owner stringclasses 14
values | repo stringclasses 14
values | id int64 116k 4.61B | issue_number int32 1 181k | author stringlengths 1 39 | body stringlengths 1 262k | created_at timestamp[us]date 2000-06-06 02:40:44 2026-06-02 20:16:18 | updated_at timestamp[us]date 2000-06-06 02:40:44 2026-06-02 20:22:56 | reactions unknown | author_association stringclasses 5
values |
|---|---|---|---|---|---|---|---|---|---|
vuejs | core | 811,634,263 | 3,521 | HcySunYang | Please check the RFC about `expose` https://github.com/vuejs/rfcs/pull/210. Also, you can ask questions on the [forum](http://forum.vuejs.org/), the [Discord server](https://vue-land.js.org/) or [StackOverflow](http://stackoverflow.com/tags/vue.js) first. | 2021-04-01T04:46:02 | 2021-04-01T04:46:02 | {} | MEMBER |
vuejs | core | 814,130,800 | 3,547 | edison1105 | @yyx990803
What he was trying to say was probably this：
Volar display an error.
| 2021-04-06T13:43:41 | 2021-04-06T13:50:24 | {} | MEMBER |
vuejs | core | 814,577,532 | 3,548 | HcySunYang | Here is a workaround https://codesandbox.io/s/nostalgic-water-cjnrw?file=/src/ControlButton.vue, see the `ControlButton.vue`
Edit: https://github.com/vuejs/vue-next/issues/1745#issuecomment-669936498
> Also, if you are trying to check for presence of a slot, you should do !!$slots.default in the template (don't c... | 2021-04-07T03:42:00 | 2021-04-07T04:23:11 | {} | MEMBER |
vuejs | core | 810,715,743 | 3,498 | HcySunYang | This should be by design, we need @yyx990803 input. | 2021-03-31T02:44:02 | 2021-03-31T02:44:02 | {} | MEMBER |
vuejs | core | 810,716,779 | 2,897 | HcySunYang | Fixed in https://github.com/vuejs/vue-next/pull/3501 | 2021-03-31T02:46:57 | 2021-03-31T02:46:57 | {} | MEMBER |
vuejs | core | 812,584,718 | 906 | stuta | I finally remembered that this may be a Vite problem. And it is, so all is ok.
There are a lot of things that do not work in Vite but do work in real build. | 2021-04-02T15:43:39 | 2021-04-02T16:11:26 | {} | NONE |
vuejs | core | 812,963,166 | 3,538 | HcySunYang | What is the purpose of this? | 2021-04-04T03:14:07 | 2021-04-04T03:14:07 | {} | MEMBER |
vuejs | core | 813,908,715 | 3,545 | posva | The width and height attributes must be integers without unit, in pixels: https://developer.mozilla.org/en-US/docs/Web/HTML/Element/img#attr-width. It's different from the css attribute. Only `<img alt="Vue logo" src="./assets/logo.png" width="100" />` is valid. | 2021-04-06T07:50:40 | 2021-04-06T07:50:40 | {
"heart": 1
} | MEMBER |
vuejs | core | 814,554,276 | 3,532 | imsunhao | close #3550 | 2021-04-07T02:26:23 | 2021-04-07T02:26:23 | {} | NONE |
vuejs | core | 814,583,318 | 3,548 | HcySunYang | The root cause of the problem is the same as https://github.com/vuejs/vue-next/issues/1745, I am thinking of a solution... | 2021-04-07T04:02:59 | 2021-04-07T04:02:59 | {} | MEMBER |
vuejs | core | 809,983,554 | 3,288 | pikax | > How would you differentiate a prop passed as undefined from a missing prop?
What would be the use case for an optional prop that wasn't passed in to the component?
```html
<my-comp/>
<my-comp :foo="undefined"/>
<script>
props : {
foo: String
}
</script>
```
That code should behave the same.
--- ... | 2021-03-30T07:28:05 | 2021-03-30T07:28:05 | {} | MEMBER |
vuejs | core | 810,024,527 | 3,517 | tweichart | @posva thanks for the hint to the other issue. Should this be added to the documentation and also be mentioned in the Breaking Changes somehow? | 2021-03-30T08:26:31 | 2021-03-30T08:26:39 | {} | NONE |
vuejs | core | 810,330,972 | 3,206 | yyx990803 | Hmm, I didn't realize `findIndex` is so much faster in Chrome and Safari. V8 and JavaScriptCore probably have some dedicated fast paths for it. It's slighlty slower in Firefox but seems to be worthwhile.
There are many places in the codebase where I was intentionally using `for` loops for performance reasons - we ma... | 2021-03-30T14:57:07 | 2021-03-30T14:57:07 | {
"+1": 4
} | MEMBER |
vuejs | core | 811,119,276 | 1,600 | martinbean | @LinusBorg Do you disagree…? This PR was opened last July, and has had multiple comments from people for _months_ saying how this has completely broken their apps.
I literally cannot upgrade to Vue 3 from Vue 2 because of this, and despite people trying to get some feedback on it, it’s gone unanswered. And I would d... | 2021-03-31T14:37:25 | 2021-03-31T14:38:44 | {} | NONE |
vuejs | core | 812,390,416 | 3,530 | vue-bot | Hello, thank you for taking time filling this issue!
However, we kindly ask you to use our [Issue Helper](https://new-issue.vuejs.org/?repo=vuejs/vue-next) when creating new issues, in order to ensure every issue provides the necessary information for us to investigate. This explains why your issue has been automatic... | 2021-04-02T07:35:38 | 2021-04-02T07:35:38 | {} | CONTRIBUTOR |
vuejs | core | 813,787,045 | 3,519 | CarterLi | It was unfair.
At least for `for loop vs Array.some` test case, `Array#some` only [tests element present in the array](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/some#description). In your benchmark:
```js
var array = new Array(1000);
```
which generates an array w... | 2021-04-06T03:02:06 | 2021-04-06T03:17:22 | {} | CONTRIBUTOR |
vuejs | core | 813,792,959 | 3,543 | HcySunYang | Early similar issues https://github.com/vuejs/vue-next/issues/3275, https://github.com/vuejs/vue-next/issues/3270 | 2021-04-06T03:20:58 | 2021-04-06T03:20:58 | {} | MEMBER |
vuejs | core | 814,600,829 | 3,326 | tjk | Bump? Also is there something that could be done so that acorn produces better output for the syntax error since it doesn't map to visible source? | 2021-04-07T04:58:56 | 2021-04-07T04:58:56 | {} | CONTRIBUTOR |
vuejs | core | 811,670,602 | 2,483 | KaelWD | Done. | 2021-04-01T06:11:35 | 2021-04-01T06:11:35 | {} | CONTRIBUTOR |
vuejs | core | 811,849,884 | 1,001 | Awolwe | `onclick="void(0);"` to App.vue helped to me | 2021-04-01T11:42:35 | 2021-04-01T11:42:35 | {
"-1": 3
} | NONE |
vuejs | core | 811,931,600 | 3,418 | oleksiikhr | Yes, I also have such a problem. This `watch` for `:ref` only works on the penultimate push.
https://codesandbox.io/s/vue3-refs-q6g6p?file=/src/App.vue | 2021-04-01T14:04:38 | 2021-04-01T14:04:38 | {} | NONE |
vuejs | core | 812,226,182 | 3,524 | rashfael | Seems a bit harsh to completely stop updating by default and adds quite a bit of uncertainty when trying to recreate an error. I also found it quite horrifying that the app just stopped completely without any indication that this wouldn't happen in production. | 2021-04-01T23:14:47 | 2021-04-01T23:14:47 | {} | NONE |
vuejs | core | 813,817,175 | 3,541 | HcySunYang | Hint.
When calling `unshift/shift/splice`, etc., these methods will manipulate the original array. In the worst case, that will call the `set` trap for the array integer key **n** times, `n` represents the number of array elements, which is a time-consuming operation.
| 2021-04-06T04:45:05 | 2021-04-06T04:45:05 | {} | MEMBER |
vuejs | core | 764,509,468 | 3,079 | posva | if the code is being fixed there must be a test case that was failing before | 2021-01-21T09:39:52 | 2021-03-30T08:52:57 | {} | MEMBER |
vuejs | core | 810,357,702 | 3,399 | pikax | > @pikax looks like there's a test failing after the merge
Seems like typescript can't infer the component type from:
```ts
defineComponent({
components: {
A: {
// No typing because component definition is quite lenient
},
D: defineComponent({
//fully typed component
})
}
... | 2021-03-30T15:31:11 | 2021-03-30T15:31:11 | {} | MEMBER |
vuejs | core | 810,725,538 | 2,353 | HcySunYang | @unbyte Can you fix the conflict? this needs to be rebased now. | 2021-03-31T03:11:52 | 2021-03-31T03:13:16 | {} | MEMBER |
vuejs | core | 812,213,263 | 3,525 | yyx990803 | Same: you should either fix the error (errors within your control), or be handling the error (errors out of your control) with `app.config.errorHandler`. It's best practice to always configure `app.config.errorHandler` if you think your app may throw unknown errors in production. | 2021-04-01T22:35:06 | 2021-04-01T22:35:24 | {} | MEMBER |
vuejs | core | 812,566,318 | 906 | stuta | Hello.
I have my own pretty simple router. If I have mode='in-out' then everything works, but I would need 'out-in' so that both old and new components do not show at the same time.
My router calls the server for dynamic component template and js code. I create js using new Function(). Then I add the component ... | 2021-04-02T14:57:22 | 2021-04-02T14:57:22 | {} | NONE |
vuejs | core | 812,845,520 | 3,418 | bazuka5801 | @LinusBorg in most projects, public components this "specific conditions" used. It should be fixed. | 2021-04-03T10:21:09 | 2021-04-03T10:21:09 | {} | NONE |
vuejs | core | 812,602,337 | 3,526 | Norserium | Just in case, if someone encounter a similar problem. I have found at least a temporary solution.
The reason of this behavior is clearly explained in the links above. The computed properties are binding to the current instance. You can see it in the [source code](https://github.com/vuejs/vue-next/blob/master/packag... | 2021-04-02T16:26:34 | 2021-04-03T11:04:01 | {} | NONE |
vuejs | core | 813,388,587 | 3,539 | yyx990803 | - SFC is a unit of composition. It is what makes HMR, scoped styling and future syntax like `<script setup>` straightforward and even possible.
- Relying on Babel means you need an integration for every tool you use anyway. And Babel is *slow*.
- SFCs have integration for all the tools you've mentioned and they alrea... | 2021-04-05T13:19:02 | 2021-04-05T13:22:23 | {
"heart": 1
} | MEMBER |
vuejs | core | 813,909,274 | 3,545 | posva | Duplicate of https://github.com/vuejs/vue-next/issues/2801 | 2021-04-06T07:51:20 | 2021-04-06T07:51:20 | {} | MEMBER |
vuejs | core | 814,635,132 | 1,074 | bhabgs | > `config.globalProperties` are meant as an escape hatch for replicating the behavior of `Vue.prototype`. In setup functions, simply import what you need or explicitly use `provide/inject` to expose properties to app.
有没有什么方式能过改善这种写法？
``` ts
import { defineComponent, getCurrentInstance } from 'vue';
export de... | 2021-04-07T06:18:23 | 2021-04-07T06:18:23 | {} | NONE |
vuejs | core | 810,183,454 | 3,513 | posva | @HcySunYang Maybe I could manually add the compiler hints to the render functions to vue router. | 2021-03-30T12:31:03 | 2021-03-30T12:31:03 | {} | MEMBER |
vuejs | core | 811,080,802 | 1,600 | martinbean | If this PR had some form of official input and hadn’t been open for nearly a whole year, then poor @CodeDaraW wouldn’t have to rebase their changes 😅 | 2021-03-31T13:45:56 | 2021-03-31T13:45:56 | {} | NONE |
vuejs | core | 813,791,824 | 3,543 | HcySunYang | This seems inevitable, maybe we need to provide an appropriate warning message for this | 2021-04-06T03:17:21 | 2021-04-06T03:17:21 | {} | MEMBER |
vuejs | core | 813,845,480 | 3,519 | edison1105 | thats weird.
https://www.measurethat.net/Benchmarks/Show/12451/0/for-loop-vs-arraysome-vs-for-of
> We use Array methods not because they are fast, but because they are easier to write and result in cleaner code. For library uses, I do think plain for loop are better for performance reasons.
I dont think so.
s... | 2021-04-06T05:57:56 | 2021-04-06T05:57:56 | {} | MEMBER |
vuejs | core | 813,887,179 | 3,519 | syuilo | Basically, simpler code tends to benefit more from optimization by the engines.
So (even if the performance is temporary worse now), this PR should be worth it.
JFYI: https://github.com/vuejs/vue-next/pull/402#issuecomment-609079328 | 2021-04-06T07:16:48 | 2021-04-06T07:16:48 | {} | CONTRIBUTOR |
vuejs | core | 813,902,199 | 1,600 | HcySunYang | @CodeDaraW Thanks~ | 2021-04-06T07:40:37 | 2021-04-06T07:40:37 | {} | MEMBER |
vuejs | core | 578,531,507 | 656 | yyx990803 | This is actually because the use of in-DOM templates. The `innerHTML` is normalized by the browser and converts all attributes names to lowercase before Vue's compiler can handle it.
i.e. `onUpdate:modelValue` is converted to `onUpdate:modelvalue`, so the event names won't match.
When using in-dom templates, even... | 2020-01-26T18:55:33 | 2020-01-26T18:55:33 | {} | MEMBER |
vuejs | core | 578,667,286 | 660 | posva | @jods4
> If you don't fix it, you should document this more clearly, with a red box in the docs and create a warning in DEV mode if you detect this situation (property exists but is not a ref).
I'm asking you the questions to get a better understanding, never said this was or wasn't going to be fixed 🙂 .
To ... | 2020-01-27T09:42:38 | 2020-01-27T09:42:38 | {} | MEMBER |
vuejs | core | 578,707,608 | 661 | posva | Applying the `inert` attribute might be the desired behavior for a modal in the scenario you specified, but it's not well supported (https://developer.mozilla.org/en-US/docs/Web/API/HTMLElement) which means we cannot always rely on it.
I checked against current Vue 2 version, and the behavior is consistent regarding... | 2020-01-27T11:33:05 | 2020-01-27T11:33:05 | {} | MEMBER |
vuejs | core | 579,827,250 | 671 | yyx990803 | @lbogdan I think there is a force of habit here. What I've found is that in a lot of cases people are actually doing this:
``` js
created() {
this.fetchData(this.id)
},
watch: {
id: fetchData
},
methods: {
fetchData(id) {
// ...
}
}
```
Where with immediate by default this becomes:
``` ... | 2020-01-29T16:00:25 | 2020-01-29T16:00:25 | {
"+1": 2
} | MEMBER |
vuejs | core | 581,073,230 | 683 | sys-keys | https://github.com/vuejs/vue-next/blob/74cb207c2296c9865ba42e1539077f4bb68929f0/packages/runtime-core/src/apiWatch.ts#L156-L164
Seems to be caused by setting oldValue before first run to undefined, then checking against the result of the getter which is also undefined so no runs will take place until the value is ch... | 2020-02-01T21:55:59 | 2020-02-01T21:55:59 | {
"+1": 3
} | NONE |
vuejs | core | 581,414,347 | 688 | dsseng | sure! | 2020-02-03T13:30:38 | 2020-02-03T13:30:38 | {} | CONTRIBUTOR |
vuejs | core | 581,816,266 | 692 | posva | Hey, I didn't check this specific PR but I've sen you've done a bunch of PR before and that's great, thank you!
When fixing a bug, it would be perfect if you could reference the issue it's solving by including the text `fix #<number of issue>` (e.g. `fix #684`) in the description of the commit (the description is t... | 2020-02-04T09:24:06 | 2020-02-04T09:24:06 | {} | MEMBER |
vuejs | core | 578,676,752 | 660 | jods4 | >I honestly don't see how a newcomer will figure out how refs work without reading documentation about it because they cannot just come up with the idea of:
Well, I did? (obviously I read some docs)
This is feedback from a new Vue user: I tried to grab an element with `ref` and I had a bad time getting it to work. ... | 2020-01-27T10:08:06 | 2020-01-27T10:11:06 | {} | CONTRIBUTOR |
vuejs | core | 578,732,769 | 660 | jods4 | > Did you read this
I did but unfortunately I found it only after figuring out what was the problem on my own.
That documentation is good but the `ref` caveat is really only mentionned once in passing:
> [...], if a VNode's ref key corresponds to a ref on the render context, [...]
So it's there but it's easy ... | 2020-01-27T12:52:07 | 2020-01-27T12:52:07 | {} | CONTRIBUTOR |
vuejs | core | 579,046,355 | 661 | Justineo | You can enforce `pointer-events: none` on leaving in userland. This at least prevents user interactions from pointer devices. | 2020-01-28T02:11:09 | 2020-01-28T02:11:09 | {
"+1": 1
} | MEMBER |
vuejs | core | 579,127,571 | 664 | LinusBorg | Can't run your repo right now but: have you tried it with an updated verion of `@vue/sfc-compiler` as well? Your current lockfile still refers to alpha-1:
```yaml
"@vue/compiler-sfc@^3.0.0-alpha.1":
version "3.0.0-alpha.1"
```
Just like vue-template-compiler in Vue 2, this dependency has to be kept in sync wi... | 2020-01-28T08:08:06 | 2020-01-28T08:08:06 | {
"rocket": 1
} | MEMBER |
vuejs | core | 579,129,648 | 664 | veaba | > Can't run your repo right now but: have you tried it with an updated verion of `@vue/sfc-compiler` as well? Your current lockfile still refers to alpha-1:
>
> ```yaml
> "@vue/compiler-sfc@^3.0.0-alpha.1":
> version "3.0.0-alpha.1"
> ```
>
> Just like vue-template-compiler in Vue 2, this dependency has to b... | 2020-01-28T08:14:48 | 2020-01-28T08:15:54 | {} | MEMBER |
vuejs | core | 579,300,930 | 668 | yyx990803 | See 78beed2 | 2020-01-28T15:29:08 | 2020-01-28T15:29:08 | {} | MEMBER |
vuejs | core | 580,252,048 | 671 | lmiller1990 | [I opened an RFC](https://github.com/vuejs/rfcs/pull/125) as suggested; I'll close this so we can keep discussion in one place. | 2020-01-30T13:27:57 | 2020-01-30T13:28:28 | {} | MEMBER |
vuejs | core | 578,662,849 | 660 | jods4 | @posva Do you find the behavior of the first repro above intuitive for a newcomer?
I ended up in that situation (more involved, but basically this root cause) and I lost a lot of time figuring out why I didn't get the element. It was not a nice experience.
Moreover I didn't try but I expect this to not work eithe... | 2020-01-27T09:31:38 | 2020-01-27T09:31:38 | {} | CONTRIBUTOR |
vuejs | core | 578,739,385 | 661 | jods4 | > Applying the inert attribute might be the desired behavior for a modal
The modal was an example. It's pretty much desirable on every element leaving DOM.
If you have a list with a button (delete, edit, doesn't matter) on every row, you don't want the command to trigger when an item is animated out.
Removed e... | 2020-01-27T13:11:18 | 2020-01-27T13:11:18 | {} | CONTRIBUTOR |
vuejs | core | 579,830,395 | 671 | lbogdan | Yes, that's exactly the use-case I use `immediate: true` for. But pretty much all my other imperative watchers ("do something when some state changes") are non-immediate, and would probably break in subtle ways if suddenly made immediate. I could go through my code bases and come up with examples, if that helps. We can... | 2020-01-29T16:07:03 | 2020-01-29T16:10:19 | {
"+1": 1
} | NONE |
vuejs | core | 580,044,539 | 673 | jods4 | @yyx990803 Sure, can do. Maybe we should define some kind of target size?
I can make it look as bad as I want since `track` allocates in `O(n)` and my proposal in `O(1)` (for ideal pattern that fit a full iteration, random access is still bad). | 2020-01-30T01:37:56 | 2020-01-30T01:37:56 | {} | CONTRIBUTOR |
vuejs | core | 580,787,905 | 671 | pimlie | @lbogdan I've always disliked `useEffect` because for some reason its never immediately clear to me what it means. An effect has a cause but what exactly is the cause here? Watch (or observe, listen) is much more clear to me in terms of what to expect.
Eg if you 'watch' your children then you make sure that when they ... | 2020-01-31T15:49:30 | 2020-01-31T15:49:30 | {
"+1": 1
} | NONE |
vuejs | core | 581,413,169 | 688 | posva | Let's keep #687 and discuss differences (using a Symbol) there since it was opened before. | 2020-02-03T13:27:42 | 2020-02-03T13:27:42 | {} | MEMBER |
vuejs | core | 581,496,537 | 673 | jods4 | Playing with this benchmark got me thinking.
The beauty of having explicit reactivity is the flexibility it gives.
It is not something you can do when everything is magically observed (e.g. Angular).
I think it would be nice if:
- Vue exposes easy-to-use, simple apis that work well enough for most cases.
- Wh... | 2020-02-03T16:24:25 | 2020-02-03T16:31:46 | {
"eyes": 1
} | CONTRIBUTOR |
vuejs | core | 581,887,868 | 692 | underfin | Thanks a lot.I will be attention to this. | 2020-02-04T12:27:27 | 2020-02-04T12:27:27 | {} | MEMBER |
vuejs | core | 583,354,326 | 701 | jiangying000 | And I found there are 75 empty lines with leading white spaces in snapshots.
| 2020-02-07T11:38:06 | 2020-02-07T11:38:06 | {} | CONTRIBUTOR |
vuejs | core | 583,891,007 | 707 | yyx990803 | > Should compiler be removable (server-renderer built w/o compiler, just as in Vue client renderer)?
I don't think so since there are no size concerns for npm packages.
> Should I rewrite existing SSR tests with templates to simplify them (except, of course, render/ssrRender ones)?
This is unnecessary - certai... | 2020-02-09T20:43:58 | 2020-02-09T20:43:58 | {} | MEMBER |
vuejs | core | 584,973,972 | 716 | ShenQingchuan | ## For Optimization
I also found @jods4 give an suggestion of optimization for full traverse in issue #673
I think we found a way for that in my PR solution, because every time we wanna traverse all the array ( Map , Set as well ), whatever we use traditional `for` loop, `.forEach` method, we would definitely use... | 2020-02-12T01:38:30 | 2020-02-12T01:38:30 | {} | CONTRIBUTOR |
vuejs | core | 585,557,924 | 719 | ShenQingchuan | Do you mean that this is supposed to be like this ?
```ts
export type WatchCallback<T = any> = (
value: T,
oldValue: T | undefined,
onCleanup: CleanupRegistrator
) => any
``` | 2020-02-13T05:27:39 | 2020-02-13T05:27:39 | {} | CONTRIBUTOR |
vuejs | core | 585,562,246 | 719 | jacekkarczmarczyk | The problem with this solution is that `oldValue` can be unnecessarily `undefined` for consecutive calls, but to be honest I have no idea if there is a different way to solve it. | 2020-02-13T05:46:09 | 2020-02-13T05:46:09 | {} | CONTRIBUTOR |
vuejs | core | 577,948,428 | 629 | backbone87 | shouldnt this be documented here https://github.com/vuejs/rfcs/blob/master/active-rfcs/0009-global-api-change.md ? | 2020-01-24T01:00:11 | 2020-01-24T01:00:11 | {} | NONE |
vuejs | core | 578,754,698 | 660 | posva | > Yeah but I'm using `reactive` to avoid declaring several `ref` and having to use `value`.
It's definitely weird to have to include one `ref` in the lot because of how it's used in the template.
Personally I didn't expect someone to return the object returned by `reactive`, in which scenario having `aRef: null` ma... | 2020-01-27T13:49:29 | 2020-01-27T13:49:29 | {} | MEMBER |
vuejs | core | 579,824,347 | 671 | lbogdan | Re: your first point: while it makes sense that having different APIs for watchers between Composition and Object API might be confusing, in my personal experience, I use `immediate` much less than not. So while the breaking change could be documented (and maybe even addressed by a codemod?), having to explicitly speci... | 2020-01-29T15:54:32 | 2020-01-29T15:55:42 | {} | NONE |
vuejs | core | 581,948,580 | 696 | yyx990803 | Fixed in `vue-loader@16.0.0-alpha.3` | 2020-02-04T14:56:46 | 2020-02-04T14:56:46 | {
"+1": 2
} | MEMBER |
vuejs | core | 583,249,745 | 701 | yyx990803 | The tests are passing, what is this fixing? | 2020-02-07T06:11:59 | 2020-02-07T06:11:59 | {} | MEMBER |
vuejs | core | 583,891,248 | 701 | yyx990803 | Snapshots are generated automatically by Jest - I don't see a point in "fixing" them if the tests are passing. | 2020-02-09T20:46:27 | 2020-02-09T20:46:27 | {} | MEMBER |
vuejs | core | 585,562,809 | 719 | KaelWD | Presumably you'd be checking that anyway with a type guard so `undefined` would be removed.
Also if `options.lazy === true` then `oldValue` should be `T` | 2020-02-13T05:48:28 | 2020-02-13T05:50:11 | {
"+1": 2
} | CONTRIBUTOR |
vuejs | core | 577,713,480 | 629 | yyx990803 | It is removed. For TypeScript, use [defineComponent](https://composition-api.vuejs.org/api.html#definecomponent). | 2020-01-23T14:48:39 | 2020-01-23T14:48:39 | {} | MEMBER |
vuejs | core | 578,533,391 | 656 | yyx990803 | After the fix:
- For a prop named `fooBar`, the emitted events should always be consistent and use `update:fooBar`.
- When using the component:
- If using non-DOM templates: prefer `<child v-model:fooBar="xxx" />`
- If using in-DOM templates, `<child v-model:foo-bar="xxx" />` will also work. | 2020-01-26T19:17:01 | 2020-01-26T19:17:28 | {
"+1": 5
} | MEMBER |
vuejs | core | 578,744,326 | 660 | posva | > the ref caveat is really only mentionned once in passing:
That's not true, it appears in all three code samples as well. The paragraph that mentions it is also the only explanation. That's rather hard to overlook.
> When you say: "this.$refs.box with no property box declared anywhere" that applies to Vue 2, rig... | 2020-01-27T13:24:14 | 2020-01-27T13:24:14 | {} | MEMBER |
vuejs | core | 578,988,217 | 661 | jods4 | I can't think of a scenario where it's not the right thing (can you?), but you can add a prop `noInert` just in case.
Chrome/Edge/Opera support is sometimes enough. It's better than nothing. A polyfill exists but is far from trivial.
I'm looking at creating a derived component that adds this in `onLeave` before c... | 2020-01-27T22:34:54 | 2020-01-27T22:34:54 | {} | CONTRIBUTOR |
vuejs | core | 579,762,617 | 670 | cexbrayat | @yyx990803 I opened the issue in this repo for visibility, but maybe you prefer it in `vue-template-compiler` if I'm not mistaken on the root cause. Let me know if I need to move it and if I can help to fix it. I'm not sure about the best way: release a `vue-template-compiler@3.0.0-alpha`? Just relax the check? | 2020-01-29T13:48:19 | 2020-01-29T13:48:19 | {} | MEMBER |
vuejs | core | 580,045,728 | 675 | jods4 | I figured since the API already existed I could forego a full RFC.
Do you want me to open a proper RFC or is there no chance you expose those two?
I kind of think they meet (1) and (2) since:
(1) a tiny fraction of users would build an optimized collection; but it could benefit many users reusing such a collecti... | 2020-01-30T01:43:20 | 2020-01-30T01:43:20 | {} | CONTRIBUTOR |
vuejs | core | 580,212,551 | 671 | pimlie | Want to chime in that currently I often rely on watchers not being triggered immediately because I put client-only code in them when I know that a change wont ever happen on the server (eg because a change can only happen after user interaction). | 2020-01-30T11:35:31 | 2020-01-30T11:35:31 | {} | NONE |
vuejs | core | 580,623,734 | 682 | dependabot-preview[bot] | One of your CI runs failed on this pull request, so Dependabot won't merge it.
- [ci/circleci: test](https://circleci.com/gh/vuejs/vue-next/2264?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link)
Dependabot will still automatically merge this pull request if you amend it and your test... | 2020-01-31T07:49:09 | 2020-01-31T07:49:09 | {} | CONTRIBUTOR |
vuejs | core | 580,768,831 | 671 | lbogdan | @yyx990803 I too think that's a much better compromise.
In regard to naming:
- `autorun` - sounds like something that only runs once, at initialization
- `runAndWatch` - too many words, hard to remember (the order)
- `effect` - I think that's the best one: short and expressive; also it's kind of similar to `use... | 2020-01-31T15:02:06 | 2020-02-02T13:19:18 | {
"+1": 1
} | NONE |
vuejs | core | 581,134,958 | 671 | lbogdan | My reasoning for `effect` is that you're defining a (side-)effect of state changing. I'm not a native speaker either, but for me, both `watch` and `observe` are semantically passive actions, meaning more like "watch state for changes" than actually "do something when state changes". I've been using `watch` for so long ... | 2020-02-02T13:20:39 | 2020-02-02T13:21:11 | {} | NONE |
vuejs | core | 583,425,376 | 684 | CyberAP | I personally would find it very useful for hooks to not contain any information of what they are related to and use them directly, like this:
```html
<FooBar @mounted="onFooBarMounted" />
```
This goes on par with the component options and does not conflict with them since they're prefixed with `on`.
I don't... | 2020-02-07T14:54:58 | 2020-02-07T14:54:58 | {} | CONTRIBUTOR |
vuejs | core | 583,889,133 | 710 | yyx990803 | 👍 This is indeed better, thanks! | 2020-02-09T20:25:23 | 2020-02-09T20:25:23 | {} | MEMBER |
vuejs | core | 584,975,003 | 716 | ShenQingchuan | > I can make it look as bad as I want since track allocates in O(n) and my proposal in O(1) (for ideal pattern that fit a full iteration, random access is still bad).
I am thinking about if this goal has already achieved, but I do not have any experience of Vue's benchmark 🤣
| 2020-02-12T01:42:29 | 2020-02-12T01:42:29 | {} | CONTRIBUTOR |
vuejs | core | 578,497,576 | 657 | mesqueeb | @yyx990803 thanks so much! | 2020-01-26T12:31:39 | 2020-01-26T12:31:39 | {} | NONE |
vuejs | core | 579,291,193 | 668 | vue-bot | Hello, thank you for taking time filling this issue!
However, we kindly ask you to use our [Issue Helper](https://new-issue.vuejs.org/?repo=vuejs/vue-next) when creating new issues, in order to ensure every issue provides the necessary information for us to investigate. This explains why your issue has been automatic... | 2020-01-28T15:10:06 | 2020-01-28T15:10:06 | {} | CONTRIBUTOR |
vuejs | core | 581,572,185 | 690 | dsseng | Sorry, didn't notice that. | 2020-02-03T19:14:41 | 2020-02-03T19:14:41 | {} | CONTRIBUTOR |
vuejs | core | 583,891,168 | 708 | yyx990803 | thanks! | 2020-02-09T20:45:38 | 2020-02-09T20:45:38 | {
"+1": 1
} | MEMBER |
vuejs | core | 579,116,075 | 664 | veaba | alpha 4 =>
alpha 3 =>
| 2020-01-28T07:30:01 | 2020-01-28T07:30:01 | {} | MEMBER |
vuejs | core | 579,427,154 | 663 | veaba | Also too:
```html
<li v-if="!authObj.isAuth&&authObj.oAuthUrl"></li>
``` | 2020-01-28T19:57:18 | 2020-01-28T19:57:18 | {} | MEMBER |
vuejs | core | 580,044,144 | 675 | yyx990803 | First of all, thanks for these suggestions. However, please beware that:
- Making more APIs public increase the maintenance surface and makes them subject to semver version constraints.
- As you have noted, this is a very niche API that only caters to a very small group of users. The criteria for adding public AP... | 2020-01-30T01:35:57 | 2020-01-30T01:36:29 | {} | MEMBER |
vuejs | core | 582,779,745 | 698 | dependabot-preview[bot] | One of your CI runs failed on this pull request, so Dependabot won't merge it.
- [ci/circleci: test](https://circleci.com/gh/vuejs/vue-next/2412?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link)
Dependabot will still automatically merge this pull request if you amend it and your test... | 2020-02-06T07:52:35 | 2020-02-06T07:52:35 | {} | CONTRIBUTOR |
vuejs | core | 583,730,907 | 677 | dsfx3d | Is this issue open for PRs? | 2020-02-08T12:09:47 | 2020-02-08T12:09:47 | {} | NONE |
vuejs | core | 584,060,324 | 701 | jiangying000 | @yyx990803 I just split the changes into several commits
code changes (aae6baab817e78d468eae5321a77585e233aabc3)
snapshot generated by `jest -u` (1a6ed4901cfdb6183ff76742079ac5d0b5af8e0c) | 2020-02-10T10:40:44 | 2020-02-10T14:30:36 | {} | CONTRIBUTOR |
vuejs | core | 584,585,058 | 716 | ShenQingchuan | # Solution's description
In this unit test case:
```ts
it('should observe for of iteration', () => {
let dummy
const map = reactive(new Map())
effect(() => {
dummy = 0
// eslint-disable-next-line no-unused-vars
for (let [key, num] of map) {
key
dummy += num
}
})
... | 2020-02-11T11:10:42 | 2020-02-11T11:10:42 | {} | CONTRIBUTOR |
vuejs | core | 581,168,364 | 671 | yyx990803 | Please use the [RFC](https://github.com/vuejs/rfcs/pull/125) for further discussions so that we can have it all in one place, thanks! | 2020-02-02T19:29:45 | 2020-02-02T19:29:45 | {} | MEMBER |
vuejs | core | 578,688,470 | 660 | posva | > Well, I did? (obviously I read some docs)
That's a bit contradictory: you didn't figure it out if you read the docs 😄 . Did you read this https://vue-composition-api-rfc.netlify.com/api.html#template-refs or something else? I don't think we have proper docs on this yet, so it's something we can gather feedback fr... | 2020-01-27T10:38:44 | 2020-01-27T10:38:44 | {} | MEMBER |
vuejs | core | 578,694,764 | 659 | posva | The classes name changed:
- enter-from
- leave-from
Using those will make it work with enter. It was working with leave because by default things have an opacity of 0 and a translate of 0 🙂
See https://github.com/vuejs/rfcs/blob/master/active-rfcs/0018-transition-class-change.md
| 2020-01-27T10:55:53 | 2020-01-27T10:55:53 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.