In the olden days, a FOSS (Free/Open Source Software) project typically had:
-
Jan Wildeboer 😷:krulorange:replied to Q last edited by [email protected]
@0leil And I will apologise for my brutal language (I am dutch, though that is not an excuse that allows for misinterpretation. I'll grab a beer at the beer garden now and will start to translate this rough thread to a shiny blog post, taking into account everything I have learned from all of you sharing their critique and thoughts! @lclapp
-
@kevingranade @jwildeboer The issue is that most FOSS projects are underfunded, understaffed, overworked or its devs/maintainers on the brink of burn out or - most likely - all at the same time. Having more contributors and especially retaining them is crucial, so it's not always easy to say "please, more contributors" and then shun people away with "yes but not by changing our contributor workflow". If we keep hearing from different people we lost potential contributions because the contribution workflow was unacceptable to them, we have to decide whether the loss of their theoretical contribution was worth it. I don't have the answer.
-
Jan Wildeboer 😷:krulorange:replied to Q last edited by
@0leil Burnout in FOSS projects is real and must be addressed. One part of reducing those risks is to focus, focus, focus. I personally find any for4m of chat system a distraction. It uses up my time in a way that rarely results in code for the project. So my radical position is: no chat stuff at all. If that means losing potential co0ntributors, so be it. They may not be the people I want anyway. WRT underfunding, understaffed: A complicated topic that can't be solved in toots @kevingranade
-
@miah @jwildeboer Sure, IRC runs on anything, but it still doesn't work for most people. I say this as somebody who's spent over a decade on IRC.
To get the very, very basic, not at all optional feature "message history", you need to run a bouncer. Which is a tall order for most people.
The amount of people on computers too slow for Matrix is much smaller than the amount of people who can't/won't run a bouncer.
-
@muvlon @jwildeboer As stated earlier. Message history is not a required feature. You simply never go into a social event and ask people for their discussion history for the evening.
I don't know why people using irc think this should be the norm. If you want message history, there are frequently logs for heavily used channels.
Let me ask this question, do you read _every_ line of text that was written while you were absent? And then I'd ask "Do you _need_ to read every line?"
-
Jan Wildeboer 😷:krulorange:replied to Miah Johnson last edited by [email protected]
@miah Exactly. Chat, no matter in which way it is implemented (IRC, Discord, Matrix, WhatApp (though, gasp ;)) should never be critical to any project. It is definitely helpful to solve stuff fast, but it should not go beyond that. The real work (and time) should go to documentation, FAQ AND CODING @muvlon
-
Dave Lane 🇳🇿replied to Miah Johnson last edited by
@miah @jwildeboer with Matrix - yeah, it's a bit more involved, but you can join from *any* Matrix instance, if you have found one you trust. It's #libre technology, so it's consistent with libre dev. Discord is a non-starter for libre projects in my opinion. I agree with you Jan - wrote this a while back: https://davelane.nz/notslack
-
Wanjareplied to Jan Wildeboer 😷:krulorange: last edited by
@jwildeboer @miah I never said it should be *critical*. But it not being critical is simply not enough of a reason to make it arbitrarily *worse* for everyone except nerds like me who run their own servers.
No, I don't ask everyone in a social even about the discussion history, but if I could have that context without bothering everyone, of course I'll take it! And this is way more valuable than being able to use it for a retrocomputing hobby.
-
@muvlon @jwildeboer "worse" is subjective.
Lots of popular channels have fully online and searchable chat histories. If you want that, its a perl script away from running on your own server and channel. This technology has existed as long as I've used IRC.
Equally, you'll also meet people who do not _want_ logs for privacy reasons. You cannot please everybody.
IRC lets you handle it any way you want by being as basic as possible.
-
Jan Wildeboer 😷:krulorange:replied to Dave Lane 🇳🇿 last edited by
@lightweight Sure. That can be done. But it's another barrier to entrance that will keep more people out than in. The more I think about it, the less I see the need for chats in a project. It has a lot to do with my personal experiences where the chat is filled with people, well, chatting about a lot of things unrelated to the code I'm interested in. That's why I personally prefer to talk in issues and pull requests @miah
-
@miah @jwildeboer Sure, it's subjective. But when I introduced people to IRC communities (again, I tried for 10+ years to make IRC happen), the universal response was "wait, I need to run a bouncer if I want message history?". Practically speaking, by insisting on IRC today, you're driving a ton of people away.
-
Jan Wildeboer 😷:krulorange:replied to Wanja last edited by
-
Jan Wildeboer 😷:krulorange:replied to Jan Wildeboer 😷:krulorange: last edited by
-
Klaus Steinreplied to Jan Wildeboer 😷:krulorange: last edited by
@jwildeboer
Do you include issue tracker/bug/whatever as part of the source code repository? Traditionally those were separate.Because I think we should _have_ some issue system somehow attached to a source code repository without relying on a central third party server (like github/gitlab/…), and we don't have one.
-
Jan Wildeboer 😷:krulorange:replied to Klaus Stein last edited by
@Lapizistik That's where #Fediforge comes into play (as mentioned later in the thread). We are on our way to better solutions, IMHO.
-
scunneen (he/him) 🚋replied to Jan Wildeboer 😷:krulorange: last edited by
@jwildeboer Why does using Github exclude people?
-
Jan Wildeboer 😷:krulorange:replied to scunneen (he/him) 🚋 last edited by [email protected]
@scunneen I said "excluding a lot of potential contributors", not people. As contributor you must have a GitHub account and thus accept Github's terms and conditions. If you don't agree to those T&C's (which includes giving them the right to use your code as input for their AI), you cannot contribute.