Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • 20:28
    dpc commented #294
  • 20:24
    zicklag commented #294
  • 20:21
    dpc commented #294
  • 20:15
    zicklag commented #294
  • 20:15
    zicklag commented #294
  • 10:12
    kornelski closed #296
  • 10:12
    kornelski commented #296
  • 10:12
    kornelski commented #296
  • Apr 02 23:52
    dpc commented #296
  • Apr 02 23:50

    dpc on master

    Bump min rust version (compare)

  • Apr 02 23:23

    dpc on master

    Upgrade deps (compare)

  • Apr 02 23:20
    dpc commented #296
  • Apr 02 23:12
    dpc commented #295
  • Apr 02 23:08
    dpc commented #238
  • Apr 02 22:31
    zicklag synchronize #294
  • Apr 02 18:27
    kornelski edited #296
  • Apr 02 18:27
    kornelski edited #296
  • Apr 02 18:26
    kornelski opened #296
  • Apr 02 12:28
    chrysn opened #295
  • Apr 02 12:12
    chrysn commented #238
Vitaly Shukela
@vi
There is libgcrypt-config --libs although, which gives -lgcrypt.
/usr/lib/x86_64-linux-gnu/pkgconfig/libgcrypt.pc is present in the Debian Sid package, but absent in Debian Buster one.
matrixbot
@matrixbot
chrysn hm, installation per cargo install doesn't work for me either, checking what exactly i did a few days ago
Vitaly Shukela
@vi
Trying workaroud: RUSTFLAGS="-C link-arg=-lgcrypt" cargo install cargo-crev
Vitaly Shukela
@vi
The workaround seems to have worked.
matrixbot
@matrixbot
chrysn same here -- the odd thing is that i didn't do that when i installed cargo-crev a few days ago ?-/
matrixbot
@matrixbot
dpc I don't think I've been touching anything there recently.
matrixbot
@matrixbot
chrysn these are most likely not crev's faults, libssh-rs's
Vitaly Shukela
@vi
Shall an issue be filed to libssh-rs?
matrixbot
@matrixbot
chrysn probably yes, given that the issue I filed (better defaults for LIBSSH2_SYS_USE_PKG_CONFIG) obviously doesn't cut it
matrixbot
@matrixbot
laurencega Hey what do you guys think about using Github users as identities? Currently there's just the native CrevID (ed25519 key + url (I think)) and planned support for PGP identities
laurencega Seems like Github identities could be a viable identifier - it would provide easy discovery of default proof repositories, already supports hosting a set of public keys for a user (not just one), can be identified by human-readable usernames and pretty much every developer has an account. Thoughts?
matrixbot
@matrixbot
dpc What are github identities?
dpc Is there anything else than just github url?
matrixbot
@matrixbot
laurencega nah pretty much just the Github username
laurencega it's not a special thing, I just didn't know what else to call it haha. It's just your public Github account
laurencega See any Github user's registered public keys:
https://developer.github.com/v3/users/keys/
dpc It's not much different from http://github.com/username/crev-repos.git etc.
dpc And it's a little bit of special-casing, for the monopolist. :D
matrixbot
@matrixbot
laurencega Well, it's not much different in terms of the repo discovery (do you mean "crev-proofs.git"?), but it provides simpler key/identity management and a lower barrier to entry. It also decouples identity from a single key, increasing portability
laurencega and yes, adds to Github's monopoly 😅
laurencega but that's each individual's choice if we offer as an extra option, right?
laurencega if it's a viable and convenient method of identification, I think it's worth considering
dpc I don't understand how key management situation changes?
dpc We can not have github users not generate keys etc. Other users need to be able to eg. copy someone else's proof and put it into their repo etc.
dpc Or eventually have aggregators and and services that filter proofs etc.
matrixbot
@matrixbot

Ralith > <@laurencega:matrix.org> nah pretty much just the Github username

github usernames are not stable

matrixbot
@matrixbot

laurencega > <@ralith:ralith.com> github usernames are not stable

That's true! Good point. All Github accounts do have a stable ID though, so I suppose that's what would be used. At least you could write like "trust @username" and it could resolve to the ID.

laurencega > <@dpc:matrix.org> I don't understand how key management situation changes?

Yeah, keys still need to be generated. But, a user can re-use their existing Github keys if they have them.
They can also use multiple keys for the same identity. I think with current crev IDs the only way to do this is now is making trust chains to your other keys, right? A Github account collapses that chain into a unified identity

laurencega I don't think this stops someone copy someone else's proof into their own repo? What implications are you thinking it may have for aggregators and such?
laurencega it would just mean a proof is signed by one of the keys associated with a user's Github account. Doesn't have to even be in a git repo on Github, still the same decoupling as there is now
matrixbot
@matrixbot
chrysn how do you generally expect crev to be pronounced? "see-rev"? "krev"?
matrixbot
@matrixbot
dpc I hope people will tell me. I only ever type it. :D
Laurence Armstrong
@LaurenceGA
My assumption was always "krev" 🤷‍♂️
matrixbot
@matrixbot
chrysn By the way, I've finally presented crev at our meetup on Monday. Not much feedback gathered, as we're still getting to grips with the remote modes, but I hope to have attracted a few more users.
chrysn Slides at https://christian.amsuess.com/presentations/2020/crev/ , just in case anyone wants to make more of them. (On their own without the talk track, they're probably only useful to people already familiar with crev, who might adapt them.)
matrixbot
@matrixbot
Danilo (@dbrgn:matrix.coredump.ch) Would someone of you be up to presenting crev at a virtual Rust Zürich meetup? https://www.meetup.com/de-DE/Rust-Zurich/
Danilo (@dbrgn:matrix.coredump.ch) (First web based meeting will take place in a week.)
Danilo (@dbrgn:matrix.coredump.ch) (Afterwards we don't yet have any plans)
matrixbot
@matrixbot
chrysn Danilo: I'm by no means a crev expert (just happen to be the first one using it in the Vienna meetup), but hey I have a 15-20' talk at hand, so if you'd like to have a presentation and dpc can't make it, I could come by.
chrysn (Can do German or English, depending on what's common in your meetups)
Danilo (@dbrgn:matrix.coredump.ch) chrysn: that sounds pretty cool :) I'll suggest it to the organizers (I'm only partially involved with the meetup). english would be preferred, zürich has a lot of expats.
chrysn By the way, as you'll have your first Jitsi meetup just as we did last in Vienna: Worked OK for us with about ten participants; it really pays to check the "participants start muted and without video" option.
Danilo (@dbrgn:matrix.coredump.ch) Yep, that's the plan for us as well :)
matrixbot
@matrixbot

dpc > <@chrysn:chat.weho.st> By the way, I've finally presented crev at our meetup on Monday. Not much feedback gathered, as we're still getting to grips with the remote modes, but I hope to have attracted a few more users.

Awesome!

dpc > <@chrysn:chat.weho.st> Slides at https://christian.amsuess.com/presentations/2020/crev/ , just in case anyone wants to make more of them. (On their own without the talk track, they're probably only useful to people already familiar with crev, who might adapt them.)

How do you get such an awesome presentation format?

matrixbot
@matrixbot

dpc > <@chrysn:chat.weho.st> Danilo: I'm by no means a crev expert (just happen to be the first one using it in the Vienna meetup), but hey I have a 15-20' talk at hand, so if you'd like to have a presentation and dpc can't make it, I could come by.

I think you should totally do it.

Myself, I'm stuck in an apartment in San Jose in shelter-in-place with my whole family (kids running wild) in San Jose and trying to do a good dayjob work remotely.

matrixbot
@matrixbot
dpc Can I share this presentation around? :)
matrixbot
@matrixbot
chrysn the presentation format is hovercraft (it converts a restructuredtext source to impress.js), which i've tried out for the first time. i'm a sucker for plaintext versionable source formats, and so far did my presentations in latex, but css does make "but i want it that way" easier to realize than latex. (on the downside, it makes creating pdf slides ... dunno, impossible? hard?, -- and that's occasionally a requirement, especially when doing lightning talks where slides are spooled for easy switchover.)
chrysn sure, share it around -- by link, or with any modifications (it's cc-by-sa licensed as per last page, and the source rst is linked there as well)