Have you tried creating a python package of the script? I believe at minimum you would require an empty "_init_" file and toml file. After creating the package, you should be able to run your script from any directory.
Tailwind’s responsive lessons (sm:, lg: and so forth.) are based totally on viewport width, not the width of a figure field. So even in case u set w-[370px] on a div the browser viewport is still huge, which means that sm:grid-cols-2 and lg:grid-cols-three hold to apply. That’s why u see grid-cols-3 on computing device-Tailwind is asking on the browser length, not ur simulated container length. If u want to truly simulate devices inside a element, u’ll need to either: Use browser DevTools responsive mode (max straightforward), Wrap the preview in an <iframe> and manipulate its width, or Enable Tailwind field queries (v3.2 ) so patterns respond to the field size rather than the viewport.
They added an easy copy button to the channels setting modal. Just go to the channel, then just click on the channels name at the top. Make sure you are on the about tab inside that modal and then at the bottom it says Channel ID: xxx and there is also a quick copy button.
The main issue isn’t Tailwind itself, it’s that responsive logic (`sm:/md:/lg:`) is hard-coded in every component. That couples breakpoints directly to JSX and makes global updates painful. The scalable fix is to introduce a thin abstraction layer:
Layout primitives (`Grid`, `Stack`, `Container`) → take props like cols/gap and generate the responsive classes.
Variants (`tailwind-variants`/`cva`) → define reusable styles and states (tone, density, padding) instead of repeating class strings.
Shared Tailwind preset → centralize screens, colors, spacing; change them once, all apps update.
Container queries → when layout depends on parent width instead of global viewport.
Design tokens in CSS variables → theming and density modes without JSX refactors.
This way you keep Tailwind’s power but gain central control, consistency, and scalability.
- **Docs:** usage recipes + “do/don’t” examples; not encyclopedias.
- **Migration plan:** 1) add preset & primitives, 2) migrate worst offenders, 3) enforce lint rules to block new raw utility soups.
- **Monorepo ops:** publish `@org/tw-preset` + `@org/ui`; version them and document breaking changes.
- **Ergonomics:** ship default responsive presets (e.g. `cols={presets.twoUpToThree}`), global density modes.
- **Edge cases:** SSR determinism, iframe/microfrontend theming via CSS vars, always expose `className` escape hatch.
Bottom line:
Stop scattering `sm:/md:/lg:` everywhere. Move them into controlled wrappers, centralize tokens, and let Tailwind do the heavy lifting behind clean APIs. That’s how you keep responsive design flexible and scalable without drowning in utility chains.
This is a feature that should be added immediately. Rider is crazy to navigate. I use the solution view to scope to my project. But because I am using Unity, Rider also sees all of my plugins and extension code. Even though I can't edit any of it.
Everyone knows that adding a feature to a feature request for a large company will mostly get it swept under the rug.
It would be incredibly helpful to just have a window on the left pane that displays only the methods of the class I have open. It should also be able to scope to parent/child classes, interfaces and other linked functions. I know of the structure window but its useless also adding in all the variables. It even positions the variables in the order they occur in the file. I get variables, function, variables, functions that I need. In this huge list. What a mess.
I’ve used the TYPO3 extension "pictureino" to handle this. It automatically generates responsive images in the frontend while keeping the cropping set in the backend. The nice thing is that it works completely automatically, without any extra configuration.
This is absurd but I found that if I launch the app using Xcode (Build & Run), the menu items don't appear. But if I double-click it from the Finder, then it works. This is on macOS 14.
It's possible that when the method with the break statement was executing, its thread got interrupted and another one started executing, which increased the measured execution time of the method. In that case, it would be helpful to try executing each method multiple times and then finding the average runtime for each method. An example of this is available at 300x slower program with while.
If you want to know why this behavior is occurring, then maybe it is because you can't declare two variables with the same name in Javascript? Is there a reason why both values need to have the name MY_FAV?
Note that even though currently multi-objective tuning is not supported in mlr3, there are similar situations where these multi-objective problems appear and pareto optimal solutions that represent the best trade-off between (2 or more) objectives have been proposed (as there are many algorithms on findings "knees" of the (multi-dimensional) pareto front - see a nice review in this article).
In a recent feature selection example, I implemented a very simple 2d knee point identification method, to find the pareto point point with as minimum selected features as possible, while retaining as high performance as possible, see this mlr3 gallery post.
The command you're attempting to run tells your system to delete the com.apple.quarantine attribute from the file, but the error message that you're getting says that it doesn't exist, which means it was already deleted or was never added. Either way, you can safely skip running that command, as the desired outcome has been reached!
Since Safari 10, the debugger console has been greatly improved and supports console logs and breakpoints in dedicated workers. Service workers can be debugged by going to Develop > Service Workers.
Thanks Psionman! Exactly what I needed to create a wx.Image from a PIL image (I didn't actually need a bitmap). The 1st line of the function might be simplified(?) to wx_image = wx.Image(*pil_image.size)
When choosing the best cloud solution for handling JSON requests, the right option depends on your project’s scale, performance needs, and integration requirements. JSON (JavaScript Object Notation) is lightweight, easy to parse, and widely supported, making it the standard for modern APIs.
Popular Cloud Options:
AWS Lambda + API Gateway: A serverless choice for quickly processing JSON requests without managing infrastructure. Great for scalability.
Google Cloud Functions: Ideal if you’re already in the Google ecosystem. It handles JSON efficiently and integrates with Firebase and BigQuery.
Microsoft Azure Functions: Offers robust JSON handling, especially for enterprise-level applications with strong security needs.
Key Considerations:
Scalability – Can the service handle sudden spikes in requests?
Latency – JSON parsing should be fast for real-time applications.
Ease of Integration – Look for services with SDKs and REST API support.
Cost-effectiveness – Pay-as-you-go serverless models are often budget-friendly.
For businesses that need tailored IT solutions beyond just cloud hosting, exploring resources like TTS can be helpful. They provide practical insights and services for integrating digital technology into real-world business needs, ensuring your infrastructure supports growth and flexibility.
Pro tip: If you’re just starting, try serverless platforms first—they’re low-cost, easy to manage, and scale automatically with your JSON requests.
This works well because embeddings + LLM can capture meaning and handle fuzzy matches.
What a Knowledge Graph (KG) / Graph-RAG adds
Structure: You can connect rules → groups → keywords → example sentences.
Disambiguation: Add explicit links like “6th day ≠ 7th day” or “evening ≠ night” so the system knows how to separate similar rules.
Explainability: Easier to show why a sentence matched a rule (“matched Rule X because of keyword Y and example Z”).
What it doesn’t solve
New rules: A KG can flag “unmatched sentences,” but creating a new rule is still an expert job.
Accuracy: If embeddings + LLM already work well, a KG won’t suddenly make results much better.
Maintenance: Building and updating a KG for 300+ rules takes work/
Don’t replace your current pipeline.
Use a small KG as an extra layer for disambiguation and explanations.
For new rules, cluster unmatched sentences and let experts decide.
A KG/Graph-RAG can help with clarity, disambiguation, and trust, but it won’t replace what you already built. Think of it as a way to organize and explain results, not as a magic accuracy booster.
In Python 3, the standard division operator (/) always performs "true division" and returns a float result, even if both operands are integers and the division results in a whole number.
It is a Firefox issue. Firefox remembers old data entered into the input fields and replaced those after reloading. Currently looking for another solution...
#edit:
Seems like adding autocomplete="off" to the form is doing the trick.
Your expected value is not possible since the getResponse() method returns an array of arrays (a two-dimensional array), just like @Wicket said.
When you submit a form, the FormResponse object contains all your responses. To get the answer for a specific form item, like a CHECKBOX_GRID, you use the getItemResponses() method on the FormResponse object.
The key to a CHECKBOX_GRID is that the getResponse() method returns an array of arrays, not a simple string or a flat array of strings. Each inner array corresponds to a row in the grid and contains the column titles of the selected options for that specific row.
For example, a grid with rows "X-Small," "Small," and "Medium," and columns "White" and "Navy."
If you selects "Navy" for the "X-Small" row, selects nothing for the "Small" row, and selects both "White" and "Navy" for the "Medium" row, the getResponse() method would return: [['Navy'], [], ['White', 'Navy']] or as string Navy,,White,Navy.
The first inner array ['Navy'] represents the selections for the "X-Small" row.
The second inner array [] is an empty array and represents the "Small" row, where no checkbox was selected. This is the correct way to handle unselected rows, not by returning null or an empty string ''.
The third inner array ['White', 'Navy'] represents the selections for the "Medium" row, showing multiple choices within a single row.
As a Java developer, I strongly recommend using ID instead of <tableName>_id in case of primary key.
When working with @OneToOne, @ManyToOne or @ManyToMany properties, you need to specify the @JoinColumn annotation with the values of the "name" and "referencedColumnName" properties. When they have the same values, it can be very confusing. Although I understand that by default for a primary key, referencedColumnName is not used, but sometimes it is inconvenient and takes a little more time.
The error means Document AI can’t find the processor version you’re asking for.
In your code you are mixing project identifiers and using a processor version that doesn’t exist.
A few things to check:
1. Make sure you use the same project everywhere. Document AI accepts either the project number ( 466668368501) or the project ID (inspired-ether-458806-u7), but you must use the same one consistently.
2. If you don’t have a custom processor version, don’t pass processor_version_id. Just build the resource name like this:
python:
name = client.processor_path(project_id, location, processor_id)
For security, the Auth schema is not exposed in the auto-generated API. If you want to access users data via the API, you can create your own user tables in the public schema.
What you’re seeing is Anaconda automatically activating the base environment in your shell. That’s why every time you open PowerShell or VS Code terminal, it starts with (base) in the prompt.
You don’t need to uninstall Anaconda — you can simply tell it not to auto-activate:
conda config --set auto_activate_base false
After running this command once, restart PowerShell/VS Code and it will open in the normal terminal without (base) showing up.
When you do want to use conda, you can still activate it manually with:
conda activate base
or switch to any other environment you’ve created.
In VS Code specifically, also make sure you’ve selected the Python interpreter you want (Ctrl + Shift + P → Python: Select Interpreter). That way it won’t keep defaulting to conda if you don’t want it to.
This way you can keep Anaconda installed for data science projects, but still have a clean, fast PowerShell/VS Code terminal for your everyday Python work.
I also found myself needing a hook that runs after new refs are fetched from the remote, no matter if I merged them into a branch or not.
Given the lack of post-fetch hook, I made a python script to simulate it. It works by wrapping the ssh command and call the post-fetch hook after a fetch happens.
Making sure that the views and view models in your subregion implement INavigationAware or IConfirmNavigationRequest will allow Prism to automatically call their OnNavigatedFrom/OnNavigatedTo methods during navigation. This is a more elegant way to take advantage of Prism's built-in navigation and region lifecycle management. To handle the subregion lifecycle consistently with the parent view and make your code cleaner and easier to maintain, think about utilizing scoped regions or navigation-aware actions rather than manually deleting views.
So, the solution ended up to be quite trivial. The only thing I missed is adding the appState.target property to the parameters of the loginWithRedirect method. So in case you're facing the same issue, do this:
Currently, you're using a Hardcoded Attribute Mapper in Keycloak. This mapper does not extract dynamic values (such as the user ID) from the identity provider token. Instead, it assigns predefined static values to user attributes after a successful login.
For example, if you configure a Hardcoded Attribute Mapper for the email attribute with the value [email protected], then after a user logs in via an identity provider like Twitter, the user's email attribute will be set to [email protected].
If you want to map the user ID or other token claims dynamically, you should use a "User Attribute Importer", "Claim to User Attribute", or "Attribute Importer" mapper — not the Hardcoded one. I am not sure if those mapper are available in the keycloak 26.x.x version and keycloak provide an options to create a own custom SPI.
0 minutes seems like a riskier length of time to allow for one token to be active. There is some drift with tokens, you don't need to enter it within 30 seconds, but half an hour is not among any recommended lengths of token expiration time.
I agree with @mice here. He isn't talking about storing the passed hash, but a sha256 hash of it. Using heavy hash functions, versus lighter hashes, reduces the random password length required for the same security against brute force cracking.
Yes, if you could reverse hash the sha256 hash stored on the server to the 'random' 32 byte binary output of the heavy hash, you could use that as the password, but that isn't feasible with current technology. The alternative would be to start with the actual password candidates and calculate both the heavy hash and the sha256 hash a gazillion times until you find the one that produces the stored hash. This could be feasible if the password is weak, but that wouldn't be the fault of the system.
In short, I see nothing wrong with this idea if implemented correctly, and sufficiently strong passwords are chosen. In practice though, it only allows you to shorten password by a couple of characters for equivalent security, so is it worth it?
Stick with @Enumerated(EnumType.STRING). The values (PENDING_REVIEW, APPROVED, REJECTED) are stable and not business-configurable. Renames can be handled via a one-time DB migration if it ever happens. The extra complexity of lookup tables or converters isn’t worth it unless requirements change.
I edited the dev.to page, clicked Preview, then Save. Then the images appeared again. This shows that the pages needed to be regenerated this way for some reason.
Including model.safetensors.index.json will solve the problem locally but if you are using huggingface repository (Private Space provide you with that much space to save the model - 100 Gb as of now), You will still face the error of loading the model, somehow huggingface repository is treating it as a safetensor file (having stack icon alongside). And transformers library by default look for the same file name convention. Even after trying multiple attempt still facing the issue.
I had such problem in Laravel 9. Solution - check your version Bootstrap. After installation of correct version bootstrap problem was solved. I hope this help you
I tried many solutions but nothing worked for. I added one more uses feature and now the app is installed in all the devices, whether the device has NFC or not
This error for me was due to upgrading a package dependency. Once downgraded, the error went away. Added back, the app crashed again via the dreaded "Lost connection to device. Exited." message. It took a while though because the dependency package was upgraded along with some other packages and SDK-related stuff.
I'd suggest one of two things; 1) start disabling packages, or 2) create a new app and add your app into it, piecemeal.
The dependency in my case (every case would likely be different) was the http package (v1.5.0). Once downgraded back to v1.4.0, the crashes stopped.
I've had this happen with other packages, though, also, like routing packages, and they can take hours to days to debug because the error can happen in some asynchronous call that happened dozens of debugging steps before the actual crash, and the logs give no help whatsoever.
Lost connection to device. Exited. | GL figuring it out. :P
CMake does not naturally supports different C/CXX compilers in CMake Project. However, it does support "subdirs". Which means CMake understand that different project have completely different toolchains.
Just make an independent CMake for the MCU firmware and independent one for the SBC.
The third CMake will be your root CMake, it should add the others as subdirs and make the joint.
ترجمه: «و قطعاً شما را به چیزی از ترس و گرسنگی و کاهش اموال و جانها و ثمرات ابتلاء (امتحان یا آزمایش) میکنیم؛ و مژده ده به صابران، همان کسانی که چون مصیبتی به آنان برسد، گویند: ما از آنِ الله (سبحانه و تعالی) هستیم و به سوی او بازمیگردیم.»
چند روز قبل؛ زمینلرزهای قدرتمند ولایات مشرقی، بهویژه کنر و ننگرهار و مناطق اطراف را لرزاند. این حادثه تعداد زیادی را شهید و شمار زیادی را زخمی و بیخانمان ساخت. داغ این حادثه قلوب همه ما را سوزاند. از بارگاه پروردگار سبحانه و تعالی میطلبیم که شهدا را در فردوس برین جای دهد، بر مجروحان شفای عاجل ارزانی کند و بر دل بازماندگانشان صبر جمیل نازل فرماید.
رسول الله ﷺ در حدیث مبارکی که از صهیب (رض) در صحیح مسلم روایت شده است فرمودند: «عَجَبًا لِأَمْرِ الْمُؤْمِنِ، إِنَّ أَمْرَهُ كُلَّهُ لَهُ خَيْرٌ، وَلَيْسَ ذَاكَ لِأَحَدٍ إِلَّا لِلْمُؤْمِنِ؛ إِنْ أَصَابَتْهُ سَرَّاءُ شَكَرَ، فَكَانَ خَيْرًا لَهُ، وَإِنْ أَصَابَتْهُ ضَرَّاءُ صَبَرَ، فَكَانَ خَيْرًا لَهُ.» ترجمه: شگفتانگیز است حال مؤمن! زیرا همه کار او برایش خیر است، و این جز برای مؤمن نیست: اگر خوشی به او برسد، شکر کرده و این برایش خیر است؛ و اگر سختی به او برسد، صبر میکند و آن نیز برایش خیر است.
آری! هرچند این مصیبتها تلخ و سنگیناند، اما برای اهل ایمان دریچهای برای صبر، بازگشت به پروردگار و بیداری دلها میباشند. اینگونه حوادث به ما یادآوری میکنند که دنیا گذراست و آنچه باقی میماند ایمان و اعمال صالح است.
با وجود دعا و صبر، امت باید بیدار گردد که دولتهای ملی قادر به ادای مسئولیتهای اساسی نیستند. سالها حاکمیت این دولتها بر افغانستان، با وجود سرازیر شدن میلیونها دالر خارجی و جمعآوری مالیات داخلی، نتوانست زمینه اسکان امن و تدابیر لازم را برای مردم فراهم سازد. در حالی که بر دولتها لازم است وظایفشان را بهگونه تخنیکی و عملی انجام دهند: نصب دستگاههای هشداردهنده زلزله و ایجاد شبکههای اطلاعرسانی فوری، آموزش مردم در برابر حوادث، ساخت منازل و تأسیسات مقاوم، از جمله وظایف حیاتی است که متأسفانه دولتهای ملی در طی سالیان متمادی از انجام آن عاجز بودهاند. به شکلی که مردم در ولایات دور دست در مناطق غیر استندرد و زلزله خیز و حتی درون درههایی که قبلا دریاچه بوده است و استحکامی ندارد، مسکن دارند.
این ابتلای الهی فرصتی است تا بیش از پیش به سوی وحدت و همبستگی برویم و در غم و درد یکدیگر شریک شویم. بدون شک، امت زمانی امت واقعی خواهد بود که فکر و احساس مشترک داشته باشد. از الله متعال مسئلت داریم که شهدای این حادثه را با نور رحمت خویش بپوشاند، مجروحان را شفای عاجل عطا کند و امت اسلامی ما را از مصیبتها و پریشانیها حفظ نماید.
Currently, we are developing an ACS-to-ACS call solution for one of our customers, and while the call placement is working perfectly fine, we have encountered an issue when trying to add a PSTN number to an existing ACS-to-ACS call. Specifically, we are faced with error code 400. Our solution utilizes the Frontend SDK (@azure/communication-calling) to create a peer-to-peer call. Could you kindly provide your insights on how to resolve this issue? Your assistance would be greatly appreciated.
Reasons:
Blacklisted phrase (1): appreciated
RegEx Blacklisted phrase (1.5): how to resolve this issue?
I'm facing the same issue, but I'm not using react-native-image-crop-picker at all, it's even failing for fresh new RN project.
then I found out that my system ruby has some problem. I removed system cocoapods, switched to use rbenv , and reinstall cocoapods again with ruby from rbenv.
i have same issue with you.
we can not send gmail in railway as a trail user
So i have create a api to send like a third-service using pythonanywhere
and in railway, just call as an api to pythonanywhere
Hope it help you
Sorry all for raising an issue where none existed. My previous code was also treating slices as views, but just by dumb luck -- because of the specific programming logic -- I was not reading the changed elements of A and hence not aware of the problem. The input used to construct A changed recently, making it so that I was seeing the changed elements now, and I thought this was new behavior. Please consider this closed.
thanks to all who contributed! It took me a while for me to realize but the actual error has nothing to do with the code itself, but rather which file the auto-grader was checking. In other words, my file name had a typo in it so the terminal command I was using (check50 cs50/problems/2022/python/outdated) was checking the wrong file over and over again. Appreciate all the responses, guess moral of the story is be careful with your file names LOL
https://xpornium.net/embed/dragon intelligence and my chatterbate profile on my phone with my chatterbate profile picture of my things so far so good 👍
Remember that &(&1 * &2) is just "syntax sugar". The following are equivalent:
1- Anonymous Function (fn ... -> ... end)
Enum.reduce([1,2,3,4], fn a, b -> a * b end)
2- Capture Anonymous Function (&...)
Enum.reduce([1,2,3,4], &(&1 * &2))
3- Capture Named Function (def ... do ... end)
Enum.reduce([1,2,3,4], &multiply/2)
# ...
def multiply(a, b) do
a * b
end
The Capture Anonymous Function syntax (#2) is nice and concise but as @coderVishal pointed out, mostly useful for short, easy-to-understand expressions
If you have a more complicated expression -- or (as you discovered) if you need more than one captured function in a single line -- then it is time to switch to longer fn-style anonymous functions (#1) or make regular functions and use them by name (#3)
In @sobolevn's example, he used #2 for the outer Enum.reduce and then #3 for the inner multiply
Enum.map([[1,2],[3,4]], &Enum.reduce(&1, fn(x, acc) -> x * acc end))
I have a similar issue with the CloudFormation Linter extension, when you rename a file - it complains the original file is missing and there’s no way to get rid of it, except for closing and opening up the workspace folder.
I’m still trying to figure out how RooCode is able to add an “Explain with Roo” item when you right click on something in the Problems panel, so I can raise PR to implement.
So the issue was that the frame being sent out is the heartbeat frame, and if there is no NMT master to ack it, then the stack isn't treating the frame as being sent anyway, so the buffer is overflowing with the heartbeat messages. If a sniffer is active on the bus, the warnings go away. I ended up editing the code to change that warning to "warn once" so I don't keep getting that spammed out when only one CAN device is active on the bus.
I was running into this same issue. The channel was filling up and the reader was stuck. In fact, I came back to this SO thread several times hoping someone had an idea. I finally realized the logic immediately following the read operation (within the read loop) was blocking the next read. In your particular example, I would ensure the logging and DoSomethingAsync is not blocking.
So, that's my solution and I foresee it might be useful for others facing similar issues on Mac. I recently switched from Linux and everything seems to be so buggy here ;)
I have experimented with different versions of this gem for some time and found out that the next patch version works just fine for me. So, as a temporary solution, cause updating the whole project's dependencies would be a huge step, I did the following:
# Checking the current gem installation path
bundle show grpc
<myhomedir>/.asdf/installs/ruby/3.2.4/lib/ruby/gems/3.2.0/gems/grpc-1.74.0-arm64-darwin
# Installing the next patch version of the gem
gem install grpc --version=1.74.1 --platform=arm64-darwin
# Removing the old version
rm -rf <myhomedir>/.asdf/installs/ruby/3.2.4/lib/ruby/gems/3.2.0/gems/grpc-1.74.0-arm64-darwin
# Faking the new version to act like a previous version via a symlink
ln -s <myhomedir>/.asdf/installs/ruby/3.2.4/lib/ruby/gems/3.2.0/gems/grpc-1.74.1-arm64-darwin <myhomedir>/.asdf/installs/ruby/3.2.4/lib/ruby/gems/3.2.0/gems/grpc-1.74.0-arm64-darwin
It all started to work now.
Feel free to suggest your workarounds for the situations like that.
Okay turns out the issue is something related to commas in the URL not playing nice. Not sure about the specifics but replacing raw commas in the URL with their encoded version (%2C) makes everything as normal.
Visual Studio now provides a mechanism to do this automatically called public project content. In your library project properties, set the Public Include Paths property to a list of paths with shared header files. Once done, these directories will be automatically added to Include Paths in all projects referencing your library project—no need to include or configure anything there, just having a project reference is enough.
I have the same problem. You can let the messages come from the bottom with
display: flex;
flex-direction: column-reverse;
for the container. But I want the first message appears at the top. If the area ist full than the all next messges appear at the bottom and pull the previuos messages to he top. With this code all messages appear unfortunatlly at the bottom
