After working in technical support for a few years I've developed an art for troubleshooting complex problems while managing the customer throughout the process.
I've decided to document my strategies and customer management process in my article.
patrique.gumroad.com/l/sslji
@GuillaumeCleme@tobias_petry Some optimizations perhaps but let's talk about the thing no one talks about. Ease of troubleshooting. In my opinion this is a nightmare to isolate the system layers which adds to the complexity of troubleshooting and issue resolution.
@tobias_petry Technically I see some potential in terms of transport optimization, prefetching, logical coupling, etc. but it also blurs the lines between MVC or MMVC patterns and looks like it could get out of hand pretty quickly.
PHP did the markup and server code mixing 20 years ago and collectively decided its a bad idea as it becomes unmaintanable pretty fast.
Now that gets further complicated by also mixing with frontend-stack complexity...
Can somebody explain to me why that is a good idea now?
Taking some time off.
Decided to come out to MIT.
Never been here but I like sitting in on random classes to get some new inspiration for whatever I'm working on -- you can just walk in lol.
Just got out of differential equations.
Architecture or nano technology next lfg.
@therealweissman I love meat loaf but the one I make at home isn't always that great. Can you make a video around making good simple meat loaf that we can enjoy at home? I would really appreciate that.
@GuillaumeCleme@thdxr In my opinion any enterprise solution will have some centralized logging. It's very rare that you can simply query that log and look for information about what's happening in the system without performing some sort of transformation on the log data.
@thdxr What I dislike the most about JSON logs is that in theory they work, but they then need transformations, indexing, and query tools to serve any real purpose. Fragmentation is also a pain.
am i the only one who doesn't like JSON logs?
i prefer if they're human readable with simple `key=value` tags if i need to include structured data
i also don't like logging libraries because every project uses a different one
tell me why i'm wrong
@marcba Hopefully there's some foreign key constraint that protects you unless you add CASCADE to the command then you're fucked. Maybe start applying elsewhere
I had a customer today mention how one of the features of the product is not doing what they thought it would do.
It's doing what is documented but it's not doing what it's not documented.
Has any other #developer run into this?
@BishoysBytes The issue also occurs if activate or modified fails what do you then?
This happens in AEM when the sling repository service fails (it's the one that performs sling repoinit scripts) then all services above it cannot start and then your left with a running jvm that's unusable
@BishoysBytes Now if Service A has data that it stores as cache that comes from Service B this needs to be handled or else you need risk returning stale data.
Typically you would listen to OSGi life cycle hooks like activate/modified or deactivate to hahdle these scenarios.
The more I work with #osgi the more I realize that the concept of hot reloading applications doesn't work when dealing with hundreds of interlinked services.
Developers need to consider how their services handle being dynamically reloaded. This is complexity is challenging.
@GuillaumeCleme@gitpush_gitpaid Once you pick it, try to stay with it for other projects. This allows your team members to be interchangeable.
Having to learn a new language and the software increases the time it takes for engineers to ramp up and be able to deliver on requirements.
@gitpush_gitpaid From my experience, don't worry about picking the newest and flashiest stack, pick what works well and allows you to test assumptions quickly.
Also something that's often overlooked, pick a stack that will make it easy to hire and onboard new team members!
@GuillaumeCleme A lot of times I hear well you can just use the search function to find the documents you need.
My immediate thought is how am I supposed to craft a search query for a document that I don't know exists and nor what keywords it contains for me to search for.
At the exact moment when you think "I need to document this", the next best question is where and how will it be best consumed.
We live in a world where content lives in 100s and 1000s of different locations. Content is only as useful as the ways we have to find it.
@GuillaumeCleme@SamtLeslie What ever happened to YOLO-ing it? If you forget, well, enjoy your forced day off (not that I've forgotten my laptop on purpose before)
@SamtLeslie No way. That's just setting yourself up for "did I put it in the car already or am I forgetting something?". I use a "staging area". Everything that goes in/out of the house is staged there in plain sight 😅.
@PabloVegaBehar@rabernat@itamblyn You sir have hit the nail right on the head. When I was in school you could literally google the question on your assignment verbatim and find the answer. That's how old it was.
@GuillaumeCleme@DThompsonDev The Java enterprise phase definitely got a lot of enterprises hooked. That being said I'm wondering what Rust will look like in the next decade.
@DThompsonDev Good call out for Java and Spring Boot. A large percentage of enterprise technology is built using Java. If you're looking for a job, those skills will certainly help.
If I had to start learning programming today I would learn:
- HTML
- CSS
- Javascript
- React
- SQL
- Java
- Spring boot
Why?
- I've done a lot of research into my area & It is in demand
- Enterprise level companies use it
- If the goal is employment, this would work for me.