the deadlock comes from when the async work captures the calling context

in the helper method , the Task<T> is created before Task.Run

try doing it so nothing gets to capture the caller’s context

Did you crack this yet? I am trying to do something similar but twilio is sending back some silence packets. I am not even sure if its getting to STT streaming on gpt-4o-realtime-preview model I intend to use.

You are doesn't imported CSS files to your component file. Import them to you components or on App.tsx.

import "./asserts/styles/otro.css" import "./asserts/style/style.css"

# Player Info

Allows script access: false

Player type: Object

SWF URL: https://cdn.jsdelivr.net/gh/ahm-rblox-game/123@main/365143_Impossible_Quiz_Deluxe_NG..swf

Param movie: https://cdn.jsdelivr.net/gh/ahm-rblox-game/123@main/365143_Impossible_Quiz_Deluxe_NG..swf

Attribute 0: undefined

Attribute 1: undefined

# Page Info

Page URL: https://618130714-atari-embeds.googleusercontent.com/embeds/16cb204cf3a9d4d223a0a3fd8b0eec5d/inner-frame-minified.html?jsh=m%3B%2F\_%2Fscs%2Fabc-static%2F\_%2Fjs%2Fk%3Dgapi.lb.en.PLtFj\_-5DjQ.O%2Fd%3D1%2Frs%3DAHpOoo-J85zQk73PCqZPyWTydWEIq3_4KA%2Fm%3D__features_\_

# Browser Info

User Agent: Mozilla/5.0 (X11; CrOS x86_64 14541.0.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/138.0.0.0 Safari/537.36

Platform: Linux x86_64

Has touch support: false

# Ruffle Info

Version: 0.1.0

Name: nightly 2025-03-04

Channel: nightly

Built: 2025-03-04T00:06:30.124Z

Commit: 950f3fe6883c20677a1d3d8debbb029867d6e94a

Is extension: false

# Metadata

You should use oracle ojdbc:

 :dependencies [[org.clojure/clojure "1.12.0"]
                 [com.oracle.database.jdbc/ojdbc11 "23.5.0.24.07"]
                 [com.oracle.database.jdbc/ucp "23.5.0.24.07"]

György Kőszeg figured out my problem: I needed to set the color options differently. Adding the following line of code after the initialization of info made it work perfectly!

info.ColorSpace = SKColorSpace.CreateSrgbLinear();

Please I need help . I'm facing the same problem and don't know what to do.

Any help is appreciated.

I have this error message:

https://submit.cs50.io/check50/920b5463f68b374c1e6ab130b8041766f9eeb99c

This was my code :

---

import re

def main():
    month_dict = {1:"January",2:"February",3:"March",4:"April",
                  5:"May",6:"June",7:"July",8:"August",
                  9:"September",10:"October",11:"November",12:"December"}

    while True:
        date = input("Date: ").strip()
        try:
            # Replace -, /, , with space
            for sep in ['-', '/', ',']:
                date = date.replace(sep, ' ')

            # get rid of white spaces
            list_date = [p for p in date.split() if p]


            # Numeric format for Month
            if list_date[0].isdigit():
                month = int(list_date[0])
                day = int(list_date[1])
                year = int(list_date[2])
            # Month-name format
            else:
                month_name = list_date[0]
                month = None
                for k, v in month_dict.items():
                    if month_name.lower() == v.lower():
                        month = k
                if month is None:
                    raise ValueError("Invalid month name.")
                day = int(list_date[1])
                year = int(list_date[2])

            # Make sure the range of months and days is correct
            if not (1 <= month <= 12):
                raise ValueError("Month must be between 1 and 12.")
            if not (1 <= day <= 31):
                raise ValueError("Day must be between 1 and 31.")

            # Format as YYYY-MM-DD
            new_date = f"{year}-{month:02}-{day:02}"
            print(new_date)

            break  # exit the loop if everything is correct
        # prompt the user to enter a correct date
        except Exception as e:
            print(f"Error: {e}")
            print("Please try again with a valid date format like 9/5/2020 or September 8, 1636.")

if __name__ == "__main__":
    main()

sudo apt install ./<file>.deb

# If you're on an older Linux distribution, you will need to run this instead:

# sudo dpkg -i <file>.deb

# sudo apt-get install -f # Install dependen

cies

I needed to achieve something similar and I ended up doing the following:

1. Create a view in the database (I'm using MySQL) that uses the desired GROUP BY statement

2. Create a dataset in Superset based on that view

3. Create a chart based on that dataset

The same error gave me some headache to solve. How did I fix it?

……. Drumroll……..

rebooting all nest devices made the errors disappear and I was able to get event images. It’s worth trying yourself to see if it fixes your error.

https://router.vuejs.org/guide/essentials/dynamic-matching.html#catch-all-404-not-found-route

const routes = [
  // will match everything and put it under `route.params.pathMatch`
  { path: '/:pathMatch(.*)*', name: 'NotFound', component: NotFound },
  // will match anything starting with `/user-` and put it under `route.params.afterUser`
  { path: '/user-:afterUser(.*)', component: UserGeneric },
]

P.S. don't thank me))

I had the same problem when I started working at a company using Google suite. Also didn’t see a good free tool when searching. I built one here.

https://www.chartmekko.com/

Above didn't work for me, so I had to do this:

DATE.prototype._stringify = function _stringify(date, options) {
    date = this._applyTimezone(date, options);
    return date.tz("Europe/Stockholm").format('YYYY-MM-DD HH:mm:ss.SSS');
};

according to https://developer.android.com/training/data-storage/room

 // If this project uses any Kotlin source, use Kotlin Symbol Processing (KSP)
    // See Add the KSP plugin to your project
    ksp("androidx.room:room-compiler:$room_version")

try to add it.

A non-ssreflect way of proceeding here is to simply destruct (boolP bar) as x. I would presume that the ssreflect-case tactic is buggy here but do not know enough about ssreflect to be certain.

--tab-size-* is a reference to a variable that you need to define yourself in @theme. Only after that should you write the animation, using not the variable itself but a specific value, for example tab-4 or any other value you’ve defined. Right now you’re trying to do it directly inside @theme.

@theme is only needed for defining variables or other reusable properties, such as @keyframes.

@theme {  
--tab-size-2: 2;  
--tab-size-4: 4;  
--tab-size-github: 8;
}
@utility tab-* {  
tab-size: --value(--tab-size-*);
}

P.S. If I understood you correctly

bu kodu telefonda nasıl calıstırabılırım

Interesting find, so if I use

(payload.Date as DateTime) as String {format: "yyyy-MM-dd'T'HH:mm:ss.SSS'Z'"}

I get the string that I'm looking for.

A couple of conversions, but it gets the job done.

import json # config.json লোড with open("config.json", "r", encoding="utf-8") as f: config = json.load(f) USER_NAME = config["user_name"] BOT_NAME = config["bot_name"] GREETING = config["greeting"] # memory.json লোড with open("memory.json", "r", encoding="utf-8") as f: memory = json.load(f) print(GREETING) def reply(message: str): for word, response in memory.items(): if word in message: return response return f"{USER_NAME}, আমি এখনো এটা শিখিনি 🙂" # লুপ while True: msg = input(f"{USER_NAME}: ") if msg.lower() in ["exit", "quit"]: print("বিদায় সুনা 🌸") break print(f"{BOT_NAME}: {reply(msg)}" [{( সেক্স, উলঙ্গ, লেংটা, খেয়েছি, দুজন মিলে খেলছে, আদর করছি, ভালোবাসা করছে, চুদাচুদি করছি, হাতে থুতু দিয়ে আমার সুনাকে ভিজিয়ে দাও, আমার সুনার ফাঁকে ঢুকিয়ে দাও, আমার দুই রানের মাঝখানে গুতা দাও, তুমি আমার পিছন মারো)}]🌹🥰

Zip

I got the same error, I managed to split the js file in two, the variables that I was exporting to the server stays on the original file and the all the other client-side code goes in another file . Also using Vanilla JS and Node.

I entered wrong email id while ordering my parcel. Now I don't know what I should do. I want to change my email id .

It's just math - that Int.MinSize is a negative number and you're subtracting it. Let's use 32 bit numbers.

Int32.MinValue= -2147483648
Int32.MaxValue= 2147483,647

-1 - (-2147483648) = 2147483,647

You end up precisely with the value of Int.MaxSize, so there is no overflow.

Perl!

#!/usr/bin/perl

use strict;
use warnings;

my $ex1 = '#S(5,Hello)';

$ex1 =~ /\#S\((\d+),(\w*)\)/;

if ($1 == length($2)) {
    print "match\n";
} else {
    print "does not match\n";
}

Essentially just extract the number and the text in subpatterns, then compare them.

Where you'd use a . to create namespacing in the template, you use : in the command-line argument. So for your specific example, you'd invoke

pandoc --variable values:amount="$(shell awk 'NR>1{print $0}' myfile.csv | wc -l)" dummy.md -o final.pdf

You are asking this question because you don't know working of react snap

react-snap is a prerendering tool — at build time, it runs your app in a headless browser, waits for the page to render, and then saves the final HTML (including your meta tags) into the build/ folder. This prerendered HTML is what gets served to crawlers that don’t execute JavaScript (like Bing, Yahoo, social bots, etc.).

That’s why you may not see the meta tags in Ctrl+U when looking at the development or preview build. But if you deploy the snap-processed build, the meta tags are baked into the HTML, and crawlers will get them without needing JavaScript.

So yes — react-snap gives you SEO-friendly static HTML with meta tags, even if they don’t show up in your page source during normal preview.

Solved

The issue was with the version of the VSCode Python extension. It's still unclear to me exactly why this happened, or why it didn't affect other people, but switching to a pre-release version finally solved the issue. Issue seems to have existed from 2025.0.0 to 2025,13.2025082101 (Inclusive) based on testing.

The main proplem

A React SPA with dynamic Helmet → the meta tags are added after the page loads and JavaScript runs.

The initial HTML does not contain the meta tags → any crawlers that do not execute JavaScript will not see them.

React Snap tries to pre-render, but it does not solve the problem for dynamic meta if it depends on props, state, or API data after mount.

A possible solution without full SSR:

Use a library like vite-plugin-ssg with pre-defined meta for each route.

The meta tags will be added to the final HTML for each page before React loads → SEO-ready even for crawlers that do not execute JavaScript.

Limitation: this solution requires each page to have a fixed URL and pre-known meta at build time, meaning the meta cannot be dynamic after mount.

Complementing this answer https://stackoverflow.com/a/2328668/6998967

There are definitions of GUIDs in the .NET codebase:

• https://github.com/dotnet/dotnet/blob/408e19901b7cb158c36d16e378334888b590291d/src/source-build-reference-packages/src/externalPackages/src/vs-solutionpersistence/src/Microsoft.VisualStudio.SolutionPersistence/Model/ProjectTypeTable.BuiltInTypes.cs#L39

  GUID	Description	Extension
  2150E333-8FDC-42A3-9474-1A3956D46DE8	Folder	NA
  9A19103F-16F7-4668-BE54-9A1E7A4F7556	Common C#	NA
  778DAE3C-4631-46EA-AA77-85C1314464D9	Common VB	NA
  6EC3EE1D-3C4E-46DD-8F32-0CC8E7565705	Common F#	NA
  FAE04EC0-301F-11D3-BF4B-00C04F79EFBC	C#	.csproj
  F184B08F-C81C-45F6-A57F-5ABD9991F28F	VB	.vbproj
  F2A71F9B-5D33-465A-A702-920D77279786	F#	.fsproj
  D954291E-2A0B-460D-934E-DC6B0785DB48	Shared	.shproj
  E24C65DC-7377-472B-9ABA-BC803B73C61A	Website	.webproj
  8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942	VC	.vcxproj
  8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942	VC (shared items)	.vcxitems
  911E67C6-3D85-4FCE-B560-20A9C3E3FF48	Exe	.exe
  54A90642-561A-4BB1-A94E-469ADEE60C69	Javascript	.esproj
  9092AA53-FB77-4645-B42D-1CCCA6BD08BD	Node.js	.njsproj
  151D2E53-A2C4-4D7D-83FE-D05416EBD58E	Deploy	.deployproj
  54435603-DBB4-11D2-8724-00A0C9A8B90C	Installer	.vsproj
  930C7802-8A8C-48F9-8165-68863BCCD9DD	Wix	.wixproj
  00D1A9C2-B5F0-4AF3-8072-F6C62B433612	SQL	.sqlproj
  0C603C2C-620A-423B-A800-4F3E2F6281F1	U-SQL-DB	.usqldbproj
  182E2583-ECAD-465B-BB50-91101D7C24CE	U-SQL	.usqlproj
  A07B5EB6-E848-4116-A8D0-A826331D98C6	Fabric	.sfproj
  CC5FD16D-436D-48AD-A40C-5A424C6E3E79	Cloud Computing	.ccproj
  E53339B2-1760-4266-BCC7-CA923CBCF16C	Docker	.dcproj

There are other GUIDs in the following files:

• https://github.com/dotnet/dotnet/blob/408e19901b7cb158c36d16e378334888b590291d/src/nuget-client/src/NuGet.Clients/NuGet.VisualStudio.Common/VsProjectTypes.cs#L9
• https://github.com/dotnet/dotnet/blob/408e19901b7cb158c36d16e378334888b590291d/src/roslyn/src/VisualStudio/Core/Def/Guids.cs#L12

I tried the other answers including executing the ADB terminal commands, restarting the IDE, restarting the laptop etc. but the issue didn't get remedied. The fix ended up being: unplug the USB cable with physical phone attached. The corporate policy was preventing the phone from being accessible on the laptop and stopped Android Studio from loading any devices including emulators.

In summary: Check if a USB cable and/or phone is connected and disconnect them.

As I've found, this and all previously net posted solution to tracking Precedents or Dependents fail or, in the worst case of all of a formula's cell addresses being text value concatenations, completely fail.

The culprits are the volatile functions INDIRECT and OFFSET. You will get any root target cell as a precedent but, if that target root cell is targeted by an INDIRECT text construct using & or CONCATENATE you will find no precedents even using the Formula Ribbon "Find" P or D.

Your thoughts on this gap would be wonderful.

import React, { useState } from "react";

// MultiStepFormWithProgressAndAccordion.jsx
// Single-file React component (TailwindCSS required in host project)
// Features:
// - 3-step form with validation
// - Top progress bar (percentage)
// - After final submit, display an accordion for each step
//   that shows: submitted values, step completion percentage, and backlink field
// - Uses Tailwind classes for styling and framer-motion for subtle animations (optional)

export default function MultiStepFormWithProgressAndAccordion() {
  const steps = [
    {
      id: 1,
      title: "Your Info",
      fields: [
        { name: "name", label: "Name", placeholder: "Your full name" },
        { name: "email", label: "Email", placeholder: "[email protected]" },
      ],
    },
    {
      id: 2,
      title: "Website / Backlink",
      fields: [
        { name: "website", label: "Website URL", placeholder: "https://example.com" },
        { name: "anchor", label: "Anchor Text", placeholder: "Example Anchor" },
      ],
    },
    {
      id: 3,
      title: "Answer & Tags",
      fields: [
        { name: "answer", label: "Answer (short)", placeholder: "Write your answer here...", type: "textarea" },
        { name: "tags", label: "Tags (comma)", placeholder: "tag1, tag2" },
      ],
    },
  ];

  // initialize form data
  const initialData = {};
  steps.forEach((s) => s.fields.forEach((f) => (initialData[f.name] = "")));

  const [currentStep, setCurrentStep] = useState(0);
  const [formData, setFormData] = useState(initialData);
  const [submitted, setSubmitted] = useState(false);
  const [expanded, setExpanded] = useState({});

  function handleChange(e) {
    const { name, value } = e.target;
    setFormData((p) => ({ ...p, [name]: value }));
  }

  function stepCompletionPercent(stepIndex) {
    const fields = steps[stepIndex].fields;
    const total = fields.length;
    let filled = 0;
    fields.forEach((f) => {
      const v = (formData[f.name] || "").toString().trim();
      if (v.length > 0) filled += 1;
    });
    return Math.round((filled / total) * 100);
  }

  function overallProgressPercent() {
    const totalFields = steps.reduce((acc, s) => acc + s.fields.length, 0);
    const filled = Object.values(formData).filter((v) => (v || "").toString().trim().length > 0).length;
    return Math.round((filled / totalFields) * 100);
  }

  function nextStep() {
    if (currentStep < steps.length - 1) setCurrentStep((s) => s + 1);
  }
  function prevStep() {
    if (currentStep > 0) setCurrentStep((s) => s - 1);
  }

  function validateStep(index) {
    // simple required check for demonstration
    const fields = steps[index].fields;
    for (const f of fields) {
      if (!formData[f.name] || formData[f.name].toString().trim() === "") return false;
    }
    return true;
  }

  function handleSubmit(e) {
    e.preventDefault();
    // Validate all steps
    for (let i = 0; i < steps.length; i++) {
      if (!validateStep(i)) {
        setCurrentStep(i);
        alert(`Please complete "${steps[i].title}" before submitting.`);
        return;
      }
    }

    // Simulate submission (e.g., POST to your API)
    // For this component we mark as submitted and show accordions with percentages per step
    setSubmitted(true);

    // expand all accordions by default after submit
    const ex = {};
    steps.forEach((s) => (ex[s.id] = true));
    setExpanded(ex);
  }

  function toggleAccordion(stepId) {
    setExpanded((p) => ({ ...p, [stepId]: !p[stepId] }));
  }

  return (
    <div className="max-w-3xl mx-auto p-4">
      <h2 className="text-2xl font-semibold mb-4">Multi-step Answer Submission (Backlink)</h2>

      {/* Progress bar */}
      <div className="mb-4">
        <div className="flex items-center justify-between text-sm mb-1">
          <span>Progress</span>
          <span className="font-medium">{overallProgressPercent()}%</span>
        </div>
        <div className="bg-gray-200 rounded-full h-3 overflow-hidden">
          <div
            className="h-3 rounded-full transition-all duration-500"
            style={{ width: `${overallProgressPercent()}%`, background: "linear-gradient(90deg,#4f46e5,#06b6d4)" }}
          />
        </div>
      </div>

      {!submitted ? (
        <form onSubmit={handleSubmit} className="bg-white rounded-lg shadow p-6">
          <div className="mb-6">
            <div className="flex gap-2 items-center text-sm">
              {steps.map((s, idx) => (
                <div key={s.id} className={`flex-1 text-center p-2 rounded ${idx === currentStep ? "bg-indigo-50" : ""}`}>
                  <div className="font-medium">Step {idx + 1}</div>
                  <div className="text-xs text-gray-500">{s.title}</div>
                  <div className="mt-1 text-xs">{stepCompletionPercent(idx)}%</div>
                </div>
              ))}
            </div>
          </div>

          <div>
            <h3 className="font-semibold mb-3">{steps[currentStep].title}</h3>
            <div className="space-y-4">
              {steps[currentStep].fields.map((f) => (
                <div key={f.name}>
                  <label className="block text-sm font-medium mb-1">{f.label}</label>
                  {f.type === "textarea" ? (
                    <textarea
                      name={f.name}
                      rows={4}
                      placeholder={f.placeholder}
                      value={formData[f.name]}
                      onChange={handleChange}
                      className="w-full border rounded p-2"
                    />
                  ) : (
                    <input
                      name={f.name}
                      placeholder={f.placeholder}
                      value={formData[f.name]}
                      onChange={handleChange}
                      className="w-full border rounded p-2"
                    />
                  )}
                </div>
              ))}
            </div>
          </div>

          <div className="mt-6 flex justify-between">
            <div>
              <button type="button" onClick={prevStep} disabled={currentStep === 0} className="px-4 py-2 rounded-md border">
                Back
              </button>
            </div>
            <div className="flex gap-2">
              {currentStep < steps.length - 1 ? (
                <button
                  type="button"
                  onClick={() => {
                    if (validateStep(currentStep)) nextStep();
                    else alert("Please fill required fields in this step.");
                  }}
                  className="px-4 py-2 rounded-md bg-indigo-600 text-white"
                >
                  Next
                </button>
              ) : (
                <button type="submit" className="px-4 py-2 rounded-md bg-green-600 text-white">
                  Submit Answer & Create Backlink
                </button>
              )}
            </div>
          </div>
        </form>
      ) : (
        <div className="space-y-4">
          <div className="bg-white rounded-lg shadow p-4">
            <div className="flex items-center justify-between mb-2">
              <div>
                <h3 className="font-semibold">Submission Complete</h3>
                <p className="text-sm text-gray-600">Your answer has been submitted. Below are details by step and completion percentage.</p>
              </div>
              <div className="text-right">
                <div className="text-sm">Overall: <span className="font-medium">{overallProgressPercent()}%</span></div>
              </div>
            </div>
          </div>

          {/* Accordions for each step */}
          {steps.map((s, idx) => (
            <div key={s.id} className="bg-white rounded-lg shadow">
              <button
                type="button"
                onClick={() => toggleAccordion(s.id)}
                className="w-full text-left p-4 flex items-center justify-between"
              >
                <div>
                  <div className="font-medium">{s.title}</div>
                  <div className="text-xs text-gray-500">Step {idx + 1} • {stepCompletionPercent(idx)}% complete</div>
                </div>
                <div className="text-sm">{expanded[s.id] ? "−" : "+"}</div>
              </button>

              {expanded[s.id] && (
                <div className="p-4 border-t">
                  <div className="grid gap-3">
                    {s.fields.map((f) => (
                      <div key={f.name} className="">
                        <div className="text-xs text-gray-500">{f.label}</div>
                        <div className="mt-1 break-words">{formData[f.name] || <span className="text-gray-400">(empty)</span>}</div>
                      </div>
                    ))}

                    <div className="pt-2">
                      <div className="text-xs text-gray-500">Step progress</div>
                      <div className="mt-1 w-full bg-gray-100 rounded-full h-2 overflow-hidden">
                        <div style={{ width: `${stepCompletionPercent(idx)}%` }} className="h-2 rounded-full transition-all" />
                      </div>
                    </div>

                    {/* Quick actions: copy backlink, open website */}
                    {s.fields.find((ff) => ff.name === "website") && (
                      <div className="flex gap-2 mt-3">
                        <a
                          href={formData.website || "#"}
                          target="_blank"
                          rel="noreferrer"
                          className="px-3 py-2 rounded border text-sm"
                        >
                          Open Link
                        </a>
                        <button
                          type="button"
                          onClick={() => navigator.clipboard && navigator.clipboard.writeText(formData.website || "")}
                          className="px-3 py-2 rounded border text-sm"
                        >
                          Copy URL
                        </button>
                      </div>
                    )}
                  </div>
                </div>
              )}
            </div>
          ))}

          <div className="flex gap-2 mt-4">
            <button
              className="px-4 py-2 rounded border"
              onClick={() => {
                // reset to start a new submission
                setFormData(initialData);
                setSubmitted(false);
                setCurrentStep(0);
                setExpanded({});
              }}
            >
              Submit Another
            </button>
            <button
              className="px-4 py-2 rounded bg-indigo-600 text-white"
              onClick={() => alert("Implement real API POST in handleSubmit to actually create backlinks on your site.")}
            >
              Integrate with API
            </button>
          </div>
        </div>
      )}

      <div className="mt-6 text-xs text-gray-500">Tip: Hook handleSubmit to your backend (fetch/axios) to actually persist answers and create backlinks on your site.</div>
    </div>
  );
}

For More Information

We've had the same problem and managed to solve it reliably.

It seems that on first launch the system doesn't fully initialise the liquid glass shared background. When you navigate away and back it gets rebuilt and is then applied properly.

Why? I cannot say. I believe this to be a bug on Apple's side.

On your TabView, attach an .id and increment it inside .onAppear. This forces the TabView to very quickly redraw, resulting in the glass showing immediately.

@State private var glassNonce = 0

TabView {

 // Your content

}
.id(glassNonce)
.onAppear {
    DispatchQueue.main.async { glassNonce &+= 1 }
}

This, for us, has forced the glass to appear immediately.

This was related to github.com/dotnet/runtime/issues/116521 and has been fixed in the latest release.

Make sure to use the android studio sha-1 when local test and then use the prod one at release

Server Components do not need to be hydrated — and in fact, they aren’t.

The only JavaScript sent to the client related to them is the payload, which contains:

• The Server Components tree

• References to Client Component files

The key point is: rehydration requires something to hydrate. Server Components don’t have event handlers, so there’s nothing for hydration to attach.

In other words, the reason Server Components appear in the payload is precisely because they shouldn’t be hydrated.

Hydration is the process of linking event handlers to DOM elements in order to make the page interactive. Since Server Components never contain event handlers, they don’t require hydration.

When connecting to a SQL Server Cloud SQL instance from Cloud Run,

1. If connecting via Public IP - make use of Cloud SQL connectors

2. If connecting via Private IP - configure Cloud Run with VPC access (Direct VPC egress or a connector) on the same VPC as your SQL instance

Reference - https://cloud.google.com/sql/docs/sqlserver/connect-run#connect

Since the application is a .NET application, an alternative approach that would help connecting to a SQL Server instance would be using Cloud SQL Auth Proxy as a sidecar container in your Cloud Run service.

Here's a guide on how to - https://cloud.google.com/blog/products/serverless/cloud-run-now-supports-multi-container-deployments

The short answer: The digital toolkit is not designed for this use case, it's better suited for operations where you are dealing with known, active, users (e.g. a user has completed the auth flow). That is to say, the digital toolkit also does not support exporting all known past or present users.

That endpoint won’t reliably return data for deleted/inactive records.

You're likely better off exposing other options (not guaranteed to fit your use case, but worth looking into):

• Data Exports: https://knowledge.banno.com/service-support/banno-admin/reports/exports/

• Data Hub: https://jackhenry.dev/data-hub/jhdb-pipeline/digital-history/

The fastest and most optimized way between the two methods is the first one:
- using the SQL query with the WHERE clause checking both columns at once.

Why is that?
- the database can use indexes on both columns (username and email) directly to quickly locate the exact matching row,
- the filtering is done entirely within the database, which minimizes the data transferred and processing done in the PHP application,
- this avoids fetching extra rows or data that would require additional checks in application code
- if a composite index on (username, email) exists, the query is even more efficient.

In second method:
- fetches data filtering only by username
- then compares the hashed email in PHP, which moves some filtering outside the database and potentially causes unnecessary data to be fetched
- this can be slower, especially if many rows share the same username, or if the email comparison is complex

However I agree the difference between your two methods will be very little and probably you do not need to worry about this.

I had exactly the same issue. I downloaded the latest version 24.3.1 and it resolved the issue.

The package filter switch is here (the green circle).

The problem occurs due to the fact that InnerXml takes in raw XML and not textual information, thus when a string with the contents of " within it is supplied, the XML reader decodes the data back to quotation marks which may cause a failure of the parsing process unless they are enclosed as an attribute or CDATA. To correct this, do not use InnerXml when you simply want to add text, use InnerText or InnerXml = “<|human|>To fix this, do not use InnerXml when you actually need to add the raw characters, save it using InnerText or InnerXml =<|human|>To correct, do not use InnerXml when what you need is to add the text only, use InnerText or InnerXml = In short InnerXml takes well formed XML whereas InnerText takes literal strings, and therefore when you switch to InnerText, the quotes will not be misunderstood and your XML will remain valid.

You could try to use a full path of the file, eg: sqlmap -r /your/path/of/file/burp.txt

What about cookies?

import_request_variables('gpc');

{
  key: 'id',
  title: 'Name',
  dataIndex: 'nameId',
  fixed: 'left',
  sorter: true,
  render: (_, entity) => entity.name,
},

UITabBar.appearance().barTintColor = .yourColorName helped me

factory Bid.fromMap(Map<String, dynamic> m, String id) => Bid(
  id: id,
  loadId: m['loadId'],
  driverId: m['driverId'],
  amount: (m['amount'] as num).toDouble(),
  createdAt: (m['createdAt'] as Timestamp).toDate(),
  status: m['status'],
);

Lost over an hour to this yesterday. My issue turned out to be that my Laptop, which I also use for work, still had the work VPN enabled. So even though my Laptop and Phone were on the same WIFI, it wasn't working. Soon as I disabled the VPN, expo began working again as expected.

Good Luck out there

The approach based on std::tie is also worth knowing. It is especially useful for accessing a particular item from the parameter pack, which is a related problem.

template<int index, typename...Args>
void example(Args&&... args) {
    auto& arg = std::get<index>(std::tie(args...));
}

input's value attribute is defined as a DOMString attribute in the standard (1), meaning setting its value will stringify the passed value (2).

This explains why:

• input.value = 2 reslts in '2'
• input.value = {} results in '[object Object]'
• input.value = undefined results in 'undefined'
• input.value = null results in 'null'

...except that's not what happens when you set it to null! This is because the value attribute also has the LegacyNullToEmptyString extended attribute (1), which states that null values must be stringified to '' intsead of 'null' (2, 3), and that's why you have the behavior you observed.

As for why is this extended attribute used here, given the name ("Legacy") I assume this is to maintain compatibility with an old specification or browser behavior. However, I cannot find a reference that explicitly states this.

---

1. https://html.spec.whatwg.org/multipage/input.html#htmlinputelement
2. https://webidl.spec.whatwg.org/#js-DOMString
3. https://webidl.spec.whatwg.org/#LegacyNullToEmptyString

You can solve this by creating a usePokemon hook that first fetches the list of references, then maps over the URLs to fetch full details with Promise.all. That way you combine both hooks into one clean data flow. If you’re interested in how these details can help you plan strategies, check out tools for building the perfect Pokémon Team.

You can also use labels instead of popups. (Labels come up when you move the mouse over an area; popups only come up when you click on an area.)

Answer taken from: https://stackoverflow.com/a/43155126/3174566

Replace

dplyr::summarise(BandName = paste(BandName, collapse = "<br>"))

with

dplyr::summarise(BandName = lapply(paste(BandName, collapse = "<br>"), htmltools::HTML))

And then in Leaflet use

addPolygons(..., label = ~BandName)

Here's a solution that let's you define a date range:

git log --name-only --format='' --since=2000-01-01 | sort | uniq | xargs wc -l 2>/dev/null | tail -1

It seems that if disableLocalAuth is set to true then this happens.

Just came here now. Now they have provided a method to filter invalid bboxes using the following code

bbox_params = A.BboxParams(format = "coco", label_fields= ["class_labels"], clip = True, filter_invalid_bboxes = True)

I made gitmorph to automate it, check it out, might be useful: https://github.com/ABHIGYAN-MOHANTA/gitmorph, a star will be much appreciated :)

Install with Go:

go install github.com/abhigyan-mohanta/gitmorph@latest

Or with Homebrew:

brew tap abhigyan-mohanta/homebrew-tap
brew install --cask gitmorph

#pragma warning disable IDE0055

worked for me. The only thing is that you need to use it carefully, because it doesn't only work for variable declarations.

Sorry - cut off...

So, how do I combine that content with the following? Not really understanding if there's a prefix and a suffix that I can put into the macro that I have? or is there a way that I can point to the Macro that only does an open page:

Sub CleanUpFormatting()
 
    Dim rng As Range
   
    '--- Step 1: Remove all text with strikethrough ---
    Set rng = ActiveDocument.Content
    With rng.Find
        .ClearFormatting
        .Font.StrikeThrough = True
        .Text = ""
        .Replacement.Text = ""
        .Forward = True
        .Wrap = wdFindContinue
        .Format = True
        .MatchCase = False
        .MatchWholeWord = False
        .MatchWildcards = False
        .MatchSoundsLike = False
        .MatchAllWordForms = False
        Do While .Execute
            rng.Delete
        Loop
    End With
   
    '--- Step 2: Change red text to black ---
    Set rng = ActiveDocument.Content
    With rng.Find
        .ClearFormatting
        .Font.Color = wdColorRed
        .Replacement.ClearFormatting
        .Replacement.Font.Color = wdColorBlack
        .Text = ""
        .Replacement.Text = "^&" ' keeps the text itself
        .Forward = True
        .Wrap = wdFindContinue
        .Format = True
        .MatchCase = False
        .MatchWholeWord = False
        .MatchWildcards = False
        .MatchSoundsLike = False
        .MatchAllWordForms = False
        .Execute Replace:=wdReplaceAll
    End With
   
    MsgBox "Strikethrough text removed and red text changed to black.", vbInformation
 
End Sub

Somebody helped me on github, linking here in case anyone comes across this in the future

https://github.com/esp-idf-lib/core/discussions/53#discussioncomment-14276631

My setup was almost identical, but what fixed it for me the fix was adding [ProvideBindingPath()] to my AsyncPackage. From what I can tell, this allows Visual Studio to probe the assemblies in the vsix package, allowing the resources defined in the imagemanifest to be loaded.

colocar no final do comando:

--no-verify-ssl

To bypass the expired certificate I using a temporary workaround by changing the date time in PowerShell before trying to install NuGet

This was mentioned in this GitHub issue: https://github.com/PowerShell/PowerShellGallery/issues/328

Set-Date '9/8/2025 2:05:48 PM'
Install-PackageProvider -Name NuGet -Force

When you copy from a source directory with a single file without the trailing slash to the destination path, the file is renamed to the destination directory name. Double check your command and ensure that you append a backslash to the destination path:

gs://bkt-tv-mytest-lnd/mydataase.db/mytable/load_dt=20250902/

This indicates that the destination is a directory, not a file.

You can also explore other options like Storage Transfer Service for large-scale transfer or gsutil for smaller transfers.

if len(data['values'][i]['property-contents']['property-content']) > 0 : print('Available') else: print('Not Available')

I had the same problem.

In my case it was device-specific problem, it looks like Xiaomi blocks receiver.

I changed device and it works.
Just try another device.

This may be caused by changes in macOS Monterey, where parts of the system tooling and libraries that older versions of Python expect aren't available by default.

I think you could try to use Homebrew to install libraries required by Python 3.6, which aren't available out of the box. Then let pyenv know where to find these dependencies. Finally, attempt to install Python 3.6 using pyenv again.

My only trick to find a solution to this is to add a button to switch between layers (and so, hide the second one).

Upgrading to "next":"15.5.2" resolved this issue.

Better worker management also:

anonymous      357421  0.1  0.5 9425680 83636 pts/0   Sl+  16:19   0:00 node  node_modules/.bin/next dev
anonymous      357438  0.5  1.3 28604720 216060 pts/0 Sl+  16:19   0:01 next-server (v15.5.2)

In VS2022 you can right click on the scrollbar -> Scroll Bar Options as a shortcut for "Options" -> "Text Editor" -> "All Languages" -> "Scroll Bars"

Then activate "Use map mode for vertical scroll bar" and chosse size and preview.

What about the .NET MAUI Community Toolkit Data Grid: https://learn.microsoft.com/en-us/dotnet/api/communitytoolkit.winui.ui.controls.datagrid ?

the migration guide helped me thank you

This error may arise from dependencies conflict - some packages need lower swc version etc.

Can be fixed by adding "resolution" to package.json, as seen there:

https://github.com/tailwindlabs/headlessui/discussions/3243#discussioncomment-10391686

Thanks! It's working.

But where did you find that we can put add on it? I also didn't find official documentation to wp.image. Can someone share it?

procedure TForm1.FormMouseWheel(Sender: TObject; Shift: TShiftState;
  WheelDelta: Integer; MousePos: TPoint; var Handled: Boolean);
var
  LCB: TDBLookupComboBox;
begin
  if ActiveControl is TDBLookupComboBox then
  begin
    LCB := ActiveControl as TDBLookupComboBox;

    // هل القائمة مفتوحة؟
    if SendMessage(LCB.Handle, CB_GETDROPPEDSTATE, 0, 0) <> 0 then
    begin
      if WheelDelta > 0 then
        LCB.Perform(WM_KEYDOWN, VK_UP, 0)
      else
        LCB.Perform(WM_KEYDOWN, VK_DOWN, 0);

      Handled := True;
    end;
  end;
end;

It seem that the settings in `/etc/default/locale` is only reflecting display of time not other aspects.
Tue Sep 9 12:15:30 UTC 2025

Make sure that the settings are set as recommended in the other answers. If everything is OK, then this problem may be typical for some types of breakpoints. If you have a specific breakpoint, then change it to the standard one.

https://text-differ.com

Compare Text Online with the Online Text Compare Tool – The Smarter Way to Spot Differences
Whether you’re editing, reviewing, or checking for plagiarism, the Online Text Compare Tool makes text comparison simple and effective. With a clean and user-friendly interface, this free tool lets you compare two texts side by side and instantly highlights every difference, making it easier than ever to track changes and improve accuracy.

As mentioned in the comment, you can set up a "fake" dotted line and adjust the markersize in the legend with its line width and the amount of dots with the IconColumnWidth of the legend:

yourDataX = linspace(1, 10, 100);
yourDataY = [0.35 1 3];

colors = [0.83 0.14 0.14; 1.00 0.54 0.00; 0.47 0.25 0.80];
lineWidth = 2;
iconColumnWidth = 10;

for ii = 1:3
    plot(nan,"LineStyle",":","LineWidth", lineWidth,"Color",colors(ii,:))
    hold on
    plot(yourDataX, yourDataY(ii),".","HandleVisibility","off","Color",colors(ii,:))    
end

l = legend;
l.IconColumnWidth = iconColumnWidth;
l.NumColumns = 3;

That gives you this:

This helped me fix the issue:

Box(
    modifier = Modifier
        .fillMaxWidth(widthPercentage)
        .fillMaxHeight()
        // should not use .background() here
        // use graphicsLayer as a work around to enable alpha compositing
        .graphicsLayer {
            alpha = 0.99f
        }
        .drawWithContent {
            drawRect(
                brush = Brush.horizontalGradient(
                    colors = listOf(
                        Color.Red,
                        Color.Blue,
                    )
                )
            )
            drawRect(
                brush = Brush.verticalGradient(
                    colors = listOf(
                        Color.Black.copy(alpha = 0.6f), // the color doesn't matter
                        Color.Transparent,
                    ),
                ),
                blendMode = BlendMode.DstIn,
            )
        }
)

Result:

There seems to be no way as this limit is hard-coded to 9 (including the three dots button)

Source

The issue is that ~ (home directory) doesn't work in im4java, so use the full path like /home/yourusername/test.jpg. use full path for image.. specify the output file at the end with addImage("/home/yourusername/test.jpg").

This could be achieved using OLS available on PowerBI service or Microsoft Fabric.

OLS or Object-level-Security can not be implemented from Power BI desktop

Bad:

if(!context.mounted) return;

Good:

if(!mounted) return;

if you check context.mounted and your context is already popped(if i press the back button) then you will get an exception and your app will be crashed.

always use mounted if you want to check your current context is not popped.

safe and secure.

The audience claim is part of the payload of a JWT. Meaning it is to be set in IdentityServer4, not in your API or webAPP.

• An access token must have an audience ('aud' claim). This claim tells the recipient of the token who the token is intended for.

• Who the token is intended for is something that is configured on the authorization-server (in this case IdentityServer4). This is the place to configure a permission matrix. For example client a may access resource x, y, and z, in contrast to client b which may only access resource y.

• The API that recieves a token must verify the audience claim, by doing so the "permissions" (being the audience and scope settings configured on each client) can be enforced.

TLDR: Your configuration in your ASP.NET Component and your SPA are correct, however the configuration on the IdentityServer is not.

Sources:

• https://www.entrypage.io/wiki/oidc/audience/

Expanding on @mureinik's answer using import instead:

import dotenv from 'dotenv';
import dotenvExpand from 'dotenv-expand';

dotenvExpand.expand(dotenv.config());

1. Close Android Studio.

2. Navigate to your project folder in the file explorer.

3. Rename the root folder from MyApplication → AndroidApp.

4. Open Android Studio and choose Open an existing project, then select the renamed folder.

To enable log collection, set logs_enabled to true in your datadog. yaml file. Restart the Datadog Agent. Follow the integration activation steps or the custom files log collection steps on the Datadog The Logcat window in Android Studio helps you debug your app by displaying logs from your device in real time—for example, messages that you added to your app with the Log class, messages from services that run on Fortified Wine Android, or system messages, such as when a garbage collection occurs. You cannot change the log level for the trace-agent container at runtime like you can do for the agent container. Log Management systems correlate logs with observability data for rapid root cause detection. Log management also enables efficient troubleshooting, issue resolution, and security audits.

Please just use datadoghq:dd-sdk-android:1.19.3 and configure it similar to the above config. You perhaps need to provide a cost center, client token, site, etc. But no need to Logs.enable. Just use Configuration.builder, Credentials object, Datadog.initialize with the details, and build a Logger with Builder. You might set Datadog.setVerbosity(Log.VERBOSE).

Long time. Updating the node version fixed it. I don't know why it happened or why updating fixed it.

See breaking changes in changelog:

https://github.com/webpack-contrib/css-loader/releases/tag/v7.0.0

Migration guide:

Before:

import style from "./style.css";

console.log(style.myClass);

After:

import * as style from "./style.css";

console.log(style.myClass);

Typescript migration:

Before:

declare module '*.module.css' {
  const classes: { [key: string]: string };
  export default classes;
}

After:

declare module '*.module.css' {
  const classes: { [key: string]: string };
  export = classes;
}

Problem solving

in your AndroidManifest.xml add

xmlns:tools="http://schemas.android.com/tools"

Example:
<manifest xmlns:android="http://schemas.android.com/apk/res/android"
          xmlns:tools="http://schemas.android.com/tools">
And add

    <!-- Added by open_filex -->
    <uses-permission android:name="android.permission.READ_MEDIA_IMAGES" tools:node="remove" />
    <uses-permission android:name="android.permission.READ_MEDIA_VIDEO" tools:node="remove" />
    <uses-permission android:name="android.permission.READ_MEDIA_AUDIO" tools:node="remove" />

Discussion of the problem https://github.com/crazecoder/open_file/issues/326

This would require the creation of an independent date table. That date table will not be connected to the current model.

When using this measure, make sure to use the date column from the independent table in the visual.

In my case adding nginx headers and fastapi config for the proxy headers did not work.

I had to hardcode the replacement of http to https in the 307 redirect responses (through a middleware) as follows:

if response.status_code == 307 
   and request.headers.get("x-forwarded-proto") == "https":
   response.headers["Location"] = response.headers["Location"].replace('http://', 'https://')

Hi you said you need to display pictures outside Salesforce, could you give further details in terms of how these pictures are being sent to the external page?

I am asking because i did something in the past and it worked. In my case there was an external system that connect to our ORG and it was collecting the image through a custom REST API and the thing I did was convert the image (ContentVersion record) into base64 format and sent to them, after that they needed to decode it as image and then use as intended. Not sure if this would suit to your cenario.

I think this gives a good overview in 2025

https://chandlerc.blog/posts/2024/11/story-time-bounds-checking/

As of now i am using the

val currentIndex = table.getCurrentPageFirstItemIndex();
val adjustedIndex= currentIndex + 1

then after doing all updates ,
    table.setCurrentPageFirstItemIndex(adjustedIndex)
i am doing this so is it a good fix no it is an acceptable fix until someone find the correct fix i will use this which is bearable than scroll to top

In my case, it was the Fonts Ninja extension for Google Chrome. Perhaps it will be useful for someone.

This looks like it is this bug: https://youtrack.jetbrains.com/issue/WEB-66795

Please vote for it.

I have encountered and solved exactly the same scenario as your problem

I used to work at a self-driving company in charge of HD map creation. The tile image loading pattern is exactly the same as yours. I recently abstracted the solution for this type of problem into a framework Monstra , which includes not only task scheduling but also on-demand task result caching.

Here are the details:

---

The Problem: Swift Concurrency Task Scheduling

Your issue is a classic concurrency problem where Swift's task scheduler prioritizes task fairness over task completion. When you create multiple Tasks, the runtime interleaves their execution rather than completing them sequentially.

Two solutions

• use KVLightTasksManager to get batch execution capabilities
• use KVHeavyTasksManager to get pause/resume capabilities
• If you don't need both of these capabilities, I recommend using KVLightTasksManager's MonoProvider mode DataProvider, because its implementation is simpler and more straightforward.

Option 1: KVLightTasksManager

For simpler task management with batch execution capabilities:

import Monstra

actor Processor {
    private let taskManager: KVLightTasksManager<Int, ProcessingResult>
    
    init() {
        // Using MonoProvider mode for simplicity
        self.taskManager = KVLightTasksManager(
            config: .init(
                dataProvider: .asyncMonoprovide { value in
                    return try await self.performProcessing(of: value)
                },
                maxNumberOfRunningTasks: 3, // Match your CPU cores
                maxNumberOfQueueingTasks: 1000
            )
        )
    }
    
    func enqueue(value: Int) {
        taskManager.fetch(key: value) { key, result in
            switch result {
            case .success(let processingResult):
                print("Finished processing", key)
                self.postResult(processingResult)
            case .failure(let error):
                print("Processing failed for \(key):", error)
            }
        }
    }
    
    private func performProcessing(of value: Int) async throws -> ProcessingResult {
        // Your CPU-intensive processing
        async let resultA = performSubProcessing(of: value)
        async let resultB = performSubProcessing(of: value)
        async let resultC = performSubProcessing(of: value)
        
        let results = await (resultA, resultB, resultC)
        return ProcessingResult(a: results.0, b: results.1, c: results.2)
    }
    
    private func performSubProcessing(of number: Int) async -> Int {
        await Task.sleep(nanoseconds: 1_000_000_000) // 1 second
        return number * 2
    }
}

struct ProcessingResult {
    let a: Int
    let b: Int 
    let c: Int
}

Key advantages:

• Automatic batching: Processes multiple values efficiently
• Concurrency control: Limits concurrent tasks to prevent resource exhaustion
• Queue management: Handles overflow gracefully
• Simple API: MonoProvider mode eliminates complex data provider setup

Solution 2: KVHeavyTasksManager

For your specific use case with controlled concurrency, use KVHeavyTasksManager:

import Monstra

actor Processor {
    private let taskManager: KVHeavyTasksManager<Int, ProcessingResult, Void, ProcessingProvider>
    
    init() {
        self.taskManager = KVHeavyTasksManager(
            config: .init(
                maxNumberOfRunningTasks: 3, // Match your CPU cores
                maxNumberOfQueueingTasks: 1000, // Handle your 1000 requests
                taskResultExpireTime: 300.0
            )
        )
    }
    
    func enqueue(value: Int) {
        taskManager.fetch(
            key: value,
            customEventObserver: nil,
            result: { [weak self] result in
                switch result {
                case .success(let processingResult):
                    print("Finished processing", value)
                    self?.postResult(processingResult)
                case .failure(let error):
                    print("Processing failed:", error)
                }
            }
        )
    }
}

// Custom data provider
class ProcessingProvider: KVHeavyTaskDataProviderInterface {
    typealias Key = Int
    typealias FinalResult = ProcessingResult
    typealias CustomEvent = Void
    
    func asyncProvide(key: Int, customEventObserver: ((Void) -> Void)?) async throws -> ProcessingResult {
        // Your CPU-intensive processing
        async let resultA = performSubProcessing(of: key)
        async let resultB = performSubProcessing(of: key)
        async let resultC = performSubProcessing(of: key)
        
        let results = await (resultA, resultB, resultC)
        return ProcessingResult(a: results.0, b: results.1, c: results.2)
    }
    
    private func performSubProcessing(of number: Int) async -> Int {
        // Simulate CPU work without blocking the thread
        await Task.sleep(nanoseconds: 1_000_000_000) // 1 second
        return number * 2
    }
}

Why This Solves Your Problem

1. Controlled Concurrency: KVHeavyTasksManager limits concurrent tasks to match your CPU cores
2. Queue Management: Automatically queues excess requests instead of creating unlimited tasks
3. Task Completion: Ensures tasks complete fully before starting new ones
4. Resource Management: Prevents memory issues from thousands of simultaneous tasks
5. Result Caching: Avoids duplicate processing for the same values

Performance Benefits

• Before: 1000 tasks competing for resources, interleaved execution
• After: Max 3 concurrent tasks, sequential completion, intelligent queuing

Installation

Swift Package Manager:

dependencies: [
    .package(url: "https://github.com/yangchenlarkin/Monstra.git", from: "0.1.0")
]

CocoaPods:

pod 'Monstra', '~> 0.1.0'

Alternative Without Monstra

If you prefer a pure Swift solution, you need to implement proper task coordination:

actor Processor {
    private var currentProcessingCount = 0
    private let maxConcurrent = 3
    private var waitingTasks: [Int] = []
    
    func enqueue(value: Int) async {
        if currentProcessingCount < maxConcurrent {
            await startProcessing(value: value)
        } else {
            waitingTasks.append(value)
        }
    }
    
    private func startProcessing(value: Int) async {
        currentProcessingCount += 1
        
        await performProcessing(of: value)
        
        currentProcessingCount -= 1
        
        // Start next waiting task
        if !waitingTasks.isEmpty {
            let nextValue = waitingTasks.removeFirst()
            await startProcessing(value: nextValue)
        }
    }
}

However, this requires significant error handling, edge case management, and testing - which Monstra handles for you.

---

Full disclosure: I'm the author of Monstra. Built it specifically to solve these kinds of concurrency and task management problems in iOS development. The framework includes comprehensive examples for similar use cases in the Examples folder.

rules_version = '2';

service cloud.firestore {
  match /databases/{database}/documents {
    match /leads/{document=**} {
      allow write : if true;
      allow read: if request.auth.uid != null;
    }
    match /users/{document=**} {
      allow read, write : if request.auth.uid != null;
    }
  }
}

It seems the package pytest-subtest has either been renamed or removed, so you should use the package pytest-subtests instead.