SoylentNews is people
SoylentNews
This is a post that I have suspected that I was going to have to write since late December last year.
You will now know that SoylentNews.org is closing down on 30 June but things have not been standing still behind the scenes since we first became aware of NCommander's decision at the end of last week. In fact, it has been a very busy weekend.
A small group of existing staff are looking at alternative possibilities for a 'replacement' site to keep the flow of stories going and allowing discussions to continue. This is a big task, especially in the 38 days remaining in which to try to achieve it. There are several possibilities which spring to mind, Pipedot for example. I have reached out to Bryan but have not yet received a response. However, things as not as straightforward as they seem. The pipecode is written in Php-5 which some of you will realise is no longer supported. We do not want to become dependant on old software which cannot be maintained into the future; that lesson has been taken aboard and reinforced by NCommander's explanation regarding his decision announced today. There are other options but at the moment it is still a search for what is available out there today which also appears maintainable into the future.
But the first thing we need to know is "Is there still sufficient interest in having a discussion site such as ours?" Do you, the community, still want to have your daily dose of stories and the ability to exchange views with many others on this site? Are there any community members who would be willing to join us in trying to establish such a site? Your views are crucial to everything that we do over the coming days and weeks. So please let us know what you think about whether a site is still required with all the alternative technology available today that simply didn't exist 9 years ago. What form should a new site take? What changes to how we operate are essential for you to continue to remain interested in the future site?
Of course, it cannot be a mirror image of what we have today - which many will see as a good thing! But I hope that we would be able to transfer existing accounts, usernames and passwords directly to any new site that we create. We would also have to start with a relatively simple site and build on that over time.
At the end of the day we would have to restart the voluntary subscriptions but not immediately. We can raise some funds to see us get established without the requirement of a financial commitment from the community. Subscriptions were always sufficient in the past and I don't see why that would not be the case in the future too. The fact that we currently have enough to keep this site going until next year bears witness to that. We have also found that we can significantly reduce our running costs based on our current community rather than being ready for a major stream of new members which never materialises. I have no grandiose ideas of becoming a huge site employing our own journalists but just a community that enjoys the discussions as we have been doing for several years. Nevertheless, we would also be trying to build on our existing community which is beginning to happen on this site now that things have settled down.
So don't hold back - let us know what you think.
(Score: 5, Insightful) by VLM on Monday May 22, @04:59PM (12 children)
I was going to write a big long post about modern parliamentary informatics and modern containerization architecture and stuff like that, but I'll just toss out some ideas first:
1) Before replacing, think about defining. What really makes /. and SN is a feature list including (but not limited to) psuedonyms, moderation, enduser story suggestions, threaded discussion, sorted by time, 'news', etc.
2) Wider range of software available now. Lots of 'social' software has been written since the 90s. You might be able to replace 95% of this site in a social/interactional sense with a docker install of mattermost or a trello-alike or redmine or something. Is there anything missing at a social level from putting "sn" into a private gitlab install, for those familiar with private gitlab installs imagine we shitpost inside projects in the issues section and stories are new issues and the mods are the only ones permitted to post new issues? Its an interesting mental exercise to see if the goals of SN could be achieved with a FOSS containerized solution like Mattermost. Or just how little extra "helper" would have to be written on top of Redmine to make it "do what SN does although it would look a bit different". For those who don't know Mattermost is pretty much a FOSS slack clone with a built in devops playbook tool that SN (probably) wouldn't need but how much modification of the SN concept would be required to put SN on MM? Imagine a MM where every "SN story" is a new channel and we wipe or lock channels over 48 hrs old. Or replace all of slash code with, essentially, one of a bazillion blogging software 'kits'? I have to admit I find the idea of a private gitlab instance acting as "SN" sounding pretty appealing, or at least its interesting to think about how it would work (or not). The point is you replace the system administration load with "docker pull" more or less. Well probably "docker-compose up -d" but you get the idea.
3) Some things have changed socially and moderation on the internet in 2023 is just political groupthink alliances based on demographic membership mostly. I don't think one number does it in 2023. Some sort of 3-d point of post quality as a scalar vs political compass vector or something. Or BRUTAL banhammer of conversational political discussion and stick to tech only. "How bout that X86-X architecture?" and ruthless censorship of all discussion of the plandemic or trump derangement syndrome, both diseases still plaguing our civilization. In a world of easy deployment do we need a central site with "politics" vs "tech" tags? The people with TDS can just go to the TDS site; the people wanting to talk 64-bit X86-S arch can go to the tech site. You could federate with links to the pages. I'm just saying a monolith tech and monolith social group is not mandatory anymore; could loosely federate tech with monolith social, or otherwise change the focus. But the days of the monolith might be over; there might never be another Yahoo home page or /. home page.
4) as a commentary on repeating expecting different results being insanity, I would suggest not reimplementing the same architecture and design but this time using the latest hotness in the small details or you'll be having the same discussion in the future. Seems like the outside world learned a lot about loosely coupled architecture and container based distribution and REST apis and stuff like that. So redoing "SN with exactly the same overall design now using Python and mariadb" isn't going to help long term.
(Score: 4, Interesting) by JoeMerchant on Monday May 22, @05:17PM (9 children)
There are many ideas that _might_ make a better community.
What we have is what made the community we have - I for one vote to try to continue what we have, maybe start "something better" if you've got the time & energy to run it, but do that in parallel and see what happens, actual outcomes involving people are hard to predict.
Україна досі не є частиною Росії Слава Україні🌻 https://news.stanford.edu/2023/02/17/will-russia-ukraine-war-end
(Score: 2) by VLM on Monday May 22, @05:36PM (8 children)
We're not necessarily disagreeing, well, not entirely. For example imagine this deep behind the scenes architecture instead of privately running one's own mysql cluster with our own DB schema and DB accessing code:
The backend is a private 'free' community edition gitlab instance with its REST API as seen at
https://docs.gitlab.com/ee/api/rest/ [gitlab.com]
The database for gitlab as a docker container is a 'solved' problem as is backup, replication, parallelization, auth, logging, etc. The gitlab people took care of all that. To make it SN-alike, create gitlab projects for "SN topics" and gitlab issues for "SN stories" and our comments are threaded discussion in those GL issues.
Now I know that doesn't LOOK like SN, but 'someone else' can write the web frontend talking to the gitlab REST API to format to look like SN, complete with rounded corners and no edit button for posts LOL. Or people can log into the private gitlab; its not a 'bad' UI although it doesn't look like SN.
As a private gitlab install, the admins can just dump the existing SN user database into this private GL instance. In fact I imagine a rather small (well.. large) Python script could dump the historical SN database into a gitlab instance.
Now that's not the only way to do it, but I was just tossing out the idea that having threaded discussion on a web forum doesn't necessarily imply running one's own mysql cluster or writing one's own API. The gitlab people (and, many others) have done a find job of taking care of all of it up to a nice REST API for threaded discussion.
Of course this is the classic "I've made this someone elses problem" to write a SN styled and appearing frontend to the GL REST API. I'm not a FE guy. Also this is hardly the "only" way to do this, although GL would work...
(Score: 2) by JoeMerchant on Monday May 22, @05:45PM (5 children)
>'someone else' can write the...
At work, I'm the 'someone else' who tends to pick up the crap that nobody else understands well enough to get it done in a reasonable timeframe. Not that I know anything about the crap, either, but I can usually figure it out.
Ever notice that "work" is a four letter word?
Україна досі не є частиною Росії Слава Україні🌻 https://news.stanford.edu/2023/02/17/will-russia-ukraine-war-end
(Score: 2) by VLM on Monday May 22, @05:52PM (4 children)
LOL I hear you but its a classic glass half full vs glass half empty thing; if someone else with very deep pockets has already written a full REST API with modern backend for threaded moderated discussions, may as well use it rather than try to write/maintain our own backend. At least "someone else" wouldn't have to write/maintain a free backend written by someone else.
Best part about standards is people like them so much they keep creating new ones, so if a private gitlab "issues as SN stories" is unpossible, well thats fine, do FOSS mattermost with "channels as SN stories", or FOSS project planning like redmine with "project issues as SN stories". Or generic ticketing systems with "trouble tickets as SN stories". Plenty of interesting backend options.
(Score: 3, Insightful) by JoeMerchant on Monday May 22, @06:02PM (3 children)
>if someone else with very deep pockets has already written a full REST API with modern backend for...
Yeah, there's also a very common choice between: "what does it take to build this thing from scratch" vs "what does it take to configure somebody else's idea of an extensible, maintainable, flexible framework and how many of their legacy bugs and regressions are you going to have to deal with while they issue constant updates and possibly architecture shifts in the name of progress?"
Some of my favorite FOSS comes from "dead" projects which haven't been maintained in years, because there's no need to: they just work.
Україна досі не є частиною Росії Слава Україні🌻 https://news.stanford.edu/2023/02/17/will-russia-ukraine-war-end
(Score: 2) by VLM on Monday May 22, @06:26PM (2 children)
My crazy semi-serious idea: Redundant array of inexpensive threaded discussion backends.
Its all just REST APIs, why not have the SN-appearing frontend talk to gitlab, MM, redmine, and others all at the same time LOL. They all have moderated threaded discussion articles, kind of.
Sometimes crazy is fun?
(Score: 2) by JoeMerchant on Monday May 22, @07:15PM (1 child)
Prototype it up on a Raspberry Pi and see how it flies!
Someone else asked for hard numbers on actually traffic loads / peaks, etc. I wonder how far we are from being hostable on a home internet connection with a NUC i5 or something similar?
Україна досі не є частиною Росії Слава Україні🌻 https://news.stanford.edu/2023/02/17/will-russia-ukraine-war-end
(Score: 3, Interesting) by RamiK on Monday May 22, @09:50PM
I've mentioned it a couple of times in previous discussions but Lemmy does threaded conversations and moderation: https://github.com/LemmyNet/lemmy [github.com]
There's a LOT of running instances already: https://join-lemmy.org/instances [join-lemmy.org]
Deployment is spinning a docker/ansible instance: https://join-lemmy.org/docs/en/administration/administration.html [join-lemmy.org]
Supposedly takes 150MB RAM so go ahead and give that raspi a try and see how much abuse it can take.
compiling...
(Score: 2) by driverless on Monday May 22, @11:56PM (1 child)
I'm perfectly happy with Soylent as it currently stands, it fixes the various eternal bugs in the other site and does the job. It's easy enough to dream up a more cromulent alternative, but then someone has to implement it and maintain it for the rest of eternity.
Perhaps a starting point might be for the admins to post a brief summary of what's involved in running the site now, to give others an idea of what they'd be getting into. That is, how much ongoing maintenance and effort is required to keep things running if someone else were to take up the reins?
(Score: 2) by driverless on Tuesday May 23, @12:01AM
Replying to my own post, the original Shutdown post, which I was still in the process of reading, already answers this question, so ignore the above.
(Score: 5, Informative) by janrinok on Monday May 22, @05:48PM
I was surprised to read the detail of pipecode - it is designed as a syndicated story source where different sites take the stories and display them, allow their users to comment on them, and the comments are syndicated also. However, we need further information. How does one control unwanted abuse, or what is stopping comments from one site spamming another different site? There may be, and probably are, answers to these questions but without more information we are really guessing.
We are not intending to replicate what we have today. But that is the very reason why I am seeking the community input as to what is now possible. Your comment has given me much to investigate. Yet even that is also putting the cart before the horse - we are also trying to ascertain how many of our own staff want to stay with the project, how many have (quite deservedly so) had enough and are happy to move on, and how many new staff might we find so that we can judge what is even possible in the future. Ultimately though, it will depend upon if there is sufficient community interest and support to make it all worthwhile.
I wanted to catch people before they decided that there was no longer any point in sticking around and they started to drift away.
Nevertheless, I thank you for your comment. You have opened even more avenues of exploration when I thought my task list was quite full.
(Score: 3, Interesting) by dalek on Monday May 22, @10:30PM
There's some real truth in what you said, but it's a shame you had to mix in some language to incite people such as "plandemic or trump derangement syndrome." That terminology isn't helpful. What you're really saying is to separate the tech from the politics, but that's mostly what happened. The tech was the front page stories and the politics was in the journals.
You have a very valid point about people tending to upvote and downvote based on agreement or disagreement, and that's true with respect to politics. Many other topics aren't prone to controversy, so the moderation is more likely to be based on the quality of the comment. A lot of the political discussion here was already separated into journals, but some of them generated a lot of controversy, and there were many blatantly bad faith AC comments. I would support banning the discussion of politics when there isn't a plausible link to technology. Yes, that goes for journals, too.
User-created discussions has been a feature of Slash and Rehash for as long as I can remember. Before journals were added to Slash, there were hidden SIDs where users could discuss anything they wanted. I don't believe there was an index of the hidden SIDs, so journals added a way to coordinate those discussions. Some of these discussions weren't helpful, and Slashdot actually had two hidden SIDs called trolltalk: https://slashdot.org/comments.pl?sid=20721 and [slashdot.org]https://slashdot.org/comments.pl?sid=31337 [slashdot.org]. One of those is still active to this day, even though I don't think it's possible to create new hidden SIDs. The name of those hidden SIDs describes exactly what they were, which was discussions about trolling Slashdot. There have to be some limits, but I don't think this feature should ever go away. It can be a place for users to ask tech questions, discuss stories that didn't make it to the front page, talk about sports, or just share life updates. I support allowing anything that's legal other than politics.
I also don't think Slashdot should be described simply as a news aggregator. Its origins were as Rob Malda's blog, where he shared whatever was interesting to him and wrote his own original content at times. He also spent a lot of time tinkering with the code pretty much right up until when he left the site. It wasn't a generic tech site, and it really focused on stories that were of interest to nerds, specifically those who ran the site. Slashdot isn't that interesting to me now, and a large part of that is because it feels like a generic tech site. There are plenty of places to read about things like NFTs and Meta, if that interested me. But there are very successful YouTube channels based on whatever is of interest to the creator, whether it's retro computing, how hardware works, gaming (sim racing is very cool, though I am very bad at it), or just about anything else. A nerd blog can still work as long as people know it exists, and that last bit might be a challenge. Even so, a site like MLB Trade Rumors [mlbtraderumors.com] is mostly a news aggregator, and it's extremely successful. There might not be a lot of comments on every story, but it still gets a huge amount of traffic. There's a lot of opportunity for a site like SN to work, provided people know about it and that toxicity stays out.
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest just whinge about SN.