After having a good clearout and backup of data at third-party services, all thanks to GDPR pressure, along with recently realising the good side to clearing out life’s baggage, I’ve just set up a Twitter bot, called Harold.
Harold is the creation of Adam Drake and can be found here, where you need to do the following to set it up:
# Have go installed and paths set up, nav to $GOPATH/src
git clone https://github.com/adamdrake/harold
go get github.com/ChimeraCoder/anaconda
go get github.com/Sirupsen/logrus
go get github.com/mmcdole/gofeed
go get github.com/spf13/viper
vim config.toml
# Then copy and amend contents from readme - including feeds
# Visit https://apps.twitter.com/app/new and follow instructions (needs phone no set up on account)
vim keys.env
# Copy key info, create and copy token info as per readme
source keys.env
go build main.go
# Run in tmux on your server/PC for Harold to stay awake
./main -botmode -config /path/here/config.toml
# Or delete timeline first, if over 200 tweets, to save having to re-run bot
# (potentially live tweeting from feeds each time, flooding followers)
./main -deletetimeline
# Leave to run in tmux for auto cleaning/tweeting goodness!
Why would you want to do this? I can’t add much more to what Vicky Lai has to say on the subject over at her blog, but in summary it’s that the me who tweeted 10 years ago is not the same me as now. The profile Twitter creates is a) out of date and b) misleading. You can see what it thinks you like in your account settings. It’s a little unnerving. And that’s just what it knows that it wants to show us.
Plus there’s the automation aspect, which is just a little exciting when you think about it. Effortless social sharing and conversation-starting. It’s much less real than actual socialising, but there’s something of value there, IMO.