I’m in a slack chat for a couple of days that is quite busy (coordinating 3d-printing face-shields etc.) and I agree that the threading is usefull for not flooding the chat on a very busy server.
At the same time I found the threading is “in the way” of just having some idle conversation.
It feels more constrained.
As in: asking myself before writing if whatever I want to write is worth making a new thread.
One other thing I don’t like is that I have to click on the replies and scroll in a seperate area to see what was writen. I find it easier to just skimm down a classic un-threaded chat.
Should a chat system steal features from a forum?
For example: Files. I find the file view of a Slack channel most unhelpfull. Its just a jumble of anything that was uploaded by anyone.
On the plus side, you can upload something.
But now you have 2 places where people can upload files. is a chat the right place for that or would that better fit a forum?
I personally just don’t like slack very much and would prefer a more chat-focused messaging platform.
And use the forum for anything more complex.
We can add one bot that will pull all logs from channels and store then on S3 bucket that will be available as S3 static website. It will cost cents per month to store the searchable archive ( 2.5 cents per GB per month) and I expect not much of traffic there. So if this is the option I can host it on my AWS account.
TL;DR Give us a couple of days to evaluate the alternative.
As a bit of story, there was a slack in the past created by the someone in the community, but no everyone at that time preferred to use it. Hence we remain with IRC and Gitter as main channel.
We don’t want to invest energy in maintaining an self-hosted installation. For example this forum is hosted by discourse servers and there is already some moderation tasks to be done.
Although I personally feel comfortable with Slack since I have already multiple workspaces, and I can accept a limited history in Slack; I undersatand that is not the case from everybody.
Zulip free tier also has a limit history. But they seem to offere the Standard plan for open source projects.
I have no experience with Zulip but it seems worth a try.
I like that other projects could benefit from this. I created crystal-db channel in Gitter in the past but there was almost no activity; maybe a discoverability issue.
Again, gives us a couple of days to evaluate the aspects discussed here.
I am here with @asterite
Choosing something mainstream allows you to fulfill the primary goal and that is to grow the community.
You target people that already have an account somewhere and dont need to create one just because of some “obscure open source project”
I don’t follow what you mean? @asterites proposal to use slack would mean everyone had to create a new account (and that even through an invite-link hack).
To clarify: I don’t mind having to create a new account. There’s probably no way around that anyway.
Especially easy to use would be if it works with federated authentication (like Github login). AFAIK slack doesn’t do that (but zulip does).
My choice of Slack is that I already have it and I prefer to have as less messengers as possible so Slack will be preferable for me, all other niche messengers will be used less likely by me.
This is my main reason for wanting Slack. It’s the tool people are most likely to be using and so having Crystal right there with their other Slack may encourage a lot more discussion
What, have Slack changed something? As far as I am a aware, Slack accounts are bound to workspaces. You actually have a different account on every workspace. I’m pretty sure that’s still the case. I have different login emails and passwords for every workspace. Not saying this is bad per se, just worth noting. Slack makes it easy to switch between workspaces when you’re logged in on the same client, so it’s not a huge a usability issue. Still, I have to separately login to all workspaces on each client.
A lot of people using something like Slack is not a strong argument.
The same can be said for Crystal.
Why the language is open to the community and not just openly released, a la Android/Chromium, or proprietary?
I am not the creator of the language, but I guess they believed FOSS would have more benefits than proprietary. Manas could have perfectly make profit on this language.
It is logical to favor other projects whose people have done the same choice.
Having said that, Slack and Zulip are quite similar at the end.
Because it should not be about ideology but about getting more users. Even if they are not likely to contribute to the stdlib or compiler they might create one or two libs that enlarge the ecosystem.
FOSS does not really work that well. At this point it benefits the corporations more than it benefits small developers like us here.
Profit was and is never on our mind. We make Crystal because we like it. Open source is more fun than proprietary, and you get to grow faster (more people can contribute).
But Slack works already just fine, and in my opinion much better than the alternatives.
If you want us to contribute to open source… we are already doing that! With Crystal. But there’s no need to always use open source. It’s not like we are selling our souls to Slack.
@ganymedes to get even more users, Crystal should be sold and backed by a big corp.
Open Source is not blind ideology, keep it mind that using other Open Source products is also a form of publicity. Usually Open Source softwares are using each other, it is a big community.
I just say Slack, Zulip, Mattermost and even Matrix are not far away from each other.
Just try them, you’ll see (Matrix is a bit special).
Is the only reason for choosing Slack is “I already know it”? If we “learn” an other one, this argument is no longer valid. All have a browser version, the “I already have the Slack app installed (which is in fact a web-view in a browser)” is also invalid.
Look how big of a time waster this is. You can have whatever the mainstream option is and then create bridge bots to satisfy those that dont really like those mainstream options.
I would not mind more corporate Crystal. We’ve become to have unrealistic expectations from software that is created by a bunch of volunteers.
We started to compare things that have hundreds of millions of dollars in backing with things that simply dont.
Let’s see if Crystal could get the Zulip standard plan, as @bcardiff said.
It would be great not to have free plan limitations, especially the 10 000 messages history.
Whatever happens, please don’t turn into Godot where there are a zillion different communities. I think having one main forum and a place for chatting is ideal, JMO.
I think it is already a big task to learn a small still growing language, it’s ecosystem, its tools. Then asking to use install a new chat is another thing they have to do. Nearly everyone uses Slack for work (whether they like it or not). So using another Slack workspace is a great way to keep it right there where they can check it and stay involved throughout the day.
There is also precedence here. Lots of other communities use Slack despite its flaws to great success. There may be some good reasons to use Zulip, but I think the primary goal here should be how familiar it is, how easy it is to get started (GitHub, invite link, etc.), and not introducing yet another new thing so there is less friction.
I’m pretty strongly in favor of Slack at this point for those reasons. We can add extensions to Slack for chat history or whatever else if we really need it