search

SparkNFT / SparkLink

The SparkNFT fontend repository.
3 stars 7 forks source link

Some customers encountered the problem of not being able to mobilize their wallets to publish during the publishing process. #192

Open Erreth-Akbe opened 2 years ago

Erreth-Akbe commented 2 years ago

What's going wrong?

A user encountered a problem that the publishing operation could not continue. image Console and Network erro record: 0d42dbf20126f13247bb4ff46f06e97 image The user switched two PCs, one mobile phone, and tried wifi and mobile networks in the network environment, but they were all stuck at this step uniformly. After the release and decryption is completed, the wallet is not activated, or the decryption is not completed. The additional information is: When the user tries to encrypt a file of about 200mb, it will cause the webpage to crash, that is, the AES encryption process should be started normally.

reference may be useful: https://stackoverflow.com/questions/46468104/how-to-use-subtlecrypto-in-chrome-window-crypto-subtle-is-undefined

How to reproduce?

We can not reproduce this bug in our machine. Or we just find the cause of this bug and adjust the user's environment to suit our product.

Expected behavior

The publish process should work smoothly.

Context

refactor branch.

SantanaSP commented 2 years ago

Do you believe we got an alien who had a good sight just now? He can even figure out what the content in your blurred "Network" screenshot is!

Erreth-Akbe commented 2 years ago

Do you believe we got an alien who had a good sight just now? He can even figure out what the content in your blurred "Network" screenshot is!

The content in the network is not very important, because the content in it does not change before and after the error is reported. I wrote in the statement above.

Erreth-Akbe commented 2 years ago

Do you believe we got an alien who had a good sight just now? He can even figure out what the content in your blurred "Network" screenshot is!

I'll replace the screenshots of this, probably in 1-2 hours.

SantanaSP commented 2 years ago

I think it's almost like a network error or a hard problem to solve since we can't reproduce its behavior in our own environment. Do you record the network activities from before the site is loading? And have you noticed anything strange been returned by the server?

Erreth-Akbe commented 2 years ago

I think it's almost like a network error or a hard problem to solve since we can't reproduce its behavior in our own environment. Do you record the network activities from before the site is loading? And have you noticed anything strange been returned by the server?

I don't quite agree that this is a network bug. If you are more sure, I will go to him to take a screenshot to see if there is any problem in the loading. You can follow the console error report to see what kind of environment this problem usually occurs in. This is likely to be a problem caused by a supply chain that is not supported in a certain environment.

Erreth-Akbe commented 2 years ago

4c3db1c2a26f3ada73df3bbc5ed6fc3

Erreth-Akbe commented 2 years ago

539fe978947635e7d0ea150e5940dde

Erreth-Akbe commented 2 years ago

We want to find a way to adjust this costomer's environment so that he can use our Dapp

SantanaSP commented 2 years ago

I think it's almost like a network error or a hard problem to solve since we can't reproduce its behavior in our own environment. Do you record the network activities from before the site is loading? And have you noticed anything strange been returned by the server?

I don't quite agree that this is a network bug. If you are more sure, I will go to him to take a screenshot to see if there is any problem in the loading. You can follow the console error report to see what kind of environment this problem usually occurs in. This is likely to be a problem caused by a supply chain that is not supported in a certain environment.

But since we cannot reproduce it in our environment and follows the principle of "unacceptable as uncorroborated evidence", I think it's more rational to treat it as an occasional problem first. Just think that, if it's the fault of one of our dependencies, why do the other instances just function well?

SantanaSP commented 2 years ago

I noticed that there're no XHR requests in your Network screenshot, which means the activity between the client and the server has not been recorded. Under such an information insufficient circumstance I'm afraid I can't guess what's the true criminal, the only thing we can do is to get a new Network report or blame the supply chain endlessly.

Erreth-Akbe commented 2 years ago

I think this kind of supply chain library error should be a relatively common problem, I mean crypto.subtle this supply chain library. So according to the results of my current research, it is very likely that this user does not have https support, so what I want to know is how can I determine that his current environment is a non-HTTPs environment? Why are all three of his devices turned off https? How do I call him back to https? How do we enable https by default?

I noticed that there're no XHR requests in your Network screenshot, which means the activity between the client and the server has not been recorded. Under such an information insufficient circumstance I'm afraid I can't guess what's the true criminal, the only thing we can do is to get a new Network report or blame the supply chain endlessly.

SantanaSP commented 2 years ago

The Rollup will automatically generate the vendor bundle using polyfills or other replacements to replace the built-in module of nodejs, including https. So I think it's most likely not an issue of https.

SantanaSP commented 2 years ago

I believe that the bottom layer network functionality package is using browser field in their packages to help the bundler to accomplish such work. Here is the specification for it: https://github.com/defunctzombie/package-browser-field-spec

SantanaSP commented 2 years ago

I insist on acquiring complete information about the context before we can dig deeper into the supply chain. If you noticed that even a request to the server hasn't happened, it's most likely what's going wrong with his chrome configurations, since the chrome profile sync mechanism could spread anything wrong to all the devices of a user.

SantanaSP commented 2 years ago

Oh I see he's using Firefox, which version of Metamask is he using? Does he syncs any his data across devices?

Have you tried to use WalletConnect to reproduce it?

Erreth-Akbe commented 2 years ago

~Oh I see he's using Firefox, which version of Metamask is he using? Does he syncs any his data across devices?~

Have you tried to use WalletConnect to reproduce it?

Yes, there was same error at mobile client . It is so strange.

SantanaSP commented 2 years ago

~Oh I see he's using Firefox, which version of Metamask is he using? Does he syncs any his data across devices?~ Have you tried to use WalletConnect to reproduce it?

Yes, there was same error at mobile client . It is so strange.

Do you mean you can reproduce it with WalletConnect? If so, this issue can be taken into agenda.

Erreth-Akbe commented 2 years ago

~Oh I see he's using Firefox, which version of Metamask is he using? Does he syncs any his data across devices?~ Have you tried to use WalletConnect to reproduce it?

Yes, there was same error at mobile client . It is so strange.

Do you mean you can reproduce it with WalletConnect? If so, this issue can be taken into agenda.

no, he can reproduce. It is weird.

SantanaSP commented 2 years ago

Yeah since I'm not familiar with WalletConnect actually there are no other things I can do to solve it. Of cause under the promise that you record the network activities correctly.