• 0 Posts
  • 390 Comments
Joined 1 year ago
cake
Cake day: September 27th, 2023

help-circle



  • By SPA I mean “single page application” it’s currently the dominant approach and powered in a large part by technologies like react and node. I’m not certain when it started precisely… with technology it’s more a case of “rising to prominence” rather than “first happened” I think it probably really started going around 2014 with HTML5?

    SPAs are still pretty hot but they’ve waned in popularity due to overuse and general complexity. Essentially your website becomes a single page that just swaps out what’s shown to the user as they “navigate” between different parts of the site. When well done this can make a site incredibly responsive, but it’s often quite poorly done and responsiveness can end up blocked by server requests anyways.




  • Nah, a lot of old tech. I used to work on shit like this… loading all your images (including the fucking rounded corners for IE) into a sprite… setting up caching, using prefetching and inlined CSS/JS for critical path stuff.

    There was a whole industry around web performance in the days that a customer might be trying to download your site over their 256 kbps connection.

    It’s neat tech and I miss fiddling with it. I honestly found it a lot more fulfilling than the SPA era of web design.


  • Honestly? Pretty fucking awesome if you get it configured correctly. I don’t think it’s super useful for production (I prefer chef/vagrant) but for dev boxes it’s incredible at producing consistent environments even on different OSes and architectures.

    Anything that makes it less painful for a dev to destroy and rebuild an environment that’s corrupt or even just a bit spooky pays for itself almost immediately.



  • Ah see, at my current company I’m employee 2/120 so I have adamantly advocated for this.

    It’s definitely a hard fucking sell though, nobody outside of the developers wants to invest in maintainability, bug fixing or infrastructure upgrades - you just need to force that shit through with clout. One thing I’ve found that helps is to try and form a technology steering committee that can try to advocate for the necessary investments. Approaching a problem as a group or talking to your manager about setting aside dedicated time to figure out which issues are most pressing can be quite effective. There is usually a trust barrier to overcome to allocate that time though.


  • You know what attribute of a code base is more valuable than any other? Maintainability.

    Every project ever should emphasize maintainability over performance, cleverness, etc… because this is the true long term cost of code. If it’s difficult to understand why or how something is how it is then you’ll pay a lot more to bugfix it and improve it over time.

    This attribute is opposed (mostly) to flexible prototyping so a really good senior dev will be able to transition a greenfield project into one that’s structured for long term usability.



  • Eh. Sublime and vagrant run on windows and the machines are better value than Macs.

    I’ll stick to a windows host with Linux rather than feed Apple ridiculous money for dongles that do shit that should be built in. Multiple display port out and a built in ethernet cable or death - I actually need a laptop that’s portable.

    I just checked and it looks like the latest MacBook Pro has a single hdmi port and three USB-C ports… so I’ve got my power cable, my mouse, my keyboard, and my ethernet cable dongle… already at negative one ports. Then I’ve got two monitors on display port to somehow cram into a single hdmi port - and apparently the processor only supports a single external monitor unless you get the MacBook Pro Pro or MacBook Pro Max… that’s impressively shit.

    All for 4,649 CAD - I can buy so many more ports on PC with a ridiculous amount of power for 4.5 thousand. I don’t mind spending my employer’s money, but I want to spend it on shit that’s useful for me.





  • If you’re talking about a service like copilot and your employer won’t buy a license for money reasons - run far and run fast.

    My partner used to be a phone tech at a call center and when those folks refused to buy anything but cheap chairs (for the people sitting all day) it was a pretty clear sign that their employer didn’t know shit about efficiency.

    The amount you as an employee cost your employer in payroll absolutely dwarfs any little productivity tool you could possibly want.

    That all said - for ethical reasons - fuck chatbot AIs (ML for doing shit we did pre chatgpt is cool though).



  • As someone who writes high throughput PHP code I can confirm that it’s much more about technique than language capabilities (though in an embedded setting things with dynamic GCs are simply unusable unless static memory management can be enabled with a compiler switch).

    For most projects you’d be much more rewarded for focusing on tools/framework/libraries available for the different languages (since that’s where most initial effort will go) and then build up any missing functionality as needed ontop of that base.

    Most languages can do pretty much anything these days. The technical advantages are much smaller than the impact the right approach will have… it’s one reason that I hold “maintainability” as the most important attribute of a project.