Why do you need AI to determine if someone is logging in from a different location?
People have built things like that ~10 years ago with good old if statements.
I built a system that analyzed login patterns to deter account sharing once. It was about 40 lines of Python composed of a tiny bit of logic and a database lookup. It literally took 30 minutes to create.
Having private repos is nice and all but CI is equally as important.
With that said, Gitlab + Gitlab CI is free to use and is a perfect match for solo developers or small teams with private projects, without having to invest in any additional services or infrastructure.
Where as on Github, if you have private repos, you can't use Travis CI for free, so now you have to choose between Azure pipelines (which is more limiting than Gitlab CI) or use some other free service like CircleCI which is also more limiting than Gitlab CI and in both cases you need to integrate third party tools into Github where as on Gitlab, it's all bundled in and ready to go.
I am really curious why the Gitlab CEO didn't talk about that. Any developer deploying web apps knows that CI is an essential tool.
I used to use a Blue Yeti for screencasting but got rid of it. They tend to perform horribly in untreated rooms because they pick up everything.
You can usually get a much better result with a dynamic mic for any type of recording where you're not moving around and it's just you talking. Not only will you reduce background noise, but you'll also get a great tone. A $70 AT2005 dynamic mic (which has both USB and XLR connections) outperformed the standard Blue Yeti in my case by a huge amount.
The interesting thing about that is most people do this with credit cards.
You're paying massive interest (usually) for a company to track all of your purchases. Of course you get something of value in return (money that isn't yours for a short period of time) but the concept is the same. You're paying out of pocket to have a company harvest and profit off your personal information.
I'm concerned that more people aren't upset over that.
The article mentions that it runs Linux through a VM but that's not going to be very good in terms of performance.
GalliumOS has been around for years. It's a fork of xubuntu that's optimized for Chromebooks.
I've been running it on my Chromebook for 2 years. It's something you boot into directly, so no VMs are required (and you can choose to dual boot into ChromeOS if you want to keep it around).
A lot of freelance platforms have been doing this for years but it's not a reasonable solution.
You can't measure development productivity based on trackable "focus" and "intensity" scores because a lot of that happens inside of your brain.
I might decide to just stop what I'm doing and do 50 push ups while thinking about a problem, and then afterwards spend 10 minutes doing nothing from a camera's POV. In my mind, I'm churning through really complex data models and trying to make sense of it all which is absolutely focusing on the work at hand.
I use this one course hosting platform called Thinkific and pay them $100/month.
I reported a bug 4 months ago that allowed people to purchase the wrong course when they had 2 check out windows open. It's still not fixed because it's not considered a priority. So a critical billing bug which results in customers buying the wrong course isn't considered important by their decision makers.
So while developers are making these mistakes I have a feeling a lot of times they don't get fixed because someone at the management level has an even worse understanding of what needs to be addressed.
I've been using a standing desk for about 18 months and it's really good to break up sitting sessions.
I stand for about 12 hours a day, but it's not just standing in place. Standing naturally gives you a reason to be active. You can easily stretch your legs and move around. Sometimes after completing a programming task or similar, I'll just shadow box for 30 seconds because "why not".
The TL;DR is, a real world web application might be able to load and get updated in 5 seconds in development on native Linux (imagine a large Rails app with hundreds of assets), but the same exact app ran through WSL takes 32 seconds before you can see your changes. That's straight up unusable.
Personally, I just run VMWare in Unity mode. At the end of the day it lets me run graphical Linux apps with floating windows that feel and act as if they were running in Windows. The performance is excellent (it feels nearly identical to a native Linux experience with none of the WSL downsides). If anyone is interested, I put together a screencast and have a few screenshots showing how to do it at https://nickjanetakis.com/blog....
If price is any bit of a concern then I recommend installing GalliumOS (an xubuntu based Linux distro made for Chromebooks) onto a Toshiba CB35 2015 edition.
It has a 1080p display, weighs less than 3 pounds and is plenty fast for development work. The best part about it is it only costs $350 with a 128GB SSD modification.
If you're interested I have a full write up and review on it here:
https://nickjanetakis.com/blog...
Slashdot Mirror
Why do you need AI to determine if someone is logging in from a different location?
People have built things like that ~10 years ago with good old if statements.
I built a system that analyzed login patterns to deter account sharing once. It was about 40 lines of Python composed of a tiny bit of logic and a database lookup. It literally took 30 minutes to create.
Having private repos is nice and all but CI is equally as important.
With that said, Gitlab + Gitlab CI is free to use and is a perfect match for solo developers or small teams with private projects, without having to invest in any additional services or infrastructure.
Where as on Github, if you have private repos, you can't use Travis CI for free, so now you have to choose between Azure pipelines (which is more limiting than Gitlab CI) or use some other free service like CircleCI which is also more limiting than Gitlab CI and in both cases you need to integrate third party tools into Github where as on Gitlab, it's all bundled in and ready to go.
I am really curious why the Gitlab CEO didn't talk about that. Any developer deploying web apps knows that CI is an essential tool.
Check out https://namesilo.com./ .com domains are $8.99 all the time and it includes free private registration for life.
It's $1 more per year than Cloudflare except it's available now and includes free email forwarding (which Cloudflare does not).
I used to use a Blue Yeti for screencasting but got rid of it. They tend to perform horribly in untreated rooms because they pick up everything.
You can usually get a much better result with a dynamic mic for any type of recording where you're not moving around and it's just you talking. Not only will you reduce background noise, but you'll also get a great tone. A $70 AT2005 dynamic mic (which has both USB and XLR connections) outperformed the standard Blue Yeti in my case by a huge amount.
This snippet isn't telling the full story.
While basic is limited to 1 screen, it's also limited to only SD streaming.
Standard has HD
And now this new premium has 4k HD
But prior to this plan change, standard members could stream 4k video. Now they require you to pay about 20% more to get what you previously had.
The interesting thing about that is most people do this with credit cards.
You're paying massive interest (usually) for a company to track all of your purchases. Of course you get something of value in return (money that isn't yours for a short period of time) but the concept is the same. You're paying out of pocket to have a company harvest and profit off your personal information.
I'm concerned that more people aren't upset over that.
Games have had paid loot boxes for a long time.
I once knew a guy who spent over $10,000 on getting randomized loot playing the North American version of Pangya during the mid-2000s.
Or 5 million CI servers.
The article mentions that it runs Linux through a VM but that's not going to be very good in terms of performance.
GalliumOS has been around for years. It's a fork of xubuntu that's optimized for Chromebooks.
I've been running it on my Chromebook for 2 years. It's something you boot into directly, so no VMs are required (and you can choose to dual boot into ChromeOS if you want to keep it around).
Details on how to set it all up can be found at https://nickjanetakis.com/blog....
I run all sorts of real development work loads on a $350 Chromebook and it's rock solid.
A lot of freelance platforms have been doing this for years but it's not a reasonable solution.
You can't measure development productivity based on trackable "focus" and "intensity" scores because a lot of that happens inside of your brain.
I might decide to just stop what I'm doing and do 50 push ups while thinking about a problem, and then afterwards spend 10 minutes doing nothing from a camera's POV. In my mind, I'm churning through really complex data models and trying to make sense of it all which is absolutely focusing on the work at hand.
I use this one course hosting platform called Thinkific and pay them $100/month.
I reported a bug 4 months ago that allowed people to purchase the wrong course when they had 2 check out windows open. It's still not fixed because it's not considered a priority. So a critical billing bug which results in customers buying the wrong course isn't considered important by their decision makers.
So while developers are making these mistakes I have a feeling a lot of times they don't get fixed because someone at the management level has an even worse understanding of what needs to be addressed.
I've been using a standing desk for about 18 months and it's really good to break up sitting sessions.
I stand for about 12 hours a day, but it's not just standing in place. Standing naturally gives you a reason to be active. You can easily stretch your legs and move around. Sometimes after completing a programming task or similar, I'll just shadow box for 30 seconds because "why not".
If anyone is curious, here's how I built my standing desk for about $50 https://nickjanetakis.com/blog....
The biggest problem with WSL is that the mount performance is unusable for every day web development.
There's been a pending issue on GitHub for a year at https://github.com/Microsoft/B....
The TL;DR is, a real world web application might be able to load and get updated in 5 seconds in development on native Linux (imagine a large Rails app with hundreds of assets), but the same exact app ran through WSL takes 32 seconds before you can see your changes. That's straight up unusable.
Personally, I just run VMWare in Unity mode. At the end of the day it lets me run graphical Linux apps with floating windows that feel and act as if they were running in Windows. The performance is excellent (it feels nearly identical to a native Linux experience with none of the WSL downsides). If anyone is interested, I put together a screencast and have a few screenshots showing how to do it at https://nickjanetakis.com/blog....
If price is any bit of a concern then I recommend installing GalliumOS (an xubuntu based Linux distro made for Chromebooks) onto a Toshiba CB35 2015 edition. It has a 1080p display, weighs less than 3 pounds and is plenty fast for development work. The best part about it is it only costs $350 with a 128GB SSD modification. If you're interested I have a full write up and review on it here: https://nickjanetakis.com/blog...