|
smackfu posted:To change the subject, anyone have any advice on getting better at grooming stories? We have grooming sessions and we go through all the stories and everyone nods their heads and says they are good and then they start their story and are instantly blocked by some question or missing bit of info. Seize the means of story production from the manager class and return it to developers.
|
# ? Sep 27, 2019 12:14 |
|
|
# ? May 10, 2024 20:11 |
|
So I wanted to put some of our internal team tools projects on a git server, any thread approved recommendations?
|
# ? Sep 27, 2019 15:10 |
|
Gitlab is pretty easy to set up on prem.
|
# ? Sep 27, 2019 15:13 |
|
For a small team that does not require a full blown solution like gitlab, I found that you cannot beat gitolite in terms of simplicity and low server requirements. Gitolite does one thing and one thing only: provide you with a remote git repo. If you need issue tracking, reports, stuff like that then gitolite is not for you.
|
# ? Sep 27, 2019 17:07 |
|
Monkey Fury posted:Seize the means of story production from the manager class and return it to developers.
|
# ? Sep 27, 2019 18:46 |
|
Woof Blitzer posted:So I wanted to put some of our internal team tools projects on a git server, any thread approved recommendations? Gogs is nice.
|
# ? Sep 27, 2019 20:25 |
|
sinky posted:Gogs is nice. May want to switch over to Gitea. There was a big fight within Gogs a couple years ago and everyone except the project founder left and started Gitea as a fork. Since then Gogs has effectively gone stagnant, with only the founder still making any significant contributions, and those have been tapering off. The contribution charts give a good illustration of what happened: Gogs, Gitea
|
# ? Sep 28, 2019 00:31 |
|
necrobobsledder posted:Sadly, that’s how we got Atlassian products. Management at my firm have begun to worship at the alter of Jira as we are completing our migration to it (from TFS) at the end of the month. Nearly all questions about process problems end up with an answer of "because we will be using Jira" as a final solution - along with a fetishistic obsession with the idea that *everything* needs a Confluence page. Based on prior experience with both, I'm not sure that they are the panacea.
|
# ? Sep 28, 2019 07:53 |
|
Progressive JPEG posted:May want to switch over to Gitea. Thanks, gogs seems to be way behind.
|
# ? Sep 28, 2019 11:13 |
|
Does anyone have dead simple flow charting software (web based ideally) they like to use? I just want to draw out our new complicated onboarding flow and share it with my team
|
# ? Sep 28, 2019 14:11 |
|
Eeh, there's always draw.io?
|
# ? Sep 28, 2019 14:25 |
|
prom candy posted:Does anyone have dead simple flow charting software (web based ideally) they like to use? I just want to draw out our new complicated onboarding flow and share it with my team We've been using the Mermaid integration in Gitlab's wikis (it also has a VSCode extension) to make a bunch of intra-team diagrams and it seems to be quick and easy. For anything more advanced, yEd?
|
# ? Sep 28, 2019 14:30 |
|
prom candy posted:Does anyone have dead simple flow charting software (web based ideally) they like to use? I just want to draw out our new complicated onboarding flow and share it with my team I've become a fan of PlantUML. Layout is mostly out of your control, so you can focus on getting the work done instead of making sure every line is straight and cursing when you need to find room for a new object.
|
# ? Sep 28, 2019 18:19 |
|
Turambar posted:I've become a fan of PlantUML. Layout is mostly out of your control, so you can focus on getting the work done instead of making sure every line is straight and cursing when you need to find room for a new object. This looks awesome thanks!
|
# ? Sep 28, 2019 18:39 |
|
Anyone experienced with codility tests? By coincidence, two of the companies that I applied to have send me a link to use this platform. I don't care about the time it costs me, I just want to know if anyone done it before and can tell me how it works. Do you think this gives a reasonable display of your skills?
|
# ? Sep 28, 2019 23:19 |
|
Keetron posted:Anyone experienced with codility tests? By coincidence, two of the companies that I applied to have send me a link to use this platform. I don't care about the time it costs me, I just want to know if anyone done it before and can tell me how it works. Do you think this gives a reasonable display of your skills? Some of the teams at work use it. The results seem okay, the questions are bigger in scope than normal white boarding nonsense, but still the same sort of algorithmic trivia.
|
# ? Sep 29, 2019 00:32 |
|
A bad question is a bad question regardless of the tools used to ask it.
|
# ? Sep 29, 2019 01:27 |
|
Awesome, it means I will suck at it. Time to look at some other people's solutions to see what to expect. Call it preparation for non job related code exercises because I don't mind a coding challenge as long as it allows me to showcase how to do my job.
|
# ? Sep 29, 2019 07:27 |
|
At my job it's being used to filter out people who can't even do basic coding skills (while still being senior according to their CV). After the codility test you get a larger test where you actually need to build a small application completely according to your own ideas. That takes a fair bit of time, which sucks. However, it does mean we only need to review the applications of people who have already demonstrated they can perform easy coding tasks.
|
# ? Sep 29, 2019 09:09 |
|
Yeah, I am just worried as I am unfamiliar with it. Looking up some other people's solutions will sooth my worries. Main thing for me is that I like to work test driven and the fact the tests are hidden rubs me the wrong way.
|
# ? Sep 29, 2019 14:13 |
|
Sagacity posted:That takes a fair bit of time, which sucks. However, it does mean we only need to review the applications of people who have already demonstrated they can perform easy coding tasks. After some pushing (from me and some of our devs) we started paying people for their time taking coding tests (though we do this after initial review of resumes so that we aren't paying people that we have no chance of hiring), but I wonder if that is becoming any more common? Having to invest a bunch of unpaid hours to apply for a job skews your applicant pool in unfortunate ways.
|
# ? Sep 29, 2019 14:56 |
|
https://twitter.com/dhh/status/1177974040432037888?s=21 Our third developer is starting in 3 weeks and I can't decide if now is the time to formalize a bunch of processes or if it'll just add needless overhead. We're disciplined in some ways (good story/ticket tracking, good test coverage on the back-end) and undisciplined in others (no code review/PR process, no CI, poor test coverage on front-end apps). We are wildly productive (even considering my earlier post about procrastinating) but it does feel like Wild West poo poo sometimes.
|
# ? Sep 29, 2019 15:33 |
|
prom candy posted:no code review/PR process, no CI, poor test coverage on front-end apps You should be doing this anyway (well, as long as you have peers able to code review you should be doing that too). It's a short term drag that saves time and money in the long run as long as your code isn't going to just get thrown away, since the bugs you find and fix while developing the thing are a lot less expensive than the ones you find in production.
|
# ? Sep 29, 2019 15:54 |
|
CI and test coverage aren’t “large company” practices. Definitely set them up ASAP.
|
# ? Sep 29, 2019 16:42 |
|
Yeah I want to use this milestone to fix some of our poo poo for sure. It feels like we're just always trying to move as fast as possible but of course that's not sustainable and even less so once we add team members. What I really want is full e2e tests of every feature but it feels so time consuming to set up and maintain.
|
# ? Sep 29, 2019 17:04 |
|
Your absolute first priority is introducing a culture of code review. It doesn't matter how fast you ship to production, if the code you're shipping is unreviewed you will constantly find yourself fighting fires and fixing bugs.
|
# ? Sep 29, 2019 17:06 |
|
prom candy posted:What I really want is full e2e tests of every feature but it feels so time consuming to set up and maintain. Doing that in the short term is probably counterproductive, but you can definitely come up with a plan to improve test coverage. Figure out what unit and integration tests you want and start doing those. Save the full e2e tests for when you have more people and more bandwidth.
|
# ? Sep 29, 2019 17:28 |
|
prom candy posted:What I really want is full e2e tests of every feature but it feels so time consuming to set up and maintain. Basically, implement a test automation pyramid now, else you end up with an icecream cone and that is a Bad Thing. https://martinfowler.com/articles/practical-test-pyramid.html
|
# ? Sep 29, 2019 17:38 |
|
prom candy posted:Yeah I want to use this milestone to fix some of our poo poo for sure. It feels like we're just always trying to move as fast as possible but of course that's not sustainable and even less so once we add team members. Code reviews should be your #1 priority, CI #2. Add e2e tests when you're done those, just bear in mind that is you're going to be frequently ripping things up and rewriting them the time spent maintaining the test infrastructure may not yet be with it. I don't know how much breaking things is involved in your company's moving fast.
|
# ? Sep 29, 2019 17:39 |
|
Eh, I’m of the opinion team consensus on a practice is much more important than any one practice. Doesn’t matter if one person tests with TDD, another guy does some yolo coding, and another person screams about code review. All practices degrade poorly with less / poor practice of it. The best teams I’ve worked with just argued less overall and the codebases tended to be easier to deal with IMO.
|
# ? Sep 29, 2019 17:55 |
|
Keetron posted:You don't want a full e2e suite, you want to test stuff at the right level and every feature testing should not be at the e2e level. Yes, this. Large suites of E2E tests do not provide value, full stop, end of story. At best they provide no value because they're just never run/not run frequently/run frequently but completely ignored. At worst they provide immense negative value because they require nonstop hand-holding to keep running reliably and provide nothing but false positive and false negative signals.
|
# ? Sep 29, 2019 18:04 |
|
Code review is important for more than just fixing bugs. With more people on a team, you need code review to enforce standards, keep track of what others are doing, and most importantly for everyone to know what's going into the shared codebase. Having your new person do code review of the rest of the team is also a good way to get them up to speed on what's going on with your project.
|
# ? Sep 29, 2019 22:02 |
|
For the devops / SRE folks, according to the Terragrunt / Gruntworks folks with hundreds of thousands of lines of Terraform large E2E test suites are probably more important than unit tests due to the massive underlying interactive components of side effects bonanza under the hood of cloud services. For example, regressions are quite common and service limits are oftentimes discovered by users rather than specified in practice. We do a mix of 40/60 unit / integration testing and it works well for us although I think we’re really crummy at input validations because our test suite driver is basically Jenkins up and down while we’re nauseous while writing our 10k+ lines of Jenkins bastardized Groovy
|
# ? Sep 29, 2019 22:47 |
|
Do yoy guys work more as generalists or on a single role (front end or backend or devops)? In my current role I have been doing BDD fullstack mainly Java Spring and React. We have been writing our own automation code and doing our own manual testing, as well as supporting a ~19 year old legacy register software application on any bugs. We monitor our current builds in prod and do our own CI. Does anyone else work like this? I have mostly only done React in my career and I feel really overwhelmed with all the moving parts at my current job.
|
# ? Sep 30, 2019 02:18 |
|
I used to work like that, I’ve found it’s prevalent at small companies. I work at a medium company as a dedicated BE engineer and it’s so much nicer.
|
# ? Sep 30, 2019 06:03 |
|
Where's the tipping point at which point I can truly pretend to be a developer? I'm a security engineer who automates everything, builds custom apps, and deploys them to kubernetes. But I'm still fielding tickets for security work too. Is this what modern engineering is or should I be hoping to at some point pivot over to pure dev? Is it number of languages known? Number of deployed apps? Popularity of apps where others stop expecting you to do other work? I feel like I would get laughed out of a programmer job interview but have no idea what else it would take for that to not happen.
|
# ? Sep 30, 2019 06:14 |
|
Judge Schnoopy posted:Where's the tipping point at which point I can truly pretend to be a developer? Congrats, you are 110% devops.
|
# ? Sep 30, 2019 06:34 |
|
Judge Schnoopy posted:Where's the tipping point at which point I can truly pretend to be a developer? The point where you learn one language that you can develop some application in and you feel you can bluff your way into a development role. I think you are there and all you need is marketing. Basically you are an engineering managers wet dream due to the large skillset, writing application code is only one part of being a developer and you have such strong foundation in many of the other parts that you'll get hired fast. Consider this: while they were out writing yet another CRUD app, you were honing your automation skills. And hell, writing a CRUD app is so easy I am confident you can do that already or at least learn how.
|
# ? Sep 30, 2019 07:00 |
|
Judge Schnoopy posted:
What’s a security engineer?
|
# ? Sep 30, 2019 13:04 |
|
|
# ? May 10, 2024 20:11 |
|
Our firewalls are split between two teams. Devices are owned by network, ACLs / security config is owned by security. We have a fairly in depth microsegmentation deployment as well. My role as security engineer is maintaining these configs, auditing them, and providing access for new apps as appropriate. I have 5 cronjob-based apps to complete routine tasks / config valuation / health checks. I'm building a new platform that will link the ACLs to the microsegmentation system to reduce ticket load. My company refuses to acknowledge the success of these projects and instead throws me to the dogs when one change caused a 2 second downtime. I'm considering moving on up to a dedicated devops / developer job but am hitting some hard imposter syndrome about it.
|
# ? Sep 30, 2019 13:28 |