Irrationally, this seems preferable to me to simply holding it hostage indefinitely. It always bothered me that they just stored all my old messages, but held them for ransom at a price I could not possibly afford to pay (something like $12k a year for my stupid little chat board that has $0 a year in revenue). It's better in my mind for them to delete my messages to free up space, because at least that's a reason I can wrap my brain around.
They probably realized after all these years that most teams on their free plan have no intention of ever upgrading, especially if they haven't done so after a full year of use. No point in taking on the costs to store all their data forever.
The pricing is insane. Maybe it makes sense for some businesses (we have it at work for sure), but I'm in 9 Slack workspaces currently, and only 3 of those are connected to any sort of business use.
Slack doesn't seem to have any pricing options for a general community, only for clearly defined business groups with clear onboarding/offboarding procedures. They charge by the user, by the month, so any kind of publicly open group seems impossible to fund.
Slack has never been a solution for hobby groups and open source communities and the like, but people love to keep using it as such and then blame the pricing model for not working for them.
Plenty of people use Discord, for better or worse. There are lots of slack communities that have little to no activity as well, so that’s an unfair judgement.
> Starting 26th August 2024, we’ll begin deleting messages and files more than one year old from free workspaces on a rolling basis.
Also note that you already can’t see messages older than 90 days on a free workspace, but they’re there if you upgrade. This change means if you do upgrade, data older than 1 year will be gone.
At larger companies, data retention policies that delete old data are pretty standard for platforms like this. A lot of them can harbor conversations that can be taken out of context or were never meant to be public, so it makes sense from a liability standpoint.
That said, there can also be a large body of fruitful technical knowledge sharing that happens on these platforms that should be retained for the good of the company.
The real rub is throwing the baby out with the bathwater. How can the signal be isolated/retained/copied easily so the noise can be purged? This is especially difficult for very old data that is spread across innumerable channels whose value could be lost and the team left unawares.
A traditional forum/message board is much better for archiving that fruitful technical knowledge. How many of us have found answers to a problem thanks to a forum post from 10 or 20 years ago?
I've never liked Slack/Discord for anything other than instant messaging, and always thought of it as ephemeral. If it's anything of significant importance, it should be documented in a place that is meant, by design, for documentation.
> If it's anything of significant importance, it should be documented in a place that is meant, by design, for documentation.
The challenge we have found is that amazing technical documentation is often born from ephemeral conversations. I would love a solution that eliminated the toil required to clip out those comments into a long-term store.
The flip-side to the challenge is that moving our team from Slack to another platform better suited to long-term technical Q&A requires a significant cultural shift that may not be practical.
> The real rub is throwing the baby out with the bathwater. How can the signal be isolated/retained/copied easily so the noise can be purged? This is especially difficult for very old data that is spread across innumerable channels whose value could be lost and the team left unawares.
I think you need a culture of summarising this type of info & saving somewhere that isn't ephemeral (e.g. README, team wiki). Else even with the history maintained, it's still a pain to try to a) play pattern matching with Slack's search engine and b) weed out all the false positives of old info that's no longer relevant.
This feels like the type of thing that maybe could be farmed out to a trained LLM in the future — "summarise today's discussion on X in channel Y and post into the wiki".
This seems like one of the few really compelling reasons, for a particular type of person, to push return to office. Fewer paper trails to get retained, with in-person conversations…
Keep, beyond legal requirements, for something as loose as Slack seems unusual to me. Having worked for some very large companies they tend to be very worried about purging email, Slack, etc., within months, before there's a chance for a subpoena.
Live chat is something worse than a commodity. There’s no way to innovate or do anything interesting at all. AIM solved the chat channel perfectly and Skype completed voice chat.
Since then the game has been: use investor money to pay for servers and collect users, then sell them to some big company.
An actual profitable chat service, like, as a business, can’t exist I think. It is impossible to compete with free service provided for “free” (free as in barnyard animals don’t have to pay for their dinner).
Lest this come across as smug or whatever, I’m personally eating from the Discord trough. It is what it is.
A chat server may be a commodity, but good user experiences, mobile apps, video calls, file storage, third-party integrations, support etc. definitely aren't.
Automatic infinite retention is a social failing, we should be selective about what we choose to keep because the rate at which we're creating it is so large.
It's not that hard to click 'save'. The old needs to make room for the new.
I think automatic infinite retention and public access to public data should be a basic human right. And we should be making lots of copies of all that rapidly-created data as well.
Yeah, totally. If it's private then it's up to whoever owns it. I think my experience is colored by using Slack for a group that's very much intended to be a public record: https://futureofcoding.org/community
Perhaps a better way to state my position is: I think publishing should be a ratchet. Everyone can choose what they want to make public. But once it's public that's it, there's no undo and no do-over. This isn't ideal in some situations, but on balance seems like the greater good. Stuff gets added to the commons of human knowledge, and we do our darnedest not to let stuff get removed once it's been added.
Some random anecdotes from my time on the Internet:
20 or so years back, I joined The WELL. I only lasted a couple months. It was impossible to have an interesting conversation without longtime members popping into the thread to "helpfully" point out that this had already been discussed back in 1988, 1993 and twice in 1997, with links to the relevant threads. I figured, if The WELL wants to be an information archive rather than a social space, that's fine, but Wikipedia is less expensive, better-curated, and easier to search.
About 15 years back, I gave up on contributing to Stack Exchange, and about 5 years back I largely stopped clicking on SE links in my search results. Similar story. The culture of preservationism meant that the site tended to favor pre-existing information and opinions. Back in the day, it just made the site uncomfortable for newcomers. More recently, it's also undermined the site's merit as an knowledge trove due to a tendency to favor out of date information.
Also about 15 years back, I became an admin on a Web forum. I observed firsthand that having a searchable record of posting history was a social negative that largely served to fuel flamewars. More Internet-savvy users were very adept at searching through the post history to dig up dirt on each other, hunt for evidence of people's opinions changing over time in an effort to try to call each other hypocrites, etc.
The rise of social media has only made this worse; Twitter mobs' ability to dig through a target's history for ill-considered or easily-misconstrued tweets that they can use for character assassination purposes is truly a sight to behold. And it has created a situation where more level-headed, continent people tend toward circumspection or even avoiding posting altogether.
On the professional side of things, at work I find that the usual corporate information troves - Confluence, ticketing systems, and group messaging - are rarely useful as an information trove, precisely because we hoard everything. It makes actually finding the information you're looking for in these systems darn near infeasible. Sturgeon's Law ("90% of everything is crap") is only true in the short term. In the longer term, 90% of what remains is compost, because it only ever had ephemeral value. Digging the reasoning behind a 5 year old technical decision out of a 10 year old Jira system is like looking for a chocolate chip in a cesspit.
Unfortunately, we've collectively painted ourselves into a corner in this department. Once upon a time, we curated information because we simply didn't have the physical space to keep all that paper indefinitely. Cheap digital storage and keyword search has tricked us into thinking that there's no cost to storing everything, though, so over the past quarter century or so we've gotten rid of all the secretaries, librarians and archivists.
My understanding was the Slack only allowed you to access X days back of content on a free plan (this page seems to indicate it was 90 days?). IIRC that’s a change (happened a while back) from when they used to allow access to the last 10K messages (which of course meant your time-based access depended on the number of messages you sent).
On the surface this just looks like Slack saying “we used to keep everything in case you upgraded but now we just keep a year back”.
Functionally it’s no different if you never planned on paying. If you did plan on paying then now you won’t get all your messages from the past back, just a year.
I wonder if Discord will ever do something like this. We're coming up on them storing messages from 10 years ago, and they're all still accessible and searchable(!) including attachments.
> To improve our services. We use your information to help us understand how users interact with our services, what features or products users may want, to develop features that make Discord safer and better to use, or to otherwise understand and improve our services.
> you can limit certain types of processing of your information:
> Limit our ability to use your data to improve our services. We offer certain settings that allow you to limit the information we collect to help us understand how users use our services, what features or products they may want, or to otherwise help us improve our business. If you turn off these settings, we will stop collecting and using certain event and log information as described in these settings.
It has never made sense to me that we want an infinite log of chats. The content is unorganized, and low quality. If something is so important you should put it in an email, wiki, PDF, etc, where the information is a lot more dense and easy to sort through. I look at stuff like imageboards, which aggressively delete old content, and can't help but feel that stuff like discord is over-engineered. If content older than 30 days was pruned or only the last 1000 were kept I think 99% of users would not care.
It also majorly creeps me out that everything I've said in what's supposed to be a casual conversation is searchable forever. Not that deleting old messages would necessarily change anything, but I don't see keeping them as a good feature.
Discord is one of those places that makes me realise the failings of the GDPR. For all the talk about fines of '4% global turnover' for non-compliance Discord seems to get away with not having to delete data?
I mean, I thought after 2 years you're supposed to delete the data, assuming the account is inactive. Google started doing that and I imagine if your old long-forgotten account got hacked and there was messages/attachments from 2+ years ago they would be fully liable as data is not supposed to be held for longer than reasonable. 10 years can be considered unreasonable in most circumstances.
But I cannot fathom why they want to pay for this, it's basically an unlimited storage file hosting platform they must be hosting TB's of probably useless data.
> Discord is one of those places that makes me realise the failings of the GDPR. For all the talk about fines of '4% global turnover' for non-compliance Discord seems to get away with not having to delete data?
What data are you expecting Discord to delete?
Not everything is PII.
‘Personal data’ means any information relating to an identified or identifiable natural person (‘data subject’); an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, an identification number, location data, an online identifier or to one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of that natural person.
I assume if I can log into my account and read old messages they are stored in a way that can be considered 'personal data'. For example, if they store my username and IP address with the messages then that is PII. Or even if they store a hash, if that can be traced back to 'me' then that would be PII.
As for messages, how can one know if each and every message may or may not contain PII? Or an attachment? You can't, which is why the best approach is a cautious one:
Delete all data from inactive accounts after 2 years, or even earlier to limit liability. The GDPR dictates you only store which is strictly required, messages from 10+ years ago from accounts that have not logged in for years have no right or nesscessity in being there
The deletion would be removing your user data but not past messages, like how current deletion work. Past messages get a new user ID shared amongst all deleted messages globally - 456226577798135808 - but attachments and content stays.
I don’t see a problem. Things like slack or discord should only be used for chat, primarily. Trying to find old messages is painful. It’s the wrong it for such job.
Why not include this data in their AI training models? Personally, I was irritated after that quiet 'opt-out' via email to prevent your corporate slack from being used in their ai training models change, recently. I guess they can double dip? Have you pay the pennies for data retention and use your corporate communications to train the next things they will sell you?
Because most companies genuinely don't value training on user data in that way.
It just isn't that valuable, even without the huge amount of negative publicity attached to doing that.
The cutting edge AI labs are leaning much more into high quality data (licensed from the Associated Press for example) and synthetic data, which it turns out is a huge part of Claude and Microsoft's Phi series.
But conversations in Slack aren’t your average webpage. Minus the channels used for automated messages/memes, a lot of in-depth, quality conversations happen on Slack on a large variety of topics
Can you imagine the storm of bad publicity that would emerge the first time some company has details of an internal strategy leaked because some chatbot ended up parroting those details back to a competitor?
I'm one of those people who constantly delete all their messages and data held on third-party services, so automating the removal of this actually appeals to me, but I know many will be upset by a change like this. I wonder how much money this is really going to save them, and at what cost in (potential) customer goodwill.
I'm actually surprised Slack isn't actively harvesting unused free instances, to be honest. I know of at least one group that would create a free instance per-event that would then sit idle after (presumably to avoid having to deal with security issues/key admins deciding to not participate).
