Yeah, this is the shortest export window I've ever seen by a long margin. There's a high chance that there will be customers who won't notice this before the end date.
I had high hopes for Quill and their high quality messaging app, but there is no way I am going to use it given that it is going to be locked and transformed into a Twitter product in the future. No thanks and no deal to that.
But it seems very obvious that Twitter is competing with Discord, and Slack. So that has been admitted.
Looks like company tanked and this is an acquihire, hopefully getting some investors some money back. No way there's any product/customer value here if they turn it all off 4 days after the announcement.
Quill had some big-time investors like Index Ventures, General Catalyst and Sam Altman. Often such failing companies are bought out as a special favor to them to maintain good industry relations.
This appears to be the MO for Altman. That’s what happened to his first [and only?] company he founded. Funny this can be called special favors. Cronyism seems more honest and apt.
But what we don't know is how long Jack and Parag have been thinking of Parag as the new CEO. It's perfectly possible that he had significant influence here if Jack has been grooming him as CEO for a while.
On the other hand, I just saw that "Quill was founded by the former creative director of Stripe, Ludwig Pettersson". [1] And who was Stripe's CEO, but Jack. So this could easily have been Jack arranging an acquihire for a pal.
Stripe was founded by Patrick and John Collison. I’m guessing you’re confusing it with Square, another payments company, which was co-founded and still run by Jack Dorsey.
Sorry. My mistake absolutely. Thanks for the correction. Not the first time I've accidentally mixed up the two finance companies named for common one-syllable words starting with s and ending with e. I'll be more careful next time.
> Quill is joining Twitter!
> Quill will be shutting down
Seems more like Quill's staff is joining Twitter than anything. Clearly Quill itself isn't joining anything as it's getting turned off and all data deleted.
First Sphere (https://www.sphere.me), now Quill.
Twitter is obviously planning something in this space, but to me, Twitter's magic is in the ability to really consume a lot, from experts, in a short space of time (240 character brevity!).
In depth conversation, spaces, voice, don't really belong on Twitter in my opinion, but I'm sure Twitter knows better than me... I have no skin in that game!
A lot of people use Twitter DMs and complain about that experience having room to improve (e.g. being able to search DMs). Maybe some of these acqui-hires (and that's what they seem to be) will be toward improving DMs?
I'm a customer of a company where I recently requested a relatively basic feature. They said they were putting together a team to implement it. That was bad.
Twitter, on the other hand, seems to need to buy a company in order to improve a basic feature. That's horrible!
What's up with management culture in tech companies!?
Buying a company is often a very efficient way of getting something built. The price may be lower than you think and they get a team that's already worked together and isn't bogged down by internal Twitter politics. If the executive running the acquisition can protect the team from those politics they can get something built faster than they would otherwise, and at $10m Twitter revenue per day, getting something out the door faster has a lot of value.
Twitter doesn't really understand what made it so successful. What did it start as but a blogging service limited to 140 characters? That's why they are so resistant to change. So they're looking for a company that has a user experience that the trust, not just spinning up a crew to spit out a feature.
Idk maybe I'm still sour after they killed "We are hunted" 10y ago, but I feel Twitter acquired a number of companies and yet remained virtually unchanged for years.
They acquired WAH to merge it into Twitter Music that they later killed. They bought Vine and wasted its potential, idk if Periscope is still alive or not (also acquired).
Maybe this time it'll be different - their recent acquisitions of Revue and Chroma Labs (which I guess became Twitter Spaces) seems to be kind of working so far
This happened with another Twitter acquisition Smyte, where in that case they shut down the api immediately [1]. It seems like it’s some property of Twitter’s acquisition process?
I feel bad whenever I read about this because I’m friends with a Smyte founder and I know they absolutely didn’t want things to play out the way they did. I had hoped Twitter learned their lesson.
> It seems like it’s some property of Twitter’s acquisition process?
What if there are serious security vulnerabilities though? I could see having a process where founders don't turn over their source code until midnight when the deal closes, and then the acquiring company gets the option to shut it down immediately if there are serious and unfixable security vulnerabilities discovered between midnight and 10am when the press release goes out.
I'm not saying this hypothetical process is a good one, much less that this is what happened, but I could certainly imagine things playing out this way.
The modern SV startup treats users the way that a coal miner treats a mountain or an LBO guy treats a middle market company. Just a resource to be exploited as they go from Engineering Manager at Salesforce to VP at Twitter and given nothing but the middle finger for their troubles.
I have known both EMs at Salesforce and VPs at Twitter, and believe me they return the favor.
The idea that SV startups aren't staffed by ruthless business people is one of the sillier thoughts still percolating around. YC/HN and similar orgs are the child soldiers/shields of VCs. There isn't anything edgy or innovative about it, and hasn't been since the 90s.
Yes it does. You don’t necessarily have a right to the other side of the conversation, but your messages are your own data and you have a right to export them.
That said, gdpr is untested in the context of data being deleted before it can be retrieved.
Yea, I think GDPR is a bit of "national government trying to govern international commerce" and not doing it very well, yet it's still at least there in theory.
Oh I was thinking more about the inability to export, or related to the data portability aspect, but I also don't have a very deep understanding of GDPR.
Not really. You could enforce that as an option on certain threads but you could simply have it be open with a single switch
I tried it out and it was a weird Slack alternative that was just that little bit of clunky where you couldn't use it on a daily basis. I actually feel some of the ideas were sound. I wouldn't have minded a thread-only topic where people post a bug at the top and discuss that particular bug below. It's hard to explain but there was just something about navigating all of it that made it just slightly more painful than was bearable. I'm sad to see it go because it could have made for a decent alternative to slack but it's so dumb that it shuts down so soon
> We started Quill with the goal of increasing the quality of human communication
We started Quill with hope someone buys us one day.
> Together with Twitter, we will continue to pursue our original goal — to make online communication more thoughtful, and more effective, for everyone.
We sold our data to Twitter so they can make more money.
> We’d like to thank everybody who has used Quill — if you came on board during our beta, or if you just sent your first message last week. We can’t wait to show you what we’ll be working on next.
We can't wait to use the data you supplied us for free to get even more money from Twatter.
Which may help explain why they're shutting down less than a year after launching. 10% improvements aren't enough to drive switching, especially against network-effect products like Slack. But having taken $14.5m in investment, you'd think they'd try a little harder.
I have no idea what makes people use Slack. We demoed it at our company, and most people hated it. I guess it's one of those things people start using when there's a handful of employees and then they just get stuck with it.
The worst feature about Slack is its "threads" feature. You're effectively manically thread chasing or you just have to learn to not care about past conversations.
The other annoying thing was input lag when typing sometimes. It was completely unpredictable when it happens (wasn't pc load, browser wasn't that busy, but i didn't check the profiler; internet was fine). It would come and go on its own. Not everyone had it either, and it seemed to be isolated to Linux machines.
We demoed Zulip, but some people didn't like that one because the UI is... It looks like brutalist design applied to UI. Most of us were ok with it even if it's a bit "ugly" (it's not ugly, but I'd call it an acquired taste). Quill was supposed to fill this gap for us, but the CTO said that the $15/user/month price is too steep. Lol.
We settled for the mediocre of both worlds: Mattermost. It's not great, but it's not shit either. You can't scroll in it (well you can, but you don't want to because it's buggy) and the search is crappy, bite you can usually find what you're looki for after a few tries.
It's crap but it's better than email for remote team communication. And once you're in it's a pain in the arse to change to a different service, so people don't.
Odd. We generally love Slack. I think you have to use it a little while for everyone to adopt sane norms for it, though. Among them is being quick to join and leave channels. A sales person has a question for me? They summon me to their channel where we chat for a bit. All done? I leave it so that I don’t see the rest of their chats. By now, if I see a conversation at all, there’s a 90% chance it’s relevant to me.
Not saying that it’s the best possible solution, just that we’ve been pretty happy with it for the last several years.
The leaving and joining channels is an interesting idea. I guess that kinda simulates Zulip's topics/threads feature to a degree. I guess that wouldn't have been too horrendous to do. I kinda do this with Mattermost, but I always feel awkward just leaving, so I leave after a few days usually lol.
Aside: I love the threads/topics implementation in Discord. However, a lot of people don't want to use Discord for work (I'd probably make a separate account myself; I don't like job stuff seeping into my play time).
Yeah, that’s the “norms” part. After a while you get a feel for how long you need to stick around after you’re no longer part of the conversation.
All that said, I love how Mattermost is coming along. Open source + on prem is a huge win in a lot of settings. For instance, I don’t think you could use Slack at all for HIPAA-covered talk. Last I looked you couldn’t sign a business associate agreement with them. You could use a properly secured internal service for that, though. For example, I know medical offices using Synology Chat for inside conversations like “we need to reschedule Joe Smith’s appointment”.
Mattermost and Slack are at least in the same class. Slack proved their market by being light-years ahead of Outlook, and Lynq. They took "the nerds who love their little private IRC server" to the masses.
I have used Slack at a few companies and it has worked fine for us. No reason it should work for you as well, of course; everybody's different.
FWIW, the thread-chasing thing is not a problem for me because toward the top there's a sidebar element that says "Threads" that lights up when there are new comments in threads.
I've used it only on Linux and haven't experienced input lag. Not sure what the story is there.
That Threads element only lights up if a thread you've participated in (or have been mentioned in) has activity. If you work in an org that uses threads and you go on vacation for a day or two, you're going to have to spend your first morning back looking for threads to read. You can't know how far back to scroll back either because someone could have replied to a month's old thread.
Threading should not have been added, IMO, without a way to optionally show all thread comments in a single view as they are delivered, as though they were regular chat messages (with a thread annotation).
Huh. We use threads. I've taken time off. I felt no obligation to go back and read every possible thread. Why would your coworkers expect anybody not in the thread to see sudden participation in a month-old thread? Mine sure don't.
It's not an area where I would want to try to claw out a niche, when the 9000 pound gorilla is essentially giving away their product to everyone with an Office 365 subscription.
For sure. It's very much "where elephants dance" territory. Heck, I heard Slack's founder, Stewart Butterfield, on a podcast where he talks about the rich-get-richer dynamic that made doing Slack pretty easy once things got rolling. I can't imagine trying to compete against two competent companies with ~infinite money and a desire to own the space.
Lots of criticism of the short export window, but no one knows how many customers they have. Quill was early stage, trying to get into a competitive market, and they've exited in what looks like an acquihire. From this I'd interpret that they had low enough traction that the business couldn't continue. They may only have 10s of active teams, and it's fairly possible that they are in good contact with all of them and helping them do manual exports.
>> Can I export my team’s Direct Messages (DMs)?
> No, we do not allow the export of Direct Messages.
As for this, this is very much in the "feature" category rather than the "bug" category. As much as communication on a company provided platform may not have a reasonable expectation of privacy, people still use these chats for private communication that they don't necessarily want all their coworkers knowing. Well done to Quill for taking this design approach. It may not even be technically possible if they're doing E2E encryption with DM participants.
> Lots of criticism of the short export window, but no one knows how many customers they have.
Even if they had only 1 single customer, that's not an excuse to close the service and allow exporting with just a 4 day notice. People and companies have lots of priorities, and now they've been forced to focus on this instead of whatever they were planning on doing. There should be at least 30 more days of service running, with another 30 days with the service offline but data exporting still available.
Nope, there really aren't. Picking up an entire established team that has a hierarchy, works well with each other and are subject matter experts is easily worth the savings in recruiting time and money.
>We started Quill with the goal of increasing the quality of human communication. We believe the tools we use to communicate today are not the best they can be. Together with Twitter, we will continue to pursue our original goal — to make online communication more thoughtful, and more effective, for everyone.
That's the OPPOSITE of what I feel Twitter is ... and what draws people to it.
I find it amusing that the page title is, "Twitter + Quill - Quill - Messaging for teams that focus". So the result is just Twitter, without Quill or messaging for teams that focus? That seems to be accurate.
wow - it's 20months from they first showed quill on HN (April 2020) - let's say it it had taken them 4 months prior to dev first version (Beta) - that's ~2 years to startup.
Giving user 4-days to export data, it's it's pretty similar to pulling the power cable on the server to shut it down ... unless they just had friends & family (piloting) using it, it's very odd. that's also not taking into account data protection / retention periods that business have to comply with local/regional jurisdiction laws.
having said that, concrats on exiting - acquhire or not, it's no mean feat.
Quill was never going to work to begin with. This is an easy way for the VCs to get some kind of return. Just another day in silicon valley insider dealing.
Good for Quill’s team, but not great for competition in the team collaboration space. I’d really like to see Discourse and Zulip get a much better UI to compete more directly with Slack.
We have a small team, and chose Quill over Slack because it seemed like the client felt more lightweight, and the product team would actually solicit feedback from us in DMs.
It very much sucks that I have to drop everything to move to an alternative. We'll probably go with Slack. I'm hoping you can still sideload their iPad app on an M1 Mac.
In the end, I don't think Quill was significantly better. Search in chat history was nearly useless, and a reason we were already reconsidering alternatives. But those issues aside, it was never better enough that I'd go out and tell people, "You need to try Quill! It's so much better than Slack!"
It's a shame that people will draw the wrong lesson from this. They'll think that Slack and Teams have deeply entrenched network effects, which might be true at large enterprises, but small teams like mine will gladly use a competitor that provides a significantly better experience. There might not be enough of us to justify raising $14.5 Million, but I suspect you could bootstrap something quite nice.
Maybe something has changed in the last few years, but the last time I used it, I found their Electron app to be a resource hogging pile of garbage. But that's just my opinion.
It was. They made improvements with their election app. Few years ago, it is resource hog and eat up batteries like it is candy while idling. It went from 100% to 30% within two hours. Now, it barely used up much resource and very light on battery. I recommend trying it out again, it is much better now than it was before.
It's still extremely heavy compared to any native Mac app. On my M1 machine, Slack (plus it's 4 separate helper processes) is currently using a total of 679MB of RAM. The Mail app, in comparison, is using 147MB.
It is not fair to make that comparison. Mail app is first-party app made by Apple. They will optimize it as much as possible and it only for macOS. Slack is a third-party app, run inside Electron which is cross-platform. A cross-platform that have to ensure that every platform that Slack have to account for OSes quirks. Mail app (email application) and Slack (instant messaging application) are entirely different beasts, they are not the same thing.
I'm glad that Slack only uses 600± MB of memory. Few years ago, it usually uses more than that, about 1.5 - 2 GB in my experience. Now I checked my Slack (and its services) combined memory usage is 400± MB.
From an end user perspective I think it's absolutely fair to make that comparison. Electron is a bag of garbage for the end user / machine. Native apps (mail or otherwise) perform, feel, look better than electron apps while using far less resources including power with less crashes in my and my coworkers experience.
It's not as heavy as MS Teams (which is possibly the worst software I've ever used) but it's still heavy, it "fixes" some of its memory leaks by reloading on a timer...
M1 itself is the biggest thing that changed. Being an energy hog isn't as big a deal when the CPU is so efficient that the battery lasts more than a workday almost no matter what. Age old case of increasingly inefficient software relying on better hardware to cover for it.
It was a great fit for us. The thread management was what slack was missing. We loved the promise that the system was going to develop even-better features for suggesting message groups using ML. The groups-of-channels was also a relief to the complexity of so many channels through the org. We were willing to put up with the incomplete look/feel compared to Slack. We struggled to get the whole team to switch, and now the product is closing. Sad
Our small company when with a hosted[1] Matrix Synapse server with federation turned off. Have been quite happy with it.
We use the Element client, which is yet another Electron application, but there are several other clients out there, so you would have some options to explore.
132 comments
[ 4.2 ms ] story [ 273 ms ] thread[1]: https://twitter.com/squad/status/1337398991210725379
But it seems very obvious that Twitter is competing with Discord, and Slack. So that has been admitted.
I'm not sure how they would even begin to compete with Slack. The only similarity is they have messages in the data store.
I'm sure Twitter also had that goal on their first day.
[1] https://techcrunch.com/2021/02/23/the-new-new-slack-quill/
This seems like just the result of unfamiliarity with the payments domain to me, rather than "horrible" branding.
I’ve heard someone mix up Slack and Stripe. Doesn’t mean those two are bad branding. The person isn’t someone who cares about SV or startups.
Seems more like Quill's staff is joining Twitter than anything. Clearly Quill itself isn't joining anything as it's getting turned off and all data deleted.
In depth conversation, spaces, voice, don't really belong on Twitter in my opinion, but I'm sure Twitter knows better than me... I have no skin in that game!
I'm a customer of a company where I recently requested a relatively basic feature. They said they were putting together a team to implement it. That was bad.
Twitter, on the other hand, seems to need to buy a company in order to improve a basic feature. That's horrible!
What's up with management culture in tech companies!?
@jack(ass)'s announcement that he was quitting Twitter was a tweet of... a fucking screenshot of a wordy e-mail.
They acquired WAH to merge it into Twitter Music that they later killed. They bought Vine and wasted its potential, idk if Periscope is still alive or not (also acquired).
Maybe this time it'll be different - their recent acquisitions of Revue and Chroma Labs (which I guess became Twitter Spaces) seems to be kind of working so far
> No, we do not allow the export of Direct Messages.
>> If I don’t export, will you delete my data?
> Yes. On 1pm PST, Saturday, December 11th 2021 we will delete all user data, whether or not you’ve exported it.
Fantastic service.
[1] https://techcrunch.com/2018/06/21/twitter-smytes-customers/a...
What if there are serious security vulnerabilities though? I could see having a process where founders don't turn over their source code until midnight when the deal closes, and then the acquiring company gets the option to shut it down immediately if there are serious and unfixable security vulnerabilities discovered between midnight and 10am when the press release goes out.
I'm not saying this hypothetical process is a good one, much less that this is what happened, but I could certainly imagine things playing out this way.
The idea that SV startups aren't staffed by ruthless business people is one of the sillier thoughts still percolating around. YC/HN and similar orgs are the child soldiers/shields of VCs. There isn't anything edgy or innovative about it, and hasn't been since the 90s.
That said, gdpr is untested in the context of data being deleted before it can be retrieved.
Thanks for clarifying that for me.
I tried it out and it was a weird Slack alternative that was just that little bit of clunky where you couldn't use it on a daily basis. I actually feel some of the ideas were sound. I wouldn't have minded a thread-only topic where people post a bug at the top and discuss that particular bug below. It's hard to explain but there was just something about navigating all of it that made it just slightly more painful than was bearable. I'm sad to see it go because it could have made for a decent alternative to slack but it's so dumb that it shuts down so soon
> We started Quill with the goal of increasing the quality of human communication
We started Quill with hope someone buys us one day.
> Together with Twitter, we will continue to pursue our original goal — to make online communication more thoughtful, and more effective, for everyone.
We sold our data to Twitter so they can make more money.
> We’d like to thank everybody who has used Quill — if you came on board during our beta, or if you just sent your first message last week. We can’t wait to show you what we’ll be working on next.
We can't wait to use the data you supplied us for free to get even more money from Twatter.
Looking at that page, I guess it's Slack with some small twists? And this seems to confirm that: https://techcrunch.com/2021/02/23/the-new-new-slack-quill/
Which may help explain why they're shutting down less than a year after launching. 10% improvements aren't enough to drive switching, especially against network-effect products like Slack. But having taken $14.5m in investment, you'd think they'd try a little harder.
I thought this was how startups work? Rake in some VC cash with empty promises, sell off to some big company, execs retire, everyone's happy.
Everyone but the users who need to scramble to export their data -- without tools -- by Saturday.
The worst feature about Slack is its "threads" feature. You're effectively manically thread chasing or you just have to learn to not care about past conversations.
The other annoying thing was input lag when typing sometimes. It was completely unpredictable when it happens (wasn't pc load, browser wasn't that busy, but i didn't check the profiler; internet was fine). It would come and go on its own. Not everyone had it either, and it seemed to be isolated to Linux machines.
We demoed Zulip, but some people didn't like that one because the UI is... It looks like brutalist design applied to UI. Most of us were ok with it even if it's a bit "ugly" (it's not ugly, but I'd call it an acquired taste). Quill was supposed to fill this gap for us, but the CTO said that the $15/user/month price is too steep. Lol.
We settled for the mediocre of both worlds: Mattermost. It's not great, but it's not shit either. You can't scroll in it (well you can, but you don't want to because it's buggy) and the search is crappy, bite you can usually find what you're looki for after a few tries.
Not saying that it’s the best possible solution, just that we’ve been pretty happy with it for the last several years.
Aside: I love the threads/topics implementation in Discord. However, a lot of people don't want to use Discord for work (I'd probably make a separate account myself; I don't like job stuff seeping into my play time).
All that said, I love how Mattermost is coming along. Open source + on prem is a huge win in a lot of settings. For instance, I don’t think you could use Slack at all for HIPAA-covered talk. Last I looked you couldn’t sign a business associate agreement with them. You could use a properly secured internal service for that, though. For example, I know medical offices using Synology Chat for inside conversations like “we need to reschedule Joe Smith’s appointment”.
FWIW, the thread-chasing thing is not a problem for me because toward the top there's a sidebar element that says "Threads" that lights up when there are new comments in threads.
I've used it only on Linux and haven't experienced input lag. Not sure what the story is there.
Threading should not have been added, IMO, without a way to optionally show all thread comments in a single view as they are delivered, as though they were regular chat messages (with a thread annotation).
>> Can I export my team’s Direct Messages (DMs)? > No, we do not allow the export of Direct Messages.
As for this, this is very much in the "feature" category rather than the "bug" category. As much as communication on a company provided platform may not have a reasonable expectation of privacy, people still use these chats for private communication that they don't necessarily want all their coworkers knowing. Well done to Quill for taking this design approach. It may not even be technically possible if they're doing E2E encryption with DM participants.
Even if they had only 1 single customer, that's not an excuse to close the service and allow exporting with just a 4 day notice. People and companies have lots of priorities, and now they've been forced to focus on this instead of whatever they were planning on doing. There should be at least 30 more days of service running, with another 30 days with the service offline but data exporting still available.
Let's say they have a couple customers... that should be even easier to maintain for a while.
Nope, there really aren't. Picking up an entire established team that has a hierarchy, works well with each other and are subject matter experts is easily worth the savings in recruiting time and money.
That's the OPPOSITE of what I feel Twitter is ... and what draws people to it.
having said that, concrats on exiting - acquhire or not, it's no mean feat.
It very much sucks that I have to drop everything to move to an alternative. We'll probably go with Slack. I'm hoping you can still sideload their iPad app on an M1 Mac.
In the end, I don't think Quill was significantly better. Search in chat history was nearly useless, and a reason we were already reconsidering alternatives. But those issues aside, it was never better enough that I'd go out and tell people, "You need to try Quill! It's so much better than Slack!"
It's a shame that people will draw the wrong lesson from this. They'll think that Slack and Teams have deeply entrenched network effects, which might be true at large enterprises, but small teams like mine will gladly use a competitor that provides a significantly better experience. There might not be enough of us to justify raising $14.5 Million, but I suspect you could bootstrap something quite nice.
I'm glad that Slack only uses 600± MB of memory. Few years ago, it usually uses more than that, about 1.5 - 2 GB in my experience. Now I checked my Slack (and its services) combined memory usage is 400± MB.
We use the Element client, which is yet another Electron application, but there are several other clients out there, so you would have some options to explore.
[1] https://element.io/matrix-services