Recommend Establishing Dedicated Working Group for External Contributions

[Charlie_]

Presumably the Chinese freebsd users the OP claims to represent want their changes to be integrated to a deadline in order to meet a wider project deadline, ie they are using freebsd for some other project (presumably commercial) that is being developed to a deadline and they want their new feature supported in freebsd so that they can achieve their own project deadline. Why else would there be any time pressure to integrate external contributions of this type?

I don't know if the ykla here and the ykla on bugs are the same person.

Bug List

bugs.freebsd.org

 

[T-Aoki]

That's why I don't have no Discord account.

Me, too, don't have Discord account.
But FreeBSD project has Matrix under testing.
I'm on lobby and general there, via bsd.cafe account.

 

[linimon@]

Maybe that's an unfair and very personal point of view of mine, but to me Discord is just kind of a
"social media kindergarden for gamers." That's why I don't have no Discord account.

Our particular one aims to be better. It varies between channels. I only run one of them (#project-infrastructure); in the others, I am at best an interested observer.

why not HERE?

A few years ago I personally had to cut back on FreeBSD social media/mailing lists/etc. There is simply too much for one person to absorb.

I don't think anyone is stopping anyone from trying to get folks organized here on the forums. I simply don't know enough about the social aspects here to do it myself. And, if I did, I already don't have the cycles to properly address the two areas that I try to work on (Bugzilla and the Wiki).

> As far as I remember a few years ago there was the issue to improve the handbook discussed here.

Apparently the effort stalled. IWBNI someone picked it back up. There is #documentation on the Discord but right now it is pretty much focused on the Wiki (which itself needs help).

Right now I know Alexander Ziaee is working hard on fixing up the manual pages. I don't know anyone right now who is focused on the Handbook. We certainly have "too many" non-reviewed patches pending. But this is true in pretty much every area of the project.

 

[sidetone]

My experience is that ports with one maintainer are dealt with much more quickly than ports maintained by a mailing list.

It depends. Often, a mailing list is largely neglected. Still, many lists don't have the correct kind of maintainer or even a hierarchy of whose edits are considered immediately worthy on that mailing list. Also, some mailing lists, not necessarily FreeBSD's, and sometimes upstream, have people dedicated to it by certain companies, who at least respond to serious questions and comments. A few lists on FreeBSD don't have this. Sponsorship of mailing lists by companies, where they have staff committed to them, also makes sense. This already happens or seems to happen on some software mailing lists, not necessarily on FreeBSD's.

A mailing list isn't the only way that a group can share a maintainership account through a single email, but a mailing list by a dedicated community or by company sponsorship can work in some places. Still, even for a mailing list, that doesn't have those benefits, it's much better, except in the case where a mailing list hasn't seen action in years, than simply the mailing list of ports@freebsd for unmaintained ports.

When it comes to accounts maintained by a specific software community, responses happen more quickly. These aren't necessarily by mailing list. For example when the maintainer is KDE, Ruby, Perl, Xorg or any other software group also mentioned in the maintainer email account, responses are quicker.

What if a collective group was there as a single maintainer email account, and this group spanned across NetBSD, OpenBSD and other BSD's too.

Also, various scripts in the ports tree assume that there is only one value in the MAINTAINER field, as does FreshPorts and at least one other codebase I could name.

So this is why lots of people would need to be able to share a single email account.

I have seen this handled with a comment line before or after MAINTAINER stating who is welcome to commit.

That's a good way.

As a rando that occasionally contributes patches, my main frustration is getting someone's attention to have them committed. They are usually cheerfully ignored for at least months, and sometimes years.

We are all in violent agreement, here.

Back to this, I realize that this is actually a marketing problem. When people leave or those who choose to stay are frustrated, there was a marketing failure on the product/service end. Lots of people leave silently, and don't stay around like may of us have. One publicly made complaint is estimated to have 10 times more quiet complaints of customers who left silently. Many of us won't leave. The greatness of FreeBSD's base system is enough to get people to want to try FreeBSD or stay around for a few years, but it's not enough to keep everyone.

It means, the FreeBSD Foundation needs to have more paid committers, more committees who are on payrolls of other companies, and a survey of committers and those who upload commits. Every month, a handful of source developers or committers is welcomed, though, this number needs to increase by 20 fold.

Since GIT is the way FreeBSD uses, FreeBSD needs to give attention to providing training, such as at conferences and at local FreeBSD groups. Same for the ports handbook. Maybe FreeBSD should pay someone to write a book on maintainership, aside from the FreeBSD Porters Handbook.

Having a great base system is not enough. Then, Ports is largely ignored by those who make decisions, because of how great the base system is, and the assumption that ports will automatically take care of itself. It doesn't do that enough. We also have different groups/individuals as maintainers who do their maintained ports the Linux way. Ports needs more structure as well.

There's very few pieces left of the LLVM toolchain to be completely free of GCC/GPL, so they need to pay someone to finish that up. Also, people with past accomplishments need to be rewarded, so they're not left out and not feel bitter about it.

 

[POSIX.1]

I realize that this is actually a marketing problem. When people leave, there was a marketing failure on the product/service end.

With all respects, I strongly disagree on "marketing problem". The problems in communities are not caused by "marketing", but by all-day problems on an operational level.
Marketing and advocacy are tools to cover up underlying problems of all kind for external presentation hiding problems.

Have an unbiased look at the one-way-communication of the "Foundation": Look how great we are, aren't we the greatest? Have you ever read about addressing problems in the contributing community?

If that is what you meant by "marketing problem" then yes, it's disinformation.

And BTW have you ever tried to contact someone at the "Foundation"? Do you know their email address?
And if you react on their marketing drop-ins here in the forum, did it ever happen that there was a reply?
If they are avoiding discussion here in the FreeBSD forums, are they part of the community?

 

[sidetone]

The problems in communities are not caused by "marketing", but by all-day problems on an operational level.

When you think of marketing, your idea of marketing is limited to the side you've described as marketing.

Marketing also includes, which it isn't typically thought of as, keeping customers/patrons happy. In this case, the patron is the person who submitted a patch, and got ignored. Fixing "all day problems on an operational level" for the purpose of keeping patrons who are involved satisfied or happy is still considered marketing. That needs to be addressed. For example, the person who left due to frustration, complains about FreeBSD's shortcomings or becomes a Linux advocate.

So, this is actually largely a marketing problem.

 

[bsdimp]

So some things. I just saw this, even though it's been going on a few weeks. Like Mark Linimon, I've been working for years to try to eliminate the friction in contributing to the project. We're both significantly better than 5 years ago when these efforts started (at least wrt the base system): bugs are being looked at (espeically ones with patches), the patch backlog is shrinking, puill requests land, though sometimes it takes a while. Phabricator reviews happen sometimes. 5 years ago, all intake methods were 100% dependent on a developer accidentally noticing a contribution and getting it in the tree. Now, about 10-20% of the commits to our git repo have an external source.

However, we are also significantly short of where we need to be: many of these efforts have been through herorics, not through creating a sustainable culture. It's still way too easy to lose submissions (especially on Phabricator). While we have good front lines now that we lacked 5 years ago, we also are still way too spotty for things that get past the front lines. Github isn't bad (but months-long latencies suck for the contributor), Bugzilla is pretty good for patches, though only ones that are fully or nearly fully baked: raw ones that need a lot of work still suck badly. Phabricator exists and is a big disappointment to everybody that's not already a developer. Too many contributions languish.

And our presence in various online media is spotty. I'm not a frequent forum person, since I'm a monthly or less frequent user, so I didn't see this for 2 weeks. We have some people on discord. We have several on bluesky. We had several on twitter before it became a cesspool of hate, politics and disinformation (regardless of your politics, that's why many people avoid it today). We have a decent presence on IRC We have a somewhat decent presence on Matrix. The problem, though, is divided attention: there's no unified clients for all this, and people can only check so many places. There's also a mismatch between where people with new interest go and where the old-timers that can most help hang out. Some of this is a tooling problem, some of this is a cultural problem and some of this is that "sustaining engineering" isn't many people's first choice and needs to be funded since landing random patches requiring a lof of context, a lot of judgement and a lot of connections to the domain experts that can properly review patches.

So while marketing brings people in the door, and sure we can debate that, we still need people and better coordination for external contributors, as well as finding ways to bring on more of the external frequent contributors as someone with full access. We need to find the right way to articulate the job we need done so that the Foundation can fund it (to date, they want to help, but the inability to define a job role specifically has been a problem for me getting someone working on the problem). Marketing isn't about keeping people happy: that's support/sustaining efforts. That works hand in hand with Marketing, at times, but takes different skills, different people and different tools. Pinning a sustaining engineering failure on marketing just means the problem won't get addressed. But I'm not interested in labels so much as solving the root problem: there's too much friction in contributing, and the project's limited resources could be used more efficiently with better effect.

I'd love more help coordinating efforts at making the patch / contribution intake process easier and better. While we've come a long ways from where we were, the journey has just started and the team needs more efforts from more people. Please talk to me if you can help.

 

[POSIX.1]

When you think of marketing, your idea of marketing is limited to the side you've described as marketing.

Marketing also includes, which it isn't typically thought of as, keeping customers/patrons happy. In this case, the patron is the person who submitted a patch, and got ignored. Fixing "all day problems on an operational level" for the purpose of keeping patrons who are involved satisfied or happy is still considered marketing. That needs to be addressed. For example, the person who left due to frustration, complains about FreeBSD's shortcomings or becomes a Linux advocate.

So, this is actually largely a marketing problem.

Thanks for clarification, but I'd prefer another wording when addressing real-world-problems.

 

[bsdimp]

We absolutely need young people to get involved. FreeBSD ultimately belongs to the younger generation. Without their participation, and without more external contributors becoming committers, nothing else can truly begin.

Please send a similar message to srcmgr@. They are the ones most likely to help with at least src contributions and ones that are driving the github experiment.

 

[linimon@]

Sponsorship of mailing lists by companies, where they have staff committed to them, also makes sense.

In my experiece in watching Bugzilla, most of the company-related things are either a) related to support of their products (hardware or software), or b) bugs in the base system.

I do not know of any commercial company specifically funding ports work. I really don't think this idea would work for ports.

It means, the FreeBSD Foundation needs to have more paid committers, more committees who are on payrolls of other companies, and a survey of committers and those who upload commits.

IIUC this is the first year that the Foundation has had enough funding to work on more than a few very specific projects. I think we should give them more time to see if this bears fruit. (Disclaimer: I am one of the people being funded.) Also, we need to find out if the current jump in funding will be continuing or is more of a one-time event. If it's the former, then it would be useful to consider such things.

My understanding is that the Foundation has traditionally only responded to very targeted efforts to fund. "Just fix things" has never been in scope, I think.

 

[bsdimp]

Ah, another to note.
If something like gtihub.freebsd.org that its accounts are managed by FreeBSD project and/or FreeBSD foundation, without sharing my account info with GitHub.com, is possible, I'll be happy to register and keep on contributing.

There will never be a github.freebsd.org. In so much as we will never run the on-site github tooling.

However, we're working on a forge system that you can push branches to, get reviews on those branches and have a final adjudication of the commit all in one place. We'll keep most information here project-only.

However, since we publish our repo various places, the information that's in the repo is 'shared' with others. But it's all public that anybody can get, and there's no 'private' sharing of 'extra' data by the project. Nor will there ever be.

Long term, this will happen. However, it's not specifically funded, which means timelines are difficult to commit to (and it's not me doing the work).

 

[bsdimp]

My understanding is that the Foundation has traditionally only responded to very targeted efforts to fund. "Just fix things" has never been in scope, I think.

Yes. To get funding, we need to be able to articulate a need. "Fix wifi," "fix graphics," or "fund the security officer" are all relatively well constrained needs that have actionable requests with measurable results. Also, these are all things the Foundation has funded. They have also solicited grants from others to support things like Mark's work on bug-busting, the development of 'dashboards' that harvest data from bugzilla and others to help people in the project manage the backlog, etc.

I've tried for years to narrow the scope of the "external submissions are badly broken" problem to find the right subset of the problem that would create a sustainable framework that needs only a little Foundation money each year to get the maximum benefit of the community of volunteers. There's so many aspects of what could be done (in fact, the bug busting and metrics work will help this problem), but we've still not found the right thing to fund. We need a project manager that can act as a backstop for dropped balls, who can coordinate everybody wanting to help, manage tooling improvements, manage and recommend recommendations for "forge" setups to use and who can help build community to make it sustainable. But these are about 10 different skills for just these three things, and they rarely overlap in one person in the Foundation's price range. So I've been looking for ways they can help solve this problem, but I've not come up with a good enough plan to get funding. There's certainly an openness to fund the right things, but nobody's been able to articulate something that can be funded successfully.

 

[Jose]

Me, too, don't have Discord account.

I use Discord to talk to my kids. It is a closed, proprietary platform, wholly owned and operated by a for-profit company subject to being bought by some other for-profit company at any time. I would never depend on any such for an open source project.

 

[ralphbsz]

I use the telephone to talk to my kid. It is a closed, proprietary platform, wholly owned and operated by for-profit companies, which get bought by others regularly (GTE -> Verizon -> Frontier, PacBell -> AT&T, Sprint -> T-Mobile, and so on). Yet all of the open source community depends on the same phone system. As does closed source commercial software, McDonalds, General {Motors,Electric,Atomics}, and much of the interaction between humans.

Software development is not an ideological purity test. It does not exist within a anti-business utopia, but in the real world.

 

[Jose]

I use the telephone to talk to my kid. It is a closed, proprietary platform, wholly owned and operated by for-profit companies, which get bought by others regularly (GTE -> Verizon -> Frontier, PacBell -> AT&T, Sprint -> T-Mobile, and so on). Yet all of the open source community depends on the same phone system. As does closed source commercial software, McDonalds, General {Motors,Electric,Atomics}, and much of the interaction between humans.

You have many options for phone carriers. Say T-Mobile gets bought by Google, and the latter decides to make you listen to an ad before you can talk to your kid. You can drop them like a hot rock and move on to some other provider post-haste. You can even keep your phone number nowadays, unlike in the olden days.

Say Discord decides to start sending ads to everyone in your contacts list without even notifying you. Maybe they decide they don't like the way you write, and start "enhancing" it with AI. Can you grab your contacts list and decamp to some other platform? Can you talk everyone you know into using the new alternative?

 

[ralphbsz]

You have many options for phone carriers.

For cell phones, that's only sort of true. In the US, we have a triopoly, but in some areas service from some of the carriers is awful; a friend of mine used to always complain to be stuck on AT&T, because his house had no other service. And that's in Los Altos, one of the most expensive suburbs of Silicon Valley! But you're mostly right.

For landline (copper connection), which is still the cheapest and (should be) most reliable option, in the US we have a monopoly with no choice.

For internet to the house, most areas have a duopoly: Either DSL (from the landline phone company), or cable modem (from the cable TV company). Both monopolies are upgrading their system to fiber today, but it's still a duopoly. For extra $, there are other systems (802.11 based, geostationary satellite, or Elon's Starlink), but those are usually not competitively priced.

And all these systems are closed source, closed systems (except landline phone).

Say Discord decides to start sending ads to everyone in your contacts list without even notifying you. Maybe they decide they don't like the way you write, and start "enhancing" it with AI. Can you grab your contacts list and decamp to some other platform? Can you talk everyone you know into using the new alternative?

For a relatively small group, of technology-literate people, such an open source project, it would be easy to switch. Don't like Discord? Try Signal, Telegram, WhatsApp, Reddit, Instagram chat (yes, such a thing exists), and so on. But where the problem exists is with large groups of people who are not technology literate. For a while, I was using nearly every possible chat mechanism, because some were popular in different geographies (WhatsApp seems to have about 99% market share in South America, where my parents were in a nursing home and I had to talk to the staff all the time), some in certain communities (lots of techies like Signal, for obvious reasons), some have great facilities for certain uses (such as Nextdoor for neighborhoods, to organize fire preparation, road cleanup, and summer BBQs), and so on.

But the important thing is: Most of those are "closed, proprietary platform, wholly owned and operated by a for-profit company". Their shortcomings are not because of those attributes, but because some are badly managed.

 

[tingo]

For landline (copper connection), which is still the cheapest and (should be) most reliable option, in the US we have a monopoly with no choice.

Just in case you didn't know: quite a few countries have done away with landlines. Yep, no wired connections to be had here in Norway for example. It is all cell phones.

 

[cracauer@]

Just in case you didn't know: quite a few countries have done away with landlines. Yep, no wired connections to be had here in Norway for example. It is all cell phones.

What do you connect your modem to then? Or your acoustic coupler?

 

[cracauer@]

FreeBSD has this exemplary forums. Not seldom GhostBSD, TrueNAS, Linux, and other users come here willfully ignoring the rules because they find no other place to get their problem solved as here.
Besides the mailing lists with this forums FreeBSD has its own vital "social media", according to the number of views read by many, many more people as write here.

Sometimes I get that feeling by the officials this forums is kind of a unpopular side show. Sometimes some official from the foundation drops something here, closed for reply, "here! read this!", and that's it. 'Oneway communication from above.'
Especially when I read articles under the large topic "get in touch with the comnmunity" I don't get it. Where is the community if not here? (Discord. right.)

I totally agree that those closed-to-reply posts need to stop. They just rub people the wrong way. Opening another thread for the purpose of replying has happened exactly once as I recall. Gotta stop.

The text should be inline and it should be open to replies (with quotes). If the author isn't here to defend - that's too bad. It's their choice,

 

[cracauer@]

Have an unbiased look at the one-way-communication of the "Foundation": Look how great we are, aren't we the greatest? Have you ever read about addressing problems in the contributing community?

If that is what you meant by "marketing problem" then yes, it's disinformation.

And BTW have you ever tried to contact someone at the "Foundation"? Do you know their email address?
And if you react on their marketing drop-ins here in the forum, did it ever happen that there was a reply?
If they are avoiding discussion here in the FreeBSD forums, are they part of the community?

I totally agree that those closed-to-reply posts need to stop. They just rub people the wrong way. Opening another thread for the purpose of replying has happened exactly once as I recall. Gotta stop. The text should be inline and it should be open to replies (with quotes). If the author isn't here to defend - that's too bad. It's their choice.

Contact is doable and should be no problem now that John Baldwin is on board.

 

[sidetone]

I kind of understand how they'll make an official post and leave it closed to replies. Some replies pollute it, and in the past, there have been arguments. Some intentionally waited for the weekend when the Custodians were gone, and started drama, then those threads got closed.

In another way, making an official post with no replies allowed was inconvenient. I wonder if there's some middle ground. I don't mind starting a new thread about discussion of those, but it's a bit inconvenient to the flow of the official article to the discussion. Perhaps a reply there should automatically go into a discussion thread, with the link already there, even if it's a empty link. Or maybe those need moderation just like the how-to section. Though, I think they don't like having to excessively moderate everything to where it's like babysitting to them.

 

[SirDice]

Sometimes I get that feeling by the officials this forums is kind of a unpopular side show. Sometimes some official from the foundation drops something here, closed for reply, "here! read this!", and that's it. 'Oneway communication from above.'

Ehm, no. Those are posts in "Blogs and newsfeeds".

Thread 'Blogs & Newsfeeds'

May 23, 2013

This sub-forum is filled automatically by RSS feeds from several external news/blog sources related to FreeBSD specifically, or to FreeBSD's 'open source and computing environment' (e.g. Netcraft).

It functions as an additional source for FreeBSD and FreeBSD-related news.

Since all of the articles posted in this sub-forum come from external sources and not from forum members or staff, responding to these posts will not notify the author(s) of the news article in any way.

New news/blog sources will be added over time.

• DutchDaemon
• Replies: 0
• Forum: Blogs and Newsfeeds

• 

 

[Erichans]

Those are posts in "Blogs and newsfeeds".

Being an external source, it is however a very official FreeBSD source: the FreeBSD Foundation. IIRC, it even supports these forums, financially or by allocation of other resources.

Isn't it possible to harvest this particular source and present it in a fashion open for replies?
Perhaps with the caveat that it isn't posted explicitly and therfore the indicated 'poster' will likely not offer anything further to the thread.

 

[Jose]

For landline (copper connection), which is still the cheapest and (should be) most reliable option, in the US we have a monopoly with no choice.

For internet to the house, most areas have a duopoly: Either DSL (from the landline phone company), or cable modem (from the cable TV company). Both monopolies are upgrading their system to fiber today, but it's still a duopoly.

My landline is provided by my ISP, Sonic.net, and comes into my house over the fiber line they installed. The situation is not quite as dire as you paint it.

 

[ralphbsz]

(Warning: Topic shift!)
I would LOVE to get a fiber from Sonic. They have been our "ISP" forever (about 25 years), but most of that time sadly only in the sense that they do the hosting and e-mail. They have never been able to also be our bandwidth provider. The reason is that while most of California is served by one very large monopoly carrier (PacBell = AT&T), we live in an area served by a small monopoly carrier (GTE/Verizon/Frontier), and Sonic has never been able to work with Frontier, who is a highly dysfunctional and incompetent local phone company.

We've gone through a variety of bandwidth providers, starting with 56K modem (that was via Sonic), then ISDN from the phone company (worked OK), then a local 802.11 coop (about 2 nines of uptime, meaning highly unreliable), then DSL from the phone company (about 3 nines of uptime, and always down when the power is out), and now we are using a good commercial 802.11 service (which has been nearly flawless in the 3 months we've been on it). For "landline" phone, I'm now using a reasonably inexpensive VoIP provider, so we kept our old phone number, and all the hardwired phones in the house work. Actually more reliably than before, since both our internet bandwidth as well as our house-internal network have good power protection (battery back up and generators).