25 karma · joined March 5, 2012
About tooling, we have tried some mermaid, plantuml, or other diagram generated from source tool but really nothing that is satisfying.
I'm interested as well in this subject
I agree that OpenAI and the associated generator are not good enough to a real day to day code generation.
License in the footer as well (perhaps more difficult to change if the license is copyleft)
As any open source (that want contributors and/or forks), a full readme, a quick start, a contribution guide, a documentation, etc.
Obviously, the PO is accountable for the user stories, but they are not to write them all by themselves.
Let the PO decide what is the next big feature, then pair them with a dev to split into manageable stories. Such stories will be then smaller and still have to be split in task. But at least, the user story will be small enough to be develop as is.
PO will have the last word about the big picture and the details (user facing) but the means are in dev hands.
Current work sent tech employees a blank computer and any licenses we ask. So basically, you'll have sudo if you install it.
On the other hand, we are often reminded to upgrade our stack / to check if we have been compromised, etc.
When react executes your component function, it will initialize a context (a big array) and each hook will be linked to an index in this context.
So, while being nearly pure, we can describe side effect but only for the provided context.
In another hand, if you are still maintaining this project, it can be interesting to refactor it and show your process doing it. It can be a great way to show actual useful skills for a future interview or, at least, train to do so before actually using it in a real production project.
Refactoring to a state of the art state while the project is still under development and shipped can be a valuable skill to train and show.
I understand that it is a MVP, but I cannot put job real data without warranty.
If you really need a portfolio, I suggest:
* A classical project (CRUD for website, small cli, etc) * A not so classic project (choice is up to you) with documentation about the architecture, a clear API (rest for website, man page for cli, interface for library, and so on...)
Each project should be well tests covered, should have good automation (github actions are free for public repositories for instance) and easily run locally.
Bonus point if you can use language/libraries/framework from your future employer in a way that is correct but different from their usage. (Hard to know before)
What I understand: a big project divided in 4 parts. Each part is owned by a team of 8-12. A teach lead by team seems fine. Seconded by senior that help mentoring.
How do you really define a tech lead? How did your tech leads become such? Do they really need to carry their whole "part" or only be an technical interface between their team and the others? The real questions are not a bare number, a simple heads count. But a whole definition of a "title" and what this title really mean in your environment.
Experience may vary from context.
In a former company, I had the "tech lead" title. I had the technical responsibility of a team of 8 members (including me). I took the final decision about architecture, implementation, API and other workflows to the outside of the team. We carried 40+ projects (real number is around 60). It was exhausting!
In my current company, the "tech lead" is a non-sense, we are autonomous and we trust each other. We need the CTO to define clear priority list he wants to see happen. Then, usually, someone is the "feature lead", the person who decide the architecture, the order of the different tasks to implement, who manages the attribution of the mentioned tasks. Currently, this feature lead is the more "junior" of the team. Because he took time to understand the feature, he benchmark and test the different solutions. Obviously, as other team members, I challenge his decision about details. But the final decision remains his. And I blindly follow his final word. (Even on those I personally disagree). It's not a matter of ego (I'm more experienced) or of knowledge (I teach him a lot). But really, about the feature and the time invested to understand the feature, the needs and how to integrate it in our code base.
And today, with your question, I realized that I was pretentious. I thought because I follow each pull requests, I discuss with him about the architecture and the feature, I will be able to lead the feature from a day to another. But the reality is: I was miserable. I failed to recognize simple errors. I failed to know where to look when something bad happen (during test phase). All of that, because I was not the feature lead this time and I didn't take as much time as him to really understand the feature and its components.
With your question, I realize now that he really did a great job, and I am happy to work under his lead (during this feature at least).
I repeat myself: The "tech lead" is just a title, what matters is how you define it and how this tech lead endorse the role.
Youtube-music: love vanilla YouTube music, but automatic pauses break my flow and ads are very aggressive. This electron port allow me to have an AdBlock and remove pauses. Bonus, I can use my media keys directly from my keyboard.
Links: * https://github.com/LandrixSoftware/PDFTK-Builder * https://github.com/th-ch/youtube-music
Chores, diner, check children's homeworks, sleep (I'm dreaming of a 8 hours night)
Dependency and priority are both levers to be able to choose the right task to do (perhaps adding an importance level?)
A task that depends on unfinished ones cannot be proposed by the system. Between two possible tasks, the priority one will be presented first (or perhaps a computation with priority, importance and duration?)
Edit: bad autocorrect
The complexity is to synchronize on all platforms, to have a nice and easy GUI, and manage the dependency (priority?)
Good luck with that and stay in touch if you come with something.
I remember feeling very stressed about all the different papers to deliver during that time. But it was a feeling, not a fact.
This post is only an ads.
Free open source software. Available on Google play and F-droid.
With features I like : icons, import/export, password to open, etc.
It is a shame that we don't have an organisation, a foundation about this subject.
A place, a website to tell such a risk for user privacy.
Perhaps the RGPD from EU can help?
My prime objective is to be able to go on vacation without any call of emergency, but the point is I showed that I was not necessary when I was ill and not at work for a month.
Two weeks after my return, the up management shows me the door...
Nevertheless, I think it is my best move to make my own position useless. I need peace of mind in vacation, I need autonomous coworkers, I need challenging problems to solve, and more, my whole team is happier that way.
The up management did not understand that... too bad for them.
What is important exactly in a code review depends on the team. But generally, the same items are reviewed:
* Are team conventions applied?
* Can another understand the new code?
* Is code resolving the issue?
* Can I think about a better code? Something to refactor?
Even two juniors can ask this questions and try to answer them.