I see you have problems on a real device too but I bumped into the same issue but only for Simulators and I've found out that RevenueCat has the problem with the iOS 18.4, 18.4.1, and 18.5 simulators.
Create a StoreKit Configuration file in Xcode. Use local testing instead of App Store Connect sandbox. Configure your products directly in the configuration file
I think your issue might be due to using .find method for toggling the lists as it may lead to irregularities since multiple elements have similar classnames' you could try the same using .child or .first method for child lists
In the project settings in Unity, you only specify the dependencies to third party packages. To update the version of the specified packages, simply go to the Package Manager, find the Appylar package in the list to the left and click on it. After that, to the right, you can see if there are any newer versions of the package to update to.
Assertion is not to be used unless you get into trouble. Why assert something you certain? If not certain, make it certain instead of mark assertion to hope it certain or guarantee it certain. It is not a guarantee, it is a debug aid. Guarantee is by design, and by unit tests.
for never happen code, use exception. Because u need it even in production. Using assertion for never happen code is probably wrong tool.
can someone please help in explaining why virt_to_phys cant be used in case of SMMU to get the IOVA address, as in SMMU enabled system physical address are not exposed , so can we use virt_to_phys in place of dma_map_single to get iova address.
Note:- I dont want to invalidate/flush the cache operation, system is cache coherent.
Reasons:
RegEx Blacklisted phrase (3): please help in
No code block (0.5):
Starts with a question (0.5): can someone please help in
this problem "Gradle exit code 1" comes when we connect to the mobile otherwise the flutter app works fine on the computer! and it seems that no one in the whole world could fix the problem! all answers are just talk and not working.
People who wrote flutter should find the answer for their mistakes because it is a stupid bug somewhere!
Probably not a suitable solution for most people, but switching the server to Jetty fixed the problem.
Inspired by the discussion on this question also having problems with Tomcat and HTTP/2, I tried using Jetty instead. I used a default configuration, enabled HTTP/2 and tested it with Safari and Chrome and the uploads from both were all as expected.
They both use CJS to load query-string, but with pnpm, it ends up installing query-string v9.x (which is ESM-only), causing issues. I temporarily solved it by overriding the version to v7.1.3 (which is still CJS-compatible) and installing it with pnpm, and that worked. Still, I felt that version resolution was quite a hassle…
Another potential cause is that the MQ server has reached its maximum number of allowed connections, resulting in new clients being unable to connect and receiving error 2009 (MQRC_CONNECTION_BROKEN). This is typically observed as the server actively closing the connection immediately after it is established
Struggling with the same and sorry to say the solution above doesn't work. I had a beautiful script in my package.json set up to sync the secrets to cloudflare to allow for auto deployments.
But that's just not how this works. I suspect those variables are accessible via $env/dynamic/private, which is shitty bc they aren't available at build-time and the server has to fetch them every time you use them (which is probably optimized in workers but a shitty condition anywhere else).
Idk what happens if you keep your vars in a .dev.vars file, if they get pushed to both sections, because adding one by one via the UI is not something I'm gonna do. The issue with .dev.vars is there doesn't seem to be any way to pick those up automatically with vite (which is ridiculous) so you can't develop locally, or again, anywhere else.
So the solution seems to be?? maintaining .env files and having a pre-commit script copying them to .dev.vars??? Idk how we came to this. I love cloudflare but bow, .dev.vars, really?
In that case, you will need to "resume track" of the Production release then head to the dashboard, you will see status changes need to review under "Update status", hit send for review, wait couple minutes/hours then refresh the page.
This won't be an issue that needs solving from when Moodle 5.0 is released, as one will be able to use questions from the question banks of other courses. So questions won't need to be stored outside of individual course instances (at the higher level category) to be accessed in future versions of a course or in other courses.
There is a 10 seconds time limit, so if you take more than that, you will get the error you decrive. You can check the status response, if there is no status, it's because of that.
More information on exactly what the application is would effect my advice.
If you have a moveAbsolute or moveVelocity, you can change the position or velocity, then drop the enable and re-enable. This will trigger a new move based on the BufferMode.
Alternatively you can set up a virtual axis, and then gear in dynamic to this axis, you can then alter the gear ratio positive or negative on the fly to make this update the velocity in a superfluid manner.
I was asking Claude about this. The answer Claude gave me is to go to the http request node's setting and set "Continue on Fail" to ON. This way, even if one request fails, it will still have an output, and they will stay aligned. You can then do a merge node with merge by position. The merge node takes as input the input and output of the http request node.
I am using firestore, but I couldn't get to use query counting. query.count() gives error all the time. How did you upgrade to new version of firebase? Can you help please. Much appreciated.
This muerde doesn't work and unfortunately the prep course for the CKA is broken. And they don't bother to tab complete it which is afik typical for the course.
kubectl -n kube-system exec -it etcd-\<bash: \_get_comp_words_by_ref: command not found
The slight memory increase in the second version comes from extra temporary variables and iterator handling inside the loop. While no extra heap memory is allocated, the compiler uses more stack or register space, causing marginally higher usage. The difference is very small and likely due to normal runtime variation.
Without changing your Target name or "AppName-Swift.h" names, you can simply just change the name in ONE place. That is the info.plist file -> Bundle display name. Only change it here without causing havoc in the codebase.
Ok so i found a way to turn off exceptions for C++ in clang which appears to turn any throw statements in functions I call into abort which is basically what I wanted.
The 'cat' version (print('\n'.join(map(str, ints)))) is fastest because it builds one big string and sends it to print() in a single I/O operation. That drastically reduces the overhead compared to printing each number line by line. Since I/O is the real bottleneck here, minimizing the number of print() calls makes a big difference, even if it feels unintuitive that string manipulation beats a simple loop :)
It looks like your writing your assembly using the Intel x86 syntax, its dense but try cross referencing your code you write with the intel x86 assembly manual. They obviously won't have all the examples you will need but generally they are pretty good starting reference. I would also depending on your OS get a debugger that you can step through your code to see if your program flow makes sense and help troubleshoot the errors your getting more effectively. Its tedious but so is writing in assembly. I used GDB with GEF for pretty print and some overlays. Hope this helps.
To replace default tenant name with your company logo. you need to search Company Branding.
External Identities -> User Experience -> Company branding
In company branding click on edit and then go to Sign-in form and change Banner logo.
Here is the logo which i have uploaded.
If you want to hide this then you can upload custom css in layout template that will hide this text.
Yeah I'm using a 'Return URL' configured on a button, and since 5th July (maybe earlier), after payment has been made and the customer hit's 'Return to Seller', it error's with 'PayPal We're sorry, it seems that something went wrong'.
Have you found a solution?
Reasons:
RegEx Blacklisted phrase (2.5): Have you found a solution
I was experiencing the same thing! Using expo 53 and any version of stripe @stripe/stripe-react-native. The fix for me was going from app.json to app.config.js! I don't understand why but this prevents the crash and I can use my app as normal like before. Ask ChatGPT to help with the file conversion ;)
It turns out that SendGrid—since it is my email backend (django-sendgrid-v5) and was being used when mail_admins() was being called by the Django's logging module (django.utils.log) every time a server error occurred—was taking anything that looked like a URL and replacing it with a click-tracking URL because django-sendgrid-v5's default behavior is to tell the SendGrid API to do that.
The solution was to set these variables in settings.py:
Removed the android.adservices.AD_SERVICES_CONFIG property tag from the SDK’s manifest file to prevent merge conflicts for apps that Configure API-specific Ad Services.
I found the problem. It is a hardware or a driver level issue. Because even in an ffmpeg recorded video I can hear the beeping. (My mic is also a webcam.) But if the camera is off the mic works fine. So yeah.
If your real goal is just to make a string that looks nice, try inserting a space in the template string before the formatted number:
>>> template = "The temperature is between {:d} and {:d} degrees celsius."
>>> print(template.format(-3, 7))
The temperature is between -3 and 7 degrees celsius.
I have found a workaround, which is to store the file-generating PHP script on a different domain and server (along with the JPG htaccess trick I described above).
My original problem only seems to occur when the script is on the same domain, which causes it to fetch the file directly/locally instead of fetching it like a normal URL.
To me this workaround is preferable to writing code for WordPress 'hooks' etc. which are more complicated to understand, and are liable to change and require rewriting the code over time to maintain compatibility.
What is your overall publish throughput and what is the average message size? Pub/Sub can deliver up to 10 MB/s per stream, though load will be balanced across streams so you may not see any individual stream saturated if you have many open.
What do you see for the subscription/oldest_unacked_message_age and the subscription/num_undelivered_messagesmetrics? If you don't see a backlog for the latter then your subscribers are generally keeping up with the publish throughput.
You can also configure maxOutstandingElementCount to 5000 and maxOutstandingByteCount to 5000 * 700: are your clients hitting these limits and getting flow controlled? You can check whether your streams are flow controlled with the subscription/open_streaming_pulls metric.
but see many modifyAckDeadline requests from the GCP Pub/Sub metrics and graphs which doesn't make sense to me
Pub/Sub client libraries send ModifyAckDeadline requests upon receipt of messages, as well as periodically for unacked messages to extend their leases up to the "Maximum acknowledgment extension period", so it would be expected to see ModifyAckDeadline requests even if you are acknowledging quickly.
This page has tips on monitoring and debugging subscription health and the subscription/delivery_latency_health_score metric can help you more easily identify factors contributing to increased delivery latency. If the metric does not indicate any issues with your subscription, you can create a support case so that someone can look at the subscription from the backend perspective.
It's nice that you code this in this manner, quite a beautiful implementation. I generally don't like to use struct Node for graph questions, though they are helpful for many cases, in terms of code readability. But I want to share how I write Dijkstra code, my instructor told me to always write a Dijkstra in this manner.
#include <bits/stdc++.h>
using namespace std;
#ifndef ONLINE_JUDGE
#include <D:/debug.cpp>
#endif
#define int long long
using ii = pair<int, int>;
#define F first
#define S second
#define mp make_pair
class prioritize
{
public:
bool operator()(ii &p1, ii &p2)
{
return p1.S > p2.S;
}
};
int n, m;
vector<ii> g[100100];
vector<int> dis(100100, 1e18);
vector<int> vis(100100, 0);
vector<int> parent(100100, -1);
vector<int> path;
void dijkstra(int sc)
{
dis[sc] = 0;
priority_queue<ii, vector<ii>, prioritize> pq;
pq.push(mp(sc, 0));
while (!pq.empty())
{
ii fs = pq.top();
pq.pop();
if (vis[fs.F])
continue;
vis[fs.F] = 1;
for (auto v : g[fs.F])
{
int neigh = v.F;
int wt = v.S;
if (dis[neigh] > dis[fs.F] + wt)
{
dis[neigh] = dis[fs.F] + wt;
parent[neigh] = fs.F;
pq.push(mp(neigh, dis[neigh]));
}
}
}
}
void solve()
{
cin >> n >> m;
for (int i = 0; i < m; i++)
{
int u, v, w;
cin >> u >> v >> w;
g[u].push_back(mp(v, w));
g[v].push_back(mp(u, w));
}
dijkstra(1);
// print shortest path from 1 to n is exists else print -1
if (dis[n] == 1e18)
{
cout << -1 << endl;
return;
}
else
{
// cout << dis[n] << endl;
int curr = n;
while (curr != -1)
{
path.push_back(curr);
curr = parent[curr];
}
reverse(path.begin(), path.end());
for (auto x : path)
cout << x << " ";
cout << endl;
}
}
signed main()
{
ios_base::sync_with_stdio(0);
cin.tie(0);
cout.tie(0);
int t = 1;
while (t--)
solve();
return 0;
}
Floating point numbers are stored as approximations in computers due to how binary systems handle decimals so they may not store the exact decimal value. For example, a number like 5.15 might be stored as 5.1499999999999. This can sometimes lead to some unexpected results which may be contributing to this discrepancy. The best approach in this situation is to store the result of the expression in a separate variable and then apply the Trunc() function to the stored result as seen in your second attempt above.
Just in case someone finds that none of the solutions proposed here work, I just found out that this problem happens if you're only using the cache as source like this:
var queriedObject = docOject.get(const GetOptions(source: Source.cache))
Finally found an answer! After making a copy of the entire individual project on each machine and doing a file diff comparison of them, noticed a number of orphaned files on my machine in the obj folders. So went through all the projects in the Visual Studio solution and deleted all files in the \solution\project\obj\Debug and \solution\project\obj\Release folders. After doing a new build on the solution, it worked perfectly on my machine right away without any changes to app.config or anything else in the entire solution.
To help understand what a resource in Azure AD/Entra is:
A resource is anything that is governed and protected by the Azure Entra (Azure Active directory) service. Usually, these resources are what your apps or services need to access. This document explains well how granular scopes work - Requesting scopes as a client app but as an example, a custom application/client that you built that would show the user a list of recently received mail messages and chat messages, the app would access the Microsoft Graph resource API (specifically, with the Mail.Read and Chat.Read permissions) to access user's email and chats. Each of these resources has a unique app id as well to allow for programmatic access while requesting for tokens. The image below gives you a flavor of the various resources available.
I compressed various timestamps into just 3 bytes (24 bits) for low-bandwidth IoT radio transmissions.
If you can tolerate a small error margin (e.g., ~1 second), you can encode multiple years of timestamps from a given epoch. This is ideal for embedded systems, LoRa, or sensor networks where transmission size matters.
I wrote a lightweight library called 3bTime that allows you to choose between different profiles depending on your application's needs — whether you're optimizing for precision or long-term range.
📦 Example:
10-year range → ±9 seconds error
193-day range → perfect second-level accuracy
Configurable, efficient, and designed for constrained environments.
if you do not set PrintPreviewControl1.Rows = 2 to number of pages then it will not work despite the number of times you add e.HasMorePages.
Without this, it will keep printing only one page
Just set the number of pages with PrintPreviewControl1.Rows=?
Set AUTOCOMMIT to TRUE on account level in Snowflake.
Reposting @sriga's comment, which answered the question:
If you are changing the account parameter inside the stored procedure, it won't allow. Instead you can change the account level parameter by running Alter session set autocommit=True; else you can run your python script outside the snowflake and change the session parameters as mentioned in the below code
Snowflake enforces the prohibition on setting AUTOCOMMIT inside a stored procedure. Note that changing the AUTOCOMMIT behavior outside a stored procedure will continue to work.
Have you figued this out? I have the same problem, I do have the app installed on the homescreen, I can see that the PWA is subscribed, but I dont get the notification on ios. Android works fine:
My backend logs:
2025-07-07T14:33:18.228Z INFO 1 --- [app] [nio-8080-exec-7] d.v.app.service.NotificationService : Subscribed to Push notification
2025-07-07T14:33:45.400Z INFO 1 --- [app] [nio-8080-exec-1] d.v.app.service.NotificationService : Sending Notification for type: NEW_POLL
2025-07-07T14:33:45.410Z INFO 1 --- [app] [nio-8080-exec-1] d.v.app.service.NotificationService : Found 2 subscriptions for type NEW_POLL
2025-07-07T14:33:45.416Z INFO 1 --- [app] [nio-8080-exec-1] d.v.app.service.NotificationService : Sending Notification via pushService
2025-07-07T14:33:46.143Z INFO 1 --- [app] [nio-8080-exec-1] d.v.app.service.NotificationService : Sending Notification via pushService
My sw.js, inspired by David Randoll above:
import { precacheAndRoute } from 'workbox-precaching';
precacheAndRoute(self.__WB_MANIFEST);
/**
* Fired when the service worker is first installed.
*/
self.addEventListener("install", () => {
console.info("[Service Worker] Installed.");
});
/**
* Fired when a push message is received from the server.
*/
self.addEventListener("push", function (event) {
if (!event.data) {
console.error("[Service Worker] Push event had no data.");
return;
}
const payload = event.data.json();
const notificationTitle = payload.title ?? "Varol Fitness";
const notificationOptions = {
body: payload.body ?? "You have a new message.",
icon: payload.icon ?? "/web-app-manifest-192x192.png",
badge: payload.badge ?? "/web-app-manifest-192x192.png",
image: payload.image,
data: {
url: payload.url ?? "/dashboard", // Default URL if none is provided
},
};
event.waitUntil(
self.registration.showNotification(notificationTitle, notificationOptions)
);
});
/**
* Fired when a user clicks on the notification.
*/
self.addEventListener("notificationclick", function (event) {
console.log("[Service Worker] Notification clicked.");
event.notification.close();
event.waitUntil(
clients
.matchAll({ type: "window", includeUncontrolled: true })
.then(clientList => {
const urlToOpen = event.notification.data.url;
if (!urlToOpen) {
console.log("[Service Worker] No URL in notification data.");
return;
}
for (const client of clientList) {
if (client.url === urlToOpen && "focus" in client) {
console.log("[Service Worker] Found an open client, focusing it.");
return client.focus();
}
}
if (clients.openWindow) {
console.log("[Service Worker] Opening a new window to:", urlToOpen);
return clients.openWindow(urlToOpen);
}
})
);
});
While digging into this question again, I noticed that there is a difference in how the SD card handles invalid commands while in SD Mode vs. SPI mode. In SPI mode, no matter whether the command you sent it valid or not, the SD card responds, usually with an R1 response (Physical Layer Simplified Specification Version 6.00, section 7.2.8). In SD Mode, however, it doesn't respond, and instead sets a register flag that has to be read via a separate command (ibid., section 4.6.1). This is further supported by this quote from section 7.2.1:
The SD Card is powered up in the SD mode. It will enter SPI mode if the CS signal is asserted (negative) during the reception of the reset command (CMD0). If the card recognizes that the SD mode is required it will not respond to the command and remain in SD mode. If the SPI mode is required, the card will switch to SPI and respond with the SPI mode R1 response.
So I guess the answer to my question is: If the SD card doesn't respond at all to your SPI commands, you know it's either disconnected or in SD mode.
