Thanks @furas for the hint: simply uninstalling qt6-main using `conda remove qt6-main` worked for me (conda installed/uninstalled a lot of other packages, unfortunately I didn't really paid attention and just clicked ok). Just in case someone is interested: that's what my environment looks now:
For this, you need to implement the IAutoComplete2 interface. Ask chatgp for "Implements IAutoComplete2 in c#", you will get a basic implementation, and sample how to attach it to a winform textbox.
Now, all you need to provide ACO_NOPREFIXFILTERING flag, when creating your new autocomplete.
Please find below the Excel upload format I am using to update item texts for multiple purchase requisitions (PRs) in SAP. The file contains PR numbers, corresponding line item numbers, and the desired item text to be uploaded.
PR Number PR Item Number PR Item Text
3400001460 10 8 HOLES
3400001460 20 PMI TESTING
3400001460 30 HARDNESS TESTING
3400000875 10 MECHANICAL TESTING
3400000875 20 LADDLE ANALYSIS
3400000875 30 HARDNESS TEMPERATURE
I kindly request your support in providing a VBA macro that can:
Open ME52N in SAP.
Process each PR and line item from the Excel sheet.
Update the item long text accordingly.
Loop through all entries and save changes.
This automation will help reduce manual efforts and ensure consistency in text updates.
Please let me know if you need any additional details or adjustments to the format.
Thank you for your support!
Reasons:
Blacklisted phrase (0.5): Thank you
RegEx Blacklisted phrase (2.5): Please let me know
I tried accessing the pusher api key in my Next.js project like this
process.env.PUSHER_APP_KEY
but received the same error but when I replaced the value with the litral value it worked just fine but due to obvious security reasons this cannot be used like this. First doubt was that I might be making some kind of typo but it wasn't the case, the following helped me so replace the above snippet with the below
`${process.env.PUSHER_APP_KEY}`
This solved my issue I hope helps other facing the same issue. Happy coding.
When working with Convolutional Neural Networks (CNNs), the goal is typically to perform automatic feature extraction and model training directly from images. It's not like working with traditional models where you have predefined columns or features — in CNNs, those features are implicitly learned from the images themselves.
Regarding your question "whether to freeze the model or to extract until Dense(128...)," — it's quite common to include one or more dense layers after the convolutional and max pooling layers, before the final classification output. This is because connecting the convolutional outputs directly to the output layer would be too raw and inefficient. Dense layers help consolidate and interpret the extracted features before making the final prediction.
As for "whether to use a number of folds or if that's only for machine learning," — CNNs can also benefit from techniques like K-Fold cross-validation. In my case, I usually start with a small number of epochs and use something like a 5x10 K-Fold setup to get an initial idea of model performance. Based on that, I adjust hyperparameters and experiment with different fold sizes.
Also, if your dataset is large or training from scratch is too slow, you might want to try transfer learning — it can save a lot of time and resources.
Regarding dropout layers, they are definitely useful and commonly used to prevent overfitting by randomly deactivating neurons during training. However, if dropout is too aggressive (e.g., using very high dropout rates), it can prevent the network from learning effectively. So again, it comes down to experimentation — finding the right balance is key.
In the end, there’s no single “right” way to do this — trial and error, careful tuning, and adapting to your specific dataset and task is what usually leads to the best results.
In my original post, I omitted the fact that I was saving the URI to RoomDB between creating and trying to read from it. My custom URI TypeConverter for RoomDB was calling uri.path to convert it to a string, but the path attribute cuts off the URI scheme (thanks to @CommonsWare for pointing this out in the comments). After changing my TypeConverter to use uri.toString(), when I retrieve the value from RoomDB I get the complete URI including the content:// prefix indicating the scheme. From there, it's simple to create a bitmap from the retrieved uri using ImageDecoder and ContentResolver:
val bitmapSource = ImageDecoder.createSource(contentResolver, uriFromDb)
val bitmap = ImageDecoder.decodeBitmap(bitmapSource)
This question only comes up if multiplications and divisions are to be "optimised" by shifts. Since decades, this is done by the compiler, and never performance relevant; just obfuscates the program.
So if a number shall be divided by 8, the proper type should be used and the division written. If a certain number of bits has to be moved right, use an unsigned integer; the high order bits are irrelevant in this case.
Force pseudo-tty allocation. This can be used to execute arbitrary screen-based programs on a remote machine, which can be very useful, e.g. when implementing menu services.
Multiple -t options force tty allocation, even if ssh has no local tty.
Try working with this -- it handles all data types and uses recursion to go arbitrarily deep into complex JS object structures. I got tired of looking for something which wasn't either a bloated library or too simple to deploy easily. Wanted to guard against folks poking holes in my SQL (I do use parameterization, but...), logging, server code, whatever. You can easily change the regexes, etc. to fit your needs. Just wrote it this afternoon, so would appreciate feedback if there are things I overlooked.
function sanitizeVar(jsVar, options = {}) {
const type = typeofComplete(jsVar)
if (type == 'object')
Object.keys(jsVar).forEach(key => { jsVar[key] = sanitizeElement(jsVar[key])})
else if (type == 'array')
for (let idx = 0; idx < jsVar.length; idx++) { jsVar[idx] = sanitizeElement(jsVar[idx]) }
else // string, boolean, number, bigint, null, undefined, function, symbol
jsVar = sanitizeElement(jsVar)
return jsVar // return mutated value -- end of main function
/*
helper functions -- keep in this scope
*/
function typeofComplete(jsVar) { // what the native JS typeof should do...
return jsVar == null ? 'null' : Array.isArray(jsVar) ? 'array' : typeof jsVar
}
function sanitizeElement(val) {
const type = typeofComplete(val)
if (type == 'string') return sanitizeString(val)
// return back up to sanitizeVar(); will likely recurse back here...
else if (['object', 'array'].includes(type)) return sanitizeVar(val)
else if (['boolean', 'number', 'bigint', 'null', 'undefined'].includes(type))
return val // these are safe as they are
else if (['function', 'symbol'].includes(type)) return null // never in a client request
}
function sanitizeString(str) { // sanitizing crux; add regex and options as needed
// DIAG to make sure it's working: str = str.replace(/\d/g, 'X')
return str.replace(/[^\x00-\x7F]/g, '') // remove non-ASCII chars
.replace(/[\\|`"';<>]/g, '') // remove \ | ` " ' ; < >
.substring(0, options.max_str_length || undefined) // undefined => no limit
}
}
I know the question is super-old, but I ran across it before I gave up looking for a more robust light-weight solution. Another idea is to pass regex(es) in as part of options object arg.
In addition to @fepegar answer - you probably don't want to save the images as .jpg format, since .jpg is a lossy-compression type, which means it losses important information. You should save the images to a lossless compression file types, such as .pgm for a single channel image, .png for 3 channels image, or .tiff (using python's package tifffile) for any other number of channels - which is probably what you'd want in this case.
pip install tifffile
import tifffile as tiff
tiff.imwrite(file_path, image, dtype=np.float32) # Save the image as a TIFF file as FP32
Ambulance service simulation is a critical tool used to enhance the efficiency and effectiveness of emergency medical services. In a fast-paced and densely populated city like Mumbai, simulation plays a vital role in preparing ambulance teams for real-life scenarios. By using advanced technology and training modules, Ambulance Service in Mumbai can simulate emergencies such as accidents, cardiac arrests, or mass casualty incidents to train staff in quick decision-making, patient care, and route optimization.
These simulations help identify potential challenges, test response times, and improve coordination between ambulance drivers, paramedics, and hospitals. As a result, they significantly enhance the overall quality and reliability of emergency care services in the city. With Mumbai’s complex traffic and varied infrastructure, simulation ensures that the ambulance services are well-prepared to save lives efficiently and swiftly.
// certFileEnv is the environment variable which identifies where to locate
// the SSL certificate file. If set this overrides the system default.
certFileEnv = "SSL_CERT_FILE"
// certDirEnv is the environment variable which identifies which directory
// to check for SSL certificate files. If set this overrides the system default.
// It is a colon separated list of directories.
// See https://www.openssl.org/docs/man1.0.2/man1/c_rehash.html.
certDirEnv = "SSL_CERT_DIR"
I found for Tomcat on Windows, the following worked.
If the Tomcat AppServer is configured as a Windows Service, you need to run the tomcatw.exe command and then enter the values in the fields as per below:
If the Tomcat AppServer is NOT configured as a Windows Service, you need to create file setenv.bat in the Tomcat bin folder and add the following entries:
set JAVA_OPTS=%JAVA_OPTS% -Dfile.encoding=UTF-8 -server -Xms512m -Xmx2g -Xss4m
You should see something like this in the catalina.log
The different tests try to explain the problem and suggest a fix :
faillingCallWithIso8859() : reproduces the initial problem (checking the exception raised is about xml encoding
testOKWithUTF8() checks it works if the server responses using UTF-8
testOKWithCorrectHeaderXML() tries a fix with the server side adding encoding in xml header : I think it is the ideal solution
testOkWithClientFixingEncoding() tries an other RestClient (provided by poc.java.springrestclient.ConfigRestClient#woEncodingFixingRestClient) which intercept the response made by the spring ClientHttpRequest and decode the body to re-encode it in UTF-8. this is a poor workaround because it requires to rewrite some spring code (class poc.java.springrestclient.ConfigRestClient.BodyReEncodedResponse which rewrites body like org.springframework.http.client.BufferingClientHttpResponseWrapper). But it works.
If you introduced such a RestClient in your application, you would only use it for the problematic endpoint thanks to @Qualifier adn/or @Primary.
I don't think the problem resides in Spring RestClient : in fact, it seems impossible to set up externally the encoding used by Jaxb unmarshalling : see Override declared encoding during unmarshalling with JAXB : the suggested solution is to process byte decoding outside jaxb to control encoding. I neither found in Jaxb a way to do it : jakarta.xml.bind.helpers.AbstractUnmarshallerImpl#setProperty does not support any property set (raise exceptions)
On click of your button you get the following errors in the console
The invalid form control with name=‘building_budget’ is not focusable.The invalid form control with name=‘drainage_budget’ is not focusable.
The thing that is happening is that the input fields with building_budget and drainage_budget are required, but they have display: none.
To fix this either make them optional or show the fields. After that your submit should be working fine.
Make sure to check the url that it is submitting to, as currently you dont seem to have escaped the <?php tag, leading to getting a post url that is different from the one you desire.
If you are unsure of how to check these things, look up a guide on how to use the chrome/firefox (or any other browser youre using) inspector tools and console
instead of win32gui.GetClientRect(hwnd), use win32gui.GetWindowRect(hwnd)
This will cleanup the image that you are converting to a numpy array for OpenCV. As far as the offset, you're going to end up with a boarder around the window itself because of shadowing around the window that is actually part of the window as you can see there isn't any spacing along the top of the images below.
Bigtable now supports SQL GROUP BY with aggregation functions SUM, AVG, COUNT, MIN, MAX, HLL_COUNT.MERGE... and Continuous Materialized Views to calculate aggregations incrementally into materialized views.
This document outlines the procedure for downgrading the Flutter plugin in Android Studio from version 85.0.4 to 85.0.3, including the configuration steps to prevent automatic updates.
java-sqs-listener a lightweight, dependency-free library was developed for exactly, with a design that supports seamless future upgrades. Refer to java-sqs-listener-springboot-example for a demonstration of how to integrate the java-sqs-listener library within a Spring Boot application.
I would start by making sure that API.cs and program.cs are part of different projects. The way I open projects is I pin Visual Studio to the taskbar (you should pin both of your versions). Once you've opened the correct project right right-click on that version of Visual Studio in the taskbar and pin the project to the recent files list. Then all you have to do going forward is right-click on the appropriate version of Visual Studio and left-click on the desired project.
If you want to make your two versions of Visual Studio look vastly different you can customized their icons:
I am finding the above filter function prevents the cart from having shipping added for Express checkout type payment methods including Apple Pay. Has anyone else experienced this?
Refer System.Text.Json.JsonSerializer instead of newton soft json serializer... Sorry Dinesh I had to repeat your answer as I couldn't upvote... But I tested it and it worked like a charm
class OTPData{ public string OTP {get;set} ....... }
var result = System.Text.Json.JsonSerializer.Deserialize<OTPData>
(response.ToString());
string otp = result.OTP;
string username = result.UserName;
string type = result.type;
I asked copilot itself about an API interface to send prompts and get responses and it said it doesn't support that as it's intended for integration into IDEs. It also said "If you're looking for programmatic access to AI-powered code generation, you might want to explore OpenAI's GPT APIs or similar services, which are specifically designed for such use cases."
Compression comes in two flavors: lossy, and lossless. Lossy compression can take file sizes down significantly, but some data will be lost (hence the name). Lossless compression on the other hand keeps all original data intact, but usually can't take file sizes down as much.
Due to the nature of SVGs being vector-based[^1] XML-defined[^2] graphics, unless you manually optimize them you can't get much better compression than removing the whitespace and comments in the files (if any).
If you want to maintain acceptable visual quality and are ok with them being quite small, you can rasterize them into a PNG, for example. Online tools such as CloudConvert or command-line tools such as ImageMagick can help you with this.
tl;dr: unless you need to use SVGs, convert them into tiny raster graphics.
[^1]: i.e. the image is defined as a series of curves and shapes instead of pixels. This generally results in less information per image ∴ smaller file sizes.
[^2]: i.e. it's written in plaintext, which can't be compressed without overhead.
I encountered the same issue. I disabled the antivirus and restarted my Windows PC. After that, I was able to log in successfully using Chrome. There was no need to install any additional browser
I tried again, noting that it might be interpreted as a non-member status. When selecting an account on the OAuth authorization screen, I had chosen the account with my email address, but instead of doing so, I chose my YouTube channel account created with the same email address. I did that and got a response without 403 forbidden.
So I just spent hours trying to figure this out... I was working with device emulation turned on in the dev console. The device I was using was even configured as desktop non-touch device. I was working on Edge, I will assume Chrome will behave the same way.
IF DEVICE EMULATION IS TURNED ON DRAGOVER EVENT WONT FIRE AND CURSOR WILL BE MARKED AS BLOCKED!
Fixed, for some reason the ios project was setup in "Release mode". I did the change in Xcode -> edit schemes
Then In Run Section, The build configuration was in "release", did the change to "Debug". Close xcode
then in vscode, run again:
npm start
npm run ios
After that, the app runs well again and able to send logs to debugger. And if someone don't know, I wasn't able to see the logs inmmediately, as I read before in some posts around internet about RN >= 0.77, it is necessary to have some third-party debugger, but i wanted to put it simple, just downloaded Chrome browser first, and then run the npm commands.
In simulator, once the app is launched, press "j" key, to see the option "open dev tools", with that the chrome dev tools will open and you'll see the logs.
I were about 4 days with this trouble,I hope these details can help others
Facing similar SSL failure. exceptions with Docker python:3.13-alpine
and rabbitmq-amqp-python-client library.
Apparently, while building proton package, it rely on pkgconf / pkg-config.
So after adding packages, proton.SSL.present() again works
Bonjour à chaque fois que j'essaie d'utiliser swiper avec react js ça ne fonctionne pas. Est ce que quelqu'un a une idée "swiper": "^11.2.6"
Merci d'avance
Okay, we already have an accepted answer, but let me add a few thoughts on this as well.
First of all, it's important to understand that most features in any tool or technology are introduced to solve existing challenges. This includes everything from design changes to improvements in problem-solving.
But why version catalogs?
Well, when working on projects loaded with dependencies, keeping everything healthy and consistent is a real challenge. For example, imagine a project with multiple modules—say, 3 to 10—where each one uses a different set of dependencies. Some modules might even use the same dependency, but with different versions or from different repositories. Managing this mess and keeping everything up to date can quickly become a nightmare!
Having a centralized approach to manage all these artifact references and version strings in one place is like achieving peace on earth. I say this with confidence because I’ve faced many issues while working on large, real-world projects packed with dependencies.
And that’s it—in practical terms, version catalogs help us tackle this exact problem.
Of course, for simple projects that use very few dependencies—or for quick prototypes or throwaway projects—there’s no need to worry about using version catalogs. Sometimes we create projects just to test something, and maybe they don't even have any extra dependencies or plugins. In those cases, it’s totally fine not to use a catalog.
But as your project grows, you should really consider migrating your dependency strings to a catalog.
Also, remember that Gradle is used to build environments beyond Android. Kotlin and even C++ are well-supported by Gradle!
None of the above worked for me. I have environment variables pointed to flutter bin folder. I have android APK support enabled. I have the new 2024 version of Android Studio and nothing I have tried is working. Can someone help me please?
Each variable comes from a "context". config.toml file defines the config, frontmatter in _index.md files define section and frontmatter of every other markdown file defines a page context to be used in templates. When you are referring to a variable from a context, you should mention variable name along with the context it comes from separated with DOTs, like context.my_variable or context.extra.my_extra_variable if defined after [extra].
We are going to declare titles for different contexts, and none of them will overwrite each other.
config.toml
base_url = "https://murtezayesil.me/"
title = "Murteza Y"
config.title = Murteza Y
content/blog/_index.html
---
title = "Blog"
---
# Blog
Welcome to my corner of the internet.
section.title = Blog
content/blog/my-first-post.md
---
title = "Blogging in 2025?"
---
Of course.
page.title = Blogging in 2025?
template/index.html
{% set title = "Home" %}
{{ title }} = Home {{ this.title }} = Home this is a special context that refers to that file. It can be omitted. Since you didn't specify any context, Tera reasonably assumed that you were omitting this and looked for this.title.
This might be confusing early on. Luckily, Tera provides {{ __tera_context }} variable which prints out all defined variables in all known contexts. If that is hard to comb through, ShawnPConroy on Zola's Discourse Forum shared this script which organizes them.
I encountered a similar problem as well and just solved it. There are two different presentation options: Projects Presentation and Package Presentation. I was automatically going to the Projects Presentation and selecting Hierarchical or checking this option. To view the hierarchical structure in packages, I selected Package Presentation - Hierarchical, and then I was able to see a properly structured hierarchy.
The single line and the eval are key to this. I did it on single steps (in an attempt to make it clear) and it produces some very odd results (all keys and no values)
After that, you can call the publishAndroidReleasePublicationToMavenRepository or publishReleasePublicationToMavenRepository task, depending on the configuration of your project (KMP or Android).
After that, all modules to which the plugin has been applied will be uploaded to the specified maven repository.
The client can then download any module as a dependency, and that in turn will download all other modules required for correct work
your-app-feature = { group = "YourProject.features.feature", name = "feature", version.ref = "your-proj-version" }
Advantages:
Flexibility
Minimal changes for publishing
No unnecessary dependencies for the client
Minuses:
Long publishing process
It is necessary to reduce the dependencies of modules on each other, otherwise the client will need to download most of them
There will be a huge number of dependencies in the maven repository. In our case it is 250+ at each release.
Yesterday I spent the whole day trying to fix this error. Today, I deleted Crashlytics from the Runner's Build Phases, and finally, my app built successfully. Hope this helps someone!
Your keys shouldn't be readable by everyone on your system.
Your problem should be that Mosquitto switches to a different user.
Your mosquitto.conf omits user, which defaults to the user "mosquitto".
Either ensure the certificate files can be read by that user, or configure to use a different user 👍
I'm on Arch Linux using the Linux Editor obviously.
And I had a similar issue.
I have a faulty USB hub that I use, it is mostly ok 90% of the time.
But it does this stupid thing where it like tries to detect weather you plugged something into it and show you an indicator light.
Which that feature didn't work right from the get go.
Point being turning it off fixed it for me.
Thanks, you gave me a clue as to what was wrong!
If you're sending information over the internet, get yourself a free SSL cert from browser-approved Let's Encrypt. There's no excuse not to be at least this secure in 2025.
