This is such a turn-off for projects to me. Please accept random terms of service and privacy policy of a closed-source product that will be collecting and selling your data as the only means of participation in this community.
I agree; I don't like that either. My own projects are using the IRC and NNTP instead, and I will not add Discord despite anyone else's suggestions (although there is a Matrix to bridge with the IRC). (Currently there isn't anyone else on; I have been told that it is because I do not publish binaries, but I don't know.)
I don't necessarily see something wrong with having one for a project that follows the code of conduct, but a) making it the only means isn't acceptable and I'd argue b) bridging to Discord is possibly more harmful since it can implicitly be sending data where ToS cannot be declined by community members.
Unless you mean binaries for Windows, I'd assume the lack of community is due to something else; IRC is definitely foreign to the younger crowd, but its not a hard hurdle to jump.
Personally I avoid IRC since I just don't like "chatting", email or forum is a far easier flow, socially. It also seemed rare to find anyone actually present in their channels, timing never works out. I'll still use it though if its the only option and I realllly want an answer
This program is a Linux program, and uses some GNU functions (and some other assumptions that are appropriate for UNIX-based systems but not Windows), so will likely require changing it to get it to work on Windows (although it might work on WSL, but I have not tested this).
This project has NNTP, so if you don't like "chatting", then the NNTP can be used, which is more like the email or forum.
I mentioned Windows since most linux users at some point have been exposed to compilation, where as windows users aren't likely to even know where to find a compiler.
I have never used NNTP, but now im curious to learn some more about it; growing up I had never seen it used, but i think i just missed that age
About Windows/Linux I believe you. Nevertheless, there are only a few people on the IRC/Matrix and they have not written anything there (although I have written).
About NNTP, there is the RFC documents relating to it, and some NNTP servers. Some are Usenet, although the NNTP server for my project is specific to that project. You can use any NNTP client software to access it (or just operate the protocol directly); some email software can also use NNTP.
I completely understand why gaming communities centralized on Discord. It has the most convenient and obvious interface for managing little communities of multiple related rooms, and the drop-in drop-out voice chat room system is obvious.
I have a lot of trouble wrapping my head around why so many tech and programming (especially FOSS) communities still flock to Discord, especially given how programmer-hostile the company is. Matrix is more than good enough to manage it now. The only things about Matrix that bother me right now is that presence is still disabled on the matrix.org homeserver, and the phone call metaphor for group voice and video chat rooms is clunky.
When I still did online gaming, we'd tried a couple of different tools but the last one we settled on making everyone, even the non-tech-savvy, happy was Mumble because it was lightweight, no frills, low-lag with good audio and decent detection for auto enabling and disabling voice for people (like me) that didn't like extra buttons for push-to-talk.
Mumble has long since falling out of favour. It’s missing most of what people consider critical features now including mobile support, good IM support, and the ability to instantly create new communities without dealing with hosting.
I’d say it’s likely that most discord users are mobile or at least half.
I won’t disagree. My anecdote was from 4–5 years ago. We’d assessed that Discord was too much for us specifically. I had used the Mumble Android app from F-Droid a time or two to communicate while on the go just to chat. As for the hosting, at the time I also had dynamic DNS and actual ISP and it was lightweight enough to self-host a room with a couple of people without issue. I think not dealing with hosting probably is a big issue for a lot of people. Monetarily the difference is small, but philosophically free vs. $2/mo is massive.
Mumble is great for voice chat! It’s just bad at everything else, sadly. I think modern telephony significantly profits from the ability to use video and screen share, which you just can’t do with Mumble.
This is why any telephony solution I’d build would probably be based on Jitsi Meet.
In case anyone’s bored: A channel based UI (similar to Mumble) for it would be rad.
I’m not a fan of Discord for many aspects, but everything else is even worse. Matrix, even after all these years, is still a massive pain to use. It doesn’t have a way of putting together multiple related channels (I’m aware it’s coming), which is a Discord core feature. It also doesn’t have voice rooms, which may not be central to the FOSS community’s use of it, yet extremely useful for it.
There are other organization chat solutions like Slack (proprietary), Mattermost and Zulip (I’m trying out the latter with an organization). Those are the real alternatives to Discord, not Matrix (... yet ...). Neither of them (besides maybe Slack?) has an integration for voice channels (not “calls”).
I know some projects for hobbyists that they setup facebook groups and you can only get support via their effin facebook group, damn them all... at least I have a total-fake account for things like that, so feed the system with BS.
Most of these just require forum and not a chat community channel like Discord, where everyone keeps asking the same question, because things are not easy to search or organize.
But the reason people aren't making forums, is it is not as easy for community owner to make forums (need to get cloud storage, then configure something like discourse forum, get a domain name etc.)
Same for the end user who wants to be part of multiple communities, has to remember the URLs of multiple communities, has to have multiple logins and passwords and needs to create an account in each of them.
If someone combines discord + discourse (basically an app, that has all your forums on the side bar with just one login, and people can create a new forum for anything, similar to how you can create a discord server) then I think we might end up going back to forums with more meaningful discussions and all information related to some issue in one place.
Ding ding ding, I think you hit the head of all the nails there.
I don't use Discord for anything serious but it's just very convenient to join any Discord group with a single click since I'm already logged in there.
I think we need to put aside the security practices of Discord (it's not like the discussions tend to be sensitive in any way), because I'm not aware of anything that comes close in ease of setup and user onboarding, at this point. Setting up Zulip or Mattermost, paying for hosting and then managing it all is quite a hassle in comparison.
Fair, but what I meant was that I don't care whether the security of a Discord group for software development for a certain project is horribly flawed. I'd approach such a group with zero security/privacy expectations, since there's not really any kind of private discussion going on and I'd consider everything to be public anyways.
Use Discord for anything remotely private and we can have a discussion. :)
It's not only about the security practices, it's about the visibility. I'm not a huge fan of Discourse, but when projects use it, I don't need to create an account to read, and it acts like a webpage, not an application.
Personally I think Discourse strikes a pretty good balance. It's a bit like a mailing list where it's easier to search and users can edit mistakes while having more modern features a browser can provide.
I'll preface this by saying that I'm not fond of the Github monoculture, but something like Github discussions or a forum on Github would be better for most of these things.
> If someone combines discord + discourse (basically an app, that has all your forums on the side bar with just one login, and people can create a new forum for anything, similar to how you can create a discord server) then I think we might end up going back to forums with more meaningful discussions and all information related to some issue in one place.
That's interesting, you seem to favor a "per tool" usage (all forums on one tool), while I would favor a "per project" usage. I prefer the "per project" as I don't want to have my identity always connected between projects. On the other hand, Github is almost forcing everyone to have a "per tool" usage. Since everything is already on Github, your only way to escape them is to enforce a "per tool" separation.
I actually don't favor "per tool" or "per project". Just making an observation on why project owners are gravitating towards discord for their community. It is easier to setup than to setup a forum or something.
The issue with per project forum (instead of a discord type one login for all forums), is that it creates friction. For instance I have visited obsidian.md's forums a lot for things I need. Sometimes I have noticed someone asking something for which I know the answer, but then the pain of having to create an account and then answer that person, makes me just skip it.
The reason for discord communities to be active, is purely because there is no friction.
Still need to remember URLs and then visit them on regular basis. It it not a push system, more a pull system. Communities die that way, at least at initial stages.
> privacy policy of a closed-source product that will be collecting and selling your data
I know this the general HN opinion of free & closed software, but I think this is one scenario where it doesn't ring true. On top of Discord's privacy policy explicitly disallowing/negating data selling, the CEO, CTO and CFO have all made public statements that they have no interest in the monetization of user data. The CTO has suggested that they're at least breaking even with Nitro subscriptions, and they've shrugged off acquisition attempts, instead wishing to go public at some point.
The risk of Discord performing commercial suicide via a data selling scandal appears exceptionally unlikely.
They've taken nearly $1b in funding [0] (likely over that goal now, given the last few rounds were undisclosed). Regardless of their current costs now, I doubt they'll ever make up $1b on a subscription product. And investors are likely looking for a 10x return, which means an exit of $10b. I doubt they can get there by either selling data, or by selling subscriptions. Turning down the $12b acquisition offer from Microsoft felt like a foolish move; I expect they'll get bought out by someone like Amazon for half that price.
I had the same reaction when i read it - why is this necessary? Discord is a terrible experience...even here, right off the bat, the link is expired!
I'd even argue that simply using Discord just makes projects seem juvenile, and not to be taken seriously. Its the modern equivalent of "ask your Pro question on our AOL Instant Messenger" page...yeah, no.
I like the utility of small scraps of paper with custom printed designs on them. Tangible todo lists etc have a sense of existence that is quite different to pixels on a screen.
That said, don't these printers use the same rolls of thermal paper that leaks Bisphenol A (BPA) like all the other cash register/credit card receipt printers?
I'm assuming the levels are low but if youre using them daily maybe the levels aren't low anymore. On the other hand, people who issue CC receipts are also routinely handling these as well.
"In a study published by Springer-Verlag, anyone who constantly handles BPA receipt paper is prone to be exposed to an estimated 71 micrograms of BPA which is 42 times less than the tolerable daily intake." (https://pandapaperroll.com/the-pros-and-cons-of-bpa-free-the...) Sadly no source to that paper.
I wish Little Printer had taken off and we could get such devices with built-in wifi. It's convenient to have some things printed automatically via API, e.g. TODO lists and QR codes for upcoming travel. There's still some things where paper makes more sense.
A thermal printer,... Ideally, a Raspberry Pi with Bluetooth,...Node.js 10, Yarn, Python 3.7, Docker...An iPhone for the app, "
Nearly forty years ago I implemented a till-roll printer for logging alerts on a UPS. It ran on a 6502 1 MHz SBC with 2 k bytes of RAm and 8 k bytes of EPROM. The 6502 also did all the measurements.
Before anyone complains that my old system did rather less; yes I do understand that, but it still seems to me that both the hardware and software requirements for the Tinyprinter seem rather excessive.
Wouldn't it be possible to just use an ESP32 to do pretty much all of it?
Possible, but it would take longer and require more skills. I guess the question is what one's aim is in acquiring a thermal printer. Is it to print thermally? Is it to have a Raspberry Pi project? Or is it to learn a load of electronics, a load of low level code, and the minutiae of how to control a thermal printer at the byte level?
ha, as the author of most of the code on the client side (and thus those beefy requirements), yes it absolutely is possible to use an ESP32! and that’s more along the lines of the original product that inspired this!
however, this was built to be more of a glue library - take whatever device you have (Pi, laptop, IoT fridge), and plug in whatever “printer” you have (typically a receipt printer, but friends have used a rooted Kindle, or Instagram API, etc). there’s more to it than that though, as the original protocol is generic (it was built to control all kinds of devices - printers with blob payloads are a nice side effect), so it’s also relatively easy to extend if you want to bolt on other uses cases within the same network.
it was never really optimised to be the leanest implementation, but instead easy enough for others to hack on and contribute to. JavaScript/TypeScript on a Pi feels like a reasonable compromise, with heavy emphasis on compromise! so, by focusing more on community and simplicity, it’s fun seeing the weird ways people extend it :)
Each to their own. I think I'd tackle this sort of solution (based on the skills I have) with a Raspberry Pi and a POS receipt printer:
Connect Pi to Printer. Have web-server (in lang-of-choice) listening for as POST payloads. spool the payload to a file. print file. Depending on use case, parse the payload to insert required ESC/POS codes (which seem to be a subset of the old Epson Printer ESC codes) for formatting.
Tbh... there's probably a way to have Linux present it as a network printer so that other OSs can print directly to the printer.
An ESP32 with a 160 MHz CPU and 4 MB of storage? That's totally overkill for a project like this ;-)
Jokes aside, probably any microcontroller with Bluetooth or Wi-Fi interfaces could do that kind of job without any problem. Is just processing every data frame you receive an sending them to a serial data interface, that isn't that difficult with the hardware and tools we have nowadays.
Unrelated comment, but I’m tired of the seemingly obligatory “join our discord” crap… when people used to hang on IRC it wasn’t being pushed as hard as this. What’s the point?
I wouldn't have thought to use Zigbee for printers. Now I'm wondering why Zigbee is uncommon. Instead of Bluetooth. Like for my bloodpressure cuff, scale (body weight), thermometer, etc.
Anything anywhere for label printers? I foraged online a bit; no joy.
Lots of reasons, but probably the main reason that Bluetooth (BLE anyway) has exploded is that Apple supported it and that drove adoption. Once you have a major platform adopting it and devs follow and build out SDK's, examples of usage, etc, then it becomes easier and easier for everyone else to jump on the bandwagon.
I had a VIC20, and I wanted a way to print out my programs and some output. I bought a cheap 25x80 column thermal printer and interfaced it to my VIC20. I wrote a program that showed the position of Jupiter's four large moons - Io, Europa, Callisto, and Ganymede. The calculation was based on a formula in Peter Duffet-Smith's book, "Practical Astronomy with your Calculator". They came out with a similar book after with code in BASIC, which I bought too. They were both little sprial-bound books that I have since lost thirty-plus years later on. The thermal printer was fun to print and bring snips of code to computer club meetings later in high school. I even had it print an PETSCII or revised Commodore ASCII representation of the positions of the moons relative to Jupiter. The only issue is that the thermal print would fade, so I would reprint the faded ones when I could.
I love the way this looks. Is there some barebones thermal printer that i could put in a custom enclosure to look like this? The ones i am looking at online seem to be much bulkier.
60 comments
[ 4.0 ms ] story [ 137 ms ] threadThis is such a turn-off for projects to me. Please accept random terms of service and privacy policy of a closed-source product that will be collecting and selling your data as the only means of participation in this community.
Not everything needs all the bells and whistles
Personally I avoid IRC since I just don't like "chatting", email or forum is a far easier flow, socially. It also seemed rare to find anyone actually present in their channels, timing never works out. I'll still use it though if its the only option and I realllly want an answer
This project has NNTP, so if you don't like "chatting", then the NNTP can be used, which is more like the email or forum.
I have never used NNTP, but now im curious to learn some more about it; growing up I had never seen it used, but i think i just missed that age
About NNTP, there is the RFC documents relating to it, and some NNTP servers. Some are Usenet, although the NNTP server for my project is specific to that project. You can use any NNTP client software to access it (or just operate the protocol directly); some email software can also use NNTP.
I have a lot of trouble wrapping my head around why so many tech and programming (especially FOSS) communities still flock to Discord, especially given how programmer-hostile the company is. Matrix is more than good enough to manage it now. The only things about Matrix that bother me right now is that presence is still disabled on the matrix.org homeserver, and the phone call metaphor for group voice and video chat rooms is clunky.
I’d say it’s likely that most discord users are mobile or at least half.
This is why any telephony solution I’d build would probably be based on Jitsi Meet.
In case anyone’s bored: A channel based UI (similar to Mumble) for it would be rad.
There are other organization chat solutions like Slack (proprietary), Mattermost and Zulip (I’m trying out the latter with an organization). Those are the real alternatives to Discord, not Matrix (... yet ...). Neither of them (besides maybe Slack?) has an integration for voice channels (not “calls”).
Voice rooms are coming soon fwiw.
Server specific pepe emojis are absolutely essential.
But the reason people aren't making forums, is it is not as easy for community owner to make forums (need to get cloud storage, then configure something like discourse forum, get a domain name etc.)
Same for the end user who wants to be part of multiple communities, has to remember the URLs of multiple communities, has to have multiple logins and passwords and needs to create an account in each of them.
If someone combines discord + discourse (basically an app, that has all your forums on the side bar with just one login, and people can create a new forum for anything, similar to how you can create a discord server) then I think we might end up going back to forums with more meaningful discussions and all information related to some issue in one place.
I don't use Discord for anything serious but it's just very convenient to join any Discord group with a single click since I'm already logged in there.
I think we need to put aside the security practices of Discord (it's not like the discussions tend to be sensitive in any way), because I'm not aware of anything that comes close in ease of setup and user onboarding, at this point. Setting up Zulip or Mattermost, paying for hosting and then managing it all is quite a hassle in comparison.
Use Discord for anything remotely private and we can have a discussion. :)
> If someone combines discord + discourse (basically an app, that has all your forums on the side bar with just one login, and people can create a new forum for anything, similar to how you can create a discord server) then I think we might end up going back to forums with more meaningful discussions and all information related to some issue in one place.
That's interesting, you seem to favor a "per tool" usage (all forums on one tool), while I would favor a "per project" usage. I prefer the "per project" as I don't want to have my identity always connected between projects. On the other hand, Github is almost forcing everyone to have a "per tool" usage. Since everything is already on Github, your only way to escape them is to enforce a "per tool" separation.
The issue with per project forum (instead of a discord type one login for all forums), is that it creates friction. For instance I have visited obsidian.md's forums a lot for things I need. Sometimes I have noticed someone asking something for which I know the answer, but then the pain of having to create an account and then answer that person, makes me just skip it.
The reason for discord communities to be active, is purely because there is no friction.
I know this the general HN opinion of free & closed software, but I think this is one scenario where it doesn't ring true. On top of Discord's privacy policy explicitly disallowing/negating data selling, the CEO, CTO and CFO have all made public statements that they have no interest in the monetization of user data. The CTO has suggested that they're at least breaking even with Nitro subscriptions, and they've shrugged off acquisition attempts, instead wishing to go public at some point.
The risk of Discord performing commercial suicide via a data selling scandal appears exceptionally unlikely.
Their goal is still to make money and as with any freemium product, the rules might change.
[0] https://www.crunchbase.com/organization/discord
I'd even argue that simply using Discord just makes projects seem juvenile, and not to be taken seriously. Its the modern equivalent of "ask your Pro question on our AOL Instant Messenger" page...yeah, no.
yah, no, thanks.
That said, don't these printers use the same rolls of thermal paper that leaks Bisphenol A (BPA) like all the other cash register/credit card receipt printers?
I'm assuming the levels are low but if youre using them daily maybe the levels aren't low anymore. On the other hand, people who issue CC receipts are also routinely handling these as well.
I wonder if there are recent studies around this.
https://www.possupply.com/paper-rolls/thermal-receipt-roll-p... was my first hit when looking for BPA/BPS-free paper.
"What you’ll need
Nearly forty years ago I implemented a till-roll printer for logging alerts on a UPS. It ran on a 6502 1 MHz SBC with 2 k bytes of RAm and 8 k bytes of EPROM. The 6502 also did all the measurements.Before anyone complains that my old system did rather less; yes I do understand that, but it still seems to me that both the hardware and software requirements for the Tinyprinter seem rather excessive.
Wouldn't it be possible to just use an ESP32 to do pretty much all of it?
however, this was built to be more of a glue library - take whatever device you have (Pi, laptop, IoT fridge), and plug in whatever “printer” you have (typically a receipt printer, but friends have used a rooted Kindle, or Instagram API, etc). there’s more to it than that though, as the original protocol is generic (it was built to control all kinds of devices - printers with blob payloads are a nice side effect), so it’s also relatively easy to extend if you want to bolt on other uses cases within the same network.
it was never really optimised to be the leanest implementation, but instead easy enough for others to hack on and contribute to. JavaScript/TypeScript on a Pi feels like a reasonable compromise, with heavy emphasis on compromise! so, by focusing more on community and simplicity, it’s fun seeing the weird ways people extend it :)
Connect Pi to Printer. Have web-server (in lang-of-choice) listening for as POST payloads. spool the payload to a file. print file. Depending on use case, parse the payload to insert required ESC/POS codes (which seem to be a subset of the old Epson Printer ESC codes) for formatting.
Tbh... there's probably a way to have Linux present it as a network printer so that other OSs can print directly to the printer.
Jokes aside, probably any microcontroller with Bluetooth or Wi-Fi interfaces could do that kind of job without any problem. Is just processing every data frame you receive an sending them to a serial data interface, that isn't that difficult with the hardware and tools we have nowadays.
I wouldn't have thought to use Zigbee for printers. Now I'm wondering why Zigbee is uncommon. Instead of Bluetooth. Like for my bloodpressure cuff, scale (body weight), thermometer, etc.
Anything anywhere for label printers? I foraged online a bit; no joy.
I did find USB connected devices from Brother.