Hacker News Reader

Comment by easton | Hacker News Reader

I'm somewhat surprised that Amazon didn't end up acquiring Slack, considering that the major thing they are missing compared to Microsoft is productivity tools. If they had purchased Dropbox and Slack they could've gotten a foothold rather quickly and slowed down one of Microsoft's major selling points (you already use Office, why not Azure?). Discord could be another interesting company in the chat space, as I'm sure they are considering an enterprise play (if they aren't they are insane, considering they have one of the more user friendly chat services and with SSO and a SLA they could probably charge $10 per user).

(Yes, I know Chime exists. But I don't think I've ever heard of anyone using it, and Slack is still rather popular.)

View on HN

>Yes, I know Chime exists. But I don't think I've ever heard of anyone using it, and Slack is still rather popular.

I think even Amazon realized how awful Chime is, given that they, as a company, switched to Slack just a few months ago.

i heard this, but all my meetings with amazon folk are still on chime. plus, slack audio and video capabilities are imo sorely lacking.

Amazon is a big place and I can't speak for all of it, but while IT announced that slack was "available" a few months ago, I don't personally know anyone that has actually switched to using it. Everyone still uses Chime.

That may just be a peculiarity of my department, but saying "Amazon as a company switched to slack" is definitely overstating what happened.

My org definitely has switched for the most part. But I'm sure it's a phased-rollout.

This is super org-dependent -- my entire org has completely switched to Slack, I only use Chime for meetings now (except for the occasional recruiter or someone outside my org who will ping me on Chime).

So that's the thing.

Switching to slack was brought up in our org and most people were at best noncommittal, because the option wasn't really "chime vs. slack", it was "chime vs. chime AND slack".

We'd still all need chime for meetings and interacting with any groups that hadn't switched.

It's actually pretty smooth since you can do `/chime` in a Slack channel or DM and it automatically starts a Chime call. Otherwise I just keep Chime open for scheduled meetings.

I work at a company that spends a lot of money at AWS, and one of the many support channels we have now is through shared Slack channels.

I can imagine that the main driver behind IT announcing Slack support was for their customer-facing staff to be able to provide support for their customers.

I would be interested to know more about these AWS Slack channels. Are the official? Do you get them with a particular support plan? Where can I find out more information on this?

They are officiall and we use our own slack instance. We have an internal dedicated slack channel with our aws team i(technical account manager, solutions architect etc) as well. It can be handy just for organising meetings etc but it is really hit and miss depending on your account team. We are on an enterprise plan.

It definitely helped to have a faster turn around time and provide more visibility for people.

>I don't personally know anyone that has actually switched to using it. Everyone still uses Chime.

It must be just your org/team, because a few of my friends working there in entirely different orgs had their teams switched to Slack fully within the week of announcement. Not claiming you are wrong though, because my sample size of 3 people/teams is purely anecdata.

Slack uses Chime under-the-hood for meetings. So, people are still using Chime. This was announced at re:invent today.

The video meetings are easily the worst part of Slack. Even though Slack meetings are free and integrated, everyone at my org still uses Zoom with Slack instead.

Are the video meetings worse than chat threading?

They don't work at all in Firefox, so... for any Firefox users, yes.

Threading that didn't work at all would actually be better than the threading that exists.

It may be Firefox. The video conferencing platform we use (Dekko Lynx) requires Chrome or Edge, because Firefox does not support the tech that Dekko needs for its end-to-end encryption calls (as we have been explained).

https://github.com/jitsi/jitsi-meet/wiki/Browser-support

I think my team switched over the day slack was allowed.

Know someone at Amazon in the kindle space and I say them using slack when we went to lunch today. He said his team uses it for a lot.

Weird. From my little view of the Amazon world it seemed like everyone moved within a week of availability. I always thought Chime worked fine, but it seems the kids want to wear the cool sneakers.

Chime has the best noise canceling of any service I've used. On the other hand, I've only ever used it to talk to Amazon employees at their request. That says a lot about adoption.

Chime is a great video call system, but wouldn't try for anything more. We're using slack, but still prefer chime for conferences.

Was it after Slack started using Chime as the backend for their calls implementation? https://www.geekwire.com/2020/slack-expands-amazon-partnersh...

Discord could be interesting as Amazon has already shown interest in the gaming / streaming / online community space with their Twitch acquisition.

One could copy-paste Discord, replace all occurrences of "gaming" for "enterprise" and it would be a much better business communication platform than Slack.

I don't think this is true. You can't really join/leave channels on Discord (yes I'm aware you can use roles to give permissions but it's not the same).

I've never used Slack but in Microsoft Teams there are Teams and Channels and they pretty much map 1:1 to Discord's Servers and Text Channels.

You cannot leave individual channels of a Team though you can mute them and you have even less permissions than Discord.

So maybe it isn't a 1:1 replacement for Slack but it certainly could compete in the field.

Teams is garbage compared to Discord or even Slack imo. Complete garbage UI.

This is my biggest complaint about Discord, and the reason I rarely check the few servers I'm on. There are three channels out of 80 I'm interested in. Just let me leave them!

You can mute channels and hide them, or mute sections, or mute chanells and collapse the sections they are in. There are lots of ways to get channels you don't want out of your face.

This isn't scalable. My employer has many many thousands of channels on our slack instance. Opt in is so much better than opt out at that level

You can opt-in with role-based permissions. Discord has been used by massive servers with hundreds of thousands of users, it's a bit involved to set up but it's not impossible. You can even automate a lot of it with bots.

You understand that an admin needing to set up a channel, a bot and a role is a bit more involved than any user just hitting create channel and inviting the relevant users/users searching and joining the public channel, right?

EDIT: Another example of discord's permission system being a poor match for the business use case is pinning messages to the channel is the same permission as deleting other user's messages. Additionally, the channel per server limit is 500.

And replace all instances of “Discord” with “Harmony” and you have the perfect Enterprise chat tool!

No no no, it should be called Synergy

Someone already thought of this. https://symphony.com/

Message threading would make it better though.

God please no. I would disable this in slack in an instant if I could. I dread the day it becomes available in any other platform I use.

Underrated opinion right here.

It's like they took IRC, saw it worked for businesses, and were like "how can we fuck this up, slowly, to destroy the very essence of what made chat channels useful in the first place?".

Why? It's a really fantastic feature in my opinion that turns a noisy chat room into more of a focused forum.

The issue for my Slack workspaces is that they are not noisy. So what happens is, people (inadvertently?) make threads, which then turns off notifications about that conversation for everyone else, therefore taking them out of the loop. And because the room is not noisy, they may not notice that they've missed the conversation until hours or even days have passed.

Ah that makes total sense!

They're inching towards this it seems, they've been incrementing rolling out revisions to the new reply UI for the last couple of weeks.

Google needs to buy Discord, do the gluey bits to integrate it into GSuite, shut down all its other chat offerings, and then let them do their thing.

What Google heard is: buy discord, glue it into everything and then shut it all down after 2 years.

I'm imagining a meeting a few years back, at Google:

"Ok, you know our motto, 'Don't be Evil.'?"

Heads nod.

"What if ... what if ... what if we simplify that?"

"How so?"

"Just cut off the first word, 'Be Evil.' Simple, elegant."

"I like it!"

----

Jokes aside, I think Google does a pretty good job of keeping the major products steady. Our little cohort on HN isn't a great sample for determining what "major" is.

Google Play Music, Hangouts, AngularJS, Works with Nest API....

They've killed off a lot of buds, but they've also killed some great products.

https://killedbygoogle.com/

Seems to me that either "Hangouts" (or whatever it's been most recently renamed to) is "major," and so many might argue it proves you false regarding "steady," OR it's not major, and discord would thus not be major and likewise not be kept "steady".

That's a fair assessment.

Given some of their moves with Google Workspace tying together Gmail / Meet / Chat, I doubt they could get away with this without significant antitrust issues.

They still haven’t fully closed on their Fitbit acquisition, so I’d expect they’d have a tough hill to climb for a Discord or a Slack.

Yeah, Google has a chat app addiction that’s not going to be solved by buying yet another chat app.

Relevant XKCD: https://xkcd.com/927/

Twitch already purchased Curse Voice about four years ago and renamed it to the Twitch Desktop application. It was what Discord is now before Discord existed. That was already a losing battle when Twitch made the purchase though so they ended up stripping out all of those features. It would have made more sense for them to buy Discord back then.

https://help.twitch.tv/s/article/Ending-Support-for-Voice-Ch...

> Twitch already purchased Curse Voice about four years ago and renamed it to the Twitch Desktop application. It was what Discord is now before Discord existed.

Curse Voice released in June 2015. Discord released in May 2015.

At best you could fudge the numbers and claim that Discord and Curse came out at the same time. A claim that Curse Voice was anything before Discord existed is flat out false.

Curse Voice actually released in early 2014[1][2]. I remember mainly because that year is burned into my memory from the CTO at the time snapping his fingers non-stop into microphones testing noise/echo cancellation and latency.

[1] https://www.player.one/curse-voice-vc-funded-skype-gamers-be...

[2] https://venturebeat.com/2014/06/17/curse-gets-a-million-game...

> if you already use Office, why not Azure?

My gut reasoning is that cloud infrastructure is not the same product as office productivity and the two use cases are not the same, the purchasing decisions aren't made by the same people (except at scale, I guess, I'll concede that), and the boots on the ground building shit in the cloud take the path of least resistance where that logic would never apply.

I don't really subscribe to the belief that a company should half ass a product or acquire something in a new market because they aren't already there, but I may be in the minority. Amazon has shown they can build a trillion dollar enterprise without caring about office productivity or chat, but they'll sell the shovels to anyone who wants to try and dig those holes.

First, you start with Office 365, then you put some stuff on One Drive, you add a little list in SharePoint online, then you use Teams to access all that. You move to Azure AD.

And then you want to automate or do some nice charts, and there is Power platform just around the corner...

and then you find yourself limited so you go Azure Functions... and then start one VM on Azure...

The jump from Power Platform to Azure Functions and a VM is likely not to be the same people

Power Platform is for "power users" to be able to "no code" things, this is targeted (MS Claims) as non-IT Users, as is the Rest of the office/Microsoft 365 stack

Azure (including Azure AD) however is targeted at IT Staff, dev's,sysadmins, etc. There is a clear line in how the tools are marketed, deployed, licensed and utilized

Next after that is running VDIs on either a small Citrix farm on Azure or as Desktop as a Service, since AD, SSO, 2FA ,Productivity tools and user file storage is there.

After that, you already have some firewalling and VPNs in place so adding the odd server is just a step away.

Adding dedicated lines and firewalls and rules and all bella and whistles with all four big players (AMZ, IBM, MSFT, GOOG) plus 2 ISPs adds cost fast, so you typically pick 2. MSFTbis already there for O365 so yeah, they have a head start

I kind of agree here, is Office 365 stuff targeted at devs and software companies? I sure don't care about it when I'm shopping for managed cloud services and infra to run my software.

They'd have to acquire so much more - it's AD, Office and Exchange that keep Microsoft embedded in the enterprise.

Hosted AD on Azure and Office 365 are _so good_ that nobody is coming close, not even Google.

You nailed it, Google product managers, specially in Cloud lack the Enterprise expertise. They need to get that a company with IT staff already went through updating NT to Windows 2000 and then 2008 and deploying Linux and Active Directory and Exchange 5.5, 2k..etc...many MS products which were de facto for any large Company, if you can't integrate with it you will never have a solid footstep into the Enterprise world, yeah they may use BigQuery and that's it.

My large disappointment with Google's Cloud Platform offering was lack of commitment to parity features.

I am sorry but the concept of "Beta release level" is a complete misfire/mismatch for what an enterprise needs. For example, you cannot, on one hand, advertise that you have a competitive feature (Google Functions competing with AWS Lambda), but then, on the other hand, turn around and label it beta and basically use that an an excuse for inferior reliability and inferior documentation. How can you basically GA a major piece of platform functionality, call it beta, and then ask any enterprise (that has lower risk tolerance) to use it? Enterprises don't want to invest into beta-labelled features, because they know it's an excuse for bugs and half-baked support. No wonder GCP is struggling for adoption.

("beta release" level is a conversation you have with select trusted partners that are willing to tolerate your early stuff in exchange for early access to value... it is not something that is acceptable to advertise as open enrollment to all)

Agree I heard they are changing the release process, they call it private early access now and then they will make it GA directly, no more alpha, beta and GA. But I think your point still is valid. Watching Re-invent today, just helped me realize that many Google Products are playing catch up and once launched Product managers and Eng team move somewhere else as there is no incentive to get promoted, this means your GA product may be understaffed or supported by new staff and at the end dead. Look at Cloud Datalab and the mess they have in ML world (Kaggle, Colab, AI Platform Notebooks, plans for new editor...etc)

I'm confused by your argument. Beta is the stage before GA: if it's GA, it's not Beta by definition.

This is all obsolete anyway, since GCP has announced they're moving to two stages: Preview or GA. That's it.

https://cloud.google.com/blog/products/gcp/google-cloud-gets...

Thank god they've changed that. Preview or GA is much harder line, where you can't hide behind a beta label for cycles on end. You can see the language in that release "give your business the confidence."

My point was that features were being advertised as essentially GA that were not actually GA. When evaluating the Google Cloud Platform sans-beta features, I was suddenly looking at a much more barren offering, in comparison to AWS.

Yes, thank you for the clarification. We had some problems with VPC-SC in the past. (ML training) and since it was Beta, there was no accountability. Anyways we are a AWS shop :)

That enterprise need reliability in Office land is a myth and MS is pumping out half baked systems just like Google.

Flow, Teams, Lists, Planner, Sway, Stream, Forms, Kaizala...

Sorry, but these aren't battle tested enterprise apps at all. Could as well pull random apps from the Google store.

Whenever something new releases the first reaction already is "not again"...

These apps are not why enterprise uses Microsoft though. Heck, in our 90k employees corp we don't use any of those

I am sorry but the concept of "Beta release level" is a complete misfire/mismatch for what an enterprise needs.

Anyone who uses GCP knows you type “gcloud beta ...” an awful lot. But enterprise sales are made on the golf course between people who have never and will never use the product themselves.

Hear hear, here's to products getting the job done well and reliably, untouchably. If you want to come anywhere near close, you have to put out something considerably better and I've yet to see that. Maybe Google Docs is starting to get there. Still I see expert users driving towards Microsoft Excel, for example. Because they are productive with it like you wouldn't believe.

Hosted AD on Azure or Azure AD?

While I think an MS environment can be a boon to enterprises, I think the MS cloud is very lacking, especially on performance.

Maybe because I am located in Europe, but it is not fun to work with.

We produce hardware and in dire cases we can do that with the pen and paper office solutions as long as our basic ERP solutions are working.

It would be no problem to use Linux and Libre office, it wouldn't affect productivity at all.

MS has good long term support but their cloud is certainly an exception to that. Not yet sure if good or bad.

If buying a business would put Amazon in the position of having to introduce a new end-user customer support burden, then they are unlikely to do so.

Eero devices require this customer support, but that can be merged with their IOT support team for all other in-home objects. I think Amazon made an exception to their “user support is never our priority” choice because they know it’s the only way to get a physical foothold in our homes, with Kindle and Echo, and I think the burden of Eero support was worth it to them for the mesh networking technology that has been harvested for Sidewalk.

Slack customer support cannot be easily integrated with any existing Amazon support team that I’m aware of, and does not offer a world-changing advantage to that degree.

Doesn't amazon have relatively good customer support for retail? Or are you specifically talking about AWS here?

Not in my experience, at least in the relatively-rare occasions where exceptions happen. I had an order which Amazon claimed to have been delivered but wasn't, and it took several days and multiple interactions to resolve. Just figuring out a way to talk to a human felt like an ARG.

The take away is that your issue was resolved. While it might not have been a satisfactory experience, you received resolution.

If your issue was with your Gmail account or Facebook, you'd still be streaming into the void.

The question was whether Amazon has "relatively good customer support for retail", which it doesn't in my experience.

If we move the goalpost to "have I been able to resolve problems", then yes — after 30 minutes of research into how to hack Amazon customer support in order to talk with a human, followed by multiple phone and chat interactions with generous on-hold time, I can share that Amazon support for retail technically exists and is excellent except when compared to 95% of other customer support experience I've ever had.

> If your issue was with your Gmail account or Facebook, you'd still be streaming into the void.

Ooof, I believe you. I'm Facebook-free, but maybe it's time to start looking for another email provider.

Amazon removed the ability to filter items only shipped and sold by Amazon.com. That is an anti customer move that squarely puts Amazon in the bad customer service category.

I have had good support retail side (granted, I'm not expecting much on my $20 item). On AWS I've had very good support.

Also paying user of google products and GSuite - support not great on the gsuite side, but retail order for Google Wifi etc pretty OK.

My support quality for a retail order from Google was 3 months of dead silence, followed by 10 of the replacement part all arriving at the same time

Yeah Amazon is really hit-or-miss.

I've had good experiences. But when you have bad experiences, they can be really bad. I ordered a TV on Amazon (Prime) and it was scheduled for delivery in 3 days. Because it's a large shipment truck delivery, they give you a window. Like Spectrum/Comcast. So I sat home all day and... nothing. I checked my messages and almost the exact time the delivery window was closing they left a message saying "could not deliver will try some other time" basically.

It took a month and about 15 customer service reps before a single person could tell me where my TV was. I'm not exaggerating. They had no clue what happened to a $1,000 purchase. It fell into their bureaucratic black hole like some scene from a Monty Python skit.

Maybe it was because you were trying to talk to a human in the first place?

when this happened to me, I simply filed the claim online and they resolved it, no need to talk to anyone

> ... I think Amazon made an exception to their “user support is never our priority” choice ...

Amazon is virtually the singular FAANG company that not only has customer support, but has generally very good customer support.

AppleCare and the Apple store Genius Bar are readily accessible, too.

Amazon pays under market for acquisitions while sf pays above market. Unclear if slack would ever grow revenue to the level sf is offering at, nor if shrinking set of alternative high-payers like cisco/google/other will, so cashing out while they can at such a high level makes sense.

| you already use Office, why not Azure?

Not sure why using Office would be a reason to prefer Azure over something else. Can you elaborate? Are there some useful integrations one can benefit from?

Regarding Amazon buying Slack. My guess would be it's too expensive. If you look at Amazon's acquisitions history, you would see they almost never spend more than $1B. The one exception I recall is Whole Foods which they paid ~$13B for.

Fewer vendors, unified billing/support/contract negotiations. These non-technical things are still integrations in a sense, that enterprises can benefit from.

It's nice to be able to get your email (Exchange), identity (Active Directory), productivity/office apps (Office), and underlying infrastructure (Azure), and even business intelligence/analytics (Dynamics) all from one place.

Yes it's nice. But not sure it's $25B nice...

I think the big seller/on-ramp is Azure Active Directory. Once you’re all set up to permission your shared documents in Office, and log into your corporate machines with the same accounts, those permissions are automatically available to cloud services. Authentication/authorization is built in automatically to any app you want to spin up, much more easily than syncing your ldap accounts to AWS or creating them manually, and/or manually building auth for your internal app. So internal apps and resources end up on azure, and other usages follow.

Disclaimer: I work for Microsoft, but I'm not here to tell you to use Azure - just providing an example for the use of AAD.

Kubeflow is an OSS framework for doing distributed data science / machine learning (including Tensorflow hence the name) on Kubernetes.

With the AAD integration it's just much more convenient to give a team of data scientists access to a Kubeflow Kubernetes cluster - AAD with the Kubeflow OpenIDConnect component will handle AuthN & AuthZ. Kubeflow itself sets up different Kubernetes namespaces for each authorized user upon first login.

You definitely don't need to run Kubernetes on AKS or Azure in general to take advantage of the AAD features.

Amazon would have been frustrating, but I suspect less terrible than Salesforce. What a mess.

I don’t think Amazon wants this. It seems like they are trying to have more completely automated, higher scalable services for companies to build stuff on top of.

While slack and Dropbox are customer facing and build on top of Amazon.

Theoretically that stuff is more value added, but I think their markets are smaller compared to what Amazon is going for.

If the goal is to be the buggy whip maker for the universe there’s a lot more there until they need to start vertically integrating and buying stuff that uses the buggy whips.

> as I'm sure they are considering an enterprise play (if they aren't they are insane, considering they have one of the more user friendly chat services and with SSO and a SLA they could probably charge $10 per user).

I truly hope not. I like discord for its informal, gaming, and specifically not enterprise or "professional" approach/atmosphere. I use it to chat with friends and talk in some communities (eg. programming ones).

I mean, you'd probably white label it, call it harmony as someone suggested up thread.

Joke’s on Microsoft, now Amazon can just buy Salesforce instead!

Maybe the culture at Amazon isn't a fit for Slack. Depends on the users, I guess, but it may end up as a tool that hinders productivity.

I think Salesforce compares itself to Microsoft and sees its ambitions to enter ERP, CRM and BI markets and they already have a collaboration platform with Teams, for better or for worse.

> If they had purchased Dropbox and Slack they could've gotten a foothold rather quickly and slowed down one of Microsoft's major selling points (you already use Office, why not Azure?).

Are you suggesting an enterprise is going to leave Office for Dropbox and Slack? Cause that's not happening.

Small businesses aren't nearly as attached, and the Office experience sucks on mobile and web. There's definitely a market for someone who can get that right, and I don't think that someone is Microsoft. I say this as someone who is irrevocably attached to Excel (and to a lesser extent, Word and PowerPoint), I don't think Microsoft is currently willing to cut off the entire ecosystem of software based around Office on the desktop, but if you were to make a clean break, make it cheap and have it work great on web and mobile, a new player could have a shot. There's tons of people going to Google and their suite isn't great, what if it was?

Can you explain the shortcomings of both Google suite and MSO on mobile/web you mention?

Amazon is more likely to buy Discord.

They also need email. They should buy Zimbra.

I heard people say the thing they like about discord is that it's not associated with work.

Surely S3 is a pebble compared to Dropbox?

You're probably backwards there.

Edit: in 2016 Dropbox stored 400PB (https://dropbox.tech/infrastructure/magic-pocket-infrastruct...), in 2018 S3 was storing multiple exabytes (https://www.quora.com/How-big-is-Amazon-S3-across-all-region...)

I suspect multiple exabytes is a low ball.

Plus, that 400PB Dropbox was storing in 2016 was - until around that same time, when they built their Magic Pocket system - itself stored on S3.

>You're probably backwards there.

Probably -- came back to my comment 24 hours later and realized I mixed up DB and S3 in the sentence.

I'm not as sure about that, given the extent of the chaos that ensues every time S3 goes down for half a day.

People don't put big data warehouses on Dropbox, they do in S3.

however amazon is not really a tech-first company, their main business is retail (even if a lot of their revenue comes from AWS).

Retail is the lion share of revenue but AWS is over half of profits last I checked.

Quite the other way, they're a cloud company with a gift shop.

I would say Amazon’s main business is commerce. They want to be involved everywhere people are buying/selling things. AWS & Amazon.com are coincidences of them starting up around the same time as the internet.

Before long, they’ll own the entire feature for “oh, I need to buy cereal. Alexa buy me cereal. Ok, done. <few mins later>. I’m passing by the Amazon Go store on way home, Alexa. Can I return these shoes I bought last week and pick up my cereal there. Sure.”

please stop with this meme ? meme is not tech first for whatever definitions of /tech first/ you have.

Tech first: Your main product is technology.

But then google is not a tech company either? Their main product is advertisements.

Retail is Amazons biggest revenue (raw cash) generator. AWS is amazons biggest profit (cash after expenses) generator.

You’ve got it backwards. Their main business is aws.

Reply on news.ycombinator.com