Patchwork: TypeError: id.slice is not a function

[ahdinosaur]

just started happening:

TypeError: id.slice is not a function
    at About (/home/dinosaur/repos/ssbc/patchwork/node_modules/patchcore/about/obs.js:127:65)
    at get (/home/dinosaur/repos/ssbc/patchwork/node_modules/patchcore/about/obs.js:65:19)
    at Object.name (/home/dinosaur/repos/ssbc/patchwork/node_modules/patchcore/about/obs.js:33:21)
    at Object.name (/home/dinosaur/repos/ssbc/patchwork/node_modules/depject/apply.js:15:30)
    at Object.person (/home/dinosaur/repos/ssbc/patchwork/modules/profile/html/person.js:19:26)
    at Object.person (/home/dinosaur/repos/ssbc/patchwork/node_modules/depject/apply.js:15:30)
    at Object.manyPeople (/home/dinosaur/repos/ssbc/patchwork/modules/profile/html/manyPeople.js:39:33)
    at Object.manyPeople (/home/dinosaur/repos/ssbc/patchwork/node_modules/depject/apply.js:15:30)
    at renderItem (/home/dinosaur/repos/ssbc/patchwork/modules/feed/html/rollup.js:216:78)
    at add (/home/dinosaur/repos/ssbc/patchwork/node_modules/pull-scroll/index.js:42:15)

[ahdinosaur]

fixed by 5521a9a9d761edfed6fe7d0cd2651a77a9670692 and 46d6da03cf4ba28bbedeafd78648e0111d0bc3f7, thanks @mmckegg!

[mmckegg]

yep, tis miserable 😢

Will push out a release soon!

[mmckegg]

@ahdinosaur we should track down what message is responsible and who's "fault" it was! Just for interest's sake 😼

Edit: it was %n+SBv/eEvoyPLcaVhPp228JcWEmzTNr0KTZcgrh4Q7U=.sha256

[clehner]

@ahdinosaur we should track down what message is responsible and who's "fault" it was! Just for interest's sake *

maybe this:

{
"key": "%n+SBv/eEvoyPLcaVhPp228JcWEmzTNr0KTZcgrh4Q7U=.sha256",
"value": {
"previous": "%uADMhLQkz1L7a8JKqf5xwRfEWGASbRip8vxEwPjaVAI=.sha256",
"author": "@U5GvOKP/YUza9k53DSXxT0mk3PIrnyAmessvNfZl5E0=.ed25519",
"sequence": "182",
"timestamp": 1489718720580,
"hash": "sha256",
"content": {
"type": "contact",
"contact": {
"isTrusted": true
},
"following": true
},
"signature": "0K7RqgLdGVHsXAjs2ydsFoAU1dPj5SETg65ZcIpcw9bKWe6O0/ISwZCu7tZqGuVLDbgPkkI+5FWo1fy4HfNUAQ==.sig.ed25519"
}
}

app should handle every case, of course

[mmckegg]

Are any other ssb clients freaking out about this? Or is it just patchwork-next?

[clehner]

patchfoo rendered "... follows [object Object]". don't know about patchwork-classic. could this be EOL??

[ahdinosaur]

originated from picknmix, happened when @pietgeursen, using the latest picknmix, followed @joshuavial's new identity. :sweat_smile:. @mixmix in picknmix accidentally published a DOM click event as a message, which is where isTrusted comes from.

brings up all the more reason for stronger type checking before publishing and filtering bad data from queries.

[mmckegg]

@ahdinosaur or maybe some shared vocab in patchcore. contact.async.follow etc

@clehner I think patchwork-classic actually handled filtering bad data quite well. Gonna boot it up and see.

[mmckegg]

Yep, classic is fine!

[ahdinosaur]

yeah, i think classic used ssb-msg-schemas. i'm keen to reboot something like this, so we can add type checking as we wish.

[dominictarr]

it's better to address this defensively, the renderer should always be able to validate and not display invalid things. (or something sensible) we fundamentially can't stop anyone posting invalid messages, apps just have to take it on the chin.

[mixmix]

Agree, i want a strainer over the stuff coming through my patchcore pulls
and obs

On Sat, 18 Mar 2017 01:53 Dominic Tarr [e-mail] wrote:

it's better to address this defensively, the renderer should always be
able to validate and not display invalid things. (or something sensible) we
fundamentially can't stop anyone posting invalid messages, apps just have
to take it on the chin.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/ssbc/patchwork/issues/439#issuecomment-287346145, or mute
the thread
https://github.com/notifications/unsubscribe-auth/ACitnlOGH0CxJvBRZx3cFytfYmaZXLYmks5rmoJKgaJpZM4MgHc-
.