profile
viewpoint

issue comment6eDesign/svelte-calendar

Feature: How to internationalize the calendar

43de4fea3bd9d593cd5ffd6856a202d03e5bbb88 added weekStart parameter. Thanks a lot for that.

<Datepicker weekStart=1 /> does not seem to work (due to type mismatch in date.getDay() !== weekStart). <Datepicker weekStart={1} /> works, but I'm not sure whether it is the way that parameter was meant to be supplied. I don't mind if it was, though.

jdgaravito

comment created time in a day

issue commentpzmarzly/mic_over_mumble

Add alsa support

I removed pulseaudio, are you want to add support for alsa?

Hi. While I'm not interested in writing ALSA version myself, I could merge the script if someone created a nicely working one.

Here are some links that hopefully will be helpful for someone:

  • https://www.alsa-project.org/main/index.php/Matrix:Module-aloop
  • https://www.kernel.org/doc/html/v4.11/sound/alsa-configuration.html

Disclaimer: if you are a Linux beginner reading the pages linked above: don't compile the packages from scratch, use your distribution's repositories. By the way, you probably don't want to have PulseAudio removed in the first place, though check out Jack if you need to keep latency low.

If you want a quick hack, you can just remove all pactl, paid, and pacmd stuff from the code, run sudo modprobe snd-aloop before running the script, and select correct input and output devices manually. This method should work for most programs.

japiotr123

comment created time in 4 days

push eventrmst/yoke-android

Ivan Amaro

commit sha 98738f53f51a0ea11b4a5e891342b5f347ab9f8a

pt-rBR translation (#8) by iramaro

view details

push time in 21 days

PR merged rmst/yoke-android

pt-rBR translation

Inserted translation.

+71 -0

5 comments

1 changed file

iramaro

pr closed time in 21 days

pull request commentrmst/yoke-android

pt-rBR translation

OK, merging...

iramaro

comment created time in 21 days

pull request commentrmst/yoke-android

pt-rBR translation

@medape Are these debug strings presented inside the app? If so, then they should be translated. If not, then it differs from app to an app, and I don't have an opinion about that.

iramaro

comment created time in 22 days

pull request commentrmst/yoke-android

pt-rBR translation

Thanks a lot for this.

@medape should debug messages (log_) be translated?

iramaro

comment created time in 24 days

issue comment6eDesign/svelte-calendar

Feature: How to internationalize the calendar

@6eDesign thanks a lot for this. A minor nitpick (irrelevant to me): in Slavic languages, noun forms change (due to case forms). Thus, e.g. in Polish "February 2020" is "luty 2020", but "2. February 2020" (second of February 2020) is "2. lutego 2020". Hence, I cannot use #{F} (full month) formatting option, as it would give "2. luty 2020" (otherwise I would have incorrect headers in the calendar popup). But I prefer #{M} (short month) anyway, where this problem doesn't occur. Actually a lot of programs use short month names, probably to avoid this i18n mess.

Then again, it's a minor issue, and even if I used full month names, such grammar mistakes wouldn't prevent anyone from understanding the date (it would just sound wrong in their heads). There are lot of programs that make such mistakes, and almost no one cares.

Weeks not starting on Monday would be a much more distracting problem for users, though.

jdgaravito

comment created time in a month

issue openeddenis-sokolov/remove-github-forks

Deprecated `error.code`

Deprecation: [@octokit/request-error] `error.code` is deprecated, use `error.status`.
    at RequestError.get (/usr/lib/node_modules/remove-github-forks/node_modules/@octokit/request-error/dist-node/index.js:29:17)
    at /usr/lib/node_modules/remove-github-forks/src/shouldDeleteFork.js:35:21
    at /usr/lib/node_modules/remove-github-forks/src/githubFactory.js:75:25
    at runMicrotasks (<anonymous>)
    at processTicksAndRejections (internal/process/task_queues.js:93:5) {
  name: 'Deprecation'

created time in a month

issue closed6eDesign/svelte-calendar

TypeError: weeks[0] is undefined when using Change Timezone extension

With this extension enabled, I keep getting this error, both in Firefox, and Chrome, no matter which timezone I choose:

DeepinScreenshot_select-area_20191019185351

Current real date on my PC is 2019-10-19 19:41, Saturday.

Smallest reproducible example code:

<script>
    import Datepicker from 'svelte-calendar'
    let selected
    const format = "#{l}, #{j} #{F} #{Y}"
</script>
<Datepicker {format} bind:selected />

I don't think any of my users will have this extension installed, but this behaviour shouldn't happen nonetheless.

closed time in a month

pzmarzly

issue comment6eDesign/svelte-calendar

TypeError: weeks[0] is undefined when using Change Timezone extension

I dug for a bit, and found out that there is a problem with Date.getMonth() when using that extension (it e.g. still reports September on 2019-10-01 00:00, unlike the JS API). This is definitely extension's bug, not svelte-calendar's. It looks like it could be worked around by changing these lines:

-  let date = new Date(year, month, 1);
+  let date = new Date(year, month, 1, 15);
-  let date = new Date(start.getFullYear(), start.getMonth(), 1);
+  let date = new Date(start.getFullYear(), start.getMonth(), 1, 15);

, but I don't suggest we do that - I'm leaving this just for future reference.

Looking at that extension's review page, I can see svelte-calendar is not the only calendar affected:

after a few days of use: breaks a number of sites, including flight booking at westjet.com (calendars do not work at all) and managing my Mikrotik router (all dates are incorrect by a day.)

I'll try contacting the extension's author. Anyway, the issue can be closed.

Do you have more specific steps needed to reproduce the issue you're seeing?

Sadly, not really. The bug also happens on my Ubuntu VM with clear Firefox (aside of that extension), English (US) locale. I tried setting different timezones, it made no difference.

pzmarzly

comment created time in a month

issue commentrmst/yoke-android

Button presence

Once you're done, you'll find the most current APK

Well, not if he clones --branch v0.1 :smile: . Try:

git clone --depth 1 https://github.com/rmst/yoke
cd yoke
pip3 install .
galou

comment created time in a month

issue closedrmst/yoke

Connectivity issues with Yoke over Wifi

Has anyone else experienced disconnects when using yoke via Wifi? To get good connectivity? I usually use USB tethering and connect manually, which seemed to be the only way for yoke to work reliably. Perhaps it would be sensible to get rid of the UDP over IP communication and instead use direct USB (and potentially bluetooth).

closed time in 2 months

rmst

issue commentrmst/yoke

Connectivity issues with Yoke over Wifi

By the way, when we do, we should use more bits for the axes events. Yoke expects integers in [1, 32768] and jstest-gtk always reports within [-32767, 32767]; any 15-bit or 16-bit range would be a better match than the 8-bit values we're using now, at no cost of speed

While I think that 1 byte of precision is enough for almost all cases (I'm not aware of any program that expects more precise input from a real gamepad/joystick/racing wheel), I'm not against increasing it either. If you want to, go ahead and increase it.

That's unrelated to this issue, though, so I'm going to close it now.

rmst

comment created time in 2 months

issue openedrmst/yoke-android

[minor] Hide the progress bar when there is no update in progress

The gray bar could be hidden by default, and shown only when files are being transmitted. @medape

Screenshot_20191025-160550

created time in 2 months

push eventrmst/yoke

Boris

commit sha b650c9dfe8951e3d8841a424267dafea2f1f62c0

Write manifest upon starting webserver Also, tweak error messages.

view details

Boris

commit sha 787d748559e277b9859886f32bb171651c876e2c

Publish last APK

view details

Monso pequeño

commit sha 714db6a5f7f76655fbd9abe8098e11edacb02341

Customisable socket buffer size Mostly for experiments and for measuring lag, but some power users might find it useful too.

view details

Monso pequeño

commit sha bafb12d927561628d8cd76c0531797252e2dbaf4

Fix typos

view details

Boris

commit sha ccfd7637f77314f7fcdec990a3d70900f156dce8

Changing manifest filename to manifest.json Also adding a dummy file so pip3 won't get confused when uninstalling Yoke.

view details

Boris

commit sha 8eb5a830d6ddd4ba2c03570e34ee9e219486fa2e

APK from commit d91eb99 at rmst/yoke-android

view details

Paweł Zmarzły

commit sha 7cf90f413fae5ab3f97e67697bef77f3f8a28ec2

Merge pull request #41 from medape/master JSON manifest upon creating server

view details

push time in 2 months

PR merged rmst/yoke

JSON manifest upon creating server

The first commit is necessary for rmst/yoke-android's latest PR to function. It creates a JSON file every time a web server is started. This JSON file is the first thing the APK downloads, and contains paths to every file and folder in the webserver, the total size, and the last date of modification (this last one isn't implemented yet in Android's side, it's just to inform the user before downloading a new gamepad).

The other commit is just to upload the newest APK.

+54 -15

0 comment

5 changed files

medape

pr closed time in 2 months

issue closedrmst/yoke

Can't run on Windows 10

Hi! When I try to run it on Windows 10 as Administrator, I get error:

C:\Program Files\Python37\Scripts>python yoke Traceback (most recent call last): File "yoke", line 79, in <module> dev = yoke.Device(args.name, events=Service.events) File "C:\Program Files\Python37\lib\site-packages\yoke\service.py", line 100, in __init__ self.device = VjoyDevice(1) File "C:\Program Files\Python37\lib\site-packages\yoke\vjoy\vjoydevice.py", line 32, in __init__ self.lib = cdll.LoadLibrary(lib_path) File "C:\Program Files\Python37\lib\ctypes\__init__.py", line 434, in LoadLibrary return self._dlltype(name) File "C:\Program Files\Python37\lib\ctypes\__init__.py", line 356, in __init__ self._handle = _dlopen(self._name, mode) OSError: [WinError 126] The specified module could not be found

Please, help, I want to try it! :)

closed time in 2 months

iRomanyshyn

issue closedrmst/yoke

Motion controls stopped working after Android System WebView 76 update

Fun fact: Yoke was created as no phone-as-PC-gamepad app on the Play Store used to capture device movement.

onDeviceOrientation events are never fired now

Details: https://github.com/rmst/yoke/pull/38#issuecomment-521857892

If you are affected by this, please try downgrading ASW for now and describe your results here.

From Chrome DevTools (remote device inspection over ADB):

[Deprecation] The devicemotion event is deprecated on insecure origins and will be removed in M76, around July 2019. Event handlers can still be registered but are no longer invoked since M74, around April 2019. See https://www.chromestatus.com/feature/5688035094036480 for more details.

Suggestion from Google ~announcement~ "deprecation proposal":

On a local network, you can test on your Android device using port forwarding to access a remote host as localhost.

... which sucks as phone would usually need to be connected over USB (ADB over Wifi is disabled by most vendor overlays).

Perhaps we could add a proxy server hosted on the phone (since localhost is a secure origin) but this seems like a lot of work. It would be a easier to download the files to the phone (I hope file:// is a secure origin, but am I right?).

Apparently we cannot unconditionally accept unsigned certs, even for LAN, since 2016, if we want the app to be on the Play Store (and, AFAIK, as it's usually the case with Google, there is no appeal process where we could explain it's LAN connection). Maybe we could add "certificate error" dialog that users would need to click through every time they open the app, just to satisfy Play Store algorithms.

Another option is to once again fetch orientation events in Java code and everything else in JS, but I'd like to avoid that.

Personally, I'd go for downloading files to phone storage, but I'd like to hear your opinions.

closed time in 2 months

pzmarzly

issue commentrmst/yoke

Motion controls stopped working after Android System WebView 76 update

Closed by https://github.com/rmst/yoke-android/pull/5

pzmarzly

comment created time in 2 months

push eventrmst/yoke-android

Medape

commit sha 5428ccaad2bf6dce6659fd7520089a1737206159

Fix issue #4 (hide navigation bar) Also remove unneeded lines of code.

view details

Medape

commit sha 263baef7e3b748460c9d2935f0c4a551cc3479ec

Remove more unused components

view details

Medape

commit sha 6b7401daa604f8d68a91eeec7011e4a19bd9ff93

Fix part of issue #33 in rmst/yoke Reconnect button added. Upon disconnection, webview points to about:blank and vals_str is emptied.

view details

Medape

commit sha 29494337a39cc4e3cc54df9af8681d5129914f1c

Add support for Android localisation For the moment, only the app itself can be translated. There are no strings for other languages than English, and the webview is not translatable yet. Support for RTL is added for the app, but not for the webview yet.

view details

Medape

commit sha 56f2222142dbd607408f94cb72bf34ae1d78a52b

Fix and prettify reconnect button Reconnect button is now an ImageButton, and displays a toast if not connected to any machine. (Previous button raised a runtime exception). Drawables downloaded from https://material.io/resources/icons/ licensed by Google under the Apache Licence.

view details

Medape

commit sha 5793b4a25f984b126b6c0065cc15fec57189d28b

Fix issue #39 in rmst/yoke A greeter screen checks for the layout in external storage, and webview loads that. New bug: prompt for the spinner has disappeared. To do: bundle layout with the app, and copy it from the app to external storage.

view details

Medape

commit sha f009f3b1811c065b56612d8127fdec45bccd7580

HAHAHAHA Prompt is shown now. https://stackoverflow.com/questions/43267647/

view details

Medape

commit sha ae412c80c6034e237d199384efa5e0c0a935d6c3

Simplify paths and fix grammar

view details

Medape

commit sha 746e46cff9d23c98925eabb93c29bc8660c6a373

Make linter happier Read somewhere that it's mandatory to target Android Oreo or higher to appear on Google Play. In any case, this doesn't do any harm, and makes the linter happier.

view details

Medape

commit sha 3dd9f1c6ca2cc61d2ef8aa95a1bfe3dc8e7fef32

Return to one-activity model Change of plans; app will now download from webserver again.

view details

Medape

commit sha 7dc2f5a161a83609111ed67bb16202431c7c5712

Fetch files from webserver and save them in internal storage This is not the last commit. Extra code in both Python and Android needed for this to work flawlessly.

view details

Medape

commit sha 47dff61e7fb6a9581c4abfab87fceac2bf8accd9

Create every folder and file in manifest Currently only works the first time after erasing program data.

view details

Medape

commit sha b9361396951f88e13049d678054d456e42e1d71b

Redownload files from webserver whenever, delete whole folders To do: fix textview, add progress bar, decouple layout update from layout load, make error handling saner, maybe not run AsyncTask in UI thread when I'll know what I'm doing.

view details

Medape

commit sha 7d54a6ac952c03d935b35f7f9148cdbf58655c0d

Simplify code for erasing a folder, and handling error messages differently to debug messages.

view details

Medape

commit sha 86e0bf6fff5d9944f8d5aa6c5cb80dbf67ee1506

Improve AsyncTask and layout AsyncTask is now fully finished and truly runs in the background, not ont the UIThread. Layout includes new button, progress bar and overflow menu. Errors are now visible to the user.

view details

Medape

commit sha 11d30c7028557e3c09531779aa0800fe55addf40

Fix bugs, add warning toasts, finish AsyncTask

view details

Medape

commit sha 3f621b0dfabe8fcca57c9132a090eb964875cd51

Update Gradle so we can compile with Java LTS 11. Also changed one line of code because it wouldn't compile otherwise.

view details

Medape

commit sha 8ff842676ad4c9707190e62d84d3ab93efee9141

Edit one string to be less confusing It's very confusing to have a question as the error message, and OK as the dismiss button.

view details

Medape

commit sha 8dfd782f68e6345de12c2914c649b1cc25115422

Changing manifest filename to manifest.json

view details

Medape

commit sha d91eb99f54d9a4901166e9332fdac6e5a15963b0

Tweak manual connection dialog Dismissing it should cancel it, or else, the spinner reads → new manual connection whlie disconnected, and it gets confusing.

view details

push time in 2 months

PR merged rmst/yoke-android

Grand bug fix festival, still incomplete

This is still the very first Android app I'm working on, and half my work is based on StackOverflow's answer and copy-pasting from @rmst. I'm hoping I don't forget to document anything here.

  • Yoke is now fully fullscreen. Issue #4 (hiding the navigation bar) is now fixed.
  • Two nice things from issue 33 on rmst/yoke are implemented:
    • refresh button, with an icon downloaded from https://material.io/resources/icons/
    • app now forgets the joypad state upon disconnection. It won't send the latest status anymore.
  • Moved every string (even debug messages) to values/strings.xml, to allow easy localisation in the future. I had to create very long strings and it was becoming too much.
  • Done some preliminary work on applying a uniform style to the app, but buttons are still ugly as sin.
  • Improved the manifest file.
  • We were right about file:// being a secure origin: the WebView now reads from the external storage, restoring motion controls and fixing issue 39 from rmst/yoke .
  • Added greeter screen.

Most work is needed on these two last points. The greeter screen is a separate Activity with a very simple layout welcoming the user and guiding them around errors. If it detects a layout already in your phone, it automatically skips to the real app.

It's also supposed to allow the user to copy a default layout with the click of a button, but I haven't added the code for that yet.

Special permissions are needed to write to Android/data/com.simonramstedt/files in the external storage, but not to read it. The current code doesn't include this permission yet.

I suppose the easiest way would be to bundle our layout in the APK, and ask the user to grant WRITE_STORAGE_PERMISSION, and then copy to it. Anyways, @rmst, @pzmarzly, I'd like your opinions on this. (I'd also like you to check these commits for bugs or things to improve.) I'll upload the binary for this version on Releases.

There's no need to update the yoke repo to merge this PR.

+521 -86

11 comments

51 changed files

medape

pr closed time in 2 months

issue comment6eDesign/svelte-calendar

TypeError: weeks[0] is undefined when using Change Timezone extension

I made a repo with a tiny website that shows this behaviour. I also uploaded the website there. Please let me know whether you are able to reproduce the error with it. If not, I'll try to debug the calendar myself.

pzmarzly

comment created time in 2 months

create barnchpzmarzly/svelte-calendar-crash

branch : master

created branch time in 2 months

created repositorypzmarzly/svelte-calendar-crash

Tiny page to showcase a potential bug in svelte-calendar

created time in 2 months

startedDooskington/grubbnet

started time in 2 months

pull request commentrmst/yoke-android

Grand bug fix festival, still incomplete

So I'm not too keen on trying it again

No problem.

is there anything you would improve on this PR?

Maybe renaming manifest to manifest.json, like you've suggested before? Nothing else comes to my mind, and this PR has grown pretty large already, so it would be nice to have it merged.

medape

comment created time in 2 months

pull request commentrmst/yoke-android

Grand bug fix festival, still incomplete

I think it would be nice if DownloadFilesFromURL class was moved into a separate file. I'm not an Android developer, but I heard some of them complaining that their *Activity.java files are 1000s of lines long, and that it's hard to avoid that due to how Android Java API is built (verbose, boilerplate-y, callback-based, centralized around Activity). It looks to me like you've already done some work to separate download-in-background stuff from the rest, so we could avoid that problem.

This is just a thought, though. You don't have to do it (especially if it requires non-trivial effort).

medape

comment created time in 2 months

issue opened6eDesign/svelte-calendar

TypeError: weeks[0] is undefined when using Change Timezone extension

With this extension enabled, I keep getting this error, both in Firefox, and Chrome, no matter which timezone I choose:

DeepinScreenshot_select-area_20191019185351

Current real date on my PC is 2019-10-19 19:41, Saturday.

Smallest reproducible example code:

<script>
    import Datepicker from 'svelte-calendar'
    let selected
    const format = "#{l}, #{j} #{F} #{Y}"
</script>
<Datepicker {format} bind:selected />

I don't think any of my users will have this extension installed, but this behaviour shouldn't happen nonetheless.

created time in 2 months

push eventpzmarzly/productive-time-tracker

Paweł Zmarzły

commit sha 82295bcd20602ce56a2c40600a354737893642e0

Add LICENSE

view details

push time in 2 months

create barnchpzmarzly/productive-time-tracker

branch : license

created branch time in 2 months

fork pzmarzly/productive-time-tracker

Track how much time you spend sleeping, playing and working

fork in 2 months

startedd-kowalik/Coro

started time in 2 months

pull request comment6eDesign/svelte-calendar

Fix lints

Actually, in case of the second commit, now my IDE reports warning (no-console, no-undef) but npm run lint does not. Weird.

If the same happens to you, I will restore the "disable" comments (though I think it should also be possible to use .eslintrc.* overrides for that file).

pzmarzly

comment created time in 2 months

push eventpzmarzly/svelte-calendar

Paweł Zmarzły

commit sha e5378cf22bd090c544e62dca5dc628855fe2103e

fix: allow Object.assign

view details

push time in 2 months

create barnchpzmarzly/svelte-calendar

branch : lint-migrate

created branch time in 2 months

PR opened 6eDesign/svelte-calendar

Fix lints

I'm not sure which VS Code plugin showed me the CSS lint, but it's fixed now.

I tried removing eslint-disable-next-lines. In 2 cases they were unnecessary. In others, I don't agree with airbnb-base/legacy. Having to declare a function before using it? What is this, C?

+1 -2

0 comment

2 changed files

pr created time in 2 months

create barnchpzmarzly/svelte-calendar

branch : lints

created branch time in 2 months

PR opened 6eDesign/svelte-calendar

Improve docs

They still could be better, but at least now they will mention all options (so nobody will need to look at the source code export lets or "guess" the existence of selected).

Some links were repeating so I put them into a separate section. Then I thought it doesn't look right when some URLs are along the texts and others aren't, so I moved them all.

+43 -10

0 comment

1 changed file

pr created time in 2 months

push eventpzmarzly/svelte-calendar

Paweł Zmarzły

commit sha 0225153f15a26788b3842ddd862ba28a3bea3362

chore: docs

view details

push time in 2 months

push eventpzmarzly/svelte-calendar

Paweł Zmarzły

commit sha 85c66208c84a02e63e7e030b23bbe275ddae4023

chore: docs

view details

push time in 2 months

push eventpzmarzly/svelte-calendar

Paweł Zmarzły

commit sha c5e490de00f315bd19cf98e9d32ceee60f0793a4

chore: docs

view details

push time in 2 months

create barnchpzmarzly/svelte-calendar

branch : docs

created branch time in 2 months

PR opened 6eDesign/svelte-calendar

Release the code under free license
+22 -0

0 comment

2 changed files

pr created time in 2 months

create barnchpzmarzly/svelte-calendar

branch : license

created branch time in 2 months

fork pzmarzly/svelte-calendar

A lightweight datepicker with neat animations and a unique UX.

https://6edesign.github.io/svelte-calendar/

fork in 2 months

issue opened6eDesign/svelte-calendar

Button class "popover" conflicts with Bootstrap's .popover CSS

I'm trying to move existing app to Svelte, part by part. I had 2 weird issues with svelte-calendar:

  1. borders around buttons were distorted (it didn't cover right side of them)
  2. z-index of buttons was 1060

It turns out that the class .popover is also used in Bootstrap 4.

As a workaround, I can either use a custom version of Bootstrap (as I don't use popovers), or add:

:global(.datepicker button) { width: initial !important; } /* for the first issue */
:global(.datepicker .popover) { z-index: unset; } /* for the second */

But a proper fix would be nice (just rename .popover to something else).

created time in 2 months

issue openedsveltejs/svelte

Svelte generates code referencing undefined variables like div_nodes

Hi. I'm trying to use ESLint on the final IIFE .js that Svelte generates, in order to get the no-undef lint (usage of undefined variable).

<details> <summary>.eslintrc.js</summary>

module.exports = {
  parserOptions: {
    ecmaVersion: 2019,
    sourceType: 'module',
  },
  env: {
    es6: true,
    browser: true,
    node: true,
  },
  plugins: [
    'svelte3',
    '@typescript-eslint/eslint-plugin'
  ],
  overrides: [
    {
      files: ['**/*.svelte'],
      processor: 'svelte3/svelte3',
      extends: "eslint:recommended",
    },
    {
      files: ['**/*.js'],
      rules: {
        "no-undef": "warn",
      },
    },
  ],
}

</details>

This gives me:

/path/to/bundle.js
1668:43  warning  'form_nodes' is not defined  no-undef

Offending code in bundle.js:

    		l: function claim(nodes) {
    			if (default_slot) { default_slot.l(form_nodes); }
    			throw new Error("options.hydrate only works if the component was compiled with the `hydratable: true` option");
    		},

Form.svelte is just <form><slot /></form> (I simplified it for the purpose of this report). If I add content to the default slot, or other HTML tags (children) to <form>, the issue is still there. Searching for form_nodes and _nodes in bundle.js gives me no results aside of the function above. If I change Form.svelte content to <div><slot /></form>, I get warning about undefined div_nodes.

Are <...>_nodes in the code above intended to be undefined? How can I set them? Or what else can I do to get the results I want (getting "... is not defined" errors at compile time, not runtime)?

created time in 2 months

issue commentiiuni/projektzapisy

Wyczyszczenie ciasteczek powoduje endless loop

Właściwie dalej nie wiem, skąd bierze się 100% użycia CPU. Myślę, że winowajcą może być parsowanie do NotificationsDict, ale chętnie się dowiem (choć oczywiście do rozwiązania problemu nie jest to potrzebne).

PS. created() nie musi być async, bo nie await-uje, dobrze myślę?

pzmarzly

comment created time in 2 months

issue openediiuni/projektzapisy

Wyczyszczenie ciasteczek powoduje endless loop

Błąd prawdopodobnie spowodowany pobieraniem powiadomień.

Wideo: https://youtu.be/3q23E7Oyt5A

Screenshot z ZAP-a (zaproxy) - request przed wyczyszczeniem cookies i po:

DeepinScreenshot_select-area_20190927233333

Testowane na Firefoxie, zarówno z włączonym blokowaniem trackerów, jak i bez. Wyświetlana liczba powiadomień może być wielkością strony logowania (od wyniku z ZAP-a trzeba odjąć rozmiar nagłówków).

https://github.com/iiuni/projektzapisy/blob/f1282624de12865cb54add664e3c8805026b0284/zapisy/apps/notifications/assets/components/Widget.vue#L34-L47

created time in 2 months

issue commentfreedomlayer/offst

Upgrade to latest version of the ring crate

SystemRandom implements Clone now but it doesn't help as much as I thought. It would probably be possible to get rid of arc_rng (RngContainer), and make a custom trait for it, but ring methods still expect SecureRandom. So the comment about 3 methods is wrong.

I dropped the work on fes (crypto library), but it could be completed one day (perhaps I will come back to it when I work on decentralized systems again, for now I moved to full-stack freelancing).

BTW congratulations on getting to HN front page.

realcr

comment created time in 2 months

create barnchpzmarzly/offst

branch : ring-clone

created branch time in 2 months

push eventpzmarzly/ble-ancs

Paweł Zmarzły

commit sha 1b25d4afafcbe7e6f74b244378b6379ae0b5d789

update deps

view details

push time in 3 months

fork pzmarzly/ble-ancs

A NodeJS library for Apple Notification Center Service over BLE

fork in 3 months

issue commentsveltejs/svelte

Tutorials don't work on iOS apparently

Another video - Sapper SPA routing breaks navigation - after going back, the same page is displayed, and only after a second or two the previous one is re-rendered. But it's this bad only on Tutorial and Examples. On other pages I usually get flashed by a white empty screen for about 50 ms when going back/forward (but sometimes navigation works as it should).

Rich-Harris

comment created time in 3 months

issue commentsveltejs/svelte

Tutorials don't work on iOS apparently

Yes. Here's a video: I'm tapping a checkbox with no effect (few times I managed to get text input to open). Then I go to the same code in Examples section and it just works. Using Firefox for iOS, but the same happens on Safari. This happens both with and without "Request PC version".

Same thing with buttons (e.g. in "Event modifiers" tutorial page) - nothing happens on click.

Rich-Harris

comment created time in 3 months

pull request commentutkarshkukreti/markup.rs

Add optional caching

Is there any other popular crate which does this?

I wasn't familiar with any, which is why I suggested it as an optional feature. Quick search on crates.io led me to gmp-mpfr-sys, which creates temporary files in ~/.cache/gmp-mpfr-sys, though it doesn't have to (it could use $OUT_DIR provided by cargo).

I was thinking about using target dir, but AFAIK we don't get any path to it. We would need to either:

a) assume it's $PWD/target - but it's sometimes not, b) assume it's $CARGO_MANIFEST_DIR/target - false in multi-crate projects, c) use $PWD to find a topmost directory that contains both Cargo.toml and target - which would work for 99% of use cases, but that 1% of users could get surprised

Many Rust developers already have large ~/.cargo/registry and ~/.cache/sccache dirs, so I thought writing few hundred KBs to ~/.cache wouldn't be that much of an issue if users were warned about that.

pzmarzly

comment created time in 3 months

startedjpodwys/prefetch

started time in 3 months

startedmagiclen/rocket-multipart-form-data

started time in 3 months

issue openedmagiclen/rocket-multipart-form-data

Make a release

Hi. I was investigating why cargo was downloading nickel in my project, and after a while I found out that the code here on master is newer than on crates.io. Since there is no ongoing development, could you upload the code to crates.io as version 0.7.1? Thanks in advance.

created time in 3 months

push eventpzmarzly/markup.rs

Paweł Zmarzły

commit sha a4e5034ba95eeb26c68dc3942a1936608d31cb1f

Use separate cache dirs for different crate versions

view details

push time in 3 months

PR opened utkarshkukreti/markup.rs

Add optional caching

In a project I'm working on, this decreases overall build time from between 8 and 10 to between 5 and 7 seconds.

+49 -3

0 comment

4 changed files

pr created time in 3 months

create barnchpzmarzly/markup.rs

branch : pzmarzly-patch-2

created branch time in 3 months

startedutkarshkukreti/markup.rs

started time in 3 months

PR opened utkarshkukreti/markup.rs

Show alternative parenthesis workaround in README

This code also works, but:

  • it's more flexible (you can create even larger blocks, e.g. @if {v == true} and @if {v} == true both work, versus @if *&(v) == true)
  • it looks less hack-y
  • it's easier to write
+9 -1

0 comment

1 changed file

pr created time in 3 months

PR closed pzmarzly/markup.rs

Show alternative parenthesis workaround in README

This code also works, but:

  • is more flexible (you can create even larger blocks, e.g. @if {v == true} and @if {v} == true both work, versus @if *&(v) == true)
  • looks less hack-y
  • is easier to write
+9 -1

1 comment

1 changed file

pzmarzly

pr closed time in 3 months

pull request commentpzmarzly/markup.rs

Show alternative parenthesis workaround in README

wrong repo

pzmarzly

comment created time in 3 months

PR opened pzmarzly/markup.rs

Show alternative parenthesis workaround in README

This code also works, but:

  • is more flexible (you can create even larger blocks, e.g. @if {v == true} and @if {v} == true both work, versus @if *&(v) == true)
  • looks less hack-y
  • is easier to write
+9 -1

0 comment

1 changed file

pr created time in 3 months

create barnchpzmarzly/markup.rs

branch : pzmarzly-patch-1

created branch time in 3 months

fork pzmarzly/markup.rs

A blazing fast, type-safe template engine for Rust.

fork in 3 months

pull request commentrmst/yoke-android

Grand bug fix festival, still incomplete

What do you think about downloading the files to internal storage just the first time, and then having an update button within the app?

That sounds okay. A bit weird, but I guess it would have its fans.

risk malicious code

I don't think there's much that HTML, JS and Yoke.update_vals can do in terms of malicious behavior. Perhaps it could fill all free phone's storage, but Android is still usable when this happens.

I don't know how SAF would work, but we probably wouldn't need it if we follow this approach

It would indeed be irrelevant. SAF is just an alternative for the Storage permission (SAF lets user give app permission for one folder/one file instead of all /data/media/0).

medape

comment created time in 3 months

more