Andrew Leung

We're halfway through the year, and all browsers have improved their interop score. In this post I take a look at some of the features that are now interoperable because of Interop 2024. https://lnkd.in/eQmV3iZA

36

I'm seeing a lot of these "expert answers" about frontend development and what does HTML, CSS, and Javascript do. So many of them over complicate it: HTML is content authoring. You are writing a document and should use the tools and formatting available to make it as clearly understood as possible. Most people should be able to write some HTML much like most people can post dross on social networks. However there is a big difference between a coherent, semantically correct HTML doc and some div/span catastrophe. The latter is more like the local racists latest rant with no punctuation, swimming with spelling mistakes, and they've tied themselves in knots over the fact there was once a rainbow flag in Morrisons. CSS is not just about making things "look nice". Yeah it'll make things look better but it's main purpose should be to improve usability, legibility and make your content more understandable. Finally Javascript - Javascript is usually unnecessary and I reckon 90% of the internet could do without it. A lot of the things that Javascript has introduced has made the internet worse, less stable, and less accessible. It's main purpose has been to make frontend development more like a programming discipline, rather than the design discipline it originally was. This is coming from someone who has built their career around frontend development 🤷‍♂️

17

2 Comments

Having a ton of fun writing an ORM for V2 of Case Kit. Typescript has such an incredible type system. I have autocompletion for column names, the ability to include relations, relations of relations, etc etc etc... all returning properly typed data (no classes in sight). It's been really satisfying seeing it coming together, and it's been great getting my hands dirty with some (relatively) complex generics. And to head off the obvious question… why aren’t I just using Prisma? Well there’s a couple of things I’m doing differently that I think are marginally better: there’s no code generation involved here, it’s just types (inspired by Drizzle); there’s no Rust binary you need to deploy, and it hews much closer to Postgres behaviour and syntax (the aim here is not to abstract away which database you’re using). But the real blocker to me using Prisma was that its extensions don’t support modifying queries of nested relations - a hard requirement for me due to the way I have multitenancy and soft deletes set up. It’s not shown in the video but the Case Kit ORM lets you write simple middleware that will be applied to every level of a nested query. Perfect for setting timestamps or adding clauses on tenant id or things like that.

21

Notably not on this list: Swarm, the tool many game devs are forced into because our industry tends to rely on Perforce more than most. Swarm really is a poor tool, but its better than nothing. Crucible would be my choice, but though it has some level of P4 support it's not trivial to get and keep working and doesn't fully support all P4 features. Review Board is also good, but requires you to have some infrastructure and dev ops set up, its not as plug and play as others. Or just bite the bullet and switch to Git/GitHub, and put in some art asset friendly workflows and processes. That's the direction I'm leaning these days.

4

LW-Bits is now live and @ 1.0.0 https://lnkd.in/g4RhmKXw "What is LW-Bits?" you might ask. Great question! It is a PHP Package that utilizes the Livewire functionality within Laravel to provide all of the HTML5 elements and their attributes as Livewire components.

2

I have people ask me what is the main difference between ChatGPT Enterprise and Copilot. Today it’s which model and how many characters you can use, but 2GB upload and GPT Turbo is decent, but you may need vocal and 4o and that’s an important part of the story today. ChatGPT Enterprise - 4o the latest model, GPTs. There is a 3+ month delay for any partners using their latest model it seems M365 Copilot - model GPT Turbo (that’s the latest they can use as of today) and all the enterprise integrated tooling with Outlook, Teams, PowerPoint, WhiteBoard, Excel, Forms, Copilot Studio, Power Platform, Viva, etc… also supports additional integration to graph, connectors, with a ton of options etc…

17

1 Comment

UPDATE: I fixed the link! I recently created a new model for self-hosted Multisite, using GitHub Actions and DDEV for live and test environments. Check out that matrix!! https://lnkd.in/e3tJU59M #drupal #hosting #cicd #qualitycontrol #continuousdelivery #ddev

21

7 Comments

The most common complaint I hear about serverless is that it "is difficult to test". And this is where I most strongly disagree. In fact, I will try to prove that serverless actually is much easier to test and debug than the more common solution of using a container or a local server. Why do I say this? Simple: a serverless application is built out of isolated functions. Each function is a component, that receives an input and an output, and the services it uses are injected as arguments.   If anyone remembers the basics of the Spring Framework (which is an Inversion of Control, a testing framework at its core), the concept is the same, just at the infrastructure level. With this structure it is extremely easy to write tests: all you need to do is to pass arguments, and, if you are using external services, provide mocks for them. Serverless also makes it easy to write integration tests because the deployment is fast and can be integrated into the test code itself. However, in my experience. people who are used to building with containers are upset by this experience and complain that they "can't debug it". They expect to be able to start the server (which they cannot do with functions) and tweak with the application by interacting with it, often manually playing with the UI. The truth is that they are not used to writing tests at all . They don't do that, period. So they start the whole thing, interact with it often at the UI level and attach a debugger to find out what is wrong.  Of course this is a bad practice because you do not get a repeatable test suite, and what is worse, you have to test the whole thing instead of unit tests. Unit testing an entire server is much more harder than testing functions. So the truth is: is not that serverless is hard to test, it is that you usually do not write tests for your code at all.

22

3 Comments

"I asked 100 devs why they aren’t shipping faster. Here’s what I learned" > QA processes represent the tradeoff between quality and velocity. Developers will often accuse QA of being to “nitty” and adding useless repetition to the dev cycle. Taylor, who has worked at a series of high-growth startups and now runs one, described the QA process like this: >Me creating a test spec for QA. QA finding problems (because QA will always find problems) >Getting list of problems 2 days later >Fixing merge conflicts because more code has shipped since I last pushed. Plus context switching >Back to QA* In the amount of time it takes QA to review code, underlying context might change, making the change more involved. https://lnkd.in/gZ3-TW3a

4

I find interesting the hype about counting downloads for a given tool as a metric for success. I guess it helps to show how popular something is, but I would love it if people also considered other aspects, such as project governance and project activity. Anyway, let's follow the trend (evidence in the comments): - Selenium Java downloads: 98.5M in May 2024 - Selenium active users in the last 30 days: 2.6M. Happy automation!

54

9 Comments

Time to be fair.... In past posts, I have been less than complimentary of the DLC regulations that now make it more difficult to implement SMS into an application. I understand that all sorts of organizations use (and misuse) SMS millions (billions?) of times daily. I get that the new regulations make an attempt to reduce, as much as possible, SMS Spam. All in all, it just seemed like another example of hand-cuffing the good guys while trying to put some limits on the bad actors. My experience at the time was with #Twilio... and I tried to make it very clear that my frustration wasn't necessarily with them - they were doing the best they could to handle an avalanche of organizations trying to comply with new rules that everyone was stumbling their way through. I've been using Twilio for years and can't complain AT ALL about the platform performance for my particular small-scale needs. It was the nearly impossible approval process that really soured me. After nearly a year of living without application-based SMS (email only), I decided to revisit. The application was already all wired up for the feature - it was just dormant. This time around I was able to get up and running in 3 days... 99% of which was the DLC approval - which happened today. Kudos #Twilio for suffering through the tough days and making SMS once again something that can be done with very little pain. #twilio #sms #dlcapproval

2

This is what I always told to my teams: whenever your release process requires manual intervention (such as taking hours to verify changes or deploying during off-peak hours), it indicates that your release process is fundamentally broken. This is a symptom of a low-performance team or even a company. How to escape from this trap? - Reduce number of manual test or verification by leveraging automation (living the values, what i mean here just do not having a words on the wall for automation, but you need to living the values) - Shift left. Try to leverage the shift left as much as you can. From shift left on security and shift left for health checks - Avoid code yellow manager. Sometimes the blockers is not a technical issues the blockers are coming from a manager that did not understand the points of improving the lead time for changes and deployment frequency. I called this as code yellow manager, a manager that not close to the ground, essentially they haven’t a context and hard for them to take a good decision and vision due to lack of context and understanding. - Get buy in. If this initiatives require a sponsorship from exec, then you need to synthesize the problem and pitch them.

41

6 Comments

Custom Rules for Semgrep Secrets: Solid as Hell https://lnkd.in/dxjQx-Yv

30

Great article by Ray Carnes. In my 20+ years in the software industry I've seen companies learn and relearn the basics again and again because each generation of leaders think things are "different this time" because of such and such technology. The frameworks, platforms, underlying technology changes but the principles and foundations are still the same. #softwareengineering

20

3 Comments

I started writing regularly about AI. If you're an innovators or product developer eager to apply the latest in AI, whether you're in a startup or a big company, consider checking https://lnkd.in/gGdW8PYe I won't teach you ML, but I share real-world insights from launching - and sometimes failing at - numerous ML features on a large scale.

55

2 Comments

New blog post: The first 10k games at bgammon.org, an open source online backgammon service https://lnkd.in/gmc_c2ym #backgammon #foss #freesoftware #opensource #software #gaming #blog #free

2

1 Comment

I had the chance to play with Claude 3.5 Sonnet this weekend and integrate it into my novel drafting scripts. Some notes: * The output is good. Better than Gemini Pro for my use case. I had to do only very minor edits. That saved a ton of time. See also attached sample chapter, courtesy of Claude. * Thanks to the overall quality and the big context window, I had to send fewer queries overall to the LLM to generate the text. I estimate that I can create a good-quality draft for less than 10 cents / 1000 words. Meaning, I could create the first draft of a novel (assuming 100k words) for under 10$. Very exciting!

8

2 Comments