search

DefendPDXCode / DefendPDX-TwitterBot

This is the open-source bot that runs the Twitter Bot on @DefendPDX
MIT License
6 stars 2 forks source link

Further desired specs? #5

Open jmlingeman opened 4 years ago

jmlingeman commented 4 years ago

How do you want to interact with the bot to be able to change modes, add users, remove users, etc?

I have a python IRC->Twitter bot that could be fairly easily repurposed for this. A simple web interface could also work if you had a trusted web server to run it from.

Let me know what the further required specs are and I’d be happy to contribute.

jmlingeman commented 4 years ago

My group uses IRC bots to interact with the wider internet in a secure way by SSHing into a VPS that is running an IRC server. These cost only a few bucks a month for ones with a small amount of RAM and disk space. Digital Ocean is the one we use.

If the group would be comfortable using IRC this would be quick and easy to set up. Happy to fork my code over to this project and modify it to spec.

I just don’t know whether the mods who want to run this account would be comfortable with IRC. It could also probably be controlled via something like Discord, but I have not used that API.

grinnellian commented 4 years ago

+1, would love to understand user stories further before starting to code & am happy to help in discovery/creation of user stories. I can get on some sort of call for that too if you like.

Some starter questions:

motcreux commented 4 years ago

Hey all! Getting into this project and the topic of the thread was my first question for sure.

It's been two days and we haven't heard back from DefendPDX folks, should we just try to mock something up and see if it works for them?

SSH, although it is pretty useful and stuff, isn't the easiest for everyone who would be part of the mod team to use. Instead I think the idea of a web service would be pretty straightforward assuming someone has $5 per month kicking around for a server.

One thing I think would be worth considering if we decide to proceed with mocking something up, would be that the security measures folks on the ground need to take are a lot more serious than the measures we would need to take. The worst thing that could happen here is that the service gets shut down, right? There's (imo) no sense in choosing SSH over web with concerns of nation-state monitoring, because it's a public service regardless; our real concern is far-right trolls mucking about with the service, not the NSA. While writing secure code will be important, I think we need to keep our priorities in check.

I think the idea of writing a collating twitter bot is an excellent idea, and I think it can be useful moving forward in other protests, in addition to the intended users of DefendPDX.

If we're all good with starting something and seeing if DefendPDX likes it, has the decision been finalized to go with Javascript over the original idea of Python? Tweepy vs JS' twitter library should probably not be the reason to choose one over the other. Otoh I noticed that @grinnellian seems to use a lot of Javascript, at least on Github. I'm fairly comfortable with either option, I've finished more projects in Python, but am pretty confident with JS.

If we're good with Javascript, should we just standardize on ES6?

motcreux commented 4 years ago

Again this is under the theme of "if we're good with moving forward without hearing back first".

My assumption, based on some experience, is that we should assume people will use both mobile and desktop for moderation, and try our hardest not to write something people will have to install on their phone/desktop for all sorts of reasons (although PWA would probably be fine).

Additionally, it would be preferable that users will spend as little time as possible on moderation, because this isn't the kind of thing that should become the core focus on anyone's day-to-day work (except possibly ours :smile: ).

Typically the people running this sort of system would already know each other, so the separation of accounts would probably be less important, maybe we would want to do peer-adding first and then have the addition of administrative users as a high-priority next item on our agenda?

Given the nature of how things are going in Portland, we'd probably want to get a very minimal product that we can deliver first, and then add things based on questions the users raise, right?

grinnellian commented 4 years ago

@motcreux thanks for jumping in! I've mostly done Node the last few years but at this point my JS/ES, Go and Python are about equally rusty.

I'm v. limited bandwidth but can do a meeting or pairing session this coming weekend.

Agree with jumping in and doing something fast and incremental (as I said on ESB issue).

Seems like @defendpdx has jumped in with some node stuff, but that needs to be npm init'd and whatnot. I'd love to make things collaborative/teaching. @defendpdx are you open to pairing sessions? Is there a better way to get in touch than via the issues?

Hosting on a DigitalOcean droplet or similar seems super reasonable (and is a trail blazed by ESB too). DefendPDX would have to weigh in on whether they'd want that donated (I have a DigitalOcean acct I already use for VPN and can eat a few bucks a month) or if they'd want to control hosting themselves.

On Wed, Jul 29, 2020 at 4:12 AM motcreux notifications@github.com wrote:

Again this is under the theme of "if we're good with moving forward without hearing back first".

My assumption, based on some experience, is that we should assume people will use both mobile and desktop for moderation, and try our hardest not to write something people will have to install on their phone/desktop for all sorts of reasons (although PWA would probably be fine).

Additionally, it would be preferable that users will spend as little time as possible on moderation, because this isn't the kind of thing that should become the core focus on anyone's day-to-day work (except possibly ours 😄 ).

Typically the groups behind these systems would be managed by people who already know each other, so the separation of accounts would probably be less important, maybe we would want to do peer-adding first and then have the addition of administrative users as a high-priority next item on our agenda?

Given the nature of how things are going in Portland, we'd probably want to get a very minimal product that we can deliver first, and then add things based on questions the users raise, right?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/DefendPDXCode/DefendPDX-TwitterBot/issues/5#issuecomment-665509611, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACHJCEBSSSTZZ745X4BVGZ3R57KYTANCNFSM4PIMOKEA .