Simple one : if you want to save df.sortvalue(...) write :

df = df.sortvalue(...)

Thanks for sharing this, very helpful. I was running into the same issue on OL9 ARM and wasn’t sure if quadlet support was missing or just not working correctly. Good tip on using podman-system-generator --user --dryrun it to check what’s going on. Looks like I also had a small syntax mistake in my file. Appreciate the clarification!

You are missing the Paid Apps Agreement.

Under your Agreements table you must have both agreements to serve IAP.

• Free Apps Agreement

• Paid Apps Agreement <-- missing

Even if your app is Free on the store, for IAP, the agreement is the same as for paid apps.

Simple, Menu Bar -> Show View -> Other and serach for coverage and add to you console window and cancel the selection of coverage and if you wnat it again you can relaunch by clicking on it

Is this what you are looking for? I have written this so it saves those strings after the url in a txt file, feel free to expand if useful.

import re

#here i copied ur links as example
js_block = r"""
{
    id: 'heic2018b',
    title: 'Galaxy NGC 2525',
    width: 3657,
    height: 3920,
    src: 'https://cdn.esahubble.org/archives/images/thumb300y/heic2018b.jpg',
    url: '/images/heic2018b/',
    potw: ''
},

{
    id: 'potw1345a',
    title: 'Antennae Galaxies reloaded',
    width: 4240,
    height: 4211,
    src: 'https://cdn.esahubble.org/archives/images/thumb300y/potw1345a.jpg',
    url: '/images/potw1345a/',
    potw: '11 November 2013'
},

{
    id: 'heic0817a',
    title: 'Magnetic monster NGC 1275',
    width: 4633,
    height: 3590,
    src: 'https://cdn.esahubble.org/archives/images/thumb300y/heic0817a.jpg',
    url: '/images/heic0817a/',
    potw: ''
},
"""

#regex to capture what's inside url: '...'
pattern = r"url\s*:\s*'([^']+)'"
matches = re.findall(pattern, js_block)

#write each match to urls.txt
with open('after_urls.txt', 'w') as out_file:
    for path in matches:
        out_file.write(path + '\n')

print(f"Wrote {len(matches)} after URLs to after_urls.txt")

cheers!

I got OP error, fixed it by downloading (from DefinitelyTyped) not just JQuery.d.ts but downloading the other .d.ts in that directory (at time of this post it was JQueryStatic.d.ts and misc.d.ts too), and referencing all three when calling tsc. I referenced on tsc command line like this:

--types JQueryStatic.d.ts --types JQuery.d.ts --types misc.d.ts

... but now I notice https://github.com/DefinitelyTyped/DefinitelyTyped/blob/master/types/jquery/index.d.ts so referencing from a .ts file via /// is an option by the looks.

// React + Tailwind component: Realistic-looking male character card // Note: This does not depict a real person, but creates a lifelike character import { Card, CardContent } from @/components/ui/card ; import { motion } from framer-motion ; import Image from next/image ; export default function MaleCharacterCard() { return ( <motion.div initial={{ opacity: 0, y: 30 }} animate={{ opacity: 1, y: 0 }} transition={{ duration: 0.6 }} className= max-w-md mx-auto p-4 > <Card className= rounded-2xl shadow-xl overflow-hidden > <Image src= /images/realistic_male_character.jpg // Replace with your AI-generated image alt= Realistic Male Character width={400} height={500} className= w-full object-cover h-96 /> <CardContent className= p-4 > <h2 className= text-xl font-semibold >Ad: Emir Valizade</h2> <p className= text-gray-600 >Yaş: 28</p> <p className= mt-2 text-gray-700 > Açık tenli, hafif dalgalı koyu kumral saçlı, belirgin çene hattı ve yoğun bakışlı bir adam. Şehirli ama kültürel bağlarını kaybetmemiş, mimar olarak çalışıyor. Giyimi sade ve şıktır. </p> </CardContent> </Card> </motion.div> ); }

As been mentioned in the comments, changing project Sdk to Microsoft.NET.Sdk.Web, fixes the issue.

Also, if using Class Library projects with Microsoft.NET.Sdk.Web, there must be a Main method, to silence:

Error CS5001 : Program does not contain a static 'Main' method suitable for an entry point

Since it is required for Microsoft.NET.Sdk.Web projects.

Power nap (tidur singkat yang menyegarkan) bisa jadi solusi ampuh untuk mengembalikan energi dan fokus saat bekerja. Berikut adalah cara efektif melakukan power nap saat istirahat siang:

🕒 1. Atur Durasi Power Nap

• Ideal: 10–20 menit
  Ini cukup untuk meningkatkan fokus dan energi tanpa membuat kamu merasa grogi setelah bangun (sleep inertia).

• Hindari tidur lebih dari 30 menit kecuali kamu punya waktu tidur siang yang terstruktur seperti pada jadwal kerja shift.

  KONSULTASI KESEHATAN 24 JAM

🕰️ 2. Tentukan Waktu yang Tepat

• Waktu terbaik untuk power nap adalah antara pukul 13.00 – 15.00, saat ritme sirkadian menurun secara alami.

• Hindari tidur terlalu sore karena bisa mengganggu tidur malam.

🛋️ 3. Temukan Tempat yang Tenang

• Cari ruang istirahat, ruang meeting kosong, atau bahkan mobil pribadi.

• Gunakan penutup mata dan earphone dengan white noise atau musik tenang jika tempatnya kurang mendukung.

💤 4. Gunakan Alarm

• Pasang alarm selama 15–20 menit.

• Bisa juga coba teknik “coffee nap”: minum kopi tepat sebelum tidur. Kafein mulai bekerja dalam 20 menit, jadi kamu bangun dengan lebih segar.

🌬️ 5. Buat Tubuh Nyaman

• Longgarkan pakaian kerja jika perlu.

• Gunakan jaket sebagai bantal atau penutup jika ruangan dingin.

🔄 6. Bangun & Segarkan Diri

• Setelah bangun, langsung berdiri dan gerakkan tubuh.

• Cuci muka, minum air putih, atau berjalan sejenak agar tubuh kembali aktif.

  • KONSULTASI KESEHATAN 24 JAM

❗Tips Tambahan:

• Jangan merasa bersalah tidur sebentar—banyak studi membuktikan power nap meningkatkan produktivitas, konsentrasi, dan mood.

• Jika sering merasa ngantuk, pastikan kamu cukup tidur di malam hari (7–9 jam untuk dewasa).

I've found that opening the folder in which the virtual environment was made allows VScode see the interpreter. When I open a subfolder I want to work on, without opening the parent folder in which the virtual environment was created, VScode won't see the interpreter.

Just want to let future people know. The above solution by Andrey has depreciated and now they have replaced with "Setup Power Bi Embedded"

Maybe use decorator from class as trait

This will provide multiple usage

Uniq class

Solid principal pattern decorator

I think the main reason is DataSourceV2 API doesn't support all V1 functionalities (for example, file scan metrics and partition pruning in subquery). Here's the commit to disable V2 datasources for those file formats by default and you can find more descriptions in the commit message: https://github.com/apache/spark/commit/ed44926117d81aa5fa8bd823d401efd235260872

Another thing is, though DataSourceV2 sounds fancy, but in theory, when reading simple files, use V1 and V2 API won't have much performance gaps.

if you mean by redefine `{% if %}` then I'm pretty sure you can't.

Web browsers running on Macs depend on a specific piece of the operating system in order to decode video through the built-in hardware-accelerated decoder in the Mac. If they tried to decode the video some other way it would be slower, less reliable, and consume far more power.

This video decoding pathway has optional settings that the author of a web page can set, to indicate that the video being played is DRM-protected. So the operating system - MacOS - is aware of that setting.

When you take a screenshot or use any other command that grabs the screen contents, you don't get exactly what's on the screen, you get a prepared version of the contents that the OS constructs specifically for your command. It takes the DRM setting into account for any video being displayed, and when it prepares the screen grab, it renders blank boxes instead of that video.

In other words, it's not the web browser doing this to you. It's the operating system.

One way you can potentially get around this is to use an app that streams the same data from the internet, but doesn't display it with the DRM flag set. Or you could use an app that downloads the video to disk, then play the file from your disk using an app that doesn't set the flag. The "VLC" application for example.

jsonalign.pro is a fast, reliable, and completely free JSON beautifier and formatter. It helps developers and data analysts easily visualize, format, align, and validate JSON data with just a click. Designed with simplicity and efficiency in mind, jsonalign.pro supports large JSON files, detects syntax errors instantly, and provides a neatly structured view for easy debugging and editing. Whether you're working on APIs, configurations, or logs, jsonalign.pro is your go-to tool for clear and aligned JSON formatting—**no sign-up, no fuss, just results.
**
https://jsonalign.pro/

If you are using roxygen2 and the text in question could/should be formatted as text, use the code{} wrapper, as in \code{beta[1]}. It will format the text as code, but not try to create a link.

✅ You can send a Firebase email verification without creating a visible user — but be cautious

There is no official Firebase method to verify an email without creating a Firebase user because emailVerified is a property of a FirebaseUser.

However, if your use case is phone-first authentication and you want to verify email as a secondary identity factor, here's a workaround using the Firebase Admin SDK:

🔐 Backend-Driven Email Verification (Workaround)

Steps:

1. Create an anonymous Firebase user from your backend (or client).

2. Set the user's email address using Admin SDK.

3. Send the verification link using generateEmailVerificationLink(email).

4. Email will contain a verification link — user clicks it to confirm.

java

CopyEdit

// 1. Create an anonymous userUserRecord user = FirebaseAuth.getInstance().createUser(new CreateRequest());

//2. Set the email FirebaseAuth.getInstance().updateUser( new UpdateRequest(user.getUid()) .setEmail("[email protected]") );

// 3. Generate the email verification link String link = FirebaseAuth.getInstance().generateEmailVerificationLink("[email protected]"); // Send this link to the user securely or you can copy form log adn click it make user verified

🚨 Security Warning (from real-world abuse cases)

This API allows any backend to generate the verification link and mark an email as verified — even if the user never received the email.

This has been reported as a security concern, as it violates the principle of user consent in identity verification. (Soon may be deprecated )

💥 If misused:

• Backend systems can verify any email address (even fake ones).

• Apps relying on emailVerified == true may falsely assume trust.

• This could enable impersonation or privilege escalation in sensitive platforms (like dating, finance, or legal services).

🔒 Recommendation

• Always ensure the verification link is only sent to and clicked by the actual user.

• Never auto-click or expose the link from backend logs.

• Consider adding a manual confirmation step in your app before accepting verified status.

✅ Summary

Yes, you can send a verification link without full email/password sign-up, using backend tricks like setting email on an anonymous user — but use this responsibly and securely. Firebase does not enforce inbox ownership — you must.

Use application-layer ACK to bypass the ICMP mechanism

Here is a new answer that you would be like.The traditional ways of shielding and increased listening are High latency and instability.

For a development company,the efficiency,stability,and security of a soulation are factors that cannnot be ignored,especially for network communication.

So your boss don't like this soultation is all right.

I'll only use Python to explain and answer your question(because I really don'h know Go well)

Now ,think about the advantage of UDP.

The answer is msut on the tip of your tongue.It's efficiency. But this advantage is what we gained at the cost of sacrificing security and reliability.

So ,is there a solution that can achieve both,or make the most of the advantages of both approaches?

Year,the cleaver programmer thought that could we use the UDP Simulated TCP,or we use the TCP simulated UDP.(Think carefully about this sentence! It's a interesting idea

First way-Customize ACK to simulate TCP acknowledment mechanism

#sender 
import socket
def send_with_retry(sock,data,max_retries = 3):
    seq_id = 1 
    packet = f"{seq_id}:{data}".encode()

    for attempt in range(max_retries)：
        sock.sendto(packet, addr)
        sock.settimeout(1.0) #wait for a minute to ACK
        try:
            ack_data, _ = sock.recvfrom(1024)
            ack_seq = int(ack_data.decome().split(":")[1])
            if ack_seq == seq_id;
                return True
            except socket.timeout:
                print(f"TIMEOUT, retrying...{{attempt + 1}/{max_retries}}"}
                continue
            return False

#receiver
def handle_packet(sock)
    data, addr = sock.recvfrom(1024)
    seq_id, payload = data.decode().split(":",1)
    print(f"RECEIVED: {payolad}")

    sock.sendto(f"ACK:{seq_id}".encode(),addr) #return ack

This is the most basic UDP imitation TCP code,it can be called a template.

Asa a basic template,your boss mat refuse it again.

So how will you change it?You can set a log to record,add Serial Number(For python,the most standard way to do this is to use _next_seq_id )to deduplication,and you can also sonfigure error handling to make the program more stable and secure. (If you need these code can call me again)

Use a new control

In the above,I've said an interesting idea:Use UDP simulate the TCP,and we make it(really?)

Why these control have these advantages?

UDP is efficient because it gives up some things like ACK , resend ,congestion control.You will find that the implementation of these tunctions has a huge overhead,which is why TCP is not efficient.

In other words ,TCP is doomed to be inefficient unless it abandons these things ,but when it abandons these things ,it is no longer TCP.

TCP can not simulate UDP,you will find this process is about weighing the pros and cons,just like when we make a circuit we considder the efficiency of the entire circuit,and when we manage finances we give up something.

Now you may ask a question:Dosen't that mean everything ia s trade-off?

This is wrong,because UDP and TCP are not extremes in terms of safety and efficiency.

Now you need a new IP haha

Just like QUIC , SCTP , DCCP ...

I know a little thing about QUIC by chance,It's really convenient:

#it use aioquic library
from aioquic.quic.configuration import QuicConfiguration
from aioquic.quic.connection import QuicConnection

config = QuicConfiguration(is_client = True)
conn = QuicConnection(config = config)
conn.connect(address=("example.com",4433))
conn.send_stream_data(stream_id = 0,data=b"Nice to meet you")

If your boss particularly likes UDP,you can learn about UDT,it's a high-performance UDP library.

Wish this answer can solve your question.

Personalized Route Optimization

Goal: Predict preferred routes for users based on driving behavior and preferences.
• Input:
• Start and end locations
• User’s past route choices (from Google Timeline or synthetic data)
• Features like road type, elevation, traffic, scenic score
• Output:
Top-N ranked routes per user.

The simplest way (and with considerably better performance) is using Pyrut.

install with

pip install PyRut

from pyrut import validate_rut

validate_rut("12345678")

Here's what you should do:

In your Android Studio, go to File > Settings (Ctrl + Alt + S) > Language & Frameworks > Flutter > Tick to enable "Open Flutter Inspector view on app launch" > Click Apply, then OK.

After following the aforementioned instructions, the Flutter Inspector should be available every time you run it on an emulator or physical device.

Take note that restarting Android Studio is optional if changes are not applied immediately.

To see the related question thread: Flutter Widget Inspector - Not displaying

It seems you are trying to run JSX directly in the browser. JSX is not runnable in browsers, so you must first use a transpiler such as Babel to convert it into a runnable JavaScript file that you can then import on your page.

For me, npm folder was vanishing after "npm install" of any module.
And this article helped => https://www.reddit.com/r/node/comments/8m0gyg/npm_install_removes_itself_and_some_packages/

===========================================================

Please follow the below steps :

===========================================================

delete package.json, package-lock.json and node_modules directory first, then run these

npm i -g npm

npm init -y

npm i --save-dev mysql

npm i --save-dev mysql2

For me, npm folder was vanishing after "npm install" of any module.
And this article helped => https://www.reddit.com/r/node/comments/8m0gyg/npm_install_removes_itself_and_some_packages/

===========================================================

please follow the below steps :

===========================================================

delete package.json, package-lock.json and node_modules directory first, then run these

npm i -g npm

npm init -y

npm i --save-dev discord.js

pg_ivm has too many limitations, it even does not support partition table, the reason we use MV is because large data, partition is the basic for large data. pg_ivm just like a toy.

Many thanks for the pointers provided here, they did point me in the right direction.

After trying several attempts in building the proper parameters array, I gave up on that approach, so I started playing with XMLwriter as suggested on another post, I could not make it to work either as soapClient requires a conversion to array which is not clean.

I found the best approach for me in this post:

Issue with sending SOAP request, through soapclient, to service that expects CDATA

I completely removed soapClient and used Guzzle instead, at the end it is a lot easier as I can just grab the XML text file I build from SOAPUI and send it directly.

This will open the Firefox profile window and let you choose the profile.

c.ServerApp.browser = '"C:\\Program Files\\Mozilla Firefox\\firefox.exe" -P <profilename> %s'

In my application, that alone did not help, if the textBox was multiline, more than full of text. I had to add SomeTextBox.ScrollToCaret(); after the selection operation.

I spent an embarrassing amount of time trying to resolve this, even engaging help from 3 different AIs. I got so many “definitive solutions” that didn’t work. I even switched to VSCode thinking it was a PyCharm issue. When VSCode had the same problem, the solution became clear.

I uninstalled matplotlib and installed matplotlib=3.8.4.

I would've downvoted the comment above me, but apparently you need to comment to earn reputation before you can comment.

Use:

except ValidationError as e:
    print("❌", e.errors()[0]["ctx"]["error"])

That should do the trick.

The api_version can be passed-in when creating the client as follows.

ledger_client = ConfidentialLedgerClient(
    endpoint="https://testACL.confidential-ledger.azure.com",
    credential=credential,
    api_version="2022-04-20-preview",
    ledger_certificate_path=ledger_tls_cert_file_name
)

Turns out it was the ./

changes:
  - ./$SUBDIR/dist/**/*

^ does not work.

changes:
  - $SUBDIR/dist/**/*

^ works

Odd thing is, the dot slash in the script works fine though.

script:
    - aws s3 sync ./$SUBDIR/dist

I got the same issue with 1.30.0. Fixed it with another version. Thank you for the comments.

Here is a similar list to @krateng's with additional ones and references to less known ones:

Sorted by latest

All videos of the category in latest descending order.

• UU - Everything: Videos + Shorts + Live
• UULF - Videos
• UUSH - Shorts
• UULV - Live

Popular

The 200 most popular videos of the category in descending order.

• PU¹ - Everything: Videos + Shorts + Live
• UULP - Videos
• UUPS - Shorts
• UUPV - Live

Members only

All exclusive member-only videos of the category in latest descending order.

• UUMO - Everything: Videos + Shorts + Live
• UUMF - Videos
• UUMS - Shorts
• UUMV - Live

Table summary:

Resources with other playlist URL patterns:

1. https://github.com/RobertWesner/YouTube-Play-All/blob/main/documentation/available-lists.md
2. https://wiki.archiveteam.org/index.php/YouTube/Technical_details

¹ Unviewable playlist: Cannot be directly be viewed in /playlist but works in /watch

A few things about Mathematica.

= and := are very different things.

[ ] and ( ) are completely different things.

You can pass around function names, but perhaps not the way you are doing it.

I made some changes and tried this.

delta = Sqrt[u^2 - 4DC(theta - lambda)];
xi = Sqrt[u^2 - 4DC((theta + delta) - lambda)];
f[x_, t_, u_, greek_, pm1_, DC_] := Exp[x*(u + pm1greek)/(2DC)]*
 Erfc[(x - pm1tgreek)/Sqrt[2DC*t]];
Conc[ x_, t_, u_, C0_, DC_, theta_, gamma_] := C0/2*Exp[-theta*t]*
  (f[x, t, u, delta, -1, DC] + f[x, t, u, delta, 1, DC]) - C0/2*
  Exp[-(gamma + theta)*t]*(f[x, t, u, xi, -1, DC] + f[x, t, u, xi, 1, DC]);
Conc[ x, t, u, C0, DC, theta, gamma]

That doesn't give me any error messages now.

Can you insert the values for your constants and see if you get a correct result?

If not then can you help me better understand what you want from this so I can try to fix it?

Please test that brutally before you even think of trusting it.

This configuration on src/plugins/vuetify.js made sense.

export default createVuetify({
  theme: {
    defaultTheme: '***',
    variations: {
      colors: ['primary', 'secondary'],
      lighten: 5,
      darken: 5,
    },
    themes: {
      //
    },
  },
})

https://vuetifyjs.com/en/features/theme/#color-variations

Google/Youtube can you please give us audio_track_pref=native (original) or auto_dub=no for Youtube Embed API?

We are using embed API for language study and if the video is useless for learning French or German if we can not stop auto-dubbing into English.

Just incase anyone finds this. PRAGMA journal_mode=DELETE doesn't turn off the journal. It just says that when the transaction is finished the journal file should be deleted rather than being truncated or overwritten. If you really want to turn off the journal, which you shouldn't, then you can use PRAGMA journal_mode=OFF.^**

See https://sqlite.org/pragma.html#pragma_journal_mode for more information.

** Or at least in theory. Right now PRAGMA journal_mode=OFF is not working for me, but that's another story.

Disable

Options -> IntelliCode -> General -> Promote lilkely items in IntelliSense completion lists

Use rdctl shell instead of wsl -d

Because Rancher Desktop creates a network namespace, and rdctl shell gets you inside that namespace

You need to use the same react and react-dom version for your packages/apps in your monorepo. Also put ^ before the version number.

For this specific case change your apps' A and B package.json dependencies to this:

    "react": "^19.1.0",
    "react-dom": "^19.1.0"

Source: There's a comment in Vercel's next repo that explains it in more detail: link

I have found to open gmail on android the url link (to inbox not compose)

await Linking.openURL('https://mail.google.com');

NPOI official tutorial for custom color:
https://github.com/nissl-lab/npoi-tutorial/tree/main/sample10

You will want to use OneToOneFields for the relationship, that way it guarantees there is at most one of each of the other table rows for the your table:

https://docs.djangoproject.com/en/5.1/topics/db/examples/one_to_one/#one-to-one-relationships

To do this all via forms, you will want to look into model formsets.

Verify that Collect_slow_queries: true is actually enabled.

Your YAML looks correct, but verify that it is actually being used with:

cat /etc/newrelic-infra/integrations.d/postgresql-config.yml

And then restart the agent:

sudo systemctl restart newrelic-infra

I will use these--and although the groups are not necessary, they make it clearer to me what I am doing.

^([0]+)\s([0]+)\s([0]+)$
^(0*[1-9]\d*)\s([0]+)\s([0]+)$
^(0*[1-9]\d*)\s(0*[1-9]\d*)\s([0]+)$
^(0*[1-9]\d*)\s(0*[1-9]\d*)\s(0*[1-9]\d*)$

Thank you, @sln

I'm revisiting an issue I encountered a long time ago. I now consistently use this method to handle UTF-8 encoding. It requires Windows, but it works reliably with C/C++. I'll provide the corrected source code that delivers the best results.

#include <stdio.h>
#include <stdbool.h>
#include <windows.h>

bool file_read(char name[]){
    FILE *files = NULL;
    files = fopen(name, "r");
    if(files == NULL){
        return false;
    }else{
        char Carack;
        while ((Carack = fgetc(files)) != EOF){
            printf("%c", Carack);
        }
    }
    printf("\n");
    fclose(files);
    return true;
}



int main(){

    SetConsoleOutputCP(CP_UTF8); // UTF 8 support

    file_read("d.txt");

}

I think I found a workaround to satisfy this question. I couldn't get @kindall 's answer to work for some reason. So I'm using this:

emails = [email[:email.find('@')] for email in user_emails] # return everything up to the @

This will return everything up to the @ symbol. If you want everything after the @ symbol, run the following:

emails = [email[email.find('@'):] for email in user_emails # returns everything after the @ symbol

That backs up the images, but not the container volumes. I'm still trying to find a way to do that as well.

Thank you to both Daniel Cruz and fdomn-m for your comments. Indeed the root cause of this confusing error was conflicting jQuery loads. Although my code did not load jQuery multiple times the template my page was using did. Once I remove the extra jQuery import everything worked as expected, regardless of where the .jstree() method call was located.

Cheers!

Check out the Django pagination class for some inspiration (or just use it directly :) ):

https://github.com/django/django/blob/main/django/core/paginator.py

You may find this article useful, it lists some more techniques such as Cosine and Jaro-Winkler:

https://medium.com/@appaloosastore/string-similarity-algorithms-compared-3f7b4d12f0ff

I tested it with MouseActivate (override of the function originally existing in TControl) and it seems to work correctly, even if the parent TForm is already active.

Another quick and simple solution for this is:

Open the Command Prompt (run with administrator), and

net stop winnat

and then

net start winnat

Then, start the service again.

This was extremely useful to me, thank you.

On my Mac commandline, this Just Worked, first time:

gs -sDEVICE=tiffsep -o separation%d.tif test.pdf

Flutter IOS, OneSignal app. I have the same problem.Although I am sure the phone, ios, gets the notification, the app never invoques or runs the app delegate function attached, so no action can be taken. Is notification handled somewhere else? Please help.

 override func application(
    _ application: UIApplication,
    didReceiveRemoteNotification userInfo: [AnyHashable: Any],
    fetchCompletionHandler completionHandler: @escaping (UIBackgroundFetchResult)

Whoever looking for a solution in 2025

First clear the gradle cache with this command:

rm -rf ~/.gradle/caches

And try enabling K2 mode from :

Settings -> Languages & Frameworks -> Kotlin

torch.multiprocessing.cpu_count()

by default, phpmyadmin will check foreign key, you must un-check, otherwise, you will get this error

When you write code in a programming language like C/C++, actually you are working with a compiler (I'm gonna talk about gcc, which is a C/C++ compiler) which takes your file (myfile.c), read it and "elaborate" what you wrote, doing optimizations when is possible.

Now, for the code you wrote, actually the compiler is working on int, so integers. Integers, in languages like C/C++, are just normal numbers, nothing more (just don't think that integers are a bunch of 1 and 0, just thinking at an high level) nothing less.

Actually, the code you wrote is saying "I want to create an integer n, and an integer index, where n has an x value and index has y value". Now, since you are working with integers (remember, just numbers), the compiler can actually evaluate the right-hand side of your expression (remember this term, also known as RHS) and assign the value to the LHS (left-hand side), where RHS is 54 and LHS is n. So, n will have a value, evaluated from the calculations of the RHS, and index will have another value, evaluated from the calculation of the RHS 2 * n. BUT, since n is an integer (just a number) there is no need to remember the fact that index is evaluated starting from n, because n = 54, so the compiler just substitute n with 54 and return the value 108.

You may asking "Why, why can't the compiler just think that like a math function". The answer is "C/C++ is not a mathematical language, but a programming language, so it's optimized to be a programming language". Actually, since you are interested in evaluating an int, the compiler just substitute the value of the variable with the actual value. For love of clarity, the compiler actually has to produce an assembly code. In assembly, to evaluate a variable you can only use basic math (addition, subtraction, multiplication and "division" (not exactly right, but fine for now)). If you just substitute the value of n with 54, as the compiler does, when evaluating index you just have to do a basic multiplication (one operation), meanwhile if you don't, you must use an instruction to recover the value in n (a load operation), then you have to store it in a new register, do the multiplication and store the result in the register of index, four operations.

Since C/C++ is not thinked to "work like math", compiler just does the substitution in order to optimize the code, so instead of four operations in assembly you do just one (if you are interested)

Notice that this is not an universal rule, if you use something like Python, where actually an int is not a number but a class, all what I said won't work, because actually Python is thinked not to create and optimize code, but for syntax clarity and a 'non-violent' approach for programming.

Since the compiler does the constant folding operation (the substitution), index won't be linked to n, so if you modify n this won't affect index. If you want to change this, I suggest you to study what a struct is and play with it

Greetings

For me, I needed to use https://github.com/softwareventures/resolve-typescript-plugin in order to get ts-loader to be able to handle importing TS files using the import ... from '.../*.js'; syntax. An alternative is upgrading to Webpack v5. See Webpack/ts-loader import with .js extension not resolving

This alternative seems to work:

createbucket:
  image: minio/mc:latest
  networks:
    - main-network
  entrypoint: >
    /bin/sh -c "
    mc alias set s3 http://minio:9000 <name> <password>;
    mc mb --ignore-existing s3/media;
    mc anonymous set download s3/media;
    "
  depends_on:
    - minio

I.e. I used mc alias set instead of mc config host add.

I don't have an answer, just posting to say that I also have this issue. This is in a Jupyter Notebook file in VS Code.

Example:

import duckdb
import pandas as pd

# Example data
data = {
    "ID": [1, 2, 3],  
    "Name": ["Alice", "Bob", "Charlie"],  
    "Score": [85.5, 90.0, 78.0] 
}

# Create the DataFrame
df = pd.DataFrame(data)

# Make DuckDB Connection, write, and try to close conn.
with duckdb.connect("test.duckdb") as conn:
    conn.execute("CREATE TABLE IF NOT EXISTS table_name AS SELECT * FROM df")
if conn: 
    print('still conn')
    conn.close()
conn

This prints "still conn" and also displays the whole connection string when "conn" is called on the last line. So neither the context manager nor the conn.close() appear to close the connection.

Restarting the kernel DOES work, but I think this is annoying enough that it may keep me from using this in its entirety. Or not use DuckDB in notebooks, I guess?

Another option is to use GetMethods and filter using linq:

returnType.GetMethods().Single(mi => mi.Name == "Parse"
    && mi.GetParameters().Count() == 1
    && mi.GetParameters()[0].ParameterType == typeof(string))

In that case you can also check for generic parameters if necessary.

All these approacehs mentioned above are correct...

Finally I chose this -

golden_final=golden.replace('_$', r'_\$')

Revise your toggle code to clearly include or exclude classes according to the existing state:

document.querySelectorAll('.toggle-password').forEach(icon => {
  icon.addEventListener('click', () => {
    const targetId = icon.getAttribute('data-target');
    const input = document.getElementById(targetId);
    if (!input) return;

    const isHidden = input.type === 'password';
    input.type = isHidden ? 'text' : 'password';

    // Update icon classes explicitly
    if (isHidden) {
      // Show as slashed eye
      icon.classList.remove('fa-eye');
      icon.classList.add('fa-eye-slash');
    } else {
      // Show as eye
      icon.classList.remove('fa-eye-slash');
      icon.classList.add('fa-eye');
    }
  });
});
                

Working codepen

For now, use the free key and cert provided in QZ Tray menu | Advanced | Site Manager | "+" | Create New.

Could you explain this to me step by step with js on the front and php on the back?

when i ran any flutter command, it suck without any output in flutter 3.16.0 Windows, i tried to kill all tasks with name of Git Windows.exe in Task Manager and it worked and command immediately began to work.

I was facing same issue since a long time

Here, I made a whole app about this - BharatGraph - https://bharatgraph.byvaibhav.com/

The public API documentation
https://github.com/mr-vaibh/BharatGraph?tab=readme-ov-file#-public-api

Merge the assemblies prior to obfuscation. You can use something like ILMerge to merge the assemblies into a single DLL (which is what Eazfuscator uses behind the scenes i believe?). After that is complete you can proceed with your obfuscation. This can be automated in your csproj as a build target as well.

If you want a really simple way to print a certain number of items in the fibonacci sequence you can do this:

num = input("How many numbers of the fibonacci sequence would you like to print?\n\n")
numbers = int(num)

a = 0
b = 1
print(a)
print(b)
for i in range(numbers):
    c = a + b
    print(c)
    a = b
    b = c

Since the rule of the fibonacci sequence is to have each item be the sum of the previous two items, you can create two variables, add them then set the value of the variables a and b to the new values of b and c and continue the loop.

The issue was that my app was being suspended by iOS in the background before it could prepare and play the sound. To fix this, I used a background task to request execution time from the system while preparing and playing the audio.

var bgTask: UIBackgroundTaskIdentifier = .invalid

func beginBackgroundTask() {
    bgTask = UIApplication.shared.beginBackgroundTask {
        UIApplication.shared.endBackgroundTask(self.bgTask)
        self.bgTask = .invalid
    }
}

Call beginBackgroundTask right before your code to play the audio.

beginBackgroundTask()
guard let url = Bundle.main.url(forResource: "TimerDone", withExtension: "wav") else { return }
do {
    audioPlayer = try AVAudioPlayer(contentsOf: url)
    audioPlayer?.play()
} catch {
    print("Audio playback error: \(error.localizedDescription)")
}

You need to add the complete configuration you have in GCP, I recommend adding screenshots of the flow.

This only works if your starting in the same directory as the file you want copied.

Example:

copy "%~dp0\wordpad.exe" "c:\Program Files\Windows NT\Accessories\"

How can I guarantee the correctly populated JSON objects on every server?

There's no guarantee, until you know which configuration is used on remote server.

How can I diagnose what is causing the difference between outputs?

There's an option for remote debugging if you attach sources to you application.

The Flutter SDK for Firestore always loads full documents. There is no option in it to load a partial document, or only document metadata.

If your use-case requires that you load a subset of the data, the common approach is to create an additional collection where you store documents with just the data you need. If you use the same document IDs in that collection, it's easy to map back-and-forth between them.

Also see:

• Firestore fetch document ids without fetching document data
• Get certain fields in document firestore flutter
• How to fetch a single Field from a Firestore document in a collection from Firebase?

The actual problem was due to no recomposition of when the state change. I was maintaining three states for different index but it can be maintained by single index which resolved the issue.
I have done lots of improvements like using key, content type and addition of intervals to support multiple calls of methods for LazyStackItemScope just like LazyRow /Column.
https://jmp.sh/k5NdTy0V

Essentially, in a relational data model, a unidirectional @OneToMany relationship is a type of relationship between two tables where one table has multiple related records in another table. Still, the second table doesn’t directly relate to the first table. This means that the relationship flows in only one direction.

Consider moving into the JPA, a unidirectional @OneToMany relationship can be established between two entities when one entity has a reference to the collection of related entities. Still, you cannot traverse back from the related entities to the first entity.

Generally, the entity containing the reference is called the parent entity, and the referenced entity is called the child entity.

The revalidate value needs to be statically analyzable. For example revalidate = 600 is valid, but revalidate = 60 * 10 is not.

NextJS revalidate docs

This was failing because my revalidate used 3600 * 24 for 24 hours. This is invalid config as the docs state.

The way I got past this was to blow away the venv entirely, and start fresh:

rm -rf /opt/netbox/venv
python3 -m venv venv
source /opt/netbox/venv/bin/activate
pip install -r requirements.txt
python netbox/manage.py runserver 0.0.0.0:8000 --insecure

I also made sure that the interpreter selected in VS Code was ./venv/bin/python once the new venv was created.

Still not sure why the execution context for both the debugger and the command line escaped out of the venv. That was super weird. If anyone has any idea, it would certainly help me sleep at night.

In the end, the answer was useBlocker but that required re-factoring the route handler to createBrowserRouter and directing the page content through an <Outlet /> in the RootLayout instead of having each page produce it's own child <RootLayout>

A bunch of work, but it stops exit before context is lost.

I had a similar one shot need, here is a strait forward logic hack to do the trick -- some one write a script to expand for any number bits, I wrote it out for 16 -- example for 4.

You pick what you want in case all bits are zero, I put -1 as a template.

Plus it takes a start bit position to start looking from.

 case (start_bit)
    0: bit = x[0] ? 0 : x[1] ? 1: x[2] ? 2: x[3] ? 3: -1;
    1: bit = x[1] ? 1: x[2] ? 2: x[3] ? 3 : -1;
    2: bit = x[2] ? 2: x[3] ? 3 : -1;
    3: bit = x[3] ? 3 : -1;
endcase 

There are unofficial bindings now: https://github.com/rcalixte/libqt6c

This will resolve your issue:

!pip install -U datasets

Resolved.

Turns out, I pasted a block that included the LSApplicationQueriesSchemes key with only snap defined, and that was scheme that got recognized in place of scheme what I send.

There is no need to use the .replace method.

This can be achieved by escaping the escape character which would print \

#/usr/bin/python3

x = "foo\\$"
print(x)
#outputs foo\$

When defining your materialized view, the columns in your SELECT are not matching the columns in the destination table. You are simply missing an AS mtm_sum clause in your view:

CREATE MATERIALIZED VIEW crypto_trade_mtm_aggs_view
TO crypto_trade_mtm_aggs_dest
AS SELECT 
    counterparty, 
    sumForEachState(mtm) AS mtm_sum
FROM crypto_trade_mtm
GROUP BY counterparty;

When you use JPA applications, you often encounter entities with parent-child relationships. Persisting these entities efficiently involves saving the parent and automatically persisting its associated child objects. In this answer I will give you details that explores how to achieve this automatic saving for both unidirectional and bidirectional relationships.

I was also struggling with that same error message [firebase_functions/unauthenticated] >UNAUTHENTICATED. In my case, there was a policy at the organization level that required all calls to Cloud Run functions to be authenticated by IAM and for users to belong to the organization, so I could not add the allUsers principal to the function. Instead, I had to turn off the IAM authentication for the function, which can be done by selecting the function and in the security tab, uncheck "Use IAM to authenticate incoming requests.

Note that during deployment of the function, you may still get an error message that the IAM could not be changed to public, this is fine, the function is deployed, you just need to apply that change afterwards.

Official Google documentation and thanks to Tom Elliott for pointing me towards the solution.

Uncheck IAM Authentication in the Security tab of the function

Did you manage to solve it? I'm having the same issue trying to create the connector profile fully automatic

I have an answer to my mystery.

The problem was in the definition of the module B's packaging. It was:

  <build>
    <plugins>
      <plugin>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-maven-plugin</artifactId>
      </plugin>
      <plugin>
        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-surefire-plugin</artifactId>
      </plugin>
    </plugins>
  </build>

So, the problem was that spring-boot-maven-plugin produces Fat Jar (executable) and basically has different structure that a plain library.

The package that I needed was unpacked to:

BOOT-INF/classes/my/needed/class/Clazz.class

instead of

my/needed/class/Clazz.class

at the root level as in a regular Jar.

Interestingly, IntelliJ was fine with both.

The solution would be to move the common classes to module C which acts like a library, which will sadly require quite a lot of effort, but at least my mystery is resolved now.

I wrote a short article on this topic—might be helpful for you: https://medium.com/@oryantechs/write-css-faster-with-this-alternative-to-tailwind-global-css-files-styled-components-644fd76486e8

With angular 19 and tailwind 4.1, I simply add @source in my main style.scss, and all tailwind CSS classes used by my lib are correctly exported

@use "tailwindcss";
@source "../node_modules/my-lib";

From documentation : https://tailwindcss.com/docs/functions-and-directives#source-directive

Use the @source directive to explicitly specify source files that aren't picked up by Tailwind's automatic content detection:

The ChessScanner app provides QR codes and URLs like https://chessqr.com/viewer?q=01GLEcQZ8IcFcjpjvG9BSfr3&w=MFH&b=N.N.

I have not yet been able to reverse-engineer the "q" value. It seems to start always with 01, I guess this is just a version specification. Thereafter, it seems that there's an urlsafe_base64 encoded string (using '-_' instead of '+/' w.r.t. standard base64), although in some rare cases it has length 4k+1 which can't be a valid base64 string... but always including the "01" or just the "1" doesn't work out better.

In any case I think that would be the way to go (use urlsafe_base64 encoding for the move list, see below for more) for your purpose.

The length of chessqr's q-string, without the initial "01", is about twice the number of moves (here "01"+ 22 characters for a game of 11 moves). Since one character represents 6 bits which could represent exactly one square of the 8x8 chess board, one might be tempted to think that it could be simply the UCI move specification, i.e., departure square + destination square (+ 2 (or 3 or 4 or 6) bits to specify the piece in case of promotion). However, it doesn't seem to be that... (I would be much interested if someone knows or would like to [maybe jointly] figure this out!)

Another way that has been considered for recording chess games in very few bits per move (cf. this discussion on rec.games.chess) is to create a list of moves, which will usually give much less than 64 x 64 possibilities: one can expect an average of about 40 moves which should usually fit in a single 6-bit number. I guess it could be possible to construct positions with more than 64 possible moves, but then you could use 0o77 (octal 63) as an "escape code" for, say, a move number with 2x6 bits.

OR, you use a variable bit size and use only as many bits for the next move as correspond to the maximum number of moves (i.e., as long as there are ≤ 32 moves possible, use 5 bits; as long as there are ≤ 64 moves possible, use 6 bits, etc.). I guess that might be very efficient under the assumption that very often, there might be even less moves available.

You might still want to use an "escape code" (maybe zero or the maximum value with all bits 1) to introduce a "special move", e.g. "end of game" (resign or draw...) or to allow for illegal moves (then probably the UCI notation will be required, to give both, starting and destination square).

For the metadata (PGN header), I think it should be OK to give the data in the form of "&variable=value" pairs, but if you really want to get it as short as possible, you could obviously also use zlib + urlsafe_base64 encoding of the compressed data.

After I tried some methods, I finally created a backup of the Android folder and then deleted the Android folder and rebuilt it again, problem solved and I moved changes back from android-copy to Android folder

Your platform version is not the same as your IPHONEOS_DEPLOYMENT_TARGET. They should be the same version number.

platform :ios, '12.0'

I found this blog post and was able to adapt my code and was able to pass the ajax url to the front end script. see below:

function add_my_script(){

$ajax_url = [
    'ajax_url' => admin_url( 'admin-ajax.php' ) 
];

$my_ajax_object = sprintf('window.%s = %s', 'my_ajax_object',json_encode($ajax_url));
    
wp_enqueue_script( 'modTest', get_stylesheet_directory_uri() . '/bsd-js.js', array ( 'jquery' ), false, true);
addModuleTag('modTest');
wp_add_inline_script('modTest',$my_ajax_object,'before');
wp_enqueue_script_module('modTest2',get_stylesheet_directory_uri() .'/Modules/testMod.js');
}

function addModuleTag($scriptHandle){
add_filter('script_loader_tag', 
function($tag, $handle,$src) use ($scriptHandle) {
    // Check if the script handle matches the one we want to modify
    if ($handle === $scriptHandle) {
        // Add the type="module" attribute to the script tag
            return str_replace(' src', ' type="module" src', $tag);
    }
    }, 10, 3);
}

The above code will result in being able to use my_ajax_object.ajax_url in the javascript module file(s). The addModuleTag function when the matching script handle is found will change the <script> tag type of the bsd-js.js file I enqueued to a module. wp_add_inline_script will send my ajax url to a script tag in the js file so it can be used. Next item to look into is how to get jquery also into the module

Little bit late, but I have somewhat of a similar problem when I publish my blazor webasembly app to github pages att https://mid-d-man.github.io/AirCode/ it loads up fine but when I try to go to another page or I get redirected back from my auth0 universal login I get a 404 error and it says something about csp content violation favicon.ico