I am trying to understand some differences between the peer-to-peer model and client-server model, and their uses in Git.

(1) In the client-server model, is it correct that a server process has to run as a long-lived daemon, and a client runs only roughly between sending the first request and receiving the last response?

(2) In the peer-to-peer model, does every peer process make both requests and responses to any other peer process? Or, does each peer host instead run two processes: one as a server process and one as a client process?

(3) Is it correct that we can have both the peer-to-peer model and client-server model in Git. What process(es) does each host run for peer-to-peer and for client-server? Is Git inherently peer-to-peer? Does Git simulate client-server by having one host running only a Git server process, and any other host running only a Git client process?

(4) Is it correct that Git can access a local repository by running only a local Git client process and without running a local Git server process, and Git can access a remote repository only by running a local Git client process, and running a remote Git server process locally to the remote repository?

Thanks.

To be a true JS-alternative Blazor should have a hybrid model where some pages are server-rendered while some are WASM, similar to NextJS and similar frameworks, that way it's a full framework for not only enterprise or "I don't care about SEO" apps. In this day and age it's imperative that some portion of our apps are indexed and some are client-rendered.

Anyone knows how to achieve it with either the .NET 5 or .NET 6 versions of Blazor? Or is there a way to vote on requesting this feature, to truly replace NextJS, NuxtJS, etc.?

Probably one for the FAQ. Just to be sure, can I get somebody to confirm this basic model

• Plex server indexes, manages the media. It can transcode the data out to plex clients but, it isn't an integrated client in itself: its the back-end. Its not a media player.

• Plex client. this can be a web browser session, or an app built natively for the OS you run it on. It can e.g. invoke chromecast to fling content to a third device. It is a media player. It can cache content.

• PlexAMP is a music-only plex client. It can cache content.

• Plex server does NOT have a client: when you connect to the web port, "play" happens in your browser, or app, on the host you run the client on, or where you send it. This doesn't include the HDMI and sound output on the host you run the server on unless you install a plex client.

• If you install a plex client on the server host, you can run both plex server, and a local plex client instance at the same time, and THEN you can play things on that host, through its HDMI and sound output.

Is that basically correct? I'm on a Pi4 Ubuntu system for plex server, and its how I see the world right now, happily running clients on OSX (web view) and Android phones and tablet. But, unable to use the sound out on the RPi4 I have plex server on.

I am a new user of emacs. I recently learned about emacs client-server model. So I am in doubt that my understanding might be wrong about this system. So I will try to explain what this client-server model is in terms of emacs and I seek suggestions from this community about the validity of my understanding.

So client-server model is basically used for utilizing the same running instance of emacs instead of opening up a new session everytime. How emacs does this is it creates a emacs server instance which keeps running in the background . Now if I want to to do something with emacs by using that single instance running in the background I will do so by emacsclient binary. For example if I want to open a file in gui frame (emacsclient -c filename) emacsclient will direct the emacs server to do so. I can create many instances of emacsclient which will use the same emacs server instance that is running in the background. By doing so I can do things in one session. Avoid the time cost of starting up a whole new session and open files quickly. There are different switches for emacsclient and emacs server which one can utilize to use different features provided by both of them.

Thats it. This is my current understanding of emacs client-server model and how to use it.

I bought a "cook book" for asio where the author goes through basics of the library and about basics of theory with recipies of code. I am creating a simple texture based bomberman 2D game (2-4 players) with SDL2 and got stuck in how to decide upon a way of implementing networking.

I wanted to initially go with a sync TCP server and client where everything is in the same thread. But the author says that just one "malicious client" would be enough to freeze the application. Which makes sense I guess.

What would I decide go for then? I read up on UDP and it says that it is connectionless and requires no "handshake" or moments of waiting for client to finish their action.

The book also goes through async client and server tcp model. Apparently, doing it async would solve most problems according to the book.

What do I need to read up on? Which model will fit me the best?

Thanks.

Hi all!
Does anybody have example or some information about creating time client and server models on ESP32?
I'm working with esp - idf. It has a sensor example, on\off example, but unfortunately there is no any example of time model :(
Bui it's necessary for my project - i want to share time through my esp ble mesh, so i have to create time relay. But time relay based on time client and server.
I'm stumped, please help

In the original, if someone lagged, the game just slowed down and would pause and play.

In DE, it's way worse. If someone lags your entire screen constantly freezes for seconds at a time. causes so many misplays as you can't control units or eco.

I understand if the game has to pause while someone "catches up". But why does it have to lock up everyone's screen?

I was just reading this post, which reminded me of a similar situation.

TL;DR: Sales and Marketing Director has a plan to get an edge over his competition. Ends up bleeding clients to the competition instead.

This took place a little over 20 years ago in South Africa.

I'd just joined a new company that was a technical service provider for a number of companies, including gaming (as in gambling) companies, cellular companies, banks and more. We did tech work, but the industries were vastly different and that made the work incredibly challenging and interesting.

As the new IT manager, one of the first things I did was to install a ticketing system (Request Tracker - RT, if you're interested) for people to log IT support tickets. In under a week I had enough information to start making recommendations on streamlining a lot of our technical services.

The sales and marketing director (SMD) saw the stats I had access to and got (almost sexually) excited, and demanded that we give his sales guys access to the request tracker. It wasn't a perfect fit out of the box, but over the next few weeks we tweaked it so that it became a very useful tool for them.

Some thing to understand about is is that we provided technical services to our clients, and we used special (and stupidly expensive) software that was licensed to us. The way the licenses worked is that there was a very expensive 'management' server that required a license that cost about $10 000 per year, and then you needed client licenses that cost something like $20 000 per year for 10, and $25 000 per year for a hundred. We had about 5 thousand licenses, so each individual license ended up costing us something like $30 - $100 (different license capabilities). So we ran the 'main' management server at our office, and delegated x licenses to a special management server at each client's site. We got away with this because only our staff used the client licenses, even though they were using the client licenses on behalf of a different entity entirely (it was quasi legit - had the software license owner challenged this in court they may have won, but their competitor had a significantly cheaper product with slightly less functionality that we would have migrated to so everyone just pretended we were friends and were OK with everything going on. We were by far their biggest custom

Hello rustaceans, i have written this article on using rust + grpc (tonic) to make prediction calls to tensorflow model server.

Hope, you find it useful.

Hey guys, I'm new to the sub as well as network programming in general, and I need to simulate a server and client locally on my machine using HTTP calls to communicate with each other. What I mean by client and server is that this can simply be two different executables running at the same time. Through googling, I was able to find many articles describing how to set up a rest api, but I want to simply run two executables that communicate locally via rest api (http) calls. Maybe I'm not researching correctly, but I can't seem to find a good guide on how to set this up. Any help would be massively appreciated!

My end goal is to write a piece of code that receives an HTTP call from the "client", forward to the "server", and the server sends a response back to the "client". The client and server in this case are arbitrarily named, as they are just executables for our purposes.

Edit: Apparently the "non malicious code" that was the source of this has been removed by the dev and has been made public on the kyber discord. I'd say everything's resolved.

Just a quick heads up for everyone. A Kyber admin just remotely opened this (not meant to rick roll) on everyone's PCs when we were doing a private match. After people freaked out ingame they tried to do damage control by saying "that's pretty much all we can do".

I'm not an IT expert but if they can remotely execute code on any PC then there's gotta be someone who's in the mood to destroy people's PCs by hosting a server and then downloading viruses. Don't think that any of the Kyber admins would do anything like that, but to put it mildly: This has some MW2 levels of trust issues with their servers.
(To clarify this is exactly what currently happens on COD MW2's PC Lobbys - Hackers able to remote execute basically anything on the PCs connected to the same lobby as they are)

There's always been a bit of disagreement between the client and server regarding how much health an enemy has. You'd see it occasionally where the enemy's health suddenly changes when the server update comes in and sets the record straight.

As of the Anniversary update (version 3.4.0), this became the rule rather than the exception. There is some systemic issue causing the client to consistently overestimate the damage dealt to an enemy.

This next bit is speculation. I believe this bug is responsible for the non-lethal Finishers. The client thinks that an enemy's health is low enough to be Finished when it actually isn't. It plays the Finisher animation, destroys the enemy locally, and even triggers effects that take place on Finishing an enemy. Then the enemy "respawns" as part of resynchronization when the next server tick comes in.

Download it here

EDIT: Updated guide here

Northstar is a Titanfall 2 mod that enables custom servers and allows multiplayer mods! This is huge because the official servers are still plagued by DDoS attacks, rendering the game nigh unplayable. Now we can not only play on community servers, but also have mods and custom rules!

To install it on Linux, download the latest release then drop the contents of the folder into the game files (which you can find by right clicking the game on Steam --> properties --> local files --> browse)

Then, rename NorthstarLauncher.exe to Titanfall2.exe, and rename the original Titanfall2.exe to anything else.

After that, I would suggest forcing the game to use Proton Experimental, as I had issues running it with the latest Proton stable (6.3-8). If it exists, I'd also recommend deleting the game's proton prefix folder (steam/steamapps/compatdata/1237970/), which is always recommended when changing proton versions.

Now just launch the game through Steam and it should work! You'll be greeted with a Northstar welcome message on the main menu assuring you that installation is successful. Have fun.

If you have any issues running it, you could check the linux release issues on their github for updates.

Let me make it as simple as possible. I have developed a diary application, which is split into two sections:

Backend (golang):

• Backend application needs a working instance of Mongo running.
• Backend needs to be started in order to start using the frontend.

Frontend:

• Frontend is a CLI frontend which is written in Python.
• Python application needs requests library to function.

I want the thing to be so simple that end-user had to just write appname and it should work.

Here is my situation. I'm not sure how I'm gonna distribute it. How do I make sure is running when the frontend is first accessed.

backend -> mongo backend -> golang app with libraries (mgo, goji)

frontend -> python with library (requests)

I had thought of making two RPMs; for backend and frontend respectively. But that would be too RHEL centric. I've also considered docker. But I'm not sure how would that orchestrate and how will end user interact with it.

I want an outline.

There’s Gizelle, Karen, Wendy, Shannon, Gina, Leah, Luann. I think that’s it. I’m legit excited! One of them makes her designer cry! Andy is guest judge.

Our game is completely ballsed beyond belief.

https://streamable.com/c34m4 (only need to watch the first 30 seconds)

https://streamable.com/0y9xs

What the fuck is going on?

Edit: Seems to be a dedicated server problem, listen servers have the -1hp one shot bug instead.

In my contract, it says the full amount is due on the day before of day of the event. I sent an invoice the morning after, so that my client had a clear receipt of what they owed. I don't wanna be pushy but I am waiting on my money. All my other clients have been great and always paid within 24 hrs, no problem. If you wanna hear another red flag, read further, otherwise, any advice?

UPDATE: He said he wanted to pay me after he “approved of my editing style”. I referred to the contract and said I will not move forward until I got paid. He paid the full amount and now demanded that I send him all raw, unedited photos and the personal video clips I took on my phone for my own use. I will not be doing business with this client in the future.

Other red flag: 2nd model cancelled last minute before the shoot and my client was scrambling to find a backup. Finally, they found someone who was able to drive in over an hour away on two hours notice. Makeup look was approved by my client and done by an MUA upon arrival. At the end, my client only wanted to pay the model for 1hr of work when she was there for 2.5hrs because she "took too long to get ready". The MUA and myself backed up the model. They argued for over 20 mins and finally client agreed to pay her for the full 2.5 hours.

Authoritative, or deterministic?

These games are city builders/idle clickers.

All players exist on the same "Map" but there is no input for movement. What I mean is, you can tell your troops, "March Here". But you can't tell them which route to use, and you can't have them "move left" or "Move right". The user input is restricted to "March Here" or "Recall March". What type of model is best for a game like this?

The server was shut down earlier as the news of an exploit within the game, which affects servers also, started to quickly spread, to avoid potentially any players to be affected by this or the server itself.

This affects many servers and all versions of the game until Mojang applies patches to all versions (some of which they have already done). Right now it is not recommended to play on servers that you do not trust, or use any clients/mods that have not patched this exploit.

You can read more details here and here as well as you will most likely find further information elsewhere as it affects a large part of the minecraft community.

We have patched this on our server end and made protections against the exploit being sent by players, but have decided to keep the server further offline until there are further developments, and/or that it is fully patched and that more players become aware of it.

The server should be back online later today (sometime within 16 hours from this post) unless something changes that delays this, this post will be edited if so.

Meanwhile if you play minecraft right now it is recommended to not play on servers or with players that you do not trust, and that you turn off chat, however turning off chat does not entirely block this exploit if your client/mod could possibly read the exploit from elsewhere which is why it is important that the developer of your client/mod updates against this as well.

Update:

The server is now back online, along with Mojang having applied patches for most client versions. Any custom client or mod that has not updated may still be affected, you must check this yourself.

Update 2:

Some players still felt unsafe seeing 'joke' variations of the exploit in their messages or chat, and there was a potential workaround found that affected outdated clients/mods, so for now we have disabled it being possible to send any of the characters used in the exploit in the chat or direct messages. Outdated clients or mods are still affected and possibly any 'updated' client or mod that has not done an effective enough patch, while we have made attempts to protect any outdated player from being sent this exploit there might still be other ways depending on how your client or mod works.

If you want to be completely safe you may wait for further updates across the community regarding this exploit

Hello, I've been playing on Ionia from the UK for a week now and just unlocked ranked and noticed some big differences which I think we'd enjoy on our client. Not doing long intros just jumping to the point. (I'm peak gold 2 euw btw so I can't talk about high level stuff.)

1. TFT pass levels up really fast and can be levelled up through summoners rift games. So non TF players like me get emotes quickly. There are also other passes that offer free content which is cool. As well as a new player introduction thing where u get any champ in the game and some free stuff.

Edit : haven't really played on EUW for a couple of patches. This doesn't seem unique. However it does have more progression systems (3 battle-pass systems that I could find vs the 1 I've seen on riot client)

2. Every champ is unlocked. I think this is due to an event but it feels nice that events like this exist where I can play anything I want (similar to how pros/streamers get fully unlocked accounts but u still have to buy skins).

3. New players get double xp + free champs and skins temporarily at different level breakpoints all the way to 30. This makes unlocking ranked really easy if you grind for 1-2 days.

4. In client guides. You can see popular runes/skill orders/spells/items for each champ with per role information. Great for beginners and appears in the main client home area.

5. You don't see player names in champ select. This is something I've heard high elo players request a lot, I'm just a silver/gold player on euw so it doesn't impact me but I know this one is a common request.

6. Automatically curated rune page that are pulled from progamers/high elo that you can import (think blitz.gg/mobalytics/etc). This appears in champ select.

unrelated) Less people having mental breakdowns. You'll either get people trying 500% to win at any cost or somebody will just leave the game. Much better than the passive aggressive half in half out approach I see on EUW/NA. This isn't really anything to do with the client. Also I swear these players have faker hands 9/10 times, they just have insane mechanics in bronze even if their brain is not quite installed. You will also get perma invaded by jg + supp and every camp seems to be warded 24/7.

This is a desktop application to manage customers, sales orders, invoices and so on. Currently I've split the application in three projects: the UI, services and data access. I'd like to have the UI in a client pc and the services and data access in a server. Is it possible?

I installed the server on MacOS through a Docker container using these steps. Everything seems to be working when I open the web interface from Docker. Google results (including advice from this sub) and the Next Cloud login window say to just get the server address from the browser address bar, but that’s just “http://localhost:8080/settings/admin/serverinfo” There is DNS address in the system menu of the Next Cloud web interface but entering that just times out, and I can’t find it anywhere else. Help is appreciated… 🙏

So the question for my Web Systems Architecture class is to name an example of communication that does not fit the client-server model. In my mind I am thinking phone calls don't follow it right? If I understand correctly they follow Peer-to-Peer right?

If I am not right please help me out because I really don't know.

Thanks in advance.