Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Jan 31 22:59
  • Jan 31 21:20

    prusnak on new_mem_layout

    make bootloader smaller by cha… (compare)

  • Jan 31 20:52

    prusnak on new_mem_layout

    introduce new memory layout fi… (compare)

  • Jan 31 20:23

    prusnak on master

    format: black uppercase of hex … (compare)

  • Jan 31 20:20

    prusnak on master

    firmware: process tx.branch_id … (compare)

  • Jan 31 20:08

    prusnak on master

    tests: pass branch_id in zcash … (compare)

  • Jan 31 19:53

    prusnak on master

    protob: add branch_id to SignTx… (compare)

  • Jan 31 19:47

    prusnak on master

    defs: fix permissions on png fi… (compare)

  • Jan 31 18:20

    prusnak on master

    signing: use the new hasher_Ini… (compare)

  • Jan 31 18:18

    prusnak on master

    hasher: add param_size to hashe… (compare)

  • Jan 31 18:06

    prusnak on master

    hasher: don't hardcore zcash st… (compare)

  • Jan 31 15:53
    matejcik commented #357
  • Jan 31 14:44

    matejcik on master

    debuglink: add null debuglink, … flake8: add typing imports wher… (compare)

  • Jan 31 05:31
  • Jan 31 02:24
    webees closed #72
  • Jan 31 02:23
    webees edited #72
  • Jan 31 02:23
    webees edited #72
  • Jan 30 19:46
matejcik
@matejcik
why don't you look at bridge status page at http://127.0.0.1:21325/status/
re Connect, yes, it's working normally for TT. what you're seeing is what happens when a message gets lost, and the PC reads a response to the previous query. look for errors in the emulator output that would indicate that a response to something was not returned
@Rovak that is also rather unhelpful. please look at where the tests are getting stuck. also if you're working on TRON, have a look at up-to-date test suite and modify your tests to match, in the PR they are in the old style which might not work anymore
Roy van Kaathoven
@Rovak
@matejcik i'm fully aware that screenshot is far from helpful, but i don't know any way to further debug this issue or provide you more info. I'll take a look at the up-to-date test suite. Thanks for the pointers.
matejcik
@matejcik
@Rovak you can run the test suite with -sv options to see the messages being exchanged. that should get you closer
Roy van Kaathoven
@Rovak
@matejcik are you referring to the testsuite at https://github.com/trezor/connect/tree/develop/src/__tests__ ?
matejcik
@matejcik
@Rovak no, i'm talking about the test suite in trezor-firmware/tests
can't help you with Connect test suite, sorry
i didn't even realize there's a test suite for connect o_O
my bad
Roy van Kaathoven
@Rovak
@matejcik ah ok, is the person who maintains the connect project in this channel?
matejcik
@matejcik
i don't think so
Stacie Waleyko
@staciewaleyko
Nobody is having issues with Trezor + BTC testnet are they? We're seeing some odd behavior when signing with a Trezor and can't pinpoint where it's coming from.
Stacie Waleyko
@staciewaleyko

Alrighty I think we figured it out, when I try to sign a TBTC transaction with the hash 987de9444267dd493a2342cb60608ef647c63a0197263e1f3f4dd43fbe4770a4, I get an error from Trezor about the transaction not being found.

I can see it on other explorers, however it only shows up on tbtc1.trezor.io, not tbtc2.trezor.io:

https://tbtc1.trezor.io/tx/987de9444267dd493a2342cb60608ef647c63a0197263e1f3f4dd43fbe4770a4
https://tbtc2.trezor.io/search/?q=987de9444267dd493a2342cb60608ef647c63a0197263e1f3f4dd43fbe4770a4

I see that both tbtc1.trezor.io, and tbtc2.trezor.io are in a config file for Trezor connect, https://github.com/trezor/connect/blob/179292bfe223078f042a1f2906a8b0a99e0f75bd/src/data/coins.json#L59

If tbtc2.trezor.io is down or behind, is it correct to assume that this can cause signing with Trezor connect to fail?

Nick Neuman
@nneuman
It looks like the second tBTC node for Trezor Connect is currently set on mainnet instead of testnet, according to https://tbtc2.trezor.io/
2019-10-17 at 4.10 PM.png
Elliot Cameron
@3noch
Hey I'm new to Trezor but have lots of experience with Ledger Nano. I'm trying to figure out how trezor supports altcoins. On ledger you have apps you can install for any coin you want...how do you do this on Trezor?
Alexis Hernandez
@AlexITC
here the firmware supports everything, there are no apps
Elliot Cameron
@3noch
Ah interesting.
Langers
@darrenlangley
Hi guys, I am working with the NEM team on integrating NEM2 into the Trezor firmware. Excuse my question as I am quite fresh to the project. We are developing in the core area of the firmware which I believe only supports Model T - at a high-level, what additional work is required to get NEM2 working with the model One?
matejcik
@matejcik
@darrenlangley you would have to implement code in the legacy section, and you'd have to convince us to lift the restriction that new coins are not added to the T1 legacy firmware
Langers
@darrenlangley
@matejcik thank you. We are implementing in core and will let the foundation know about the legacy section.
Marco Peereboom
@marcopeereboom
Hi. Jan Pecinovsky asked me to inform you guys that the decred project has written a bunch of code to interop with trezor. A bunch of code has been written (decred/decrediton#1491) but it hasn't been enabled yet. If you guys want to get that enabled I suggest you hop on decred riot (https://decred.org/matrix/) on the dev channel and help move it forward.
Pavol Rusnak
@prusnak
I am of course aware of that code, but it hasn't been touched for ages. Waiting until there is a Trezor support in a production Decred-team delivered wallet.
You need to do your part of the integration. We have already done ours.
Marco Peereboom
@marcopeereboom
@prusnak if we want to get this moving forward someone has got to jump on the decred matrix dev channel. The code is merged and maintained it is just not enabled by default. If someone shows and brings this up and helps verifying it i am sure mattheusd will enable it and it'll be added to the gui.
@prusnak if there is anything missing mattheusd or someone else will likely pick it up. This is a desirable feature.
Pavol Rusnak
@prusnak
@marcopeereboom maybe you are that "someone"?
katyarichenkova
@katyarichenkova
Hello. I recently got my Trezor T and I noticed that it asking for passphrase every single time I switch the coin in trezor-wallet, which is kind inconvenient since I got a pretty long passphrase. Wouldn't it better, if we have an option like 'Remember me' either within the 'trezor-wallet' or a separate screen on the device itself before/after entering a passphrase the first time?
katyarichenkova
@katyarichenkova
I know that the passphrase not being stored on trezor, but I believe if the device could store it within the context (session) until it disconnects would hurt anyone.
Jonathan Zeppettini
@jzbz
@prusnak Trezor support is working fine for ages in our wallet Decrediton.
Only thing missing is support for staking.
We'd also like to see Decred added to the Trezor web wallet, we already added support to Blockbook.
https://blockbook.decred.org:9161/
Can you advise on how we would move web wallet support forward?
Matheus Degiovani
@matheusd

Ok, I'm here. Summing up from my end.

  • Basic Trezor support has been enabled in decrediton since v1.4.0. issue decred/decrediton#1491 has been used to track full support (including staking) which is not yet supported. That's why it hasn't been closed/updated even though trezor is supported in decrediton.

  • Decred support to Blockbook has been recently added. That is a separate thing from issue 1491, is used for other software and is/was a pre-requisite to add decred support to trezor's web wallet. That is already deployed and running as jz noted above.

  • To add staking support to Decrediton there are some outstanding changes needed in decred software and on trezor's firmware for both models One and T. I demonstrated these and noted them in issue 1491.

While trezor support has been enabled and available since v1.4.0, it hasn't seen much (any?) use. At least on the chats I can recall only a single person commenting on testing it.
We're in the process of finishing up v1.5.0 of the software stack for release so I've been testing and fixing trezor nits, updating depencies, fixing some reported issues, etc
(that is v1.5.0 of Decrediton)
Matheus Degiovani
@matheusd

Most recent issue I'm trying to debug: any trezor-link lib > 1.6.0 fails to init the transport to the bridge due to a "Maximum call stack size exceeded" error.

Even v1.6.1 which has just an added optional parameter fails, while 1.6.0 works without issues.

matejcik
@matejcik
@matheusd i can't find any notes of required changes in Trezor firmware in the linked issue?
@katyarichenkova what you're seeing is a problem with the web wallet, and should be fixed (AFAIK) when Trezor Suite is released
Matheus Degiovani
@matheusd

@matejcik https://github.com/decred/decrediton/issues/1491#issuecomment-413682851

Specifically we need the firmware to understand the special stake opcodes, decode commitment addresses encoded in the relevant outputs according to decred's consensus rules, present the UI for the user to confirm they want to purchase a ticket redeemable at the given address, etc.

In that comment there's a link to the proof of concept change required in the firmware for model one, which is very obviously not safe but shows roughly what needs to happen.

katyarichenkova
@katyarichenkova
@matejcik Is there a reason behind the decision to use different wordlist for shamir recovery?
@matejcik And is there hope to see any time soon an option to convert a classic mnemonic to shamir shares?
matejcik
@matejcik
@katyarichenkova 1. we wanted a shorter wordlist in order to have 10 bits per word (as opposed to 11) and 2. we wanted each word to have a unique prefix in terms of keys pressed on the on-screen keyboard, and also you never press the same key twice. which is easier to do with a shorter word list.
@katyarichenkova converting bip39 to slip39 is problematic. for one you'll have 59-word shares. for two, it breaks passphrase: each bip39 seed+passphrase combination (incl. no passphrase) converts to a different set of shares
so it is a UX nightmare, so that's a no for seeing it as an officially supported feature anytime soon
you can of course convert it at home, and we might add the option to recover 59-word shares, basically if enough people ask for it
Pavol Rusnak
@prusnak
@matheusd Thanks for the update. Please tell this info to your users, so they can try Decred with Trezor. The feature is of no use when nobody uses it. Once it has a proven track record among your users, we can talk about adding it to our webwallet and about staking. But not before
katyarichenkova
@katyarichenkova
@matejcik Did you mean each share would have 59 words? Well, then that would be a nightmare... By the way thanks for explaining.