Tag: WebDev

I have a PowerAutomate script that sends me an email every time someone other than me modifies a file on Sharepoint folder that I'm monitoring. Today, with MS Word and Excel auto-save feature, when someone is working on a document I receive a new email every minute. I would like to add an in-between steps that would say "if last email sent was in the last 10 minutes, don't send the email" or "if the previous modification occurred in the last modification, don't send the email". How could I define this extra step ?

I am developing a lightweight local desktop application that uses an SQLite database to log events. Initially, I noticed that when inserting a large batch of rows (or doing frequent small writes), the application experienced noticeable disk I/O bottlenecks. The default journaling mode was set to DELETE. After doing some research, I switched the journal mode to WAL (Write-Ahead Logging) using the command "PRAGMA journal_mode=WAL;". The write performance improved drastically, almost by a factor of 4 or 5, and the disk activity dropped significantly. I understand that WAL writes changes to a separate file first, but why does this architectural change make such a massive difference in speed for a single disk? Doesn't the data still have to be written to the storage drive eventually?

Problem I am currently developping a keyboard layout for MacOS. The problem is that, since MacOS only seems to apply changes of a keyboard layout .bundle at log in, I need to log out and log back in in order to see the changes I made to the files withing the .bundle take effect. This slows down a lot the developping process. Ideas This post from 15 years ago already asked for a solution to the exact problem I have, and no solution seemed to have been found by the OP back then. However, @jonallard commented the following : If it's helpful: to try a keyboard while developing, I added a new file to the folder with a different filename and keyboard name, and added it in settings. That way you can at least try your keyboard without log in-out. Question I can't quite understand what exactly @jonallard s

I am trying to replace the default OSC with a custom one. I copied osc.lua into $HOME/.config/mpv/scripts/ and modified it a bit. Calling mpv --script-opts=osc=no on a video file or an audio file with a picture works as expected, the default OSC isn't visible and mine gets shown instead. However, when opening an audio file without picture, both OSCs are visible, but only mine is clickable. The config directory contains only my script. Am I doing it wrong? I tried looking for similar issues online but didn't find anything.

I would like to upload a couple files to a website, and have that site run a bash script, whenever there is a visitor. After the visitor closes the session, the site should shut down. The process should start back up the next time someone visits the page. Where can I find a site like this? I thought there were sites for everything. Asking AI was a fool's errand, it just points me to replit or other sites which are apparently all-in on AI now. My specific issue is, although I initially suspected PEBKAC, i think it could actually be a bug in the program, it doesn't behave the same way in npm that it does in the standalone executable. (npm would be the correct behavior). Of course, i would assume that npx under the hood uses the same executable, so its likely im just making a mistake somewhere. Or that my system doesnt like the executable. I dont know, thats why i wanted a replicable setup.

I have a Windows 11 PC with 32 GB of RAM, and I usually have about 10 GB to 15 GB of free, unused memory during my daily workflow. However, when monitoring performance in Resource Monitor, I noticed that the Hard Faults/sec (Page Faults) graph for memory occasionally spikes to 50 or 100, even when the available physical memory is nowhere near full. I always thought "Hard Faults" meant the system ran out of RAM and had to use the pagefile on the SSD, which slows things down. If I still have plenty of free RAM, why is Windows still hitting the disk for memory pages? Is this normal behavior or a sign of a memory management issue?

I've been experiencing random 2-3 second delays when opening certain websites (like GitHub or Wikipedia) on Windows 10 using Chrome. After some research, I tried disabling IPv6 in my network adapter settings, leaving only IPv4 enabled. Surprisingly, the websites now load instantly without any delay. What is causing this fallback delay? Is it an issue with my local router, my ISP, or how the browser handles DNS resolution for IPv6? Any insights would be appreciated.

I started using rootless podman on a Debian 13 server recently, and to work around the problem of not being able to bind services to port 80 and 443 anymore, I've made some nftables rules on the host (in an nftables.conf file) to redirect traffic from those ports to 10080 and 10443 respectively, and those are what my services get bound to. The prerouting chain handles traffic from other devices, and the output chain works for traffic from the host itself, with daddr + the host's IP added to the rules to make sure traffic to external hosts isn't affected. #!/usr/sbin/nft -f flush ruleset table inet firewall { chain inbound { type filter hook input priority 0; policy accept; } chain forward { type filter hook forward priority 0; policy accept; } chain outbound { type filter hook output priority 0; policy accept; } } table ip nat { chain prerouting { type nat ho

I'm using Kubuntu 25.10 with some laptop attached to multiple different displays. Those displays are connected to some other laptop as well. Depending on the capabilities of the laptops and monitors HDMI and DisplayPort are used. Depending on which laptop I use, I simply switch the monitor inputs as necessary. Both laptops are running Kubuntu 25.10. But only with one of them I recognized an annoying behavior: When putting the affected laptop into stand-by or locking the screen only for some time, when logging back in, the music output is set back to 100 % volume. So, sometimes radio is unexpectedly screaming at me. :-) The interesting thing is that I always have a far lower volume level and sometimes it's even screaming at me when I muted the output entirely. I've already reduced sound volume of all output devices using system settings, but one device always comes back to 100 % for some reason. I've marked that red in the attached screenshot. THOUGH: That device is NOT

In WSL2 mirrored mode, it appears that each network device existing in the host OS (MS Windows) is also visible from inside Linux. How does this work? I assume in WSL2 there is a VM running an unmodified Linux kernel (rather than a WSL-specific kernel), and that the VM is interacting with virtual devices rather than the direct hardware (since the VM and host OS seem not to interfere with each other despite their concurrent and seemingly shared device usage), so is the hypervisor simulating a subset of a low level hardware interface (and letting normal Linux device kernel modules and daemons, like NetworkManager, behave naturally)? Does WSL inject anything else into the Linux OS (e.g. during installation or startup)? (E.g., is there any reason why restarting WSL would ever be necessary, say to fix any broken networking, rather than just instructing Linux to reset its own devices?)