I have a new installation of Centos 7 running cPanel and WHM on an AWS EC2 instance. All software is the latest versions.
I used WHM to setup 5 users, which in turn creates linux users with their own home directory and their own public_html directories. Then I have enabled MATE Desktop and Tigervnc so each developer can connect to the machine in a vnc session.
So basically this setup is 5 linux users with their own apache vhosts, running their IDE on the same local server.
I then installed Xdebug using pecl and now I want to setup VSCode on each persons account to use Xdebug. I am also using opcache.
When trying to use VSCode, it seems like we can get the debugger to work, but there is no output in the console of VSCode. I installed the php-debugger extension in VSCode.
Also we seem to have a problem with VSCode not able to handle include/require statements. It always says that they cannot be found. But the files are indeed there and indeed have the correct permissions to be read/written.
I am not really sure how to configure Xdebug and VSCode to get them to work for all local developers with VSCode. Will what I am doing even work? Do I need remote connections? Do I need a DBGp proxy? Does Xdebug only allow one debugging session at a time? Do I need to install Xdebug for each user?
Can anyone provide some directions on how to set this up?
Thanks
Will what I am doing even work?
Likely :-)
Do I need remote connections?
On the Xdebug side, it does not matter whether it's "remote" or on the same machine, in both cases a TCP/IP connection is used. The "remote" in "remote debugger" was always a bad choice of words by the Xdebug developer (me).
Do I need a DBGp proxy?
You don't need it. What you need to be able to do is to have each user initiate a debugging session to their own IDE. As everything runs on the same machine, you can't just reuse ports. Xdebug always connects to the same configured port (9003 by default).
There are two alternatives:
Use a the DBGp proxy to direct incoming connections from the Xdebug port to registered IDEs. Each of these IDEs will have registered their key with the proxy, so the proxy knows where to forward the request to. You will need to use a browser extension so that each developer can set their own unique IDE key as the Xdebug session (cookie) value.
Use Xdebug Cloud, which would handle the complicated proxy set-up for you, and the developers only have to set their IDE Key (or now, Cloud key) through the browser extension again, as well as in their IDE. Only PhpStorm supports this for now. There is configuration documentation available.
Does Xdebug only allow one debugging session at a time?
No. Xdebug supports one debugging connection per PHP request, but IDEs might not accept more than one incoming connection. Both PhpStorm and VS Code's Debug Plugin don't have this problem.
Do I need to install Xdebug for each user?
No. Xdebug is installed as part of PHP, and as you've only got one PHP installation, having Xdebug available as part of the is enough.
I have a Google Chrome Extension which uses a native host. This is used only on a windows box and the extension's registry settings are added along with the installation of the native host exe. Currently the the port (chrome.runtime.connectNative) or native host is started when the extensions background script is loaded. This currently means that the native host runs whenever Google Chrome is running.
The extension is used for only 1 website "www.example.com" and so content scripts only run when a tab with this website is loaded. This means that the native host would only be needed when a tab with this website is loaded and not all the time.
I now want to also create an Edge Chromium extension and give users the option to use either Edge or Chrome. As I indicated above both extensions would be "installed" meaning the registry keys added on installation of the native host. This means the extensions would be there for both browsers.
My main question is thus when and how should the native host be started.
From this main question I have a whole host of thoughts or questions;
When the user chooses to use Edge or Chromium should I at that point be making the changes to the registry? Problem here is users can of course independently of my app install the extension.
Is there a problem just leaving things as they are. In other words if Edge and Chromium are running then there are 2 different native hosts running and my app simply chooses which native host to communicate with based on the users choice. Problem here is you can see the native host process running in task manager and it would therefore be running for a browser that the user has not chosen which might not be OK for some users.
I thought about the extension having a flag to know if it should start the native host. Problem here is how could I turn this flag on. Without the native host I am unaware of a way to interact with the extension.
There are possibly more options so happy to hear them as well.
changes to the registry
Add chrome-extension://id/ with the id of your second extension to allowed_origins in the host's manifest as shown in the documentation.
there are 2 different native hosts running
Each host is started by its respective extension and it can communicate only to that parent extension, there's no confusion.
a flag to know if it should start the native host. Problem here is how could I turn this flag on. Without the native host I am unaware of a way to interact with the extension.
It depends on what the host and the extension do. Maybe you don't need to run the host all the time or maybe you can start it only when a tab with the matching URL is loaded. Maybe you can use the new experimental onConnectNative mechanism.
I want to debug my React.js project by adding breakpoints in WebStorm rather than in my web browser.
Is it possible? If yes, how?
Run npm start to get the app running in the development mode.
You can do this either in the terminal or by double-clicking the task in the npm tool window in WebStorm.
Wait till the app is compiled and the Webpack dev server is ready. Open http://localhost:3000/ to view it in the browser.
Create a new JavaScript debug configuration in WebStorm (menu Run – Edit configurations… – Add – JavaScript Debug). Paste http://localhost:3000/ into the URL field.
In WebStorm 2017.1+
No additional configuration is needed: go to step 5!
In WebStorm 2016 (.1, .2 and .3)
Configure the mapping between the files in the file system and the paths specified in the source maps on the dev server. This is required to help WebStorm correctly resolve the source maps.
The mapping should be between the src folder and webpack:///src
If you’re wondering how we got this mapping, check http://localhost:3000/static/js/bundle.js.map file. This is a source map file for the bundle that contains the compiled application source code. Search for index.js, the main app’s file; its path is webpack:///src/index.js
Save the configuration, place breakpoints in your code and start a new debug session by clicking the Debug button next to the list of configurations on the top right corner of the IDE.
Once a breakpoint is hit, go to the debugger tool window in the IDE. You can explore the call stack and variables, step through the code, set watcher, evaluate variables and other things you normally do when debugging.
This app is using Webpack Hot Module Replacement by default and that means that when the dev server is running, the app will automatically reload if you change any of the source files and hit Save. And that works also together with the WebStorm debugger!
Please take note of these known limitations:
The breakpoints put in the code executed on page load might not be hit when you open an app under debug session for the first time. The reason is that the IDE needs to get the source maps from the browsers to be able to stop on a breakpoint you’ve placed in an original source, and that only happens after the page has been fully loaded at least once. As a workaround, reload the page in the browser.
Webpack in Create React App generates source maps of the type cheap-module-source-map. This kind of source maps do not guarantee the most precise debugging experience. We recommend using devtool: 'source-map' To make changes to the app’s Webpack configuration, ‘eject’ the app (refer to the Create React App manual to learn more).
I have a web app developed with a NodeJS + Express + GraphQL + MongoDB back-end and a ReactJS + Apollo front-end. I would like to deploy this application locally. Is that even possible?
I have come across dozens of "how to deploy to Heroku," "how to deploy to Digital Ocean", "how to deploy to Github", etc. But none that explains how to deploy locally.
Right now, I run: nodemon server for the back-end, and npm start for the front-end. I see the application running on http://localhost:3000/ (I use cors to connect the front end with the server running on port 3001).
I would like to just go to http://localhost:3000/ and see the app without having to execute the commands npm start and nodemon server. Is this possible? If so, how do I do that?
To my knowledge, our local server is not a WAMP server (our OS is Windows though). The IT department told me that it is a
[...] plain, regular old server. The address is localhost running on
port 3000. You can open up another port on 3001 if you need it. Just
drop your stuff on the C: drive and you should be good to go. I've
never heard of Node or React so I can't help if you have questions.
Any ideas? Many thanks in advance for your help!
UPDATE
There seems to be a bit of confusion surrounding what I am looking for. I am trying to deploy this locally.
Let's say, on your local computer (your laptop at home) you go to localhost:3000 on your favorite browser. Unless you are serving something to localhost in that moment nothing is going to show up, it will say "refused to connect" or something. What I want is to be able to open any machine on the network whenever I go to localhost:3000 and my react site appears and functions...does that make more sense?
I don't want this is development mode. I want a build of this project on localhost...I'm starting to think this isn't possible.
As i understood, you want to deploy it on a local server, not locally on your developing device.
I thought about doing that...but I'm not so sure IT will be okay with it always running... :(
How can you use a server if its not running? Just like WAMP (which runs apache), or whatever you got rolling there, it must be running. So, just make it a background process like slawomir suggested.
PS I dont think you understand node server properly though.
Read this to understand why node server needs reloading. After that you need to understand that no hot reload tool is perfect, and you gonna need to restart your server from time to time.
PPS I dont know what this means
[...] plain, regular old server. The address is localhost running on port 3000.
if there is a server running on 3000, youll need to change port for your server to smth else (most common is 9000)
To solve the problem you can create a startup script, which executes npm start and nodemon server. Then make sure to keep it hidden, so that your server will be always running. Keep in mind though, that any errors thrown will stop your server and unless you configure it, the server won't reload by itself.
I would try following:
build your app with the production environment variables set
get all files from dist folder and deploy them in your server
now access your app using localhost/
Maybe what you are looking for is something like ngrok which creates a socks tunnel to your localhost, effectivelly deploying from localhost, as I understand it, allowing you to access your localhost through a url like ldiuhv093.ngrok.io, or even a custom subdomain if you pay for a subscription fee.
If I have this wrong, someone please tell me!
To solve the problem first of need to create a batch file with .bat or .cmd extension and under that file add the following 2 command
nodemon server
npm start
Then follows the following steps to add it as a startup script for windows OS.
Create a shortcut to the batch file.
Once the shortcut has been created, right-click the file and select
Cut.
Press the Start button and type Run and press enter.
In the Run window, type shell:startup to open the Startup folder.
Once the Startup folder has been opened, click the Home tab at the
top of the folder and select Paste to paste the shortcut into the
folder.
Above steps are for example to create a batch file and add it as a startup script for Windows 8 and 10 users.
For better clarity or reference follows the following link.reference-link
There's no option to reload the server while keeping it running. You could, technically, have your 'main' file monitor another file for changes. This would be the file where you actually keep your sever program. Then, on changes, you discard your current logic and start executing that. That said, doing it that way would be very fragile and a very round-about way to do it. It also wouldn't fix your front-end for which you'd need a similar solution.
Instead, you could hook into your favorite editor's save event, and run those two console commands, so that every time you save, the server is automatically brought up. (Make sure to also clean up existing servers)
Run on Save for VSCode
save-commands for Atom
I know this post has been two years. But, I think the solution to your second desired outcome is to use concurrency. https://www.npmjs.com/package/concurrently.
This will allow you to do one NPM START to start two all three processes.
and to your first question, I think the solution is to add Electron to your app so you can package it to an executable application. When you start the app, your express server will start running in the background.
Most people probably don't understand why there is a need for this. Running on local server (computer) allows access to local file system and can even run SQL queries inside the proxy which would require IT involvement if hosted on outside server.
From what I have understand, that you want to deploy your app on local server that means you want to deploy it on the network that you are connected to.
Check ip from the command prompt
To deploy it locally,
Run: HOST=ip npm run start
It will get deploy on your local server. And everyone connected to the server can access the url
If this worked for you, kindly upvote
You need to do npm start There may be other ways of starting it but, all will result in the same. You can read this article on Freecodecamp on deploying on DigitalOcean. You can manipulate it to your localhost. Shouldn't be too different.FCC Tut on Deploying
I am using Red Hat Enterprise Linux Server release 6.2, and I have only terminal access (no GUI). My company has a proxy auto-config script (PAC), and that is the only way to get internet connection.
I had no problem with my windows machine, since it was easy to set it up in IE->Tools->Internet Options. But in Linux, especially without a GUI, there seems to be no way to use this PAC script (I have been Googling for a while). By the way, the script is over 400 lines of Javascript.
Anyone knows how to solve this problem?
Proxy settings are implemented differently according to the software you use. On graphical desktop environments there are setup tools to configure a PAC; browsers like Chromium and Firefox detect the current desktop environment and import the proxy settings from there; Firefox also used to offer options for manual configuration of proxies and PAC URLs.
Integration of proxy options is not a priority for developers of text-only software packages. Each one of your tools will have to be manually configured to use a proxy. For example, APT must have a proxy configuration file at /etc/apt/apt.conf.d/.
Some packages may consult the environment variables http_proxy, https_proxy, et al, for proxy configuration.
You will need to read the documentation of the packages you will use in your GUI-less Linux to learn how to configure each of them to use the company's proxy. Some packages will use protocols and ports other than HTTP(S), FTP, GOPHER, which may be forbidden by the company's firewall.
I know this is an old thread.
In my case, I did the following steps
Download the automatic configuration script and copy proxy address from that file.
export "http_proxy" and "https_proxy" in bashrc using the proxy address get from step 1
Reload bashrc file
use the PAC even in the terminal environment. it works.