On the one hand, Turing their back on pretty much everything everyone liked about it because could be seen short sighted, and it will crumble.

Or an intentional pivot. Knowing a subset is locked in and can be exploited to grow in new directions.

Either way, the shift is kind of epic. And only seems to be gaining steam.

(Regarding acquisitions of Heroic, Sendgrid, Slack, Tableau, Mulesoft, and most recently Informatica...)

For those less-familiar with the reference, the Wikipedia entry[1] tells it well:

In 2001, The New York Times wrote that "Computer Associates has infuriated clients with high prices and poor technical support." Fortune wrote, "For all its ubiquity inside the tech departments of corporate America, CA had a horrendous reputation. Where Microsoft has long been the most feared software company, the old CA claimed the title of most despised – not by competitors but by its own customers."

Detractors of CA accused it of putting newly acquired software products into maintenance mode and milking them for cash flow. The products themselves were expensive and central to what corporate IT departments were doing, and so customers found it difficult to move away from CA. As Fortune wrote, "These products made it the barnacle of corporate America: Once you had CA software onboard, it was so onerous and expensive to pull it out that few customers ever did. That led to a lot of steady cash flow – and to arrogance on the part of CA's management." Or as The Register wrote, "CA used acquisitions to grow its portfolio.... Along the way it acquired a reputation as the place decent software goes to die."

[1] https://en.wikipedia.org/wiki/CA_Technologies

I urge every user of Hacker News to read Peter Thiel's book, Zero to One. It's the definitive statement on software capitalism.

The goal, which Thiel embraces unabashedly, is to use technology to create new and unique monopolies, and once you've created them, extract as much rent as possible from the users. Obviously the users hate that part once it kicks in.

Thiel really seems to believe this is a good thing and there's a sense in which he's right: the tech industry has created more gadgets and created (or consumed?) a level of economic activity on par with industrialization itself. We have been introduced to all manner of innovations and conveniences, and the winners at this game have won bigger than anybody else.

But it is undoubtedly anti-consumer and anti-user. They give you something good, you get hooked, and then they enshittify it once you can't get out, and it's all part of the plan. Again, and again, and again, for more than 40 years now.

That's why once you're done with Thiel, you should read the GNU Manifesto. Richard Stallman identified the basic dynamics here as far back as the 1980s, and started his movement from the perspective of a user of computer systems who didn't want everything to be trapped and enshittified once again. By encouraging programmers to adopt the GNU license he aimed to prevent the rent seeking stage of this process.

Both camps succeeded partially. Thiel's camp succeeded more, especially economically. Which camp you join is up to you when you write a line of code or you use a piece of software. I personally think the world is complicated and there are elements of value in both. Regardless these are the two written works which together will give you the full context about the software industry, how it works, how it got this way, and even why modern life is the way it is.

And then you will see how it is by design for Salesforce to fuck nonprofits because it works. It was in the plan from day one. They knew. They will do it again.

https://slack.com/help/articles/201658943-Export-your-worksp...

When the org I was at moved away from Slack (due to costs) we used this method and wrote a little Python script to convert the main channels' JSON dumps into PDFs so we had a usable backup of channels.

Extensibility and integrations with learning management systems, as well as owning all your data, makes it sound like a great option in particular for an education-oriented organization.

And I imaging the AWS or GCP costs for hosting it won't be as high as what Slack wants.

Almost every special rate I have ever negotiated had specific clauses about when the rate will end, even if there was no specific date there's always something about "rate is reviewed annually" or similar.

I am constantly surprised by the number of people with "manager " in their title who don't know how to read a legal document.

The other thing is you cannot build anything sustainable by depending on the charity of a single company.

On a smaller scale it happens on a monthly basis with telecomms - almost never with rates, but they amend privacy policy and stuff - as a customer a change in the contract gives you an opportunity to say you're not accepting new contract, within certain timeframe, and walk away.

I guess this is simmilar - they told them they are changing the contract, and under new circumstances they will have to pay this and that, but they are free to walk away and pay nothing.

Still a dick move.

Slack is IRC with bells and whistles. Like yes I get that group chat is a necessity for today’s workforce. But it is still just group chat, a solved problem from a technical point of view.

So if you use an open standard, but not self hosted, and your provider tells you "pay 250k or lose all your data in 2 days", I'd say are not necessarily in a better position than they are now.

It's not impossible to migrate off of slack, but migrations take time.

Although I am not in the nonprofit tbh but maybe one day I would love to apply :>

They sound cool. Sad that bad things happen to the good people.

Slack really is slacking if they are literally asking 195k$ to a literal non profit whose helping kids/teens.

You don't want an entire generation of people who can barely operate the devices that enable and control a huge portion of their lives.

Kids will benefit immensely from being able to logically reason, and will be less afraid to repair or work around shoddy software, even if they never write another line of code in their lives.

Professional programmers dont fear kids taught to code any more than novellists fear kids taught literacy or accountants fear kids with numeracy. If anything, they know personally how important it is to learn these things.

I would love it if future folks can write their own random scripts without needing a developer to do it for them.

I would love to see more people writing software. There will always be advanced work that needs doing. There will always be larger challenges.

I want the world of the future, where every 10-year-old knows calculus and python and is incredibly capable, and then I want to see the future we get when they grow up.

You could use the same argument to stop teaching many other useful skills to kids. It's a bad argument.

(the comment you made surprised me)

It's a way to create many forms of art, solve everyday problems and automate a plethora of machines in our homes.

You sound like an accountant whining about kids learning about calculators and statistics.

(*not actually Slack just annoyed by this scheme, boo)

I just wanted to highlight this. I am so happy seeing this written down explicitly and finally.

Throughout the years I struggled so much finding relevant and accurate information about a feature of a product because it was scattered in chat channels, inadequate for providing reliable data (out of date or uncertain staleness, evolving or straight up wrong suggestions found, tangential only, patial, ...). Big names do it (Unity3D, DevExpress, ...). To make the matter worst both official support personel and power users promote its use, defend its use against critique to the last blood, despite of the obvious shortcomings and unreliability for average users. It is just the lazy excuse of providing the necessary knowledge.

If those any of those 4 screenshot snippets are of Mattermost, it's not very clear. All I see is screenshots of what appears to be Slack.

It's hard to imagine being GPT-4o.

Now you can argue choosing a Salesforce product is not a good idea and that I agree with.

Surprisingly not as much as I'd thought when they took it over. They just never adjusted pricing to remain competitive. The experience is still some of the best you can get for RoR apps. But nobody in their right mind deploying a new application today would look at their insane 10 year old dyno pricing and be like - yup - reasonable

Its inertia, its just not a priority to move them over

It really hasn’t required any maintenance at all beyond incrementing the version number.

They are starting to tighten the screws (showing admins a warning if you have over 2500 users), but it’s still looking good for a few years before I need to act on that.

Organizations need to realize that being right does not matter if you are dead.

From a Slack perspective, it seems reasonable.

At first I hated this - it was like using a chat app from the 90's! Why can't I have unlimited history like Slack? Why can't I link to chat discussions in tickets and code comments like I did at every other company I've worked at? But the enforced 10 day limit means you HAVE to properly document conversations and decisions outside of the chat platform. It completely eliminates any reliance on the chat platform - we could switch to something new tomorrow and (except for some grumbling about have to relearn a new interface) nobody would really care.

These have been quite big developer heavy companies. If companies like these don't think they can motivate the cost for Slack, I wonder if there are any than can.

Usually Microsoft was opposite: giving a lot of software for education for cheap or free to vendor lock-in people into their stack.

NOT advocating for using Teams because God please no, but Microsoft reliability us much better than Salesforce.

Imagine your landlord increased the rent by 4000% and it's due in 5 days or you're out on the street.

Sure, they have the right to increase their prices, but there should be at least a month notice for something like this.

They spent multiple paragraphs complaining about Slack, and gave Mattermost a brief mention in a single sentence. I'd enjoy hearing praise about Mattermost if they're willing to provide it as well.

I am sure that being forced to spend time on this steals time from more interesting projects.

Open standards, easy migration, and servers you pay an honest cost for. Self-hosting, perhaps even. That's where we need to go.

Teams is good at what it does and serves its niche well, however unless your daily matters are not well aligned with the particular framework Teams is designed for expect significant friction. It's not really the team size that matters, but rather how you structure your daily work.

A lot of the power of teams comes from integration with Active Directory, Sharepoint and Office. Sharing a presentation in a meeting that viewers can browse (e.g. to check back on something in a previous slide), calendar syncing with scheduling assistant, meetings scheduled in a team, meeting recordings and recaps, linking directly to a single page in OneNote, etc. are all quite powerful features, but most of the power is relevant if your organizational matters are structured more or less as a traditional enterprise and around AD/Office.

Inviting third parties or contractors can be quite a pain, especially if chat history is relevant. Meetings having their own chat can create information searchability issues. Integrating with third party tools is less straightforward and consequentially ecosystem of integrations is a bit of wasteland.

Calls are better in teams, much better to be precise than slack. We rarely used slack for calls (it had nice feature of drawing on colleague's screen) which I think is also available in teams.

I think that integration is crippled in teams but I didn't have time to experiment with it.

So overall I'd suggest: go for teams if you want to call meetings and are not using slack as a main knowledge base, as we used to in my previous company. Especially considering matters highlighted in this article

I tried running a community on it and it was a collosal failure. The onboarding flow sucks, if you want to send email logins it implicitly requires them to make matrix.org or whatever accounts (or something along those lines, details escape me), and you can have a custom server for that but it wasn't well documented and there was no canonical FOSS project for that custom server, I guess you were expected to just write your own if you wanted to truly control your whole stack.

And then, it was just high friction enough to where people wouldn't use it. Nobody downloaded the client apps other than me, even though the android one was really good, and even though you're spoiled for choice - you can even use it in Thunderbird! So everyone used the webapp, but then they'd switch computers and not do whatever you have to do to be able to read encrypted messages on the new machine, and so they'd lose all their messages and then stop participating.

And so on.

We moved the community to discord and all of our metrics have 10x'd: new users, existent user engagement, hell even revenue (we're an engineer-owned dev shop).

I really, really wish we could have made matrix work.

Slack can probably charge an extra $10/month/user for this.

I recently wrote some kubernetes charts for running Zulip for my new (smol) org, but I've ran Zulip for the last 3 years as CTO for a mid-sized AAA video game development company...

I really would recommend it over Mattermost (which was in use at another development company I was briefly a part of)

We used Zulip at a company I was at (about a decade ago) and everyone on the engineering team refused to switch from it to Slack, even when it looked like Dropbox might end the product because it was so loved (it's completely independent now so that's not been a concern for a long time).

If everyone using your software has trouble using your software (or tracking the bugfixes supposedly resolved in the never ending rewrites, rebrands, etc), maybe you should stop pushing it until it’s ready.

Every experience I have had with using Matrix has been a bad one: with the old client app, with the new client app, with the web app, trying to run the server, etc. It’s clunky and slow when it does work. It phones home to the Vector servers by default, despite being selfhosted. It’s a pain in the ass for end users to point it at a different hosted instance.

Maybe the answer is just “the whole thing, client, server, protocol - it’s all still in beta and you shouldn’t expect it to work well”. If that’s the answer, I wish people would stop recommending it until such time it works well.

you might notice it's 100% free software

now there is always the question how a company used Slack, e.g. just some ad-hoc fast communication channels like "general", "food", "events" or a in depth usage with a lot of in-depth usage, including video conferences, channels for every squad/project/sprint/whatever

but the relevant thing to realize is that there is subtle but very relevant difference between a "social network" focused tool and a work place communications focused tool

and Mastodon has a very clear focus on the former while Zulip has a clear focus on the later

The transition away from Slack's nonprofit pricing is also a key element to this story, but that is glossed over.

If they determined that Hacker Club violated some terms of the nonprofit demanding they move to regular or be kicked out seems not as bad

Obvious caveat here - the law of course must be made for monopolies.

A large group of hackers likely can figure out a way to export it all...

Eventually this leads to pressure to give them newer/better tools

Sometimes, these nontechnical users are dealing with problems as real power users that technical users may not see - there really might be a better way to do something and they may have already seen it at another company or something like that

It also happens that something might be working great but looks really dated and right or not, it can give new employees a bad impression

Still another thing is of course that sometimes someone is just throwing a hissy fit and wants something for no good reason but they somehow get the powers that be to listen to them

I’m dealing with this now - everyone is going out and buying AI tools because there is so much pressure to have AI tools and everyone feels like they are falling behind if they don’t go out and buy 10 task-specific AI tools

All that is to say that it could be that those users you referred to were facing problems that you may have been too far removed from the business to understand, it’s not a knock on you, it happens. It’s also possible they just wanted something new and shiny. The pressure to do that kind of stuff is real - I can’t imagine forcing people off of slack, for example

Yeah? cool. Just get microsoft's cloud suite, its standard across non-cool companies.

Life is not worth living bikeshedding about chat apps.

I’ve nothing against self hosting, but it isn’t necessarily cheaper than saas just because you can get amazing amounts of hardware for what amounts to a rounding error in accounting.

Everyone starts off with a price that's too low because you want a "no" from a customer to be "no, because your product isn't useful to me" and not "no, I don't have that kind of money". (Maybe this is a flaw and generalizes to generative AI. I like Github Copilot for $0/month. I would not like it for $200/month. If it costs them $200/month to run it, then there is a big problem with the business model.)

I, I just have to mention that IRC had these archives so repeat questions had a corpus to search. The walled gardens don't.

For my teams the "modern" solution is Mattermost. My (biased) feelings are that it's 10x better than free-slack and 100x better than paid.

In which US jurisdictions can employee-to-employee records (from employer-owned communication media) be denied to the employer/customer but maintained by an unrelated third party?

I desperately wish Discord worked like this. As you say, current threads just shove away conversation and it's quickly lost.

N customers * X% drop out rate * $200K > N * $5K

Then its a profitable operation for slack.

For 99.9% of nonprofits, their annual budgets are in the single digit thousands or less. A sudden $250k bill is fatal.

However the value of money is quite absolute, it's dictated by the exchange rate after all. If $250,000 is nothing more than "pretty big", then your perception is either quite distorted or the rate of inflation is much more severe than I understood it to be.

just hate it.

Not to mention that basically every scientific breakthrough achieved since 1995 was achieved using email as the *only* form of communication (other than physical letters here and there).

High time preference. The free stuff is here today, and the pain will only come much later, so I can disregard it for now.

Same with private VC/PE held companies. The board will replace the C-Suite if they aren't maximizing value.

You'd need to find a company which is huge but privately held by a group of people with only good intentions.

I'm reading this book because, well, that's the kind of place I'd like to work. I think it makes sense to get a feel for how these places think, in order to really identify job opportunities

Edit: here's a Wikipedia page on the topic https://en.wikipedia.org/wiki/Hidden_champions

Technology allowed companies to expand and centralize on a national scale, and capital pushed that to the conclusion we're at now, where there are a few gigantic players (at most) and almost all recourse against bad faith has been precluded. Nowadays if a customer is taken advantage of, they can't drive 5 extra minutes in the opposite direction and take their business elsewhere, or shame the owner in the local paper. Only impenetrable monoliths remain.

>By 2019, Deng had turned his attention to consumer goods. Pool robots, though low-profile, offered untapped potential, especially in markets like the US, where high labor costs made automation more appealing.

>“For what these machines can do today, they should cost USD 300–400,” he said. “That’s already the cap. Anything higher is just an ‘IQ tax,’ unless the cleaning function actually gets significantly better.”

The problem is that you want other people to fund your goodness.

Would make much more sense to use Discord.

[0]: https://preventcancer.org/about-us/team/

Frankly most of these tools have been at feature parity since before Covid.