I was working through some autoUpdating changes on an app of mine, and I ran across the following
Electron Native - https://www.electronjs.org/docs/latest/api/auto-updater
Userland Electron Updater - https://www.electron.build/auto-update
It seems that these two APIs are very similar, but not identical. I can't seem to find any information on what differences there are between the two and which one is the "preferred" way to do seem less Electron updates?
https://www.electron.build/auto-update
Differences between electron-updater and built-in autoUpdater¶
Dedicated release server is not required.
Code signature validation not only on macOS, but also on Windows.
All required metadata files and artifacts are produced and published automatically.
Download progress and staged rollouts supported on all platforms.
Different providers supported out of the box (GitHub Releases, Amazon S3, DigitalOcean Spaces, Keygen and generic HTTP(s) server).
You need only 2 lines of code to make it work.
Related
I'm looking for a solution to run some r scripts and display graphics on an iOs device using r.
I saw this app https://itunes.apple.com/us/app/r-programming-language/id540809637?mt=8 but we can't plot graph with it.
I just need something really basic, but which allows me to test some computations when I have an idea.
Thanks a lot.
I have tried connecting to the RStudio server using chrome on my browser but the arrows don't work so it's not perfect.
I used the app Juno, to connect to Jupyter hub running R and it worked much better. The Microsoft azure service provides free "notebooks": notebooks.azure.com
R Analyst runs R 3.1.5
Analyst has both R and Python. It also includes some fill-in-the-blanks forms for simple analyses. (e.g. a mini-SAS PROC)
I believe the developer is working on a 3.5.x update.
Apple App store rules limit the full functionality of R. Due to App Store rules, neither App allows you to install packages with compiled binaries. The default packages (all 188 of them, including the tidyverse packages, data.table, and knit) are on a system path that is not writable without jailbreaking. However, you can source directories and .libPath is functional.
Both allow you to create PDF graphics directly, and to save the displayed graph as png and jpeg (IIRC). knitr::spin and knit::stitch work if you set the dev to pdf. I've yet to test knitr::knit and its variations.
Both support document sharing via iOS files app, so you can use iOS editors like Textastic, Buffer, and GoCoEdit to write your code, and Working Copy for git support.
I am working on an electron app and was wondering which APIs from Node are included in Electron and if there are any differences.
For example, I know that fs is included, but I am wondering about other APIs like util.promisify. I haven’t found anything in the docs about the supported APIs.
Electron actually is a node.js app so the full node API is available from the used version. It is pretty much always the most recent version (at least in one month) but you can check it with
process.versions
in any Electron app
I am building a HTML/JS/CSS application that I would like to distribute packaged either as a NW.js or Electron app. I am new to both, and have read that the packaged executables can be quite large. Since the application would be free, I find that the cost of distributing it from a CDN to be too high.
I was wondering if there was any installer that allowed the "standard" parts of the executable (the Chrome components, etc.) to be downloaded as a shared library from a free CDN, as part of the app's installation / first execution?
With either one, you could host the packages on Github using the "releases" feature, which is free. This is a common approach for a lot of open source Electron applications.
Here are some examples of this being done:
yoda
Caprine
Here is the official documentation on Github releases.
There is a module for Electron designed to make this easy called electron-gh-releases.
I'm using Windows, Mac and Linux machines in my daily duties. On all machines, I program in C++ and various shells scripts. So far I've adopted the various "main" IDEs on each platform, but the diversity is irritating. I'm therefore looking into the possibility of using Sublilme Text on all platforms.
I have a setup of Sublime Text on Windows that works perfectly and would like to use the same on the other platforms also, so that when I change something in my Sublime setup on, say, my Mac, I can easily pick up the latest setup on my Windows machine the next time I'm there.
Is this possible on the 3 mentioned platforms, without getting (more) grey hair? If so, any suggestions or experiences thereof?
Many folks upload the "Packages/User" folder to GitHub (or your VCS service of choice). Then, they use Package Control to install their packages. Package control, through a settings file, will install any missing packages on a particular machine. I wrote a bit more about it here. You would then clone the git repo onto each machine, pulling updates when you decide to change something.
Alternatively, you could probably use a cloud service + symlinks to keep things auto synced, but I've personally never used it that way.
There are some plugins that are platform specific, so keep an eye out for those.
There's also the Package Package Syncing, which syncs installed packages and settings via some cloud service.
Works quite nicely, and automatically.
Has the advantage that you don't have to push/pull some dotfiles repo all the time.
No idea though whether this will work seamlessly across platforms (meaning whether all the settings will be platform-independent).
Suppose I have Rails 4 app, call it "Super SaaS". Now my client says he likes my appvery much, but he doesn't want his data to be in the cloud. So he says he would buy a licience from me to deploy "Super SaaS" on his own server. More like Atlassian Jira.
The question is: is there any secure way(in terms of protecting source code) to do that?
While you can probably package up your code as a JRuby application with JAR files and Java byte code, there are decompilers for that, so you can never presume your source to be 100% secure.
Ideally you'd offer some sort of VM appliance that the customer can install, a system image compatible with VMWare or whatever virtualization system they're using. This helps package up a fairly secure environment, but won't protect against a determined adversary trying to get your source code.
If you're giving out your code to someone in any form, compiled or otherwise, you have to have a certain amount of trust. Even compiled executables are not immune to reverse-engineering.