Closed soapdog closed 5 years ago
Changes look reasonable to me! Thanks for leading out on this @soapdog
@cryptix are you happy for this to be merged?
are you happy for this to be merged?
yup, I think so but would like to take a 2nd look into windows testing. At least simple start
and stop
usage.
@soapdog where are you putting the code? I had an enlightning idea for hostapp as a ssb-plugin yesterday night.
@cryptix since we're decoupling them, I might switch to a non-js solution for the host application. Something that builds statically. Don't know, I am on ideation phase at the moment. So far, I've just deleted the code, I haven't put it anywhere.
I am beginning to wonder if the host app needs scuttle-shell at all, it could be a "Bring your own sbot" case, in which it tries to connect to the running sbot, may it be from scuttle-shell or whoever client is running at the moment. If nothing is running, then it can try to start scuttle-shell, but it could be made more flexible than that.
I saw that you (or someone else) created a https://github.com/ssbc/ssb-webexthost repo. That might be a good thing forward. I just don't know if we really need NodeJS there, something that builds static native binaries might be better, after all, web host app can be an invisible console app.
sounds cool.
I was talking with @pietgeursen today who was saying some of the rust work might ultimately have something to offer into this space as well, so I think learning from prototyping is a good idea, as in we don't need to make things totally perfect and modular in our first pass... we just need to get some things working :)
As discussed in #15 and also:
I've removed the host app from scuttle shell. I'll find another way to install it.
I've tested it here and it appears to be working fine. I've also updated the test files but I think that due to removing the host app, the appveyor thing is now testless.