This starts off strong, but then ...
>I'm 39. I've been in technology in a professional capacity for 22 years, longer now than perhaps a few folks on HN have been alive. I wrote my first computer programs over 30 years ago, started online with local BBSs and then eWorld and then the internet.
If I were typing this in slack, this is where the first :facepalm: emoji would appear.
This is the mentality that caused the now-semi-famous HN post in the early days of dropbox, arguing that it was useless because with a combination of [ simple CLI tools that nontechnical people have never even heard of ], you could create the "same" functionality.
Say it with me: Being so simple that utterly nontechnical people can use (edit: not just 'use', but 'deploy and administer') IS important functionality. That's what made Dropbox, and it's what's made Slack. You don't have to be an IT pro to deploy it successfully. You don't have to be a programmer. You just click half a dozen buttons, all of which are helpfully color coded so that your unwillingness to read any text on the screen won't be an obstacle.
Yes, there are things you lose when you "build infrastructure out of walled gardens" -- but the reason some people prefer cathedrals vs bazaars is that they often have much lower barriers to entry, and for many people that is a worthwhile tradeoff.
> the reason some people prefer cathedrals vs bazaars is that they often have much lower barriers to entry
There's nothing inherit about these "cathedrals" that require them to be walled gardens. Slack could have been an open protocol, but they decided not to be.
But would you say that they're _not_ investing in the product? They're improving on search, their API, adding new integrations with Github, etc. Those might not be the things that matter the most to you or other HN users, but they are investments in the product.
Have you used their client?
nit: inherent.
That's a strawman - no one said dropbox was useless to everyone: this is not about how simple it is to replicate the functionality, but who controls the data. The point gp was raising is that surrendering control your data to a 3rd party company has costs most people do not consider until it's too late - and this happens time and again ("oh no, this shit again"). How many stories have you read about downtime, account bans/freezes, and companies winding down or sold with little or no notice and no way to retrieve your data? Is it a worthy tradeoff for a shiny UI?
But, they handled it well and they've never screwed over their users and they have partially resolved the long-standing file transfer problem (https://xkcd.com/949/).
So I still have those concerns, but overall it's a great service and I've recommended it to a few folks over the years.
It helps also that there are some open alternatives on the scene now, so if you make Dropbox some critical part of your infrastructure, it's feasible to switch to something else if Dropbox suddenly decides they want to focus only on the smartphone market.
Slack had the ease-of-use and feature advantages -- which I acknowledged -- but it was too young and there were no open alternatives that matched it feature-for-feature. That made it dangerous to rely on too much. i.e., I'd still use it, but I wouldn't make it the de facto communication system for a company.
So yes, a bunch of "old farts" are a good a necessary corrective in our hype-driven environment. I agree with the assessment that slack is just another walled garden and comes with all the problems those bring. I even agree with the "open protocol, optionally self-hosted" conclusion. I just don't like the conclusion (not explicitely stated by you, but definitely by others) that that open protocol should be IRC and email, and that the UI for them needs to be pure text.
Put a bit shorter: the people have chosen between $SLACK over $OLD_STUFF, so if you want to use something open and still sway them then you'll have to change what's on offer vs slack.
Converse.js looks like a prime example of doing just that, as does e.g. conversations. I'd much rather we get our shit together and build jabber into something usable and beautiful than go back to IRC and email.
I work with a guy like that. I've got a lot of respect for him, but yes, his approach to things causes a lot of headaches for other techs in the company. (He likes to send config files by email ... inline, after gzipping and base64-ing them.)
There's a middleground. If more people tried harder to rely on software they could truly own, vs SaaS, it would force more companies to release open source APIs and libraries and all but the crankiest old farts would shut up.
I like things to be easy too, but not so much when it bites me in the ass later.
Your suggestion for extending jabber seems reasonable.
Year of the Linux desktop never happened, KDE semantic desktop never happened, Python in Chrome never happened, Self never took off, Lisp never took off, we're still writing coffee like we did in the 90s. Nothing happened as much as we hoped it would and we have to just go with what we've got.
This year I'm dropping Linux at gone and going to Windows, after 15 years.
Kinda funny since I feel like windows is taking more and more weird turns since windows 8 regarding UX, UI consistence etc. From win 95 to win 7 it mostly felt like consistent evolution, but now it seems they tried to modernize different parts of the OS over night and none of the designers and devs involved were allowed to talk to each other.
I think the "fuck it I'm old, don't care anymore, just want it to work and look nice" goto solution is still the Mac, even in 2018, even after all those recent os x mess ups (I'm old, don't care, etc., get it?)
I think there's a simpler explanation: people are trying to find new business models. The old model of individuals paying tens or hundreds of dollars for software is slowly dying, and it's on full display in the iOS app store where even $5 is perceived to be a lot to charge for an app. So we need new models. The rise of SaaS is tied to that shift. in-app purchases and related psychological shenanigans are related to that shift.
The SV elders, like Peter Thiel of Palantir, now more than ever seem to favor rent-seeking business models and started selecting for those businesses 5 years ago. To course-correct, someone in technology will have to discover a new business model. I'm personally betting that we'll see a backlash and new models wherein people pay so there is no chance of data harvesting (going back from SaaS to standalone software)
I don't believe there is a dichotomy between IRC and Slack. In fact, Rocket Chat and Mattermost show that locally hosted platforms can be developed openly and serve a similar purpose.
What has been will be again,
what has been done will be done again;
there is nothing new under the sun.
The truth is that no closed platform will ever reach the universality that email has. The prevalence of various chat platforms looks like a race to replace sms. At the current rate, there may be no winners, and only losers.
I'm 39. I've been in technology in a professional capacity for 22 years, longer now than perhaps a few folks on HN have been alive. I wrote my first computer programs over 30 years ago, started online with local BBSs and then eWorld and then the internet.
I've seen a lot of stuff come and go. The tech industry as a whole has, over the last decade, adopted an especially frenzied pace that looks an awful lot like someone lost in the woods, freaking out, running in different directions trying to figure out where they've been and where they're supposed to be going.
There's nobody I can wag a finger at and say, "see, you didn't learn the last time 'round", because with a constant influx of new people, there's always a big group of folks that will adopt the latest thing because it's the latest thing, and have no memory of the last 30 years of latest things, and haven't been stung before when the latest thing fizzles out and takes a chunk of your time or infrastructure investment along with it.
Slack had advantages, sure. But it's hardly the first time that some service has come along and offered advantages over the tired old thing -- all except for a public, common protocol -- and then decided to take their ball and go home and to hell with everyone else. My staid desire for things being built on open protocols and available to people who want to self-host it doesn't come from some hardnosed ideology, it comes from years of, "oh no, this shit again."
Unfortunately, comments like yours tend to dismissively shout down the comments from the been-there-done-that folks, and so we're locked in a particularly self-defeating Eternal September that's costing the tech industry an incalculable amount of money.
In my more cynical moments, I wonder if a cadre of old farts oughtta start getting together and giving the less experienced folks the advice they want to hear, just to see how long it takes them to catch on: "building infrastructure out of walled gardens is a good idea, because they are commercial businesses and will be able to offer you more features than the slow-moving open source alternatives that haven't changed much in a long time".