what's the industry standard say about organizing C++ source files liberally into folders?
I'm getting to the point in my project where I have almost 100 different classes and I'd like to start organizing them
In Xcode it just makes fake folders so you dont have to start changing include paths.. wondering if this is commonly done in C++
I just dump all mine into a "Source" folder that lives as a peer to my Xcode project. If you're going to be using any of the current generation of hipster editors (SublimeText, VS Code, Atom, etc.), you might want put some subfolders in your Source folder. Those sorts of editors tend to have a built-in file browser rather than a containing project file, so the only organization they'll show is from the file system.
htm on
0
Options
gavindelThe reason all your softwareis brokenRegistered Userregular
Me: "So its assembly?"
Him: "Except you have some registers blocked and you have to solve the puzzles using limited instructions and limited lines."
Me: "So its shitty assembly?"
what's the industry standard say about organizing C++ source files liberally into folders?
I'm getting to the point in my project where I have almost 100 different classes and I'd like to start organizing them
In Xcode it just makes fake folders so you dont have to start changing include paths.. wondering if this is commonly done in C++
I tend to prefer organising then into folders combined with CMake organising the include paths so that the directory structure is sort of like a namespace to the file itself.
This is because I tend to try to share the exact same code between multiple projects, and I find that it's much easier to do so when each set of projects is in its own directory.
If you're doing a small project, then this is probably overkill and not necessary, but if you start doing a large project, I highly recommend doing something similar, if only just to avoid filename collisions - especially if you're using third party libraries, sometimes they'll name a file something innocuous (but reasonable!) like "buffers.h". You'll need some way of separating that out.
Me: "So its assembly?"
Him: "Except you have some registers blocked and you have to solve the puzzles using limited instructions and limited lines."
Me: "So its shitty assembly?"
Hahahaha, yes, sorta. It's from the same creators as Spacechem and Infinifactory.
what's the industry standard say about organizing C++ source files liberally into folders?
I'm getting to the point in my project where I have almost 100 different classes and I'd like to start organizing them
In Xcode it just makes fake folders so you dont have to start changing include paths.. wondering if this is commonly done in C++
I tend to prefer organising then into folders combined with CMake organising the include paths so that the directory structure is sort of like a namespace to the file itself.
This is because I tend to try to share the exact same code between multiple projects, and I find that it's much easier to do so when each set of projects is in its own directory.
If you're doing a small project, then this is probably overkill and not necessary, but if you start doing a large project, I highly recommend doing something similar, if only just to avoid filename collisions - especially if you're using third party libraries, sometimes they'll name a file something innocuous (but reasonable!) like "buffers.h". You'll need some way of separating that out.
Me: "So its assembly?"
Him: "Except you have some registers blocked and you have to solve the puzzles using limited instructions and limited lines."
Me: "So its shitty assembly?"
Hahahaha, yes, sorta. It's from the same creators as Spacechem and Infinifactory.
Also The Codex of Alchemical Engineering (Flash game you can find on Kongregate, precursor to SpaceChem) and KOHCTPYKTOP: Engineer of the People (also available on Kong, integrated circuit design.)
Dude just makes super nerdy games.
Is there a good primer for modern deployment techniques? Specifically, I'm rocking Rails on Vagrant, I don't mind catching up on docker, and I want to deploy to a droplet on digital ocean.
Is there a good primer for modern deployment techniques? Specifically, I'm rocking Rails on Vagrant, I don't mind catching up on docker, and I want to deploy to a droplet on digital ocean.
I should specify that I'm currently using https://github.com/capistrano-plugins/capistrano-unicorn-nginx, but I can't deny how complicated it is to understand the constellation of apps and config files for new developers. Plus, something about storing environment information inside the project always rubs me the wrong way.
Is there a good primer for modern deployment techniques? Specifically, I'm rocking Rails on Vagrant, I don't mind catching up on docker, and I want to deploy to a droplet on digital ocean.
I should specify that I'm currently using https://github.com/capistrano-plugins/capistrano-unicorn-nginx, but I can't deny how complicated it is to understand the constellation of apps and config files for new developers. Plus, something about storing environment information inside the project always rubs me the wrong way.
I'd gladly help you with that, but what are you currently doing exactly CI(Jenkins?), CD, heroku-like workflow, something else...? You mentioned deployment and then you mention developers, so I'm a bit unsure if you want the Repo to Production part or actual devops with environment?
A setup built around docker can go from a simple haproxy in front of containers to a lot of extra tools regarding monitoring, service discovery, etc.
Thanks everyone that is still supporting PAdev! January and July are difficult months because many renewals are up then and it wasn't looking certain that we'd have enough still around to keep the lights on, but I think we're still okay.
Anyone that is looking for a hobby / proof-of-concept dedicated host, PAdev is a shared Linode VPS that we all chip in for and that I administer on behalf of everyone. $5 / mo gets you an account and to-date we have been able to setup and accommodate every tech stack requested!
+2
Options
gavindelThe reason all your softwareis brokenRegistered Userregular
Sitting through new employee orientation a week late. I like the part where I get actual employee benefits. Tech is so cushy it's ridiculous compared to my old jobs.
Yeah, I had that situation where a server had to query an off-site SQL server before it could give its response so every click was getting delayed anywhere from 2 to 7 seconds. It was horrible. Ended up just caching all possible results at pageload to avoid the lag.
DisruptedCapitalist on
"Simple, real stupidity beats artificial intelligence every time." -Mustrum Ridcully in Terry Pratchett's Hogfather p. 142 (HarperPrism 1996)
It's a web interface right? If the server that's serving it is bogged down or a piece of shit, you can definitely have a 3+ second delay.
Plus javascript is fucking terrible, so itself could be a 1+ second delay.
Clicking has no delay and moving around the page has no delay. It's just clicking buttons.
What does the button click do?
As infidel's link said, this is a problem known to Javascript/"responsive apps".
None of that other stuff would cause a delay, because you're not necessarily firing an event. If your event actually does something other than modify the internals of the page, there's going to be a noticeable delay because javascript.
TBH, if you need speed, you should be using an actual compiled/local app anyways. Just make calls to a RESTful service and call it a day and save your sanity instead of trying to patch bugs with how websites work on mobile devices.
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
Is there a good primer for modern deployment techniques? Specifically, I'm rocking Rails on Vagrant, I don't mind catching up on docker, and I want to deploy to a droplet on digital ocean.
I should specify that I'm currently using https://github.com/capistrano-plugins/capistrano-unicorn-nginx, but I can't deny how complicated it is to understand the constellation of apps and config files for new developers. Plus, something about storing environment information inside the project always rubs me the wrong way.
I'd gladly help you with that, but what are you currently doing exactly CI(Jenkins?), CD, heroku-like workflow, something else...? You mentioned deployment and then you mention developers, so I'm a bit unsure if you want the Repo to Production part or actual devops with environment?
A setup built around docker can go from a simple haproxy in front of containers to a lot of extra tools regarding monitoring, service discovery, etc.
I'm open to more or less anything, since I'm in the process of establishing workflows and best practices.
I'll clarify what we're doing now. We're a tiny team, so I'm lead developer / sysadmin / impostor devops guy. Currently we just (git) merge to master, and someone (usually falling to me, but not always) does a `cap production deploy`. That goes up to our digital ocean droplet. There are also a couple of proof of concept projects that go up to heroku, but that's not all that important.
The main thing is that I enjoy the reproducibility I get from vagrant boxes and how that onboards developers super fast. I would love to be able to have a clean, isolated environment in staging and production in the same way. My first guess is that I should tinker with containerization (read: docker), but I'm only really familiar with it in theory.
Is there a good primer for modern deployment techniques? Specifically, I'm rocking Rails on Vagrant, I don't mind catching up on docker, and I want to deploy to a droplet on digital ocean.
I should specify that I'm currently using https://github.com/capistrano-plugins/capistrano-unicorn-nginx, but I can't deny how complicated it is to understand the constellation of apps and config files for new developers. Plus, something about storing environment information inside the project always rubs me the wrong way.
I'd gladly help you with that, but what are you currently doing exactly CI(Jenkins?), CD, heroku-like workflow, something else...? You mentioned deployment and then you mention developers, so I'm a bit unsure if you want the Repo to Production part or actual devops with environment?
A setup built around docker can go from a simple haproxy in front of containers to a lot of extra tools regarding monitoring, service discovery, etc.
I'm open to more or less anything, since I'm in the process of establishing workflows and best practices.
I'll clarify what we're doing now. We're a tiny team, so I'm lead developer / sysadmin / impostor devops guy. Currently we just (git) merge to master, and someone (usually falling to me, but not always) does a `cap production deploy`. That goes up to our digital ocean droplet. There are also a couple of proof of concept projects that go up to heroku, but that's not all that important.
The main thing is that I enjoy the reproducibility I get from vagrant boxes and how that onboards developers super fast. I would love to be able to have a clean, isolated environment in staging and production in the same way. My first guess is that I should tinker with containerization (read: docker), but I'm only really familiar with it in theory.
Well, docker is still "new" in as much as people are still at a point where they are sharing workflows and saying "hey, that works fine for us" or "fuck, this really didn't do it.". I can describe our current environment in great detail, but I think that at this point it will be a bit too much like an info dump. I suggest you take a look at setting up a development environment with docker-compose(there are a lot of dockerized ruby images out there and at docker meetups usually Ruby is very well represented, so google should be a help). From a devenv to production, there aren't too many steps. If you are going to use a single node, it may be as simple as setting up Haproxy in front of the docker containers and then building directly on host via Jenkins->SSH(how do you test right now?).
If you need more scalability or a cluster, then it becomes a fair bit more complicated, you'll need a private registry, service discovery etc.
tl;dr: Read a bit on what docker/docker image is, then play around with docker compose and feel free to shoot questions in here.
Personally I find docker quiet nice as a building block for applications.
Is there a good primer for modern deployment techniques? Specifically, I'm rocking Rails on Vagrant, I don't mind catching up on docker, and I want to deploy to a droplet on digital ocean.
I should specify that I'm currently using https://github.com/capistrano-plugins/capistrano-unicorn-nginx, but I can't deny how complicated it is to understand the constellation of apps and config files for new developers. Plus, something about storing environment information inside the project always rubs me the wrong way.
I'd gladly help you with that, but what are you currently doing exactly CI(Jenkins?), CD, heroku-like workflow, something else...? You mentioned deployment and then you mention developers, so I'm a bit unsure if you want the Repo to Production part or actual devops with environment?
A setup built around docker can go from a simple haproxy in front of containers to a lot of extra tools regarding monitoring, service discovery, etc.
I'm open to more or less anything, since I'm in the process of establishing workflows and best practices.
I'll clarify what we're doing now. We're a tiny team, so I'm lead developer / sysadmin / impostor devops guy. Currently we just (git) merge to master, and someone (usually falling to me, but not always) does a `cap production deploy`. That goes up to our digital ocean droplet. There are also a couple of proof of concept projects that go up to heroku, but that's not all that important.
The main thing is that I enjoy the reproducibility I get from vagrant boxes and how that onboards developers super fast. I would love to be able to have a clean, isolated environment in staging and production in the same way. My first guess is that I should tinker with containerization (read: docker), but I'm only really familiar with it in theory.
Well, docker is still "new" in as much as people are still at a point where they are sharing workflows and saying "hey, that works fine for us" or "fuck, this really didn't do it.". I can describe our current environment in great detail, but I think that at this point it will be a bit too much like an info dump. I suggest you take a look at setting up a development environment with docker-compose(there are a lot of dockerized ruby images out there and at docker meetups usually Ruby is very well represented, so google should be a help). From a devenv to production, there aren't too many steps. If you are going to use a single node, it may be as simple as setting up Haproxy in front of the docker containers and then building directly on host via Jenkins->SSH(how do you test right now?).
If you need more scalability or a cluster, then it becomes a fair bit more complicated, you'll need a private registry, service discovery etc.
tl;dr: Read a bit on what docker/docker image is, then play around with docker compose and feel free to shoot questions in here.
Personally I find docker quiet nice as a building block for applications.
For the foreseeable future, we'll be running a single server and scaling up the droplet as our needs increase. So it's definitely going to be a case of multiple apps on a single VPS. As for testing, we really are a micro shop, and I haven't given CI much thought compared to actually getting development going. Right now, I can just trust each developer individually to run specs before merging.
So my raytracer project is back up and running. Rather than everything being one big c file I've split it up into a BMP writer and a raytracer that you request the next ray from. Also its in C++ but I'm only using C. Next I plan to add reflections.
I'm thinking after that I should get it running in a Windows window, then investigate using CUDA to make it faster.
Antinumeric on
In this moment, I am euphoric. Not because of any phony god’s blessing. But because, I am enlightened by my intelligence.
+3
Options
gavindelThe reason all your softwareis brokenRegistered Userregular
God it feels like the tablet has like a 3 second delay when we tap anything, but everything I do makes no difference at all. I have done:
React.initializeTouchEvents(true);
var injectTapEventPlugin = require("react-tap-event-plugin");
injectTapEventPlugin();
I made all of my buttons have onTouchTap instead of onClick... All of them still have the delay. What am I missing here?
are you using the PureRender Mixin for react?
could be that whatever state you are changing with button click is actually changing state throughout the app without you knowing it, even if the props are identical. with PureRenderMixin, it makes sure ShouldComponentUpdate only fires when necessary.
So my raytracer project is back up and running. Rather than everything being one big c file I've split it up into a BMP writer and a raytracer that you request the next ray from. Also its in C++ but I'm only using C. Next I plan to add reflections.
I'm thinking after that I should get it running in a Windows window, then investigate using CUDA to make it faster.
I remember one of my friends who was going to Polytechnique in Montreal hijacked a whole bunch of lab computers and popped SSH on it, then he dispatched batches of stuff for them to render for ray tracing. I think he could ray trace a scene in 2-3 minutes.
It was pretty amazing.
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
Let's hope it's all non OCR scans cell-phone photos of printouts of the documentation in random orientations inexplicably stretched to fit into arbitrary dimensions.
Posts
I'm getting to the point in my project where I have almost 100 different classes and I'd like to start organizing them
In Xcode it just makes fake folders so you dont have to start changing include paths.. wondering if this is commonly done in C++
I just dump all mine into a "Source" folder that lives as a peer to my Xcode project. If you're going to be using any of the current generation of hipster editors (SublimeText, VS Code, Atom, etc.), you might want put some subfolders in your Source folder. Those sorts of editors tend to have a built-in file browser rather than a containing project file, so the only organization they'll show is from the file system.
There is a well proven optical wavelength attack that allows an attacker to piggy back on legitimate access.
http://store.steampowered.com/app/370360/
Me: "So its assembly?"
Him: "Except you have some registers blocked and you have to solve the puzzles using limited instructions and limited lines."
Me: "So its shitty assembly?"
I tend to prefer organising then into folders combined with CMake organising the include paths so that the directory structure is sort of like a namespace to the file itself.
So:
This is because I tend to try to share the exact same code between multiple projects, and I find that it's much easier to do so when each set of projects is in its own directory.
If you're doing a small project, then this is probably overkill and not necessary, but if you start doing a large project, I highly recommend doing something similar, if only just to avoid filename collisions - especially if you're using third party libraries, sometimes they'll name a file something innocuous (but reasonable!) like "buffers.h". You'll need some way of separating that out.
Hahahaha, yes, sorta. It's from the same creators as Spacechem and Infinifactory.
Also The Codex of Alchemical Engineering (Flash game you can find on Kongregate, precursor to SpaceChem) and KOHCTPYKTOP: Engineer of the People (also available on Kong, integrated circuit design.)
Dude just makes super nerdy games.
I should specify that I'm currently using https://github.com/capistrano-plugins/capistrano-unicorn-nginx, but I can't deny how complicated it is to understand the constellation of apps and config files for new developers. Plus, something about storing environment information inside the project always rubs me the wrong way.
Priority inversion
Why are these interrupts executing so much code?
And relying on interactions from other lower priority interrupts?!
The same interrupts that have been blocked from executing, because a higher priority interrupt is in progress?!!?
*sob*
It's like
I dunno
I just don't know any more
This is I just wha urgghzhabhurbglablghlaaaa
I'd gladly help you with that, but what are you currently doing exactly CI(Jenkins?), CD, heroku-like workflow, something else...? You mentioned deployment and then you mention developers, so I'm a bit unsure if you want the Repo to Production part or actual devops with environment?
A setup built around docker can go from a simple haproxy in front of containers to a lot of extra tools regarding monitoring, service discovery, etc.
And phone.
Edit: If you meet a brit called Ben (Harrison)-Smith there. Tell him Campy says hi!
http://steamcommunity.com/id/pablocampy
Man.
How did their product actually work?
That's not a bug, that's a feature.
Thanks everyone that is still supporting PAdev! January and July are difficult months because many renewals are up then and it wasn't looking certain that we'd have enough still around to keep the lights on, but I think we're still okay.
Anyone that is looking for a hobby / proof-of-concept dedicated host, PAdev is a shared Linode VPS that we all chip in for and that I administer on behalf of everyone. $5 / mo gets you an account and to-date we have been able to setup and accommodate every tech stack requested!
You could totally spare the $5. :hydra:
Also, noticed a couple expirations are still pending, so maybe not quite there yet.
Naw, you're not one I sweat about.
var injectTapEventPlugin = require("react-tap-event-plugin"); injectTapEventPlugin();I made all of my buttons have onTouchTap instead of onClick... All of them still have the delay. What am I missing here?
It's a web interface right? If the server that's serving it is bogged down or a piece of shit, you can definitely have a 3+ second delay.
Plus javascript is fucking terrible, so itself could be a 1+ second delay.
Clicking has no delay and moving around the page has no delay. It's just clicking buttons.
http://www.sitepoint.com/5-ways-prevent-300ms-click-delay-mobile-devices/
What does the button click do?
As infidel's link said, this is a problem known to Javascript/"responsive apps".
None of that other stuff would cause a delay, because you're not necessarily firing an event. If your event actually does something other than modify the internals of the page, there's going to be a noticeable delay because javascript.
TBH, if you need speed, you should be using an actual compiled/local app anyways. Just make calls to a RESTful service and call it a day and save your sanity instead of trying to patch bugs with how websites work on mobile devices.
I'm open to more or less anything, since I'm in the process of establishing workflows and best practices.
I'll clarify what we're doing now. We're a tiny team, so I'm lead developer / sysadmin / impostor devops guy. Currently we just (git) merge to master, and someone (usually falling to me, but not always) does a `cap production deploy`. That goes up to our digital ocean droplet. There are also a couple of proof of concept projects that go up to heroku, but that's not all that important.
The main thing is that I enjoy the reproducibility I get from vagrant boxes and how that onboards developers super fast. I would love to be able to have a clean, isolated environment in staging and production in the same way. My first guess is that I should tinker with containerization (read: docker), but I'm only really familiar with it in theory.
Well, docker is still "new" in as much as people are still at a point where they are sharing workflows and saying "hey, that works fine for us" or "fuck, this really didn't do it.". I can describe our current environment in great detail, but I think that at this point it will be a bit too much like an info dump. I suggest you take a look at setting up a development environment with docker-compose(there are a lot of dockerized ruby images out there and at docker meetups usually Ruby is very well represented, so google should be a help). From a devenv to production, there aren't too many steps. If you are going to use a single node, it may be as simple as setting up Haproxy in front of the docker containers and then building directly on host via Jenkins->SSH(how do you test right now?).
If you need more scalability or a cluster, then it becomes a fair bit more complicated, you'll need a private registry, service discovery etc.
tl;dr: Read a bit on what docker/docker image is, then play around with docker compose and feel free to shoot questions in here.
Personally I find docker quiet nice as a building block for applications.
For the foreseeable future, we'll be running a single server and scaling up the droplet as our needs increase. So it's definitely going to be a case of multiple apps on a single VPS. As for testing, we really are a micro shop, and I haven't given CI much thought compared to actually getting development going. Right now, I can just trust each developer individually to run specs before merging.
I'm thinking after that I should get it running in a Windows window, then investigate using CUDA to make it faster.
are you using the PureRender Mixin for react?
could be that whatever state you are changing with button click is actually changing state throughout the app without you knowing it, even if the props are identical. with PureRenderMixin, it makes sure ShouldComponentUpdate only fires when necessary.
PARKER, YOU'RE FIRED! <-- My comic book podcast! Satan look here!
I remember one of my friends who was going to Polytechnique in Montreal hijacked a whole bunch of lab computers and popped SSH on it, then he dispatched batches of stuff for them to render for ray tracing. I think he could ray trace a scene in 2-3 minutes.
It was pretty amazing.
All in PDFs.
Fucking fuck.
No jury will convict.