ instead of a completed feature. This puts stress on reviewers too.

[@Pariatech]

I enjoy that Andrew keep a close grip on the project and take it slow. It keep me confident that I made a good choice of language for my game. I don't want the language to drastically change under my feet. If it does I would have no choice than just rewrite it in C and have it be intemporal.

[@sporefergieboy10]

A zigger’s mind much like the programming language operates at speeds unfathomable to the average rust or web developer. When Andrew Kelley programs he literally warps the space around him instantly killing the lesser programmers in the vicinity

[@robonator2945]

I'd agree that it can be complicated, but I'd be especially curious about a project-bounty instead of individual one. In other words, "we'll give 5000 to Zig if this feature gets made, but then how that money is spent, who it goes to, etc. is up to them." I could agree that a big issue with bounties could be that people come-in just trying to get the money without caring about the project, but if the project themselves is in charge of distributing the funds that problem gets mitigated quite a lot. It could probably be further helped with tiered bounties, so a MVP implementation gets 1k, an implementation that rivals X Y or Z gets 2k, etc. which would more directly encourage higher quality code.

[@asdfghyter]

it would probably be wise to split the money to “x on beforehand and y afterwards” so they actually have the money to pay people during the development and not just afterwards

[@Makeshitjusbecuz]

Whenever there is a competition/feature/bug of anykind where time is a unnecessary factor. I tend to write shiiter code than usual.

[@ThePrimeTimeagen]

agree

[@dexterman6361]

I disliked how the startup dude was advertising zig support even * before * this issue or bounty was made, and after it was made, they were, in some places, practically screaming that there was support already. It put so much pressure on the project, I felt it was unfair to the maintainers and the project's core members. I completely agree with the take by Zig team.

[@flalspspsl6858]

"imagine putting out a video when you've already processed the thoughts" lmfao this is such a good jab at social media today

[@ethanholz3733]

This actually a great video! I loved having TJ on to discuss it!

[@Telhias]

I feel like bounties are great for small, simple, numerous and boring issues. Things that are endlessly put off for later as more exciting features take precedence. $2 here 5$ there. Something that is just enough incentive to make you finally stop procrastinating and get to work on those minute problems. To add the polish so to speak. Those are the things that tend to distinguish open source software from proprietary. The little things that add up to the pleasantness of the experience rather than the endless stream of new features and bug fixes.

[@Shneebly]

The $5k bounty to fund a major feature is an unusual edge case. Most bounties get offered when legitimate issues go ignored, so somebody puts up a few hundred dollars to motivate somebody fixing it. The idea that these small incentives systematically lead to maintainers slitting eachothers throats is absurd.

[@bigfishoutofwater3135]

The bounty has problems but do I prefer that to megacorps just buying up projects. I do think bounty programs should run through the project's governance process. The task would ideally be verified with automation so everyone participating has to meet the same minimum standard. You could also flip it around and have interested individuals or teams use some sort of bidding process to avoid duplicated effort. I'd also like to see others who want the same feature to be able to pile on to increase the payout/incentive.

[@itellyouforfree7238]

I'm just amazed at how bouncy his seat looks like

[@Slashx92]

He is on a fitness ball iirc

[@jeremykothe2847]

His points are all valid. Also valid is the counter-point: It's open source. You shouldn't expect to have full control of anything open source.

[@thiagoassisfernandes]

I would suggest group bounties, like to recive a bounty at least to people need to work to solve a problem and recive some ammount (this would totally depend on the available funds) for ex 100 2p -> 100$ = 50$ each 3p -> 150$ = 53$ each 4p -> 240$ = 60$ each that way you encourage groups and people work together, like prime, tj, thor and theo solve the zig wasix problem, get paid to do it and maybe start to do that often together as a group

[@danielsan901998]

Announcing a bounty even before the proposal is accepted is just disrespectful.

[@HyperionStudiosDE]

Apparently the guy who proposed this was banned from the Zig community whatever that means.

[@avishjha4030]

As much as I don't like the way this was done by WASMer, Andrew and his team deleting comments and banning the person doesn't show them in a good light.

[@wondays654]

@@avishjha4030makes them look horrible imo.

[@DooMWhite]

The WASMER guy doubled down in his position

[@kilianvounckx9904]

​​@@avishjha4030 I don't know about banning, but deleting comments was completely valid in my opinion. They deleted them from a github issue, which should be a discussion about technical aspects, not financial or personal as the comments were

[@sethsora7963]

And now we will have more bounties!

[@denissorn]

This is nice. I was watching this for fun, maybe some tech insight, but now I have a philosophical, even ethical question to think about.

[@adrian_franczak]

2:27 top post production

[@MonkeyKong21]

i think bug bounties are "virtuous" because it also serves as compensation/amends for someone who trips over a bug in a dependency when they're trying to do work

[@casperjuelsson6376]

what if a bounty was only payable to the foundation of a project?

[@duven60]

See bounty, hard fork, issue new (open/public) bounty for feature, collect original bounty, forward (some) of the money to pay the new bounty - rules bypassed.

[@NotAFanMan88]

Bounties I think work on the exploit side better than development side. On the exploit side, you just have to find out an exploit at that moment that is repeatable in a realistic scenario. On the development side, you're creating something that might have to be maintained / adjusted for years. On the latter, that's where the art of programming comes into play, and open source collaboration to come at a good solution for most use cases is imperative.

[@alexaustin6961]

When other entities rather than the oss project itself exert influence over it via a large bounty or something else, I have seen many people also complain about how it can draw uwanted pressure and attention to projects that are not quite ready yet. Zig for example would likely have WASIX support on their roadmap, but this just shifts that timeline up where there could be more features in the future version of Zig that make it more suited for it.

[@DexieTheSheep]

15:23 "Software you can love" is a specific term they use. That's what they named their conferences, too. Google Docs might not be the best example, since Google can read all the stuff you write, even though people do literally "love" using it, it doesn't fit their definition of software that was built for and respects the user (at least in my opinion... maybe they'd disagree)... The part about not having to be open-source is still pretty accurate, though. He goes over this in one of his talks, too, but I think basically Loris's point is that you can have Software You Can Love even if it's not open-source, although open-source is definitely a plus. Not all closed-source software is anti-user, and that's also something he goes over in the article. Specifically the quote "there is more to software you can love and that sometimes a reasonably priced, rocksolid, proprietary tool can be preferable to a janky OSS project connected to a murky business model." Just wish you guys clicked the link in the article and cleared it up, because it's a good read. Software You Can Love is like a separate category from Open-Source or Free Software, but can include those too.

[@perrinsilveira6759]

I approve this wheel of time reference!

[@ErazerPT]

I'd say it's an issue of "compartmentalization", ie, works great for bug hunting and things like writing plug-ins for systems that already have a well defined plugin "architecture". Neither will cause "side effects" to the system. You report a bug, someone that knows about the system fixes it. You write a plugin, your change is restricted to your "module" and whatever data it processes. Also deals with the rejection issue. Hard to reject a (proven) bug, you can at most reject to fix it. Hard to reject a plugin, the author (or the sponsor) can simply distribute it via an "unofficial" channel. That said, bounties for things that have "system wide" reach better put through the project's team, as they'll have a much clearer picture of how to best achieve the goal, in line with their designs.

[@robindeboer7568]

bug bounties work for finding security bugs. But I agree "bounty for supporting this feature", especially when its big, is not great for cooperation and external bounty posting isnt great for the reasons mentioned. But security bugs they just find them and point them out.

[@0.Maiden]

Baby Shark, doo-doo, doo-doo, doo-doo...

[@bruwyvn]

Mommy Shark, doo-doo, doo-doo, doo-doo, doo-doo

[@ThePrimeTimeagen]

Daddy Shark, doo-doo, doo-doo

[@ggoraa]

Grandma shark doo-doo, doo-doo, doo-doo

[@Algorythm44]

I think having a bounty for "best code after X days" would give significantly better quality code

[@user-yb4hd9wg6x]

I am trying to solve this problem and would love talking to you about my solution. For now, every dev I have talked to was amazed and instantly wanted to start.

[@saultube44]

Well, ZIG 8s getting money somehow, to sustain development, so they should pay people Meritocraticly, according to advances on the language, according to some standard scale, that should payvthe equivalent of a standard software engineer: like $75K/yr

[@telnobynoyator_6183]

Just started watching, but why wasn't it "5000 bounty for *best* support for wasix"

[@Jebusankel]

Probably the solution is routing the money through a foundation and grants.

[@thinkmore8024]

The TLDR version: Zig is just a posed as Rust.

[@metropolis10]

Need a stabilizer filter for when you have TJ on, can't stop paying attention to his bouncing.

[@Tomyb15]

Almost like a profit motive is antithetical to pretty much every human value and goals 🤔

[@stencilman5030]

There's no need to further feed Moloch.

[@thinkmore8024]

@ThePrimeTime can you please do a dramatic reading of Andrew Kelley's blogpost about why he left twitter and reddit (on his website) It's hilarious. 🤣

[@plutonium_guy]

Till this day, I can't able to open and process a text file. 😞

[@blarghblargh]

tutorials (and books) are your friend. gotta put in the time!

[@jl6723]

Why not do a Prototype Bounty and then grab everyone who made the prototype and get them in a discussion group as a part of the redemption process. So, something like $1000 for a WASIX prototype for Zig that has a minimum DoD for it working. Once completed, you will be invited into a discussion group and you will present and give thoughts on what is best and once the feature is merged into ZIG, the money is then paid out. Collaboration is encouraged by taking all of the respective approaches and comparing and contrasting them to discover a single approach that balances a lot of concerns, while still attracting talent with money?

[@iflessthan3]

wait, is TJ on a medicine ball??

[@michaelucho]

Back in April Zig entered the top 50 of the Tiobe Programming Community Index where it was ranked at #46 and as of today it has completely disappeared from the list. I really hope that Zig manages to take off in the programming world, and not just become a "flash in the pan" so to say. When there is no unity or sense of purpose in the creation of a program, things really start to fall apart.

[@0.Maiden]

How about Rust? 🙂

[@wondays654]

@@0.Maidenrust is ranked in the top 20.

[@0.Maiden]

@@wondays654 hmm, that's not bad!

[@giorgos-4515]

This doesn't fit that much in OSS settings, in some corporate sandboxed setting it would be ideal, like finding somebody to fix something for a gig.

[@itellyouforfree7238]

I would like to put a bounty on Zig itself: 50'000€ to whomever kills Zig first.

[@aftalavera]

F rust and now zig too!

[@dantervil6103]

Well sun paid programmers to write libraries for java and it paid off

[@blarghblargh]

The Zig Foudnation pays programmers to write Zig. Bounties and salaries aren't the same thing.

[@somenameidk5278]

@@blarghblargh damn i can't see any sort of open source "Foundation" without chuckling a little

[@IgorGuerrero]

I'm here to watch REACTIONS not POST-ANALISMS TJ!

[@SRG-Learn-Code]

🏀🏀🏀🏀🏀🏀🏀🏀

[@k98killer]

Theoretically, there's still nothing stopping someone from independently offering bounties for feature development in forks, then proposing those changes in a pull request to the original repo.

[@haraldbackfisch1981]

Here we realize that individual profit does not help the overall goal... why cant we realize that in broader society too?

[@pyrocentury]

Because in this situation, Andrew has the sole authority in defining what that overall goal is and the scope of this situation is contained and relatively well understood. How and by whom would the overall goal of the broader society be defined? For present and future generations?

[@haraldbackfisch1981]

@@pyrocenturyyou misunderstand, we are not fostering cooperation, So whatever we decide on being the goal we might not achieve it. And let's be real if you're not an a*hole we all have a basic shared goal - a harmonious society, that works effectively.. u just might disagree on who's allowed in, but I already mentioned a*holes

[@haraldbackfisch1981]

@@pyrocentury let me put it more bluntly - capitalism bad! And if u disagree I guess u love working in mistrust, not being able to trust that the other people that u inevitably split the workload with will do a good job bc they are incentivized not to... we aren't paid for our output but for hrs if u haven't realized. And that's a bad starting point if uve done any type of societal analysis

[@HyperionStudiosDE]

@@haraldbackfisch1981 I don't trust people will do a good job if they're not incentivized because the reality is that they don't. That's why capitalism good.

[@haraldbackfisch1981]

@@HyperionStudiosDE dawg, they aren't by capitalism either is what I'm saying and you're making it harder for people that actually wanna fix things that are not deemed profitable. U aren't even interested in exploring other concepts, u think of udssr and call it a day. U very dense.

[@BreytnerNascimento]

That means neither Rust nor Zig will replace C

[@blarghblargh]

it makes no change to the likelihood

[@janAkaliKilo]

C is just too powerful, it will outlive all of us.

[@minatonamikaze2637]

none of them are made for that purpose

[@blarghblargh]

@@minatonamikaze2637 a pseudo-mission of zig is to replace C, by enabling in-place incremental refactoring. Zig can be a good language without that mission succeeding. And it doesn't have to be an all-or-nothing

[@sdstorm]

Pff, whenever he brings in that absolute beta, I gotta leave.

[@blarghblargh]

Permanent alpha means you never ship. I like the way you think!

[@viko1786]

Is it just me or is the video and audio just barely out of sync, especially in the beginning?

[@blarghblargh]

The memes

[@rupen42]

tom is a geinous

[@developerdeveloper67]

This zig guy is clueless. "Battle royale" dynamics is what leads to great products, it's competition that foster excellence. I'm a hardcore C developer and I have taken open source libraries out of my C projects because leftist "I'm a victim tweets" from the library creators. After watching this video I will never even try this language. I don't want this guy to feel he is entitled to something because I gave his toy language a try. We don't need your language.

[@eps-nx8zg]

I don't see how anything in this has anything to do with leftism but ok.. Actually who am I kidding you are trolling lol.

[@Kane0123]

Woah. Pretty extreme stance to take against a guy who is building literally free software.

[@crancpiti]

not sure if this is sarcasm 😂

[@Kane0123]

@developerdeveloper67 prob just isn’t man enough to write assembly like a professional… relying on high level languages like some kind of soy dev.

[@Tezla0]

You*

[@jordanwright3052]

Andrew is a strange dude. Between some of the points he has in here as well as some of the reasons he listed for not participating on social media this guy is unironically gonna let communism tear down Zig and its potential. Also super weird that he feels the need to not keep this belief contained within the project itself, but goes out of his way to make it shameful for third parties like wasmer to work on their own stuff that will contribute to the future success of the project.

[@notuxnobux]

I dont know if I would call it communism, he is just against bounties (competitive task handling) and not against paying for a feature to be done. He is more in favor of the traditional way of working on tasks.

[@danielsan901998]

Andrew have demonstrated that he has the ability to create a successful programing language, if you don't like it, try creating your own programing language and lets see the results.

[@Frozander]

Wasmer bounty directly affects the current state and future of Zig, he has every right to be against it. As TJ said the implementation of WASIX is not a bug bounty where expected outcome is clear to everybody. A third party posting a feature bounty can affect the future maintainability of the project and the cohesion of the new addition to the whole ecosystem of the project. Andrew said in his third main point that they are radical and won't accept the winner of the bounty into the project but it is a real possibility that a bounty backed PR spearheaded by a third party might get bullied into the codebase. Like it or not it is Andrew's project and that's the reason Zig is in a point of interest right now, him getting dragged by 3rd parties will not make it any better than it is currently. I won't even get into the controversial topic of WASIX itself and how it has no benefit to current Zig development in the slightest.

[@blarghblargh]

omg, is it the boogeyman of the 1950s, communism!?!?! holy shit man. can't let those pinko commies infiltrate OUR SOFTWARE. they want our PRECIOUS FLUIDS, MANDRAKE! Peace on earth, Purity of essence (edit: guessing not a lot of doctor strangelove fans in the audience. quoted here because it skewers mccarthyism pretty harshly, though superficially and briefly)

[@jordanwright3052]

​@@danielsan901998 I have no doubts whether Andrew is a great engineer. His work speaks for itself. TBD whether his antics will harm the success of the project in a substantial way, but he is taking actions that may stifle its growth at a pretty quick rate.

[@FastRomanianGypsies]

If you want to run something like this it makes more sense to: - Set a range of solution acceptance dates - Rate solutions on speed, memory usage, code simplicity, and modularity, and good error handling via extensive test cases, - The top 10%-20% of participants based on their scores will get reward money that's scaled so that the top place gets 2/3rds of the pot, 2nd place 1/3rd of 2/3rds, 3rd place 1/3rd of 1/3rd of 2/3rd, or some other graduated pay scheme. - I personally would give a special reward to the person who makes the worst possible solution with the most cursed code that somehow still passes. That clown has a special place in my heart. This way it's a meritocracy where there is ~some competition for ~some financial reward that allows for ~some cooperation because it isn't simply winner takes all ASAP.

[@DominusIniquitatis]

1. I always felt kinda meh about bounties in open source, as it looks like a gamble, or a casino, or whatever. Not to mention it's a typical slippery slope that might lead to corruption instead of rewarding. 2. Damn, that dude bounces a lot. 3. "org.wasm", anyone?