Pretty much no one who uses Slack, chooses Slack. The companies we work for do. I despise Slack, yet I use it all day, every week day. It's quite the predicament.
Yep, basically don't have a choice and discussing alternatives is hopeless. We had better, once. Now we have this, in a full-blown browser runtime that doesn't integrate with the OS very well.
Actually triggers a good, 8-9 year old memory. Two places I worked at used to communicate through HipChat. It was simple, native, and did the job until Atlassian fucked it up with a massive rewrite. But even then it was still decent.
In both cases there was a massive bike shedding moment where the company engaged in a chat-war, with one half being solely devoted to Slack, and the rest of us holding back and saying why we need to switch. In both cases the Slack advocates went rogue and since they were also the loudest voices, people switched over just to get back into the conversation.
I have no idea what spell was cast on those people by Slack but, god damnit, it worked. It's just IRC in a walled garden and costing 10x as much, with shitty features you can't opt out of (like the new editor and threads, which birthed a whole cadre of slack micro-managers).
Were you using the same HipChat that I was? I recall the engineers at my employer (including me) clamoring to get off HipChat and onto Slack. Search was terrible, integrations/automation were clunky, and it was difficult to post code into chat. Slack was a breath of fresh air when we switched.
Also, I think Slack's threads are useful. You can have channels where you just post a single message for your question/issue and then have people reply in a thread, so that messages are naturally grouped together by topic instead of being a firehose of interleaved messages. Combined with good search, this makes Slack more of a valuable knowledge store than just an ephemeral chat tool.
We didn't need an all-in-one hub for realtime chat. Hipchat worked great for us, we had a separate WhatsApp group (with the founders and everyone else in there) for the social stuff. We all mingled but if you didn't want to, they'd find the preferred way to talk to you.
I don't want to name the exact place this happened but it was certainly a unique culture that I now miss. It was neither British or American.
It was the only place I've worked where I had easy and direct access to the founders, the C level, and the VPs, as a lowly senior engineer. And not because I brown-nosed my way up; we'd connect outside of work because we liked the same things.
A total tangent, but it's just to say that we never needed Slack in the first place. I doubt many places do.
> Search was terrible, integrations/automation were clunky, and it was difficult to post code into chat
Ok, so Slack solved #2 I feel and formerly solved #3 (but has regressed significantly in recent months). Not sure gaining #2 was worth the new negatives we got.
My experience with Hipchat was all negative. The ability to format text/code was abysmal. At one point it started crashing after an update and I had to download/install an older version just to use it. I have issues with Slack, but I'd use email threads before I voluntarily went back to Hipchat.
I was not aware of the « HipChat rewrite » but we used it in my previous company and it was Ok. One day a colleague told the team about Slack; I was skeptical. So 4 people gave it a shot for a few days/weeks (including me). The result was that, except the built-in bot to show random Gifs, Slack was superior for our use case in all aspects and we eventually switched everyone over, and all were happy.
This was 4 or 5 years ago. Now in my current company we also use Slack. It’s good, but not great either. Mostly because the UI is super slow when running in my desktop browser (Cliqz, Firefox-based). But it’s still better than Teams in my experience (I did not try the other alternatives).
For me this was different years ago, when Slack helped our company's people (consultancy) to find each other and share knowledge much more effectively than the previous method, mailing lists.
In a number of companies / assignments, Slack was introduced "from the trenches" and adopted officially later on. You don't get a tool to rise up from the trenches if it doesn't fulfill a need - that other, comparable tools also fill, especially now that a number of good competitors have jumped on the wagon.
I disagree - in a lot of corporate environments, it once snuck in as shadow IT to work around the then incumbent, infosec certified approved solution of less usable.
Slack has become the new corporate mandated unusable messaging application.
So the official solution was proprietary garbage. Shadow IT replaces it by an equally proprietary workaround. That workaround is the underdog so the companies that owns it actually has to allow its workers to make it good if it wants so succeed.
It's officially adopted making it the incumbent. Now the company can stop to pretend caring about quality and start digging moats. Predictably, the proprietary workaround becomes proprietary garbage. And no problem was solved after all.
Asking myself this question everyday in an org that uses Skype for Business, Teams, Slack and...yes, there’s an island of engineers and ops people using IRC sneering at all of us from the horizon. Hell we still have people struggling to set up conference bridges and video calls and have resorted to buying their own solutions and expensing them back to payroll. I’ve been given expense reports for paid accounts for Zoom, LogMeIn and someone even tried expensing $3 after buying their own phone number and writing a single serving conferencing app on top of Twilio.
I found that last one especially creative and especially worrisome if these are the lengths people are going through.
I’ve sat through four all hands calls now where “unified communication” was brought up and “infra-departmental communication” was mentioned as the leading gripe on preceding employee satisfaction surveys.
Yet here we are.
Hilarious, when you consider we sell wholesale communications platforms and services.
(When is the next “who’s hiring”? Mid-level Product Manager shackled by golden handcuffs seeks short walks to happy hour and a 4 day work week, heh)
You're not kidding! I once worked with teams of field techs for a large american cellular operator who shall not be named. I'd have 3 or 4 teams in the field, and I was at a desk pulling marionette strings in the network for them.
To communicate with them, I was issued a cellphone (natch), and when I asked if there was a preferred way to juggle the multiple teams, my predecessor said "Just keep the calls short so you minimize the chances of having to put one on hold when another calls you."
"Can I set up a conference bridge and then have them all call into it?"
"Those are expensive. No."
So apparently everyone at this enormous phone company was still living in the 80s when teleconferencing was something you paid for. I called up a buddy who set up a private bridge on his asterisk box, and I gave that number to all my field techs. It was great, I could step away for a moment and come back to find one tech giving advice to another, and in the end my buddy made a few bucks on recip comp over the thousands of minutes we logged.
Did anyone else in the whole company know this was possible?
I think you're overlooking the quality measure of the proprietary garbage as it evolves over time. Slack, in comparison to say Lync, or Skype has made huge moves towards better usability. But for a company like Slack determined to maintain market share, to control the user experience for all its users (whether they want to be users or not) is repeating the same mistakes, not necessarily the corporations signing contracts and writing cheques.
Slack is not the end solution. It's a step in the direction to being less terrible in the realm of corporate communication, and will be replaced by something else that does better in the future. For now it'll hold on whilst it still can.
I think Singularity will come before proprietary corporate communication software evolves to not be shit.
The sad part is, Slack was better in the past. It's the "digging moats" part GP mentioned that makes software shit. Corporate or otherwise. Consider: once upon a time, they supported an IRC gateway. But that was just a lure to get techies on-board, and predictably (like e.g. Google before with XMPP), once they've reached critical mass, they've shut it off.
(And I'm part of the problem too. I don't complain much, because as long as Ripcord - the not shit third-party desktop client - works, I don't have to deal with Slack's web and Electron crap.)
Email mailing lists were awesome. I could filter them by keyword and search easily, and finding one result would show me the entire chain of the conversation.
And if your corporate IT wasn’t completely nightmarish about locking it down you had multiple client options on almost every platform.
Besides, there were tons of integrations, scripts already available, and it is super easy to write your own irrespective of what scripting language you know.
Technically it sucked in that it could, when used terribly, clog up an entire network (such as the infamous NHS mail chain reply fiasco from a few years ago), but I have no clue who decided Slack was better than emails for what emails did.
Slack is better than Lync and similar corporate messaging platform. But that’s not what people tried to sell it as, or for that matter, Slack advertising sells it as. They sell it as an alternative to email.
Teams has been highly frustrating. For direct IM, it’s better than Skype/lync, but that’s not saying much.
It’s weird how difficult to use their slack competitor is. Hard to search within teams and across channels. There’s a weird interaction between itself and OneDrive/SharePoint that I don’t get the mental model.
I used to hate Slack’s UX a lot until I started to use Teams.
> I used to hate Slack’s UX a lot until I started to use Teams.
This. I was told MSFT has switched to Teams. They don't use Skype. Both my current and soon to be new employer (company bought) use Skype for business.
Skype is terrible in many ways.
Teams has a slicker UI. But it brings the horror to a whole new level. Someone mentions you passing in a group. You get a notification in email. But hey, if you aren't signed into that group, you can't see it. And you can only see like 4-5 groups at a time. So you really can't track conversations in real time.
The whole multi-teams experience ... is not well thought out. I have to switch between at least two different company teams multiple times per day, to do my work over the last 4 months.
Slack, for all its annoying mis-features, handles this workflow quite well. The UI, while somewhat simplistic, just works. Search and threading are terrible, but at least search works. Still have not figured out how to download conversations for transfer into tickets/etc.
I just want something that is simple, works well, allows me to search, to reliably enter code/screen caps/etc. Right now, though I don't like it, Slack comes close to working with some annoying misfeatures. Teams ... does not. Skype is best forgotten about.
I am in the happy situation, that my employer values employee opinions. We use zulip. It's not perfect, but clearly more useful than Slack. The one complaining most is our CEO. But he is smart enough to understand that happy coders are more essential for a small SW company than a happy CEO.
The main failure of IT or computer science is the lack of open standards and federation. 40 year old Email was the last somewhat succesful one. You can use various clients like Emacs, Thunderbird, mutt, Gmail web UI and there is no need that sender and recipient share either provider or client.
In communism it was the central planning commission that decided what products consumers want to buy. Today it's Slack that decides how people do IM at work. And Google or Apple how people use mobile apps. Facebook how you interact with your friends. I hope that these monpolies/duopolies would share the fate of communism. How do we get there?
By putting 1/2 the effort into improving hexchat or some random other IRC client and playing politics for a couple minor IRC protocol updates (public key exchange/encryption might be useful/etc).
Interesting that things have gone this way. Wasn't Slack one of the services that started out spreading via guerrilla Shadow IT, with users spinning up free Slack instances on their own and feeding penetration throughout organizations?
Having tried Rocket, Riot, and Mattermost before, it's of my opinion as well as everyone that I work with that Discord blows them away. It's so much more polished.
Discord is good. But there are a couple of things that prevent my team to use it seriously:
- Message threads
- Ability to share screens when multiple people are in audio conference
- ability to "draw" over the screen while sharing
One thing I wish slack or others had that discord has is ability to select FPS for screen share: when I'm sharing the screen or camera (showing a whiteboard) I usually care more that the image looks with better detail even if it refreshes once every 20 seconds. Almost all screen sharing always introduce compression artifacts.
I think "Go Live" on discord has rolled out to everybody now? Join a voice channel, hit the button (next to Disconnect) and pick an app or screen. Needs Nitro for good quality (otherwise you're stuck at 720p). Latency is better than most other streaming things I've tried.
100% agree on the other two, though. Encouraging people to make channels as needed helps the first, but for ephemeral discussions you wind up deciding between deleting the chat log, or living with a lot of "retired" channels. And being default opt-in for notifications makes Discord _way_ more noisy than Slack could ever be.
Discord is good but it has a bunch of drawbacks. The ones that matter the most to me:
* Can't leave a channel (without custom bot shenanigans, which doesn't scale well to having a lot of channels).
* No control of notification sounds on mobile. All notifications get the same sound and there's no option to deliver quietly.
* No threading.
* Can't collapse embeds.
* No back/forward navigation. The closest it gets is in the quickswitcher it offers the last channel you were in, but that's it.
* Their support is awful. Everything except extremely obvious bugs get the response "please vote it up at feedback.discordapp.com", and even bugs often get that response.
* No "do not disturb" settings. If notifications are on, they're on 24 hours a day.
Speaking as the project lead for Riot, I agree... for now :) The difference (relative to Discord) is that Riot is FOSS and the sky is the limit in terms of how fast it evolves. These days Riot/Web, Riot/iOS and RiotX/Android are both straightforward and fun to hack on - and we do everything we can to encourage and merge good contributions and features from the community. As per the 2020 section at the end of https://matrix.org/blog/2019/12/24/the-2019-matrix-holiday-u..., our top priority for next year is to get the UX of Riot to be as mainstream as possible (ie as good or better than Discord), and we need all the help we can get. The company which funds most of Riot’s development (vector.im) is also hiring currently for UX Designers and developers to work fulltime on this.
TL;DR: If Linux can obsolete Solaris, Riot can obsolete Discord/Slack. Come get involved and help :)
As a very heavy Matrix user, the one thing that's kept me from trying to use it for my local hackerspace is that there's not a simple way to grant access to rooms based on group membership. Within companies/larger private groups, you often (almost always!) want to grant people access to chatrooms conditionally based on membership in the group. There are few options for this right now:
- Have a bot invite everyone in the group to every room and kicks them out if they leave (noisy, not everyone needs to be in every room)
- Have everyone in the group use the same homeserver, make rooms public and non-federated (defeats much of the purpose of Matrix, doesn't let you invite people from outside the org as one-offs)
I think some of the future plans for communities would solve this?
Anyway, Matrix right now is great for both small private chats and FOSS projects; in the former case you can just handle access ad-hoc, and in the latter case the rooms are public anyway. It just seems a bit lacking at the moment for organizations that need private rooms.
You can't reuse matrix IDs (yet), as they are the unique key for your message history - it's like email; if you reuse an email address you'll start getting email intended for the original user.
Meanwhile, if you want echelon as a username, go register a domain name, run a matrix server on it, and you can have @echelon:echelon.xyz or wherever. Try not to get sued by five eyes though ;P
What do you recommend as server software for riot? I had some problems with the various bridges which the default server currently provides. Is the default server also provided with synapse?
Synapse is the best server for now, and works relatively well (resource usage has improved lots over the last year). Bridges plug into the server, and there’s typically at least one good one per protocol: matrix-appservice-irc, matrix-appservice-slack, mautrix-telegram, mautrix-whatsapp etc. We need to improve the UX for provisioning and managing bridges though; atm it’s mix of config files and bot commands.
Thank you for working on Matrix, it has been smooth sailing (after some database cleanup) for me to operate since taking over a homeserver from a frustrated operator nearly half a year ago.
Why would someone other than the server admin host the bot?
It's not ideal, but shipping a mobile and desktop client that are literally incapable of displaying the same channels in the same order is also not ideal.
That's not the same. Unless you're adding one role per channel, and then it's quite unwieldy. For example, we're not going to set up a role and have everyone else self-serve add that for a channel for an outing, a specific feature/bug, etc etc.
How do you get the right subset of employees into a channel in Slack? The same amount of work split the same way can accomplish the same thing in Discord.
Yeah imo, discord is the only conceivable competitor to slack in this space and anyone who thinks Microsoft teams is their competitor just doesn’t know what they’re talking about.
I hard a weird conversation with a group rolling out teams. They had never used instant messenger. They had never used slack. They had never used discord. They had never used irc.
It was very strange because they had no idea what Teams was used for. They shared some literature that says the features are the same, but when asked to do simple tasks like use teams to coordinate the rollout, they didn’t know how to do it.
Kind of weird. My hope is that Microsoft eventually makes it usable since it’s basically free with other purchases. But then, Microsoft killed yammer and Skype.
I tried to use ripcord for about a month. There were several features that I really liked (for example - tabs are amazing imo).
The deal breaker ultimately was that I didn't receive all of my notifications which was a problem for work. But I would love to try it again if they solve that problem.
Edit: The hotkeys and the lightweight feel to it was also amazing BTW.
Do you have a way to reproduce it, even unreliably? Or any other info? If so, could you email me cancel@cancel.fm ? The Slack server determines most notifications that the client is supposed to display -- the logic all runs on the server, and is sent as explicit events to the client. So, the only way that Ripcord would be losing notifications for standard chat messages is if the server did not send the notification events, or the connection went down for several minutes and had to be restarted without an event replay.
But if you have any lead on a way to show that notifications aren't being displayed, that would be really useful information.
I've heard of some people using Ripcord on Windows 10 having some kind of automatic suppression of notifications occur at the the OS level. Probably some automated anti-user-spam thing. In which case, the notifications would only be listed internally within Ripcord, but not in the OS notifications list. Is that the case, or were they also not visible within Ripcord's own notifications list? (This one isn't relevant if you're using Mac or Linux.)
Is there a way you could enable touch events for scrolling the chat history? I'm using a Surface and trying to touch scroll just selects messages. Other than that my first impressions are pretty good.
I'm guessing this is what needs to be handled. I wonder if Qt is already interacting with this system, and I'm handling it improperly on Win32. Does scrolling with touch gestures work for you in other types of scrollable views, such as the sidebar, or the workspace files list?
OK. I will see if I can find a way to emulate a touch device like this one on Windows for testing. I can pre-empt Qt's input system for these types of events and handle them manually in Ripcord's own application code.
Seconding, please do. I'm a paying Ripcord user working on a Dell 2-in-1 (essentially a Surface clone with a better price tag), and I miss proper touch (and pen) handling too - to the point that I started looking into fixing this on my own with AutoHotkey (I'm thinking of an always-on-top semi-transparent area I could touch / pen over, which would translate this into appropriate mouse events to send to Ripcord to make it scroll). But it would be vastly better if I didn't need to do this :).
Oh, and since I have you here: would it be possible for you to look into the group DM handling? We run plenty of group DMs at work, and I have two problems with them: one, they tend to randomly disappear from the sidebar, and two, there seems to be no way to start a group DM from Ripcord.
Friendly reminder that Ripcord is closed source shareware (free by now, $20 when it goes out of alpha). If you help this guy, he'll be pocketing the profits.
Are you seriously suggesting not to investigate an issue because the person at the other end is making money selling software they wrote? Heresy! Heaven forbid they might even fix the bug with a little information and improve the software for everyone - paying users and shareware users alike. It’s not always a zero sum game.
I didn’t know about ripcord until it was mentioned here and I’m about to pull out my credit card to buy a license just so that I might have an option for a slack client that’s not slack.
You make it sound like that's a problem. I think it's pretty great that he's pocketing the profits, to be honest. I hope he can continue to develop the app.
For me the dealbreaker was the visuals; I've been spoiled by years of good looking webapps or native (ios, macos) ones. You can't get away with making a chat client look like 90's IRC clients anymore, imo.
I absolutely don't understand why chat is such a lucrative market. All the products are nearly identical. All the most useful features of Slack were available in IRC.
There are some features that aren't available on basic IRC, such as search, history, message emoji responses, message reminders, threading, rich integrations (eg. actionable notifications, polls, etc.) and so on.
Aside from Rich Integration (which is arguably also there if you include bots/scripting), most IRC clients have search, history, emoji, and reminders. What's missing I think is a pretty UI and lots of marketing.
> > most IRC clients have search, history, emoji, and reminders.
> Keyword: most. So: not all of them. So, there will be a mismatch and a mish-mash between features.
You can't judge a protocol by the subset of UI features supported by all of its clients. That's unfair to all open protocols.
If you want to talk about UI, you should evaluate each client independently.
> At this point someone on HN suggests IRCCloud... which is exactly the same as Slack (proprietary chat-as-a-service).
Except IRCCloud still uses the open protocol, so it's easier for users to switch to a different client. (Which, hopefully, is also an insentive for IRCCloud to not fuck up things too much in the future.)
> Usability for non tech people. Not everyone in chat has technical prowess, I know IRC seems simple but its too intimidating for some.
I communicated via IRC with a lot of non-technical people in the '90s and 2000s. Most people just used mIRC and connected to the default IRC server and went from there.
It's not the end user features. It's the admin features. You can very easily read everyone's messages, see stats on who's talking to whom the most, control who has access to what channels. The customer is the business ownership, and the product is control.
Honestly, how much data would the typical company need to transition off?
Some photos, docs, code snippets? Most companies do use those features to some extent, but it should be fairly ephemeral. If you're using pinned messages as a WIKI, then you are really using the tool incorrectly.
Channels and memberships? If you made all of my company's channels disappear tomorrow, you'd be doing us a favor. We'd re-create the ones that matter within a few hours, and the other 90% didn't really need to exist anyway.
The only really true "sticky" part of Slack are webhook integrations. Some companies really do setup some degree of workflow automation around those things. But most of the time, it's just bot-spam. The CEO wants everyone to belong to a #wins channel, which gets a spam alert whenever a sales guy qualifies a lead, etc.
>Honestly, how much data would the typical company need to transition off?
In many cases, all of it.
Data retention laws require many companies retain copies of certain types of internal company communications for extended time periods. This is why some companies have policies that discussions about topics such as company finances or government contracts have to be done via more easily archived channels. But if your company doesn't have those policies and you just use Slack for everything, then everything in Slack has to be retained on the off chance that someone discussed an upcoming earnings report in your company Slack.
Rocket developed many issues with missing notifications and messages for me over time. No push notifications on mobile when a desktop is connected. Overall things fail intermittently especially when multiple clients are connected for the same user. At some point it stopped respecting some settings about the per-room notifications.
My experiences both as a Slack org owner and a Rocket org owner - I'd set aside my personal dislike for Slack to use it over Rocket ten times out of 10. The admin tools for Rocket are really sub par.
Does anyone know why Slack would have done this? I can certainly understand Slack not providing a specific mobile-web experience, but that is not what is going on here. The user specifically switched their mobile browser to "Desktop Site" (one of my favorite features of Android Chrome by the way for sites that mysteriously provide dumbed down versions for mobile that are missing all the content I actually need - looking at you, Square), so the server should just serve them desktop-optimized content. Slack has to go out of their way to prevent this. Why would they care? It has to be a very small portion of their users in any case.
"Does anyone know why Slack would have done this?"
"App installs" is a key metric driving valuations for SaaS/PaaS/etc. businesses.
Trading a useful product for dollars isn't that interesting. Collecting an open-ended, continuous stream of intel on a large population is fascinating - at least until it become generally accepted that this "intel", and the levers for manipulating people based on it, are really not that useful or effective (which is my own prediction).
Slack has the attention today, but the real poster child for this kind of behavior is Sonos - nowhere do you see such a stark juxtaposition between an actually useful product that people pay premium dollars for and pivoting the business model towards shady, sneaky, anti-customer patterns.
I wonder how the "Microsoft is gaining on Slack" narrative would be changed if we had metrics on Lync/Skype for Business users that switched to Slack or Teams.
Basically I have enormous doubts about Microsoft's growth numbers in the space, since I can't fathom that anyone who wasn't already paying for MS products would choose to use teams, and I'm curious if they've made up the ground they lost to Slack in the first place.
Eh, web traffic can be measured too. App installs aren't all that interesting if people install the app and never use it. Therefore, valuation metrics should be driven by usage (regardless of platform).
I think companies want you to install their app for a couple reasons:
- Sunk cost - I would guess people feel a lot more invested in a service after installing an app (and then being forced to make an account, in many cases)
- Mind share - Their logo will appear in your app drawer and they will be able to send you push notifications
It must be effective, because many companies push heavily for app installs. Check out mobile reddit, for instance. Although yes, I think that a lot of this is useless metric driven strategy.
You forgot the most important reason: telemetry. An app can exfiltrate an order of magnitude more kinds of interesting data on you than a webpage ever could.
And the second most important reason: push notifications.
A big reason many companies push for app installs is because it gets them out if the business of having to pay a very high Google tax for traffic. If a hotel company can convince you to download their app, the hope is that the next time you search for a hotel you go straight to their app, instead of going to Google where the hotel chain would have to pay out the nose for very competitive AdWords.
* customers might have complained that the site didn't work on mobile when the desktop site was requested; this may have led to a burden on their support team, and a manager might have decided to block usage instead of fielding support requests
* a team internally might have been spending a lot of time chasing down bugs for desktop on mobile, and a manager decided to officially not support desktop on mobile; instead of just stopping support or carefully triaging tickets (e.g. distinguishing desktop on mobile from desktop), they just decided to block desktop on mobile access
* a team might have been burdened by being required to fix all bugs that came in; instead of dealing with the problem properly (e.g. continuing to allow desktop on mobile but not officially supporting it), they officially killed support to avoid having to work on it
* there could be some KPM to reduce bug reports over some period of time; they killed access to the non-supported use-case of desktop on mobile to make numbers look better
Not really a smart decision in any of these cases, kind of stupid, honestly; but not really nefarious either and the kind of decision that can be made sometimes at a larger place with politics.
I actually think all these problems could actually be fixed more easily but just showing a "we really think you should download our app" banner in this case. Let the user know they're going off the reservation but don't explicitly block them.
As someone who once used Slack's website on mobile, I can assure you they didn't have a team spending a lot of time chasing down bugs with that configuration.
As mentioned above, this has to be a very small number of users who are sophisticated enough to bother instructing chrome to request the desktop site. So the bug burden would be small on an absolute scale, and could be ignored entirely by simply telling these advanced users the desktop-on-mobile was "as is" and not supported.
The site I work on has hundreds of millions of visitors a month. A double digit percentage of our overall traffic uses our desktop site on mobile. It’s not the same use case as Slack but I’d bet it’s bigger than you think.
Also, bug burden shouldn’t be an excuse if your market cap exceeds $10bn.
Aren't you agreeing with me? The important thing is that, as a fraction of users, the number of people asking for this is small. Therefore, given the size of the company, the resources devoted to it are relatively small.
No. I think when you are big enough, a feature with even 1% usage is big enough to dedicate resources to. Slack made $168M last quarter. 1% of that deserves at least 1 eng and 1 QA full time.
Where do you think we disagree? I am arguing that bug-fixing costs are not a good justification for Slack to ignore requests for the desktop website from mobile devices.
Benefit of the dobut Reasoning: Engineering determined that the cost of supporting a mobile site outweighs the benefit given they have a great mobile app. Then, someone decided to disable mobile access, because it better communicates that they don't support mobile web; cusutomers may still use the broken, unsupported experience, and then they'll complain about it to their support. Thus, raising the cost of maintenance while lowering customer experience.
Keeping mobile web but adding a banner that the experience isn't supported: Doesn't Work. Most people are total idiots when it comes to tech; this is 250% more true in the B2B space; this is a further 250% more true in a bottom-up corporate sales model like what Slack deploys. Unsurprisingly, but likely; many of their workspaces are shadow-IT operations created by some do-gooder who knows Excel and thus he's a "tech expert", on an isolated team within a large company who's IT department really wishes everyone would just use Teams and Sharepoint. This person is going to dodge the company's frontline IT support with any issues they run into. This person probably has a company-issued phone and thus doesn't want to, or can't, install the app.
The issue isn't about a mobile site. It's about desktop site. Mobile browsers like Chrome and Firefox have a "request desktop site" switch that makes them fetch the site meant for desktops. It introduces zero support burden if you're already supporting a desktop site, and you have to go out of your way to break this feature.
I'm past giving them the benefit of the doubt. Also, such issue can be resolved by saying in a clear and unambiguous way that "the site is not supported on mobile, proceed at your own risk", but not actually blocking it.
That doesn’t really acknowledge the point above. Tech confuses people. You can put a banner that says “this won’t work” and people will close it, then ask why it doesn’t work.
You may also simply not want to allow people to use a version of your software that doesn’t work.
Given that the hypothesis is true for vast majority of Android apps, including the vast majority of popular ones, including well-known vendors like Google and Microsoft, it's a really good default hypothesis.
It would potentially allow them to forgo the need to limit their UI to phone based with desktop upscaling. Make an app for the "best" small screen experience and the web site for the "best" large screen experience.
This strategy probably mitigates what might be a much more common complaint in the future that their "UI" sucks on mobile if they were to design for only larger screens.
This is the argument that I don't think is valid, though. When folks switch on "Desktop Site" in an Android browser, they expect that things will be squished, and that the site isn't optimized for the screen size. They just do it (or rather I just do it) because I want the full set of features, even if I need to do a bunch of zooming and scrolling around.
That is, Slack should still focus on making the website for the best large screen experience, but they shouldn't go out of their way to do somewhat sneaky stuff to detect my "Desktop" browser is really coming from my phone.
Look at their stock price. I bet there are a lot of tense board meetings happening right now. Culture flows from the top down, and if executives are panicking about the stock price, it makes sense that they'd lose focus. I mean, if you had stock at slack, it would be easy to fall into the trap of cutting costs and following the money instead of shipping an excellent product.
I see a common pattern among utility providing IT companies, that despite the core functionality can be supported by tens of people, it goes ahead and hires thousand anyways, which give raise to features that are neither core nor that anyone wanted. The company then go under as the business can’t generate enough revenue to feed the thousands of developer and management that has been brought in.
If you have a good app, keep it that way by not increasing bloat, both technically and organizationally.
Patreon comes to mind. They're essentially a members-only blog subscription company. It's beyond me what they need all those employees for. The only difficult problem they should have: dealing with fraud.
I'm guessing sales & marketing (and the usual O(log n) amount of management and administrative positions to support that).
It's one of those hidden costs of advertising being a zero-sum game. The waste only grows as you try to keep up with the competition that tries to keep up with you.
It was like 40 or 50, but yeah the points stands, it was still like two orders of magnitude smaller than any other 10+ billion dollar exit we’ve ever seen.
Agreed, and I was highlighting how post-acquisition they moved away from the core functionality and started stuffing other apps into it in a pretty predictable fashion.
Personally I’ve minimized social network use and don’t think every single app needs a status broadcast feature, but it’s popular. Not just a group chat app anymore though.
It got acquired for the trust people put in the brand, not the technology or the employees. Facebook needed a backdoor into people's contact lists and WhatsApp had a huge user base of people who authorised it to access contacts.
More boring explanation: For most people Facebook is just a platform for connecting with other people. Any messaging platform with significant amount of users is a threat for them.
A platform like WhatsApp could evolve to cover the main things people are using Facebook for. History has shown that people are not too stuck on any platform.
Huh, so apparently the Status tab is supposed to be something like Stories? I've never used it, it wasn't really advertised to me, and when I open it it's an empty page with a prompt to add a status update. From what I gather from the internet it's somewhat like the status messages from old messengers, just some text you can set?
Stickers are new, I must have forgotten that because every other chat app I use has had them for ages. They are really an example of WhatsApp moving slowly compared to Telegram and Discord.
Group video chats are something that's probably motivated by rising mobile speeds? As bandwidth becomes more plentiful and smartphones have plenty of RAM and CPU power there's no reason to restrict video chats to one person.
1. Successful software product that is beloved by its core users goes public.
2. The valuation expectations when the company goes public are based on widely optimistic growth projections where the company grows beyond its 'natural' user base (remember 'Twitter is the new Facebook' back in the day?)
3. In order to meet those widely optimistic, unrealistic projections, the company tries a whole host of "throw against the wall and see what sticks" features to attract users outside their core. This has the effect of pissing off their core users while not really attracting many new users in any case.
4. Eventually (hopefully) the company realizes it's not going to be the next world dominating superpower, learns to be content with just hundreds of millions of users instead of billions, and gets back to focusing on making its core users happy.
No - I insist that we name this after Sonos, even though they were hardly the first (and certainly not the most widespread) abuser.
As I mentioned elsewhere in this thread:
"... nowhere do you see such a stark juxtaposition between an actually useful product that people pay premium dollars for and pivoting the business model towards shady, sneaky, anti-customer patterns."
The sad thing is that if eg Twitter decides not to accept that VC money, the VCs will turn their back on Twitter and invest all of that into another Twitter clone, and that’s when competition may suddenly become an issue.
This is a problem with the current financing model of tech in NA; the product didn't arise strictly from user-sourced revenue growth, but from the hopes and dreams of investors who are attracted to the user growth. Revenue sourced from user demand is tied to features users almost certainly want, whereas investors want features that will continue growth. That's not always useful to existing users.
I don’t think they’re panicking about the stock price — why should they? They have plenty of runway and shouldn’t need to raise capital any time soon. I’d worry about their stock price at $10 or $5. At $20+, they’re still trading at large multiples above revenue.
Recently I found a bug in Slack, and I was communicating with their support over it. They asked me to check on the web in addition to the electron app. Now, the number of people who actually care enough to do this are a very small minority, but they are also very significant because they're happy to give Slack free QA. And as a member of this small minority that will reproduce a bug on multiple platforms and communicate back, I can say this seems pretty hostile to me and anyone else willing to pay money to do free QA for them.
> Yet another service that forces you to use the app for no reason
For sure there is a reason behind. It's just not a user-focused reason but a business one instead. One guess would be that the browser is a environment the user can kind of control, so they can block tracking and so on. When you're using the app, you don't have the same control anymore, so the business can receive more metrics from people that they wouldn't before.
Not saying this is good/bad, just that there is most probably a reason behind the choices they make.
Not trying to be too glib here, but we get it. That’s the problem. HN is pretty tuned in to the trend of walled gardens and creeping control popular apps and platforms exert on their users. It’s kind of the point of this article. Of course it’s helpful for the business for many reasons.
This is the result of the VC market models and the near universal buy-in to the idea limited partner's money (and "advice" that comes with it) is required for starting software companies. Nothing more.
Bootstrapping is pain, but at least you and your customers retain control long term.
It also depends on what company is doing it, apparently. Some are all for the walled garden approach when it suits their business model or they are a "fanboi" of the company/product, but despicable and evil when others do it. They're probably releasing their ire about Slack to post here on one of these companies devices right now.
> One guess would be that the browser is a environment the user can kind of control, so they can block tracking and so on.
Someone in the twitter thread says it still works on firefox for android, which means it's really only broken on chrome. Chrome doesn't have adblock, so it's not like this move results in less of their telemetry getting blocked.
Apps have been turned into a disease. So many different services keep obnoxiously pushing their app. Delivery services are extra annoying. I tried searching for a service that delivers alcohol, and found a few, but some are app-only and I couldn't type in my address to see if I was within their delivery area. Why would I download an app, and probably create an account, only to find that they won't even deliver to me?
The nagging on Reddit is extreme. And now that I have the Reddit app, if I'm viewing a Reddit thread in the browser then certain actions (but not all) will open it in the app without a prompt.
Last I checked, on the mobile web interface there is an option to always give you the desktop site, and it remembers your preference. I never see the Reddit mobile site and never need to tell the browser to request the desktop mode. It's completely usable if you don't mind zooming and panning, which I'm fine with.
>I never see the Reddit mobile site and never need to tell the browser to request the desktop mode.
There is a seemingly persistent bug where reddit fails to remember preferences surrounding mobile/desktop and old/redesign. Sometimes you will click a link and the page it delivers you to will be on the mobile site, sometimes it will take you to the redesign. It's been going on for at least a year. I use reddit a lot and I run into it at least every week. Sometimes it sticks and I have to relog to get my preferences back.
> then certain actions (but not all) will open it in the app without a prompt.
Firefox for Android prevents such attempts by default. That, plus uBlock Origin with its integrated element picker to permanently block elements make reddit bearable.
This is a bad move, especially given the constraints of their mobile app.
For example, their iOS app requires iOS 11 or later which means people with a 5c, 5, 4s and iPad 4th gen will not be able to use Slack. These may be deemed 'older devices' by the tech forward folks like those on here but you would be shocked how many are out there still today. It only presents more resistance to adopt the app at team scale, as they will probably realize soon.
Sometimes old APIs are dropped server-side and thus older versions of the apps stop working. I have an old iPad at home where I can watch YouTube from the browser but not from the app (though Safari is so old that I don't really trust that device for anything that requires any kind of security).
Thanks for confirming. I was more responding to the title claim, which seems incorrect - the desktop site option clearly still works for others. It may be a bug, or a new policy but time will tell.
I found that you have to enable desktop mode before entering slack on mobile. If you have visited slack with a mobile user agent, it will force you to download the app even if you enable desktop mode.
yeah - I read their response at the time, but it occurred to me it may have been a response in haste or a misreading (it's the desktop option on mobile browsers; mobile browsing per se isn't supported for a long time I think), or a pragmatic implied "until we've investigated" as it worked for me on testing, and others, and its probably a skeleton support crew over Xmas.
They won’t, at least not for now, I assume. They can put far more intrusive tracking inside an app compared to a web browser. There is also no way to do content blocking or ad blocking inside apps, at least not on iOS. I use Facebook in Safari - can’t be happier (well, I mean it’s Facebook, but I am not able to pursue all my friends to cancel it).
Back in the late 1990s I used to have all sorts of software in my computer which I downloaded and installed. Some of it was purchased and other free or open source.
As the time passed, more and more services became "web based" including mail, music, calendar, note taking, etc.
I was part of a small group that did not like that. My reasoning was that there was no need to hack all that functionality in the web browser as it was highly inefficient. That it was better to code the network features into the desktop clients.
Web caught on and nowadays seldomly you get a client program in the desktop.
But now in mobile, we are returning to the clinet software first approach. And for some reason we resist the installation of client software and prefer browser based functionality.
> But now in mobile, we are returning to the clinet software first approach. And for some reason we resist the installation of client software and prefer browser based functionality.
IMO for any given program or 'app' today, the better it functions offline directly correlates with the desire I have to use it natively on a given platform, as opposed to web access.
Slack is just about useless offline, hence my strong preference to use it in the browser.
Back in the 1990's you installed those applications when you wanted, upgraded them when you wanted, moved whatever data you wanted between them and in and out of the computer as you wished. Also, you were pretty certain those applications were not spying on you.
You can reverse each of those statements for a current mobile app.
The height of the desktop era was before surveillance capitalism was a thing. With web came the on-line ads, which, being a cancer[0], infected the business models of increasing amount of software. Then the mobile exploded, and the new breed of native software - apps - were designed to be user-hostile pretty much from the start, and continue to be so.
There seems to be a lot of angry nerds in here, and that basically describes me too. Hoping I can get a response from somebody with a broader perspective:
I'm in a position to switch my company over to one of the many alternatives mentioned in this thread. Has anybody here actually done that, and if I do it how angry is that likely to make my non-technical staff? If I don't want my marketing guy and my customer support team to hate me, can I still switch my company over to riot.im or mattermost?
Probably just stick with slack. Discord would be the other viable alternative (which I actually prefer).
Remember, you lose the ecosystem when you switch away, too. There are lots of plugins for slack that won’t be there on the alternatives. Even little things like seeing GitHub commits is nice.
Mattermost CEO here, to the extent it's helpful, here's a demo video of how it looks and feels importing a Slack team into Mattermost from 2017: https://youtu.be/AKqHWqrAgpk?t=249
This is when we were largely just an open source project and before raising $70M in venture capital, which has been accelerating development: https://mattermost.com/blog/category/releases/
This is why I try to use old software. It has much less feature creep and will stay relevant for the long run. If you can compile it yourself, even better.
It astounds me how far people will go in giving the benefit of the doubt to obviously user-hostile actions by Silicon Valley companies. There's no reason why Slack couldn't show a modal rather than outright blocking the website on mobile.
The choice is to use Slack or not. It's their service to offer how they wish, and your choice to use it or not. Tesla doesn't have to make a manual transmission (equivalent) car just because I'd rather drive stick, but I just won't drive a Tesla then.
Every growth-oriented startup with a popular product eventually ruins that product in the name of growth, which eventually kills the company and encourages somebody to start a new startup to build a product half as good as the original. The Silicon Valley circle of life.
This is a shame for mobile developers as well. Deep linking from one app to another can be tricky to get right. Loading slack from the Android emulator browser was an easy way to test various deep links. Not sure why they would care.
Slack is absolutely _unusable_ on a slow connection. It's not even just a matter of being patient, it flat out breaks with everything timing out.
I used to use the IRC gateway to maintain contact with some friends, and basically lost access to that social group when they closed down the IRC gateway.
IRC is far more inclusive, plus it's open and federated if you're into that.
Yes, IRC is much better, and is something I use. (IRC even works without specialized software, although it works better if you do have it (which isn't too complicated to write).)
I wonder for enterprise licenses if there is specific mention of which clients on which form factors are covered in the license. I expect and hope that customers will get a lot more specific about what they are paying for so that companies can't just turn off functionality that has been paid for.
I understand why people don't like this change (limiting options, etc), but I can see slack's POV from a business perspective. The mobile web usage might be so low that maintaining it costs more than forcing people to switch to the app. And mobile web might be a bad experience anyway (because slack is focused on standalone desktop app, mobile app, desktop browser instead), so it turns new users off with the crappy experience. But I don't have the numbers or any other insight, so it could be more simple: "we want more app usage". And even if you decide to just leave it alone, is anything truly maintenance free? (possibly)
This won't be the last thing to turn off mobile web to force-guide users to the app. And luckily for actual mobile web users, there seem to be alternatives.
The OP had a browser on the mobile device capable of rendering a "Desktop experience" and requested it. Slack went out of their way do discover that and block that user intent, rather than serve the Desktop UI that they test already in a desktop browser. The user knows they will be getting a desktop UI crammed into a mobile device, they asked for it.
And the OP is probably the exact profile of person who will request the desktop page and then open support tickets with Slack when some features break, causing them to spend hours debugging the issue.
The unfortunate reality is that it costs businesses money to support platforms. Blocking unsupported platforms probably results in an extremely small number of irritated users in exchange for saving thousands in support costs.
At which point the support folks can close the ticket and say "We don't support that platform, please use the mobile app, or ask your browser developer to better support Desktop-on-Mobile". Not supporting the platform, and actively preventing it are two different things. Especially with all the invasive power mobile apps have on your device vs being contained in a browser.
Too many apps stick their fingers into parts of the device that they don't belong. The well has been poisoned with so many apps abusing their privlege, so privacy conscious folks have to assume the worst when being pushed in that direction especially when the browser app worked without an issue (According to OP).
That's besides the point. Why block users from using it if they want to so? If it works without any extra maintenance, great. Which, apparently, it did.
Unless Slack has material amounts of support due to mobile users trying to use its web app, which I'm very skeptical about given that the app is a dumb wrapper around a web view anyway, there's no sensible reason to actively prevent those users from using it.
What likely happened instead is that someone at Slack is trying to growth hack app download numbers.
It is not possible to access your workspace from your mobile browser, we're afraid.
This isn't the first time I've seen a thinly-veiled-fuck-you attitude from them, and it's something that definitely won't make me want to start using Slack.
Maybe, but... Twitter actually has a mobile website. You can argue - and I would agree - that "request desktop site" should give you what you asked for, but there's still a difference between "here is the mobile web UI" and "here is a link to download an app": namely, one still allows you to use the site if you can't or won't install a native app.
Slack is actively locking folks out of their site here, a service that folks have paid for even... That is a significant departure from what folks might've seen in the recent past with mobile device detection.
The sentiment around Slack has changed so much so fast. It was only a few years ago when all anyone could do was rave about it. Maybe it was because on the surface it was the rare decently designed piece of enterprise software.
Is there an easy way of altering the user-agent header for my phone?
I would rather take the desktop site every time instead of being pidgeonholed into stupid m.-.com that forcefeeds me popups of how their mobile experience is so much better.
If you have Android and have Firefox, there are extensions that can do it. Otherwise, you would need something to intercept requests because Chrome and Safari on mobile don't have extensions.
There are options for persistent connections, with logs and detachment, etc. ZNC/PsyBNC/Quassel come to mind. ZNC, in particular, is extendable via Perl or Python (or TCL if you're a masochist). I hear good things about Quassel as well, but it's still early in development.
I get Android notifications if I'm detached from irssi in tmux and someone PMs, says my handle, or any number of custom triggers I set. All free & open source (other than the android phone).
Slack is IRC for millennials. This is purely about gather device specific metrics that are gathered via granted 'permissions' vs providing a service users pay for.
To be fair to the devs at slack, it does actually cost them something to re-enable this.
every change to the code base has to be checked on mobile web or things will eventually break. at that point will we see an angry tweet about not fixing that?
i develop on a web based system and we don’t support mobile on most of our site because we don’t have the resources to. if we started getting bug reports about this more frequently than we do, we may do the same. It makes our product look bad when we let someone do something we don’t QA for. We pride ourselves on rapid response to our clients requests, with this we build goodwill. we explicitly state we do not support mobile except for specific modules. even with that prior information, users sometimes try anyway and run into bugs. this experience + being told we won’t fix the bug because it’s not a supported platform causes us to lose the goodwill we’ve worked hard to build.
in my local neighborhood there are a ton of wholesale shops that even though there are maybe one or two customers a day (ie the shops are dead most of the time) they refuse to sell non wholesale. it’s not packaging, most things are easily sellable one-off. there are no regulations or laws. they choose not to sell to one class of buyers. if they allowed one-off sales, they would then have to start running their business to support doing that correctly- this means having small change, more personnel, probably introduce some security to watch the now increased traffic in their shops etc.
if slack does mobile web, they’ll have to do it like they do everything else, the “slack” way.
Zulip: https://zulipchat.com/
Rocket: https://rocket.chat/
Riot: https://about.riot.im/
Mattermost: https://mattermost.org/