Re: [Zephyr-tsc] [Zephyr-devel] Highlights from the TSC meeting during ELCE
Adding to Martí’s points here.
Can’t seem to break lines with Outlook so I will need to use another color. Apologies for that, but I blame whoever introduced HTML to this thread.
From: tsc@... <tsc@...> On Behalf Of Marti Bolivar
Sent: 29 October 2018 21:11
To: Flavio Ceolin <flavio.ceolin@...>
Cc: Perez-Gonzalez, Inaky <inaky.perez-gonzalez@...>; Nashif, Anas <anas.nashif@...>; devel@...; tsc@...
Subject: Re: [Zephyr-tsc] [Zephyr-devel] Highlights from the TSC meeting during ELCE
I'd like to discuss some counterpoints.
On Mon, Oct 29, 2018, 6:38 PM Flavio Ceolin <flavio.ceolin@...> wrote:
In all honesty I think
And I agree.
IRC is not "easy" across platforms in a modern sense of the word, unless you use irccloud (which, full disclosure, I do, after changing from ERC within emacs by way of various other clients starting with Ircle on pre-OS X Macs back in the day).
Note irccloud is not free software.
Not only that, but IRC doesn’t provide you with a permanent connection unless you pay for money in one way or another, unless you use something like matrix and that is not very pretty.
I think this statement also has some Linux bias. Zephyr is a Linux foundation project, but it's also important to be able to develop using Zephyr and collaborate with other users on all supported platforms, and convenience and familiarity do have some practical weight here.
IRC clients that look good on a modern desktop (again, other than irccloud) are lacking. (Yup, that's an opinion.) Mobile support without irccloud is also lacking.
Again, I completely agree with Martí here. While HexChat is usable, it definitely does not fit the category of “favourite messaging client” for most people. Mobile support is a pretty fundamental feature these days, and our users do not currently have access to it unless they use irccloud or a similar service.
And it's not like irccloud is exactly a household name.
No, but to be fair, it is very reliable and dependable.
Taking "everyone" by raw numbers, we'd be looking for WhatsApp, WeChat, etc. integration, and I don't think their clients can be called "well" integrated with IRC.
So the above statement seems suspect to me.
That said, preferring open and battle-tested standards is usually a good idea in an open source project, at least so long as they get the job done well enough.
And as long as they provide the functionality we need (i.e. permanent connections) for free for users. Not to mention the advanced functionality available on Slack that simply will never make it into IRC.
Slack is a proprietary de facto standard in this context, at least in the west. IRC is a venerable and interoperable open standard with usability issues and mindshare problems depending on who you're talking to. That seems to be the real crux of the matter here.
There's good arguments on either side of this debate, but I think we ought to be honest with ourselves that this is really what we are arguing about.
I think we should carefully consider what our users would like to use as well. IRC is not only a tool for core contributors, maintainers and TSC members, but also users of the RTOS. The sentence “oh, but IRC still *exists*” has come up too many times in the last few months while introducing engineers to the Zephyr project.
Now, how to actually do that is not trivial, but I’m thinking perhaps a poll amongst users of some sort?
Full of spam (and, let's be frank, some of that spam is hate speech), it's 2018 and slack is lightweight enough for all the laptops we tend to use, lacks native support for anything that isn't plain text, ...
I hope the above is some fodder for discussion on why this is not a no-brainer decision.
I hope so too. While I personally have nothing against IRC and it pretty much does the job for me, not everyone wants to set up a proxy or pay for irccloud in order to get the functionality they need.