I've made a replacement with Java Swing and I'm using it:

https://github.com/Podbrushkin/tableview-swing

I have to import the files via kubectl. After that it works properly.

kubectl create secret

I had the same problem building on windows with, the solution was to log in with Docker Desktop.

This is currently tracked here https://github.com/Azure/bicep/issues/1895

As of now (5th De 2025), this has not been implemented

You seem to ask your question on the premise that git keeps a "list of files which have been changed". It does not. Each git add updates the index, and when a later command (status, commit) needs to know what's to be committed, it is computed from analysing the HEAD commit, the index and the working tree. As a sidenote, we could have done without the defensive disclaimer about downvotes. Let people do their thing and focus on the subject matter, if I may. Agreed with user207421: XY problem here. (Btw, this is not an insult to you, it's a judgement on the question.)

There are different ways to handle this:
(I believe 1. fits you the best)

1. Use Streamlit (or look for alternatives), it lets you write pure python and it runs as a web app that you could deploy and others can use.

2. Use a Python Web Framework (Flask, FastApi, other) which will require some html knowledge (and js if you need dynamic updates i believe)

3. Or you could use Pyodide or Brython, which let python run inside WebAssembly (i think you will still need js/html in some way there)

Check the result of select hex(@value);

If the result is empty string, then check the value of parameter named 'dashcope_sk'.

If the value of parameter is empty, then set it with ur own service key of dashscope from Alibaba Cloud.

Thanks for your replies. Now I'm calm again. ;-)

@Marce Puente.... thanks, I exactly got what I was looking for by adding just a f string in your code :

import random
lst1 =[]
lst2 = []
lst3 = []
lst = [ lst1, lst2, lst3 ]

for i in range( 0, 3 ):
    for j in range( 1, 4 ):
        lst[ i ].append( random.randint( 1, 101 ))
    print( f"lst{i} = {lst[i]} ")

print( lst )

This gave me the output as below :

lst0 = [76, 35, 29]

lst1 = [85, 86, 94]

lst2 = [92, 41, 85]

[[76, 35, 29], [85, 86, 94], [92, 41, 85]]

Here I have the list names generated as well.

Why is this tooling (which we can only add comments to) and not a real general question (which we can add actual answers to, including formatted code blocks)?

did you find a way? I want to get the profile photos of users who send messages to Cloud API. I'm currently researching and will share if I find a method.

At least with PostgreSQL, you should always use the latest version.

I'll have to repeat, please read How do I ask a good question?. The question is too vague - why would you ever consider using old JDK or PostgreSQL versions? What "risks" are you talking about? There shouldn't be compatibility issues (unless you use abandoned libraries) and the old versions won't have any of the fixes and performance improvements introduced in the newer versions.

Databases especially live for decades, while upgrading them is non-trivial. Quite often it's better to migrate to new iron instead of trying an in-place upgrade. And PostgreSQL 18 introduced significant performance improvements especially around IO. At the very least, don't use an older database, to reduce the need (and risk) of migrations.

As for Java, do you have a specific library that somehow only works in the 2-year old Java 21 ? In that case the real problem is the library, which is probably abandoned. Or the newer version wasn't purchased for whatever reason.

With Java at least, it's a lot easier to migrate once you get rid of obsolete third-party libraries. Breaking changes are very rare and announced well in advance, allowing developers to upgrade their code. If you feel Java 25 (released Septemeber 2025) has teething problems you can start with Java 24 (March 2025) and upgrade later.

To begin with, you either set a column to null via migration or not, depending on the case, using ->where('column', $value); is all you need to do. For cases where you needed to exclusively check if a column is null then ->whereNull('column'); is your go-to. The framework is good enough to understand this. So, you just need to follow the right practices.

You won't get any answers if people can't even read the question. SO is an English site. Post a new question in English, or use https://es.stackoverflow.com .

Please read How do I ask a good question? or ¿Cómo elaboro una buena pregunta? before posting the new question too.

I am currently working on a Flutter project and I also getting this problem recently

Current youtube player dependency in mobile right now (mostly) won't be able play the video because they put only the id of the video like this
String url = www.youtube.com/watch?v={videoId}

Yesterday I also working on this also while trying to solve this problem, I browsed on youtube to get the embed html tag then I realized it has ?si parameter youtube_embed_html_tag

so apparently we need to provide that parameter a value to get the video loaded right. The ?si parameter can be filled with any string, here's the example

String url = www.youtube.com/watch?v={videoid}?si=j4sdf2ffh

After trying this, the player managed to load the video without any errors

n.b.: Some video may also won't load properly because some video authors may disabled the embed sharing

You are very close! The issue is in the use of 'event.target.name' in your 'switch' statement. Currently, your input elements only have 'id' attributes, but **no 'name' attributes**, so 'event.target.name' returns 'undefined', and none of your switch cases match.

To fix it:

Add the 'name' attribute to each input to match your 'switch' cases:

'''html

<input type="number" onchange="trial(event)" id="celcius" name="celcius" placeholder="Celsius">

<input type="number" onchange="trial(event)" id="faren" name="faren" placeholder="Fahrenheit">

<input type="number" onchange="trial(event)" id="kelvin" name="kelvin" placeholder="Kelvin">

Once you do that, your switch(event.target.name) logic will work properly and the other fields will update automatically.

Make sure you have your route name defined

    [HttpGet("GetDriver/{id:int}"), Name = "GetDriver"]
    public IActionResult GetDriver(int id)
    {
        // Your code
    }

Close VS → delete all *.vssscc/*.vspscc files → edit .sln and remove all lines starting with Scc → edit the project file and remove <Scc*> XML nodes → delete the .vs folder → reopen solution → rebind via File → Source Control → Advanced → Bind.

I think what you should be asking is “How can I become a more independent learner?”

I had faced the same issue and there wasn't any adblock extension, I failed to notice that since I am using Cloudflare, I'd to allow the tags in cloudflare, after that it worked.

I believe there may be an issue with your system. Please ensure that your system is functioning correctly and try running it again.

I finaly found my answer here Delegate 'RequestDelegate' does not take 2 arguments - ASP.NET Core 8 Minimal API

When TypedResults are used i must specify the possible retursn in the lambda like async Task<Results<Ok<InternalGenericEmail>, NotFound<string>>> very important here to specify the type that takes eg return TypedResults.NotFound("hi") needs NotFound<string>

public void MapEndpoint(IEndpointRouteBuilder app) => app.MapGet("api/notifier/generic/check/{id}", async Task<Results<Ok<InternalGenericEmail>, NotFound<string>>> (string id, IEmailRepository emailRepository) =>
{
  IEnumerable<InternalGenericEmail> a = await emailRepository.GetAllAsync();
  InternalGenericEmail? result = a.FirstOrDefault(p => p.Id.ToString() == id);
  if (result is not null)
  {
    return TypedResults.Ok(result);
  }
  return TypedResults.NotFound($"Die angegebne Id: {id} ist nciht vorhanden");
}
  )
  .AddFluentValidationAutoValidation()
  .WithTags("Generic")
  .WithName($"GenericEmail123")
  .WithDescription("GenericEmailDesc")
  .WithSummary("GenericEmailSum")
  .Produces((int)HttpStatusCode.NoContent, null)
  .ProducesValidationProblem()
  ;

conditionally disable <select> using readonly angular 16

 <select class="form-control" formControlName="group_id"  [attr.readonly]="!isEditMarkupFlag ? '' : null">

'uv run' makes dependencies availabe on PATH automatically.

AFTER Installing a wheel, uv is not involved, so the console script from j2lint is not installed.

Declare j2lint as a dependency in pyproject.toml so its console script is installed:

[project]

dependencies = ["j2lint"]

{project.scripts]

j2linttest = "j2linttest:main"

After building and installing the wheel, j2lint will be on path and subprocess can find it.

Ah. this one really worked for

import 'assets/fonts/PTSans-normal.js';

import jsPDF from 'jspdf';

const doc = new jsPDF();

doc.setFont('PTSans'); // set custom font

doc.setFontSize(12); // you can change size

doc.text('Hello', x, y) // and you can write your text

Verilog (IEEE 1364 2005 is the latest and final Verilog specification) does not support unpacked arrays. Unpacked arrays become legal in SystemVerilog (IEEE 1800) concept.

1. Add .aar file into plugin dir:

.
└── my-plugin/
    ├── android/
    │   ├── libs/
    │   │   └── my.aar
    │   ├── src/
    │   └── ..
    └── ...

2. Update my-plugin/build.gradle

repositories {
    google()
    mavenCentral()
    
    flatDir {
        dirs 'libs' // Add libs directory to the repository
    }
}
dependencies {
    implementation(name: 'my', ext: 'aar') // Add my.aar to the dependencies
}

3. Update package.json

{
  "files": [
    "android/libs/"
  ]
}

When you use this plugin in your project, you should copy the my.aar file to project/android/app/libs/may.aar. Remember to remind your users to do this.

Note:

If you are using npx cap open android to launch Android Studio and debug the app, remember to modify the app configuratio:

App -> Edit configuration -> Installation Options -> Deploy:

Switch Default APK to APK from app bundle

This is important, in my tests at least, the default mode causes the built debug APK to be missing .so files that might be present in .aar libraries, leading to runtime errors.

And if you are using command like ./gradlew :app:assembleDebug, it works well, I still don't understand why the default run mode in Android Studio misses .so files.

For specific pages that need to always fetch new content

1. Use the PHP headers above (or server config)

2. Meta tags are optional but not required

This ensures the user always receives the newest version of the page.

<?php

header("Cache-Control: no-store, no-cache, must-revalidate, max-age=0");

header("Pragma: no-cache");

header("Expires: 0");

?>

Valid but unspecified state. What does this mean

You may try https://marketplace.visualstudio.com/items?itemName=elong0527.vs-rtf-preview that convert RTF to PDF then preview.

Good question. I cannot imagine why that line would be needed.

This compares the i-th character from the start with the i-th character from the end.

if (str[i] !== str[len - 1 - i])

str[i] from the front

str[len - 1 - i] from the back

If they don’t match, the string is not a palindrome.

I have encountered this issue as well while using wsl on vs code.

The solution in my case was to disable the telemetry in settings.

I got the answer from here

https://github.com/microsoft/vscode-remote-release/issues/10804

I raised this question after working on the code, and I’m unsure why it was deemed not a good practice. Could you please help me understand the concerns?

Here is a long answer.

Organizing related business entity classes that represent the same business concept and share common attributes is good practice because it reduces duplication, improves consistency, and makes the model easier to evolve over time.​

Encapsulation of domain rules

When similar entities share a common base (abstract class, interface, or composition), shared invariants and validation rules live in one place instead of being copy‑pasted across classes.​

This makes it easier to reason about the business domain, because the behavior for that concept is localized and changes to rules propagate uniformly.​

Maintainability and evolution

A cohesive set of related entities is simpler to maintain than many ad‑hoc classes with overlapping fields and logic.​

When the business concept changes (for example, adding a new common attribute or changing an identifier strategy), you change it once in the shared abstraction instead of touching many scattered classes.​

Reuse and reduced duplication

Organizing entities around business concepts encourages reuse across use cases, services, and even applications within the same domain.​

Common aspects like identifiers, auditing fields, or value object types (e.g., Email, Money) can be implemented once and reused by all entities that represent that data type, reducing boilerplate and risk of subtle inconsistencies.​

Clearer domain language and structure

Grouping entities by domain meaning (rather than by technical concern) leads to packages and types that match the ubiquitous language used by the business and other developers.​

This improves discoverability—when reading the codebase, it is obvious where to find and place logic related to a given business concept, which in turn makes onboarding and code reviews easier.​

Better type safety and testability

Shared abstractions for a business data type make method contracts more precise and avoid treating everything as a generic “bag of fields.”​

Tests can target the common behavior in one place and then focus on the differences in specialized entities, resulting in leaner test suites and less chance of inconsistent behavior between similar entity types.

If you want to run Python scripts online (even heavy ones, with libraries, networking, etc.), one popular answer suggests using PythonAnywhere — their free plan gives you limited CPU time and storage, allows pip-installs, and supports network access. Stack Overflow

If you prefer something simpler for sharing/running Python directly in the browser, you might try Pybadu (http://budibadu.com/pybadu) — it gives a quick online-Python environment with minimal setup.

You can share a Python program online by turning it into a small web app (Flask/Django) and hosting it so the code stays on the server, hidden from users. If you want an even simpler option, try Pybadu (https://budibadu.com/pybadu), which lets you run and share Python projects directly in the browser.

if you look at your error stack you can see that useAuth hooks is being called from inside another hook called useAxiosAuth and before that inside fetchUsers (app/context/user.tsx) which can't be done because it ain't a component

To integrate Superset and Keycloak, use this configuration (superset_config.py):

from flask_appbuilder.security.manager import AUTH_OAUTH

AUTH_TYPE = AUTH_OAUTH

OAUTH_PROVIDERS = [
    {   
        'name':'keycloak',
        'token_key':'access_token',
        'icon':'fa-address-card',
        'remote_app': {
            'client_id':'<superset>',
            'client_secret':'<secret>',
            "client_kwargs": {"scope": "email profile"},
            'server_metadata_url': '<keycloak_url>/realms/destra/.well-known/openid-configuration',
            'api_base_url': '<keycloak_url>/realms/destra/protocol/',
        }
    }
]

AUTH_USER_REGISTRATION = True

AUTH_USER_REGISTRATION_ROLE = "Public"

The tricks are in the Flask-AppBuilder implementation:
* Use "keycloak" as provide name (Extends SupersetSecurityManager it is not necessary)
* Fix the "api_base_url" parameter
* The scope must have "openid", "email" and "profile" entries.

This links can help you understand how Flask-AppBuilder handle with keycloak integration:
https://github.com/dpgaspar/Flask-AppBuilder/blob/f4a8cfd9f31f7eb36fb7891ccf9747b7506a41d3/examples/oauth/config.py#L110

https://github.com/dpgaspar/Flask-AppBuilder/blob/f4a8cfd9f31f7eb36fb7891ccf9747b7506a41d3/flask_appbuilder/security/manager.py#L690

The image disappears on mobile because the left flex column(flex : 15%) is allowed to shrink to 0 width in small screens. flexbox on mobile compresses it completely, so the image (100% of 0) becomes invisible.

To prevent shrinking, we can set minimum width :

.fasciasx { flex-shrink: 0 ;}

or

.fasciasx { min-width: 120px; }

this one keeps the image visible on mobile screen.

Can you provide the location in the article where it talks about this? The closest that I can find is "Converting to SSA", where it only mentions the use of phi in non-looping scenarios.

@Afolabi Adedayo Thanks for your comment. Yes, I understand that. But my question is the following. Suppose that I have only added one file, foo, to the cache, by doing git add foo. I fully understand why it needs to run diff old_foo foo. But why does it need to do any comparison at all on all of the other files in the repo? Aren't all the other files in the repo irrelevant when I am only committing foo?

@user207421 I'm a professional software developer and am very familiar with the XY problem. This is not an XY problem. I am interested in all of the following: how it works, why it works that way, and how I can make it faster. But the question is primarily about the "how it works" part.

Two emails are attached to my Microsoft rewards. I sign in with one it tells me to switch n it’s a tennis match saying the exact same thing each time! What do I do

The request initiator chain marks the inspected request in bold.
Above this request are its initiators and below are its dependencies.
You read this from top to bottom.

When you hold the Shift  key and hover over the network requests the initiators are marked green and the dependencies are marked red:

And as you said the Request call stack is read from bottom to top.

@ADyson That sounds like the best solution in the long run. But I was hoping for a quicker solution, even if it's a bit of a hack. I should have specified that this isn't a high priority issue: I was just looking for something that might work with minimal rework. As far as the link to that article on relative vs absolute paths, much appreciated. I used relative paths for a while, but absolute paths were the most reliable. Thanks again!

You whack my whole family and then post about it like it's a fucking experiment while you guy us dry. Real fucking amusing.

Stay tuned. We(WoolyAI) will be releasing something very soon that will let you run PyTorch 'torch.device("cuda")' inside Docker on Mac using MPS acceleration.

Actually I should probably save myself any more explanations and just tell you to read https://phpdelusions.net/articles/paths

@ElainaTruhart Non-negative doesn't matter, but "increases as the index increases" is indeed what it means. So yes, you should be using the bisection offered by searchsorted.

Judging from an intense five-hour scour of a zillion Internet sources, it looks like v3.x.x of VLC changed (broke!) multi-item-playlist streaming. When commanded to stream a list of multiple items, only the first item makes it through to the "source" specification page, let alone to the Playlist -- and even if other items are added to the Playlist after streaming has begun, "streaming-ness" attaches only to the first item on the Playlist -- all the other items play locally, and when play circles back around to the first item (or it is reselected manually), a whole new instance of the stream is cranked up and THAT ONE ITEM is streamed again. Internet anecdotes support my impression that "this used to behave differently."

So if whatever YOU guys are talking about, simply takes a stream feed from VLC and does something else with it, and your problem is that suddenly you're only getting one item and then you lose the connection -- this is why. VLC is closing/deleting the stream -- the very ports on which you connect to the stream! -- at the end of the FIRST ITEM in the playlist.

I don't know of any other solution besides keeping a copy of VLC 2.x.x around for this purpose, and hoping all future versions of Windows (or Linux, or whatever) continue to be able to run it. I'm planning to TRY to take this up with the VLC people, but I don't have much hope: if they haven't fixed it by NOW, they might have their own reasons NOT TO.

The ones in the middle are unachievable using plain HTML / CSS you may

• Use a image jpeg / png, and artificially crop and adjust to the elements background

• Use SVG bezier calculations with a <path> tag

As far as I’m aware dbt Snowflake doesn’t support materialized views

I have just managed to open a Query Tool in the Query Tool Workspace.

First, I disconnected from my database in a Default Workspace. Then I manually entered the Database and User names in the corresponding fields in the Query Tool Workspace, entered the password, removed all rows from the "Connection Parameters" section, and after all that, pgAdmin finally allowed me to connect.

Works the same if I don't select anything or clear the "Existing Server (Optional)" field and manually enter other info.

To be honest, this Query Tool Workspace feature is kinda disappointing. I did not achieve the expected results with it, and this "bug" with the connection just annoyed me.

Git doesn’t diff every file.
It first checks cheap metadata (mtime and file size) for each tracked file. If those match the index, Git skips the file. If they don’t match, Git then reads the file, hashes its content and compares it to the index version. Only those “suspect” files get fully diffed.

Read the Chapter 2. Peer-to-Peer Communication by Means of the Selections of the Inter-Client Communication Conventions Manual. See also Chapter 3. Peer-to-Peer Communication by Means of Cut Buffers. I would also recommend to read the section about Atoms.

Everything you want to know is basically explained in the conventions.

const getUserNameFromEmail = (email) => // Arrow function hence the arrow on "=>""
{
    return email.slice(0, email.indexOf('@')); // just gets user and truncates rest of email from @ forward  
}
console.log(getUserNameFromEmail('[email protected]'));

Just in case anybody else had this problem, I was getting this when trying to install pytorch3d. Torch was installed and working properly but installing pytorch3d with pip was not possible and it kept giving me """No module named "Torch"""

I have fixed it with adding "--no-build-isolation".

The final command: "pip install --no-build-isolation git+https://github.com/facebookresearch/pytorch3d.git@f34104cf6ebefacd7b7e07955ee7aaa823e616ac#egg=pytorch3d"

Cheers

For me, using Linux, I had to list the name of the font I wanted to use in my config.json, found at ~/.config/rstudio/. I listed the one I wanted as fixed width and had no issues after. I could not find any other way to update my fonts.

Well, which IP address do you want your container to listen on? 0.0.0.0 is just the wildcard address, which means the container listens on all interfaces. As long as the port is published, you can reach it via the loopback address (127.0.0.1:5000) as well as via the host’s LAN IP address, for example 192.168.1.100:5000.

Guessing by the screenshot, the error seems to caused because the header has 7 fields and the rest of the data has 8 (probably a trailing comma). This causes read.csv to use the first field (Date) as the "row name" of the dataset.

From the help page for read.csv (to access use ?read.csv):

If there is a header and the first row contains one fewer field than the number of columns, the first column in the input is used for the row names. Otherwise if row.names is missing, the rows are numbered.

Try to verify if your original CSV files looks something similar to the csv_text I created below.

  # Header has 7 fields, data rows have 8 (trailing comma)
  # → First field (Date) becomes row names (per read.csv docs)
  # → Trailing empty field gets assigned to "Volume"

  csv_text <- 'Date,Name,Close,High,Low,Open,Volume
08/31/2015,eBay Inc,27.11,28.935,23.23,28.09,271799372,
09/30/2015,eBay Inc,24.44,27.60,23.76,26.54,267684281,'

  read.csv(text = csv_text)
#>                Date  Name  Close  High   Low      Open Volume
#> 08/31/2015 eBay Inc 27.11 28.935 23.23 28.09 271799372     NA
#> 09/30/2015 eBay Inc 24.44 27.600 23.76 26.54 267684281     NA

The best solution I can suggest is to use {readr} and read_csv which will correctly import the data while also warning you about the error. Note that I added locale to properly parse your date which seems to be in the MM/DD/YYYY format.

  library(readr)
  # The I(...) function is used only because my csv is a 'string' and not an actual
  # csv file stored in my computer.
  read_csv(I(csv_text), locale = locale(date_format = "%m/%d/%Y"))
#> Warning: One or more parsing issues, call `problems()` on your data frame for details,
#> e.g.:
#>   dat <- vroom(...)
#>   problems(dat)
#> Rows: 2 Columns: 7
#> ── Column specification ────────────────────────────────────────────────────────
#> Delimiter: ","
#> chr  (1): Name
#> dbl  (4): Close, High, Low, Open
#> num  (1): Volume
#> date (1): Date
#> 
#> ℹ Use `spec()` to retrieve the full column specification for this data.
#> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
#> # A tibble: 2 × 7
#>   Date       Name     Close  High   Low  Open    Volume
#>   <date>     <chr>    <dbl> <dbl> <dbl> <dbl>     <dbl>
#> 1 2015-08-31 eBay Inc  27.1  28.9  23.2  28.1 271799372
#> 2 2015-09-30 eBay Inc  24.4  27.6  23.8  26.5 267684281

So, your final solution should be this:

library(readr)
ebay <- read_csv("EBAY.csv", locale = locale(date_format = "%m/%d/%Y"))

Please let me know if this fixes your issue. In the future, use something like dput(head(ebay)) to output a small version of your dataset.

^{Created on 2025-12-04 with reprex v2.1.1}

You will likely need to build the application remotely. when with manylinux there is likely wheels that to match up with runtime of the azure function app. When build automation is enabled, App Service activates a virtualenv and runs pip against your requirements.txt on Linux, resolving proper wheels automatically. Research how to remote build your application with the Azure functions app and that will most likely fix your dependency issue.

C:\Users\(! NAME OF THE USER !!! )\AppData\Roaming\Litecoin

I ended up doing this in mvvm. I got that working. It's just for a demo, so not horribly important, just providing a set of options to a client.

  const obj = Slides.Presentations.Pages.getThumbnail(
    presentationId,
    slide.getObjectId(),
    { "thumbnailProperties.thumbnailSize": "LARGE", "thumbnailProperties.mimeType": "PNG" }
  );

"Slides" is not defined.

🧐 with Bezier calculations...

This is currently not possible, see Allow wiki edits from non-members in project wikis (#25177) · Issue · gitlab-org/gitlab. This ticket was opened in November 2018. It was closed in September 2025 by Matthew Macfarlane. The reason given was: "we have decided that this request will not be prioritized in our upcoming 12-24 month plans, and as such, will close this issue in an effort to reduce the noise in our product backlog." (copied 2025-12-04)

Contributor Markus Koller explained in October 2020 the current status:

Currently, write permission in wikis is hard-coded to members with role "Developer" and above.

This Wikipedia article explains it pretty well.

I just had to remove 'mongodb://127.0.0.1/basic-setup'.

This was my issue:
If the file path is over 255 characters, then System.Drawing.Image.Save will fail regardless if long paths are enabled in the registry.

So I followed this tutorial https://brentonmallen.com/posts/circular_qr_code/circular_qr/

and manage to obtain this

Example QR Code

i was using global.getDotnetRuntime(0).Module.HEAPU8.subarray to get from the pointer on c# side, but in net 10 it change to globalThis.getDotnetRuntime(0).Module.HEAPU8.subarray
i thing is a little more easy to use on javascript and leave it here if another person was using global.

Fix 1: Add !imporatnt to force the color

Fix 2: Increase CSS specificity

nav.navbar.navbar-expand-lg{

background-color: rgba(var(--bg-rgb), 0.2);

}

Fix 3: css file should load after bootsrap 

start the web browser first then the .net app

ex.:

@echo off

start "" "http://localhost:5000"
cd "C:\publish_path\"
dotnet Program.dll

Once reCAPTCHA keys are migrated to Google Cloud they are no longer considered Classic keys. All reCAPTCHA Classic keys are in the process of being migrated and will be subject to the 10k monthly assessment free-tier (see Billing information). Once migrated, there are no changes required to your integration and SiteVerify will continue to function as-is.

To make sure I understand your question, are you seeing keys in the "Classic Keys" table pictured in this documentation? If so, those keys need to be migrated. Either you can click Upgrade key, which is recommended if you'd like to choose the destination project, or reCAPTCHA will automatically migrate them to a new project shortly.

Please see more details in the Migration overview.

You could do a lot worse than read this amazing article by Stephen Toub - ConfigureAwait FAQ by Stephen Toub.

There's a section devoted to When should I use ConfigureAwait(false)?

And also I’ve heard ConfigureAwait(false) is no longer necessary in .NET Core. True?

This should answer any questions you have on the subject.

Remember second-level caching in NHibernate? Pepperidge Farm remembers...

Were you able to figure out any solution for this? I too have a web app and want to download an Excel file and read it contents to assert. This works fine locally but not on gitlab pipeline running on Linux server

The following complements @Svante's answer, by taking into account nested lists.

(defun nested-list-to-ps (lst)
  (if (listp lst)
      `(ps:array ,@(mapcar #'nested-list-to-ps lst))
      lst))

(defun example () '(1 (2 3) (4 (5 6))))
(ps:ps (ps:lisp (nested-list-to-ps (example))))
; => "[1, [2, 3], [4, [5, 6]]];"

Please help I still get this error after trying that out

You don't have permissions to access this datastore "workspaceblobstore" either due to issues with network setup or issues with access. We are unable to get more details at this time. You can try again later.Find steps to troubleshoot here.

@Philip, unfortunately,

| summarize dcount(B) by A | where dcount_B > 1;

times out for me, but based on answers to does-a-b-s-where-s-is-a-set-of-tuples-a-b-a-in-a-b, I did find a correct approach that does not time out

// T is the name of a table
// A and B are each a name of a column in the table
let IsOneToOneRelation = (T:(*), A:string, B:string)
{
    // 1) Project the requested columns by name (strings) -> ACol, BCol
    let S =
        T
        | project
            ACol = column_ifexists(A, ""), 
            BCol = column_ifexists(B, "")
        // | where isnotempty(ACol) and isnotempty(BCol) // and remove rows with empty/null values.
        // keep only unique (A,B) pairs to avoid duplicated rows skewing counts
        | distinct ACol, BCol;
    // 2) Compute the three cardinalities in one go
    let Acnt = toscalar(S | distinct ACol | count); // do not use `S | dcount(ACol)` b/c dcount is an estimation
    let Bcnt = toscalar(S | distinct BCol | count);
    let ABcnt = toscalar(S | count);
    // 3) Verdict: bijective iff |pairs| == |A| == |B|
    ABcnt == Acnt and ABcnt == Bcnt
};
print IsOneToOneRelation(TenantSnapshot, "Id", "NodeName");

enter image description herereplace last line of the text to Vishnu-24ET102372

Why is this an "open-ended question"? This has rather concrete answers I would say.

Maybe this repo can help you :)
https://github.com/edinsalimovic/SRDoubleStickyHeaderList

For a use case like this, you might want to consider opensource StyleBI. It’s designed to handle large-scale, multi-tenant analytics with high cardinality efficiently, and you can embed dashboards directly into your web app. Because it’s serverless and elastic, it can scale automatically with your data volume without incurring the same costs as continuously sampling millions of traces or sending high-cardinality metrics to Datadog. StyleBI also supports connecting directly to common data sources and applying filters like user location or application usage, making it easier to build the insights you need without overspending.

1. Please don't use images for data - use table markdown.

2. This should have been a proper Q&A - you want a concrete answer, not a best practice.

3. How is a case expression (originally incorrectly called a case statement) relevant?

In the current snippet styelsheet is misspelled, so the browser ignores it and never loads the CSS. <link rel="stylesheet" href="css/main.css">

You can refer this: The External Resource Link element

This web site's focus is programming, though, not general computer issues.

@marv51 That worked great, thank you! I must have run the publish flow when I first made the project, because it's been in there since the beginning.

We can use border="0" atttibute in the <td> tag

What about this alternative: in the aggregate constructor apply an extra event:

public GiftCard(IssueCardCommand cmd) {
    ... 
    apply(new CardIssuedEvent(...);
    apply(new CardRedeemedEvent(/*defaults*/);
}

This would avoid the replay issues and still work for both types of aggregates, wouldn't it?

Anyone reading this in 2025, msw 2 does not work nicely with projects created with create-react-app. Stick with msw 1.x.x version if your project uses create-react-app.

To answer my question (thanks to robertklep in the comments): I was not aware that apt install behaves differently compared to pip3. 'apt install' draws packages from linux distribution whereas pip3 draws packages from the supplier (in my case Pypi). I was trying to avoid venv´s because my code runs by a user and also by a service...i know there are solutions for this but it adds to complexity....

Nope, doesn't work at all. Command line says it doesn't recognize any sox, ./sox, sox.exe, etc. Thanks for nothing!

I tried https://dansleboby.github.io/PDF-Form-Inspector/ & it says 0 fields found

Depending on the literature, I sometime find q1 = (C_23 - C_32) / (4 * q_c) and sometimes the other sign around, and I'm not sure why, which would explain this conjugate issue. Could you help me please?

Part of your math is following the convention for selecting the positive root of q_c, and part of your math is following the convention of selecting the negative root of q_c.

The paper Computation of the Quaternion from a Rotation Matrix has a good explanation of the process for finding a quaternion from a rotation matrix. There are multiple formulas you can use, and some have better numerical stability than others, but the one you are using is "0.1.2 Solution of diagonal for b₁."

(Note that Farrell follows a scalar first convention, and that q_c in your code corresponds to b₁ in the paper.)

In this step, he solves the equation 4*b1**2 = 1 + R_11 + R_22 + R_33 by taking the square root, to obtain b1 = np.sqrt(1 + R_11 + R_22 + R_33) / 2. In this step, he is taking the positive root. However, it is equally valid to say that b1 = -np.sqrt(1 + R_11 + R_22 + R_33) / 2 is a solution.

He addresses this choice in the summary:

Each of the quaternions involves a sign ambiguity due to the fact that either the positive or negative square root could have been selected. This document has selected the positive square root throughout. If the negative square root is selected, then the direction of the vector portion of the quaternion will also be reversed. This results in the same rotation matrix.

I am guessing that that is where your confusion stems from: you are combining code from a source that uses the positive root to obtain the scalar component with code from a source that uses the negative root to obtain the imaginary components.

The simplest fix is to swap the signs here:

    q_vec = np.array([[C_32 - C_23], [C_13 - C_31], [C_21 - C_12]])  / (4*q_c)

Appendix: code used to test this answer

import numpy as np
import scipy.spatial


a_quat = np.array([0.1967, 0.5692, 0.5163, 0.6089])

print("Original quaternion", a_quat)
a_rotation = scipy.spatial.transform.Rotation.from_quat(a_quat)
a_matrix = a_rotation.as_matrix()
print("Matrix")
print(a_matrix)

def convert_dcm_to_quaternion(dcm):
    """
    Convert DCM to a quaternion
    """
    C_11 = dcm[0,0] #angle between vector 1 of initial frame and vector 1 of rotated frame
    C_12 = dcm[0,1] #angle between vector 2 of initial frame and vector 1 of rotated frame
    C_13 = dcm[0,2]
    C_21 = dcm[1,0] #angle between vector 1 of initial frame and vector 2 of rotated frame
    C_22 = dcm[1,1] 
    C_23 = dcm[1,2]
    C_31 = dcm[2,0]
    C_32 = dcm[2,1]
    C_33 = dcm[2,2]
    q_c = 1/2 * np.sqrt(C_11 + C_22 + C_33 + 1) #consider that scalar value != 0, i.e. not at a singularity. Use Markley or Shepperd methods otherwise.
    q_vec = np.array([[C_32 - C_23], [C_13 - C_31], [C_21 - C_12]])  / (4*q_c)
    q = np.vstack((q_vec,q_c ))
    q = q.flatten()
    return q

print("converting back")
print(convert_dcm_to_quaternion(a_matrix))
print()

"Good enough" isn't good enough when you spend a zillion hours on working around something that should be easy to do. The first thought is "what am I doing wrong?", followed by googling and looking at stack-overflow, when you have the second realization "ok, everybody is saying my thinking is wrong, and that it is just good enough, and just live with it", followed by the feeling "are these people crazy?", then "F-that" for using crappy software. I came from an era where software was very well designed, not now. What an effing mess.

remove the

finally:
    await db.close()

from the async get_db function. This should resolve.

Just in case when somebody wants to disable animation:

@State private var showPortfolio: Bool = false   
CircleButtonView(iconName: showPortfolio ? "plus" : "info").animation(.none, value: showPortfolio)