This piece is a little bit here and there and I’m not convinced. Or I am misreading the title. Is it “Business makes better open source than open source projects” or is it “Business makes better open source, versus the same team not doing business, just doing the project”.
Sure, concourse might the current best forum software, depending on your requirements, so what? This isn’t a proven track records of “projects like concourse” working out better than “projects unlike concourse”. Also look at the contenders. phpBB is ancient compared to this and afaik there was only official commercial stuff after 10 years.
Also not everyone benefits from a company not only backing, but being the major driver. It very often complicates things. I’ve seen smaller projects being more eager to accept your patches because they don’t clash with the company roadmap (often because there was no roadmap, whatever :P)
Then again I might be biased, I’ve been involved with free and open source software for 15ish years and I don’t want most of it to be business stuff. Sure, if I didn’t have a day job I might want to work on some of the things I do in my spare time while being paid, but usually it’s participating as long as I like. I lose the most important freedom: just giving up on a bug I can’t solve or doing a chore I don’t want to do. Nobody is paying me, so nobody’s the boss.
NB: I personally dislike how concourse looks and feels, but I haven’t participated in any forums in a few years. I kinda liked the phpBB style, so if concourse could be themed to look like it, I’d probably try it.
I have actually used Discourse pretty extensively, so I think I can reasonably respond to it:
Is it “Business makes better open source than open source projects” or is it “Business makes better open source, versus the same team not doing business, just doing the project”.
It’s never going to be “the same team (not) doing business.” The basis of the argument is that a volunteer project can’t have as many people, or as diverse people, as a business can. Volunteer projects only get volunteers, which means your contributors will always be pretty well-off, will vastly prefer doing things that are fun, will always be strapped for time, and will have to sacrifice their volunteer time in favor of work in ways that they wouldn’t have to if it was work.
Also, a business might be better able to perform usability tests, by using their money to pay people to participate in surveys. This is much better than expecting volunteers to notice usability problems, because volunteers are always familiar with the project while the usability testers that you really want are people who aren’t.
Sure, concourseDiscourse might the current best forum software, depending on your requirements, so what? This isn’t a proven track records of “projects like concourseDiscourse” working out better than “projects unlike concourseDiscourse”. Also look at the contenders. phpBB is ancient compared to this and afaik there was only official commercial stuff after 10 years.
phpBB is ancient and had no commercial stuff for ten years because it isn’t backed by a business.
It probably isn’t fair to judge every project by Discourse’s success criteria. The number of sites running on the Lobsters codebase can be counted on one hand, but I’m pretty sure pushcx and companions consider it a success anyway. Discourse wants to be a standard. Lobsters, and XMonad (an example that someone else brought up) does not. That’s fine.
The Discourse team judges the quality of software by “usability”, where usability is defined by friendliness to non-contributors. I like Discourse’s interface better than GitLab’s overly-complex interface or IRC’s impenetrable wall of jargon, but I honestly think Discourse’s interface is also a bit overly complex. But that’s just, like, my opinion, man.
What matters is that Discourse’s customers get to tell them what to do. If you aren’t a programmer, but you have money, that’s a terrific arrangement.
Also not everyone benefits from a company not only backing, but being the major driver. It very often complicates things. I’ve seen smaller projects being more eager to accept your patches because they don’t clash with the company roadmap (often because there was no roadmap, whatever :P)
If your goal is to be used by non-contributors, or, worse, non-nerds, then the complexity of being a company will be the least of your worries. You’re going to have to do requirements gathering. I have no idea how you would do a truly good job of that without having someone who can just tell the engineers “I don’t care what you want, the stakeholders want it to do this. Just do it.”
Then again I might be biased, I’ve been involved with free and open source software for 15ish years and I don’t want most of it to be business stuff. Sure, if I didn’t have a day job I might want to work on some of the things I do in my spare time while being paid, but usually it’s participating as long as I like. I lose the most important freedom: just giving up on a bug I can’t solve or doing a chore I don’t want to do. Nobody is paying me, so nobody’s the boss.
I’m know that feel. Solving weird technical problems can be really fun, but there’s more to designing an artifact that lots of people use than just solving the technical problems. There are companies that fail at the little stuff, too, but volunteer projects usually don’t even try.
Erlend wants people who aren’t programmers to have a say in the software that they use. You got a better way to do that than wallet-voting?
In the long run, that’s strictly better for a software endeavour meant to last and to blossom rather than wither (or to be dragged forward like a comatose horse by sheer corporate / personnel / funding force). Volunteers bring something invaluable that (on average / on the whole / excepting outlier rockstar-style specimens) salaried employees (even if the same people =) can’t bring or only rarely / short-lived: deep inshakable innate enthusiasm and self-propelled drive for the project and thus its pending task(s) at hand. This brings a completely different quality of attention to detail (helping with bug rate, robustness, polish), potentially less “rat-race/routine-boreout-inflicted carelessnesses” (helping maintainability) etc.
which means your contributors will always be pretty well-off, will vastly prefer doing things that are fun, will always be strapped for time, and will have to sacrifice their volunteer time in favor of work in ways that they wouldn’t have to if it was work.
Sure but would view that as just a different “price” than monetary. Not so neat for “product”-type software endeavours where you want to reach stage X by date Y, but for to-be-long-lived slowly-eclosing meandering evolving-rather-than-directed projects — not a bad deal. Less hustling & bustling & huffing & puffing, fewer messes & spillage =)
In the long run, that’s strictly better for a software endeavour meant to last and to blossom rather than wither (or to be dragged forward like a comatose horse by sheer corporate / personnel / funding force). Volunteers bring something invaluable that (on average / on the whole / excepting outlier rockstar-style specimens) salaried employees (even if the same people =) can’t bring or only rarely / short-lived: deep inshakable innate enthusiasm and self-propelled drive for the project and thus its pending task(s) at hand.
I can barely parse that paragraph. Please reduce the parenthetical and slash-separated phrases.
That’s also completely missing the observed facts. Volunteer software developers, especially the maintainers and BDFL’s of the world, get burnt out all the time. Volunteer free software projects often fail for external reasons just like business projects do.
And, as I mentioned before, not every useful task benefits from volunteer enthusiasm. There’s a decent amount of drudgery even in an optimally-managed project, because sometimes the work you actually need to do isn’t algorithmic puzzle-solving. Sometimes, it’s going from person to person collecting data to help plan out the next sprint’s priorities. Sometimes, it’s putting together documentation. Sometimes, it’s running the fuzzer for a week, which doesn’t really require much work at all, but definitely requires money for all the hardware you want.
Sure but would view that as just a different “price” than monetary. Not so neat for “product”-type software endeavours where you want to reach stage X by date Y, but for to-be-long-lived slowly-eclosing meandering evolving-rather-than-directed projects — not a bad deal. Less hustling & bustling & huffing & puffing, fewer messes & spillage =)
That’s why I made a point, around the middle of my comment, to say that “it’s not fair to judge every project by Discourse’s measure of success”. Writing your software for yourself with no plans to monetize it is not wrong, but it cannot, and never will be, a true replacement for the proprietary social networks in the same way that Discourse can be.
Erlend wants people who aren’t programmers to have a say in the software that they use. You got a better way to do that than wallet-voting?
I know many open source projects that work on a “someone suggests a feature and hey, we never thought of this and it makes total sense, so we’ll do that” basis. Sure that’s not always working and often enough someone from the committers must be interested, but that happens. You also can just pay someone to contribute a feature, many projects are also open to that.
Overall I’m maybe just a little confused about the language of the post that seems a little overzealous to me. You’re kind of hitting the same tone. I really hardly had any contact with Discourse (sorry for the concourse typos..) and I wouldn’t give it the status you two seem to give it. Either party could be wrong. To me, forum software is a niche market these days, it was much, much bigger (in percent of the web, not total) 10-15 years ago. You might argue it’s the Facebook of self-/paid-hosted forums and I had never heard of Facebook…
But let me stress that I have nothing against Discourse (the software, the project, the company) - cool if it works, but I simply don’t find this a feasible MO for any random project, And I’m happy not everything is commercial, but it’s cool if some people can work on open source software as a paid job - my main problem is the structure of the project and if your boss tells you what to do, I only care so much if the result is then open or closed, although of course I’d prefer open most of the time.
What’s implicit throughout the article but nowhere mentioned in the title is the target. Business makes better open source for business, I think is probably not that controversial - it doesn’t follow that business makes better open source for everyone else as well. I doubt, for example, that every XMonad user values “sustainability” and the ready availability of consultancy/support over the efficiency gains they get from a tiling WM with their own personal workflows compiled in - otherwise they’d be using GNOME. Or Emacs vs VS Code. Or … but I’m sure you can provide your own examples.
tl;dr “mass-market-ready” is not the only axis along which one might usefully and sensibly measure “better”
Exactly, that’s why I think it’s important the distintion between Free Software and Open Source, one seems to value more the freedom of their users experience, and that’s more important than any long term sustainability, it even serves as a kind of guard against bad sustainability (even if the company goes under, I will be able to use and modify their software). The other is more interested in “usage”, and the developers freedom. The user freedom vs devs freedom battle has always amused me.
It was a genuine question to someone who worked with the company for four years, I think it’s somewhat related. I just thought it was a little weird is all.
Any endeavour large enough will require work to begin, achieve, sustain
That work needs to be done by people
More people [usually] means the work is done faster
or, restated: when there is work that could be done (in parallel) by N people, and you have fewer than N people, then the work gets done slower, and/or some things (nice-to-haves) don’t get done at all
Paying people is a great way to incentivize them to do work
Businesses often pay people; non-business organizations usually do not (or pay very, very little in comparison)
So… it’s less about business vs. not, and more about people power, but, yes, business is a good way to attract and organize people to achieve something.
This piece is a little bit here and there and I’m not convinced. Or I am misreading the title. Is it “Business makes better open source than open source projects” or is it “Business makes better open source, versus the same team not doing business, just doing the project”.
Sure, concourse might the current best forum software, depending on your requirements, so what? This isn’t a proven track records of “projects like concourse” working out better than “projects unlike concourse”. Also look at the contenders. phpBB is ancient compared to this and afaik there was only official commercial stuff after 10 years.
Also not everyone benefits from a company not only backing, but being the major driver. It very often complicates things. I’ve seen smaller projects being more eager to accept your patches because they don’t clash with the company roadmap (often because there was no roadmap, whatever :P)
Then again I might be biased, I’ve been involved with free and open source software for 15ish years and I don’t want most of it to be business stuff. Sure, if I didn’t have a day job I might want to work on some of the things I do in my spare time while being paid, but usually it’s participating as long as I like. I lose the most important freedom: just giving up on a bug I can’t solve or doing a chore I don’t want to do. Nobody is paying me, so nobody’s the boss.
NB: I personally dislike how concourse looks and feels, but I haven’t participated in any forums in a few years. I kinda liked the phpBB style, so if concourse could be themed to look like it, I’d probably try it.
I have actually used Discourse pretty extensively, so I think I can reasonably respond to it:
It’s never going to be “the same team (not) doing business.” The basis of the argument is that a volunteer project can’t have as many people, or as diverse people, as a business can. Volunteer projects only get volunteers, which means your contributors will always be pretty well-off, will vastly prefer doing things that are fun, will always be strapped for time, and will have to sacrifice their volunteer time in favor of work in ways that they wouldn’t have to if it was work.
Also, a business might be better able to perform usability tests, by using their money to pay people to participate in surveys. This is much better than expecting volunteers to notice usability problems, because volunteers are always familiar with the project while the usability testers that you really want are people who aren’t.
phpBB is ancient and had no commercial stuff for ten years because it isn’t backed by a business.
It probably isn’t fair to judge every project by Discourse’s success criteria. The number of sites running on the Lobsters codebase can be counted on one hand, but I’m pretty sure pushcx and companions consider it a success anyway. Discourse wants to be a standard. Lobsters, and XMonad (an example that someone else brought up) does not. That’s fine.
The Discourse team judges the quality of software by “usability”, where usability is defined by friendliness to non-contributors. I like Discourse’s interface better than GitLab’s overly-complex interface or IRC’s impenetrable wall of jargon, but I honestly think Discourse’s interface is also a bit overly complex. But that’s just, like, my opinion, man.
What matters is that Discourse’s customers get to tell them what to do. If you aren’t a programmer, but you have money, that’s a terrific arrangement.
If your goal is to be used by non-contributors, or, worse, non-nerds, then the complexity of being a company will be the least of your worries. You’re going to have to do requirements gathering. I have no idea how you would do a truly good job of that without having someone who can just tell the engineers “I don’t care what you want, the stakeholders want it to do this. Just do it.”
I’m know that feel. Solving weird technical problems can be really fun, but there’s more to designing an artifact that lots of people use than just solving the technical problems. There are companies that fail at the little stuff, too, but volunteer projects usually don’t even try.
Erlend wants people who aren’t programmers to have a say in the software that they use. You got a better way to do that than wallet-voting?
In the long run, that’s strictly better for a software endeavour meant to last and to blossom rather than wither (or to be dragged forward like a comatose horse by sheer corporate / personnel / funding force). Volunteers bring something invaluable that (on average / on the whole / excepting outlier rockstar-style specimens) salaried employees (even if the same people =) can’t bring or only rarely / short-lived: deep inshakable innate enthusiasm and self-propelled drive for the project and thus its pending task(s) at hand. This brings a completely different quality of attention to detail (helping with bug rate, robustness, polish), potentially less “rat-race/routine-boreout-inflicted carelessnesses” (helping maintainability) etc.
Sure but would view that as just a different “price” than monetary. Not so neat for “product”-type software endeavours where you want to reach stage X by date Y, but for to-be-long-lived slowly-eclosing meandering evolving-rather-than-directed projects — not a bad deal. Less hustling & bustling & huffing & puffing, fewer messes & spillage =)
I can barely parse that paragraph. Please reduce the parenthetical and slash-separated phrases.
That’s also completely missing the observed facts. Volunteer software developers, especially the maintainers and BDFL’s of the world, get burnt out all the time. Volunteer free software projects often fail for external reasons just like business projects do.
And, as I mentioned before, not every useful task benefits from volunteer enthusiasm. There’s a decent amount of drudgery even in an optimally-managed project, because sometimes the work you actually need to do isn’t algorithmic puzzle-solving. Sometimes, it’s going from person to person collecting data to help plan out the next sprint’s priorities. Sometimes, it’s putting together documentation. Sometimes, it’s running the fuzzer for a week, which doesn’t really require much work at all, but definitely requires money for all the hardware you want.
That’s why I made a point, around the middle of my comment, to say that “it’s not fair to judge every project by Discourse’s measure of success”. Writing your software for yourself with no plans to monetize it is not wrong, but it cannot, and never will be, a true replacement for the proprietary social networks in the same way that Discourse can be.
I know many open source projects that work on a “someone suggests a feature and hey, we never thought of this and it makes total sense, so we’ll do that” basis. Sure that’s not always working and often enough someone from the committers must be interested, but that happens. You also can just pay someone to contribute a feature, many projects are also open to that.
Overall I’m maybe just a little confused about the language of the post that seems a little overzealous to me. You’re kind of hitting the same tone. I really hardly had any contact with Discourse (sorry for the concourse typos..) and I wouldn’t give it the status you two seem to give it. Either party could be wrong. To me, forum software is a niche market these days, it was much, much bigger (in percent of the web, not total) 10-15 years ago. You might argue it’s the Facebook of self-/paid-hosted forums and I had never heard of Facebook…
But let me stress that I have nothing against Discourse (the software, the project, the company) - cool if it works, but I simply don’t find this a feasible MO for any random project, And I’m happy not everything is commercial, but it’s cool if some people can work on open source software as a paid job - my main problem is the structure of the project and if your boss tells you what to do, I only care so much if the result is then open or closed, although of course I’d prefer open most of the time.
What’s implicit throughout the article but nowhere mentioned in the title is the target. Business makes better open source for business, I think is probably not that controversial - it doesn’t follow that business makes better open source for everyone else as well. I doubt, for example, that every XMonad user values “sustainability” and the ready availability of consultancy/support over the efficiency gains they get from a tiling WM with their own personal workflows compiled in - otherwise they’d be using GNOME. Or Emacs vs VS Code. Or … but I’m sure you can provide your own examples.
tl;dr “mass-market-ready” is not the only axis along which one might usefully and sensibly measure “better”
Exactly, that’s why I think it’s important the distintion between Free Software and Open Source, one seems to value more the freedom of their users experience, and that’s more important than any long term sustainability, it even serves as a kind of guard against bad sustainability (even if the company goes under, I will be able to use and modify their software). The other is more interested in “usage”, and the developers freedom. The user freedom vs devs freedom battle has always amused me.
Does Discourse have an aversion to the term Free Software even though it’s licensed under GPLv2?
Not really, but the company never strictly favored one ideology over the other.
[Comment removed by author]
I really don’t like this response, it’s not a preference, it’s a different ideology. It means different things.
It was a genuine question to someone who worked with the company for four years, I think it’s somewhat related. I just thought it was a little weird is all.
Here’s what I think:
So… it’s less about business vs. not, and more about people power, but, yes, business is a good way to attract and organize people to achieve something.