|
Master_Odin posted:Yeah! It's especially awesome how issues and pull requests and ability to have conversations and stuff are built right into git so we never need a centralized service! uhhh... code:
|
#
?
Jun 4, 2018 15:30
|
|
|
|
| # ? Jun 6, 2024 21:00 |
|
|
leper khan posted:uhhh...
|
|
#
?
Jun 4, 2018 16:27
|
|
|
BabyFur Denny posted:As if today's developers ever knew how to do git ... Seriously, boot camps and colleges need to add that and information security to their syllabi. Everywhere I've been so far, knowledge in those were pervasively lacking.
|
|
#
?
Jun 4, 2018 16:44
|
|
|
So regarding the, uh, bad culture fit at my current job, I've started looking for work in Vancouver whilenot living in that city. Next time I'm able to I'll attend meetup(s) in the city, but for now I'm just applying on Indeed and other job boards. There seems to be a lot more opportunity there than in my current city. I'm curious how I should frame my desire to relocate when I apply for jobs. I have been doing a LDR with a SO in the area for several years now, so relocating is a good idea. Should I state outrightthat I do not need compensation to relocate? What details are worth giving for relocation? And how do I manage interviewing in a new city while employed in my current city?
|
|
#
?
Jun 4, 2018 16:47
|
|
|
BurntCornMuffin posted:Seriously, boot camps and colleges need to add that and information security to their syllabi. Everywhere I've been so far, knowledge in those were pervasively lacking. I'm surprised it's not something people know. My bootcamp had us do sections on it, it was part of the knowledge we could get tested on, and my mentor would never ever let me do anything other than command line Git. bvj191jgl7bBsqF5m posted:So regarding the, uh, bad culture fit at my current job, I've started looking for work in Vancouver whilenot living in that city. Next time I'm able to I'll attend meetup(s) in the city, but for now I'm just applying on Indeed and other job boards. There seems to be a lot more opportunity there than in my current city. Speaking as someone that did this last year, part of my strategy was to imply I already had a set date I was moving there in my cover letter (big ol' lie), and to remove any details about my current location from my resume. If you can, can you use your SO's address as yours, just for the application process? Otherwise, you're going to see a dramatic drop in response rate. I am getting waaaaaay more ping backs from jobs now that I'm close to Chicago than when I lived in North Carolina. Still no response from the place I was waiting for an offer from  Sent an email on Thursday and nothing yet
|
|
#
?
Jun 4, 2018 16:59
|
|
|
leper khan posted:uhhh... Master_Odin fucked around with this message at 18:39 on Jun 4, 2018 |
|
#
?
Jun 4, 2018 18:36
|
|
|
Master_Odin posted:Which then outputs to stdout on my command line which I then take that output to send to a centralized place (like the maintainer's email address, a mailing list, GIthub, Gitlab, etc.) so that I can receive feedback from people on the PR and get it integrated into the project. Um its 2018 just lol if youre repo consistency and PR signoffs arent powered by the BlockChain.
|
|
#
?
Jun 4, 2018 18:40
|
|
|
Munkeymon posted:Does VSTS support Git repositories? We use it with TFS at work, but, well, TFS Yes, but build artifacts can't be put into Git repositories for some reason. Your options are: TFS or some UNC file path, which is ... not great.  And as far as I can tell, Releases can only be the product of published build artifacts, so either your build agents have to all have UNC file path access to some file share or you have to use TFS.
|
|
#
?
Jun 4, 2018 19:23
|
|
|
No Safe Word posted:Yes, but build artifacts can't be put into Git repositories for some reason. Your options are: TFS or some UNC file path, which is ... not great. Everything you've said is incorrect. You're talking about the build system, which has nothing to do with version control (and, in fact, the build system supports GitHub as a source already and has for some time). You can store artifacts internally, or on a UNC share with the "Publish Artifacts" task. Or you can use different tasks (or extensions from the marketplace) to push them as NuGet packages, or publish them to Artifactory, or publish them to basically anything. Or you could run some command line stuff to push the build outputs into a Git repo, if you wanted to do that. No Safe Word posted:And as far as I can tell, Releases can only be the product of published build artifacts, so either your build agents have to all have UNC file path access to some file share or you have to use TFS. Releases can use build artifacts. Or they can use Jenkins artifacts. Or any internally-hosted Git repo. Or any GitHub repo. Or any internally-hosted TFVC repo. Or a package management feed. Or a container from Docker Hub / Azure Container Registry.
|
|
#
?
Jun 4, 2018 19:53
|
|
|
No Safe Word posted:Yes, but build artifacts can't be put into Git repositories for some reason. Your options are: TFS or some UNC file path, which is ... not great. This is just wrong. Even if you are ignoring every other task or plugin available to you, that task still says "VSTS/TFS" and works fine for building for a release from git repositories within VSTS. e: Basically the only thing this task cares about is if your build pipeline is in VSTS or a UNC. You can even hook up a github repository, build it, publish to VSTS, and use the artifact in a release. The Fool fucked around with this message at 20:20 on Jun 4, 2018 |
|
#
?
Jun 4, 2018 20:18
|
|
|
No Safe Word posted:Yes, but build artifacts can't be put into Git repositories for some reason. The reasons: 1. Git can't compare binaries (not nicely, anyway). 2. Binaries bloat the repository, making it slower. 3. Dependency management exists, so there is usually no reason to keep a library on the repo. 4. There is no need to store build artifacts because a proper repository has everything you need to build your own (and it'll be guaranteed to reflect the state of your source). Include a binary and I will reject your commit with maximum prejudice.
|
|
#
?
Jun 4, 2018 20:18
|
|
|
BurntCornMuffin posted:The reasons: I dunno what kind of developer would want to put a binary in a git repo anyway. Yeah, it's an instant decline on the PR.
|
|
#
?
Jun 4, 2018 20:23
|
|
|
BabyFur Denny posted:I dunno what kind of developer would want to put a binary in a git repo anyway. Yeah, it's an instant decline on the PR. git-lfs is a thing. It replaces the binary with a text file that points to the binary. It works pretty decently after setting it up.
|
|
#
?
Jun 4, 2018 20:24
|
|
|
ratbert90 posted:git-lfs is a thing. It replaces the binary with a text file that points to the binary. It works pretty decently after setting it up. That still doesn't make putting build outputs into the repo a good thing.
|
|
#
?
Jun 4, 2018 20:29
|
|
|
The idea of a build storing binary artifacts in git is horrifying, sure, but dogmatically never putting any non-text data in git isn't a very measured response to git being bad at binaries either. For instance, if your desktop application needs a 10kb PNG file for its icon then the correct thing to do is to just commit it. Nothing will explode: git will just store a complete copy of every version of the icon you commit, which for a rarely updated 10kb file is not going to cause you any noticeable slowdown or bloat. Sometimes natively versioned binaries are faster and easier than using LFS or some other external resolve.
|
|
#
?
Jun 4, 2018 20:38
|
|
|
Sometimes a binary is not a build output.
|
|
#
?
Jun 4, 2018 20:52
|
|
|
Bongo Bill posted:Sometimes a binary is not a build output. The context that the conversation started in was putting build outputs into a Git repo. Binaries that never/infrequently change are fine as long as you LFS them past a certain point. I hope everyone can agree that every build pushing the outputs of that build into the repo is a bad thing to do, even if you're using LFS.
|
|
#
?
Jun 4, 2018 21:25
|
|
|
Speaking of Git, my team (almost entirely my boss) let branches get entirely out of hand, with what we were creating branches from, what production was, how old master is, and how insanely many branches we have (I think around 30?) Now my boss is freaking out because a unit test is failing sporadically and is making us re-build all the work we did last week so we can get a clean sprint branch for master. We already knew some of the unit tests were borked and needed to be re-done, but lord. Hes acting like he didnt know any of this was happening either, he constantly says (when things go wrong) that he cant be that down in the details and doesnt know it if we dont bring it up edit: make that 50-60 branches Shirec fucked around with this message at 21:40 on Jun 4, 2018 |
|
#
?
Jun 4, 2018 21:35
|
|
|
Here let me write your unit tests for you.code:
|
|
#
?
Jun 4, 2018 21:50
|
|
|
Shirec posted:Speaking of Git, my team (almost entirely my boss) let branches get entirely out of hand, with what we were creating branches from, what production was, how old master is, and how insanely many branches we have (I think around 30?) code:
|
|
#
?
Jun 4, 2018 22:27
|
|
|
Xarn posted:
Yeah, but this is Shirec's boss we're talking about. He's probably not merged any of them, and picks which one goes to production based on a random function seeded by the time he takes his morning poo poo.
|
|
#
?
Jun 4, 2018 23:21
|
|
|
BurntCornMuffin posted:Yeah, but this is Shirec's boss we're talking about. He's probably not merged any of them, and picks which one goes to production based on a random function seeded by the time he takes his morning poo poo. It's more whatever is occupying his mind currently. He's delegated merging to master to our team lead, but he constantly overwhelms that guy and has him chasing a million other things, and there's no one allowed to perform that function. So it goes: Only team lead and boss can merge to master Boss leaves it all to team lead Boss assigns a ton of production work to team lead, confusing instructions on what to do with branches ... Time passes ... Boss freaks out because of all the branches and doesn't know what's going on Boss freaks out because an issue we've had forever (one unit test failing) happens in front of him ... We debug and figure out issue New bunch of rules to prevent problem boss says we have I'm also probably in trouble again because I was reading docs on our Node SQL library and was trying to help with why we were dropping milliseconds. Boss gets annoyed because I "don't know what I'm talking about" edit: Honestly the part that bums me out the most is that he interrupts me constantly and talks over me until I'm quiet 
Shirec fucked around with this message at 23:56 on Jun 4, 2018 |
|
#
?
Jun 4, 2018 23:53
|
|
|
Sometimes your boss seems more terribly insecure that they have no idea how to manage people or write software than they are exhibiting outright maliciousness. When dealing with insecure micromanagers, always tell then yes to their face and couch any of your ideas as slight tweaks on their brilliant ideas. Even if they say no to your face, they may at least think about it and come back with it as their idea. If there is a obvious problem with their ideas, wait for a while and present the problem nicely packaged up for them to "solve". Not that it applies here, but that's how to get in with that sort of person until you can stab them in the back. Sometimes it's worth it if you can set them up for a hard failure with their boss while your rear end is covered and you have a solution waiting in the wings.
|
|
#
?
Jun 5, 2018 00:23
|
|
|
Shirec posted:I'm also probably in trouble again because I was reading docs on our Node SQL library and was trying to help with why we were dropping milliseconds. Boss gets annoyed because I "don't know what I'm talking about" You're not HIPPA compliant and this is the biggest worry?
|
|
#
?
Jun 5, 2018 02:20
|
|
|
geeves posted:You're not HIPPA compliant and this is the biggest worry? Hahahahahahahhahahaha. Lord. Boss gets on us about our sense of priority but will spend 3 hours lecturing us because he doesn't think we are "writing assertions" the correct way. Now he wants to drop unit testing all together because one day of debugging some weird unit test stuff isn't worth that amount of time. He doesn't give a poo poo about HIPAA compliance. Any time he has to get us to do something he knows violates it, he'll say "I have no knowledge of how you are choosing to solve this problem, but get it done this way" He's incredibly bad at this. I have no idea how other shops work but I assume it's better than this. I'm incredibly pissy, if you can't tell, because I'm now on past 12 hours of working today.
|
|
#
?
Jun 5, 2018 02:24
|
|
|
I got in trouble for having stack overflow open in a separate tab today.
|
|
#
?
Jun 5, 2018 02:28
|
|
|
bvj191jgl7bBsqF5m posted:I got in trouble for having stack overflow open in a separate tab today. What reality do you live? is this the one in which Thanos was sucessful?
|
|
#
?
Jun 5, 2018 02:52
|
|
|
If someone came to me with a question without googling it first I'd be irritated.
|
|
#
?
Jun 5, 2018 03:19
|
|
|
geeves posted:What reality do you live? is this the one in which Thanos was sucessful? One where people say "Javascripts" I'm also in charge of updating everybody's JIRA tickets now, which I found out last week after the CEO got mad at me for not making sure all of the other developers' JIRA tickets were updated. So I guess I'm a product manager now or something? bvj191jgl7bBsqF5m fucked around with this message at 05:16 on Jun 5, 2018 |
|
#
?
Jun 5, 2018 05:10
|
|
|
Shirec posted:He doesn't give a poo poo about HIPAA compliance. Any time he has to get us to do something he knows violates it, he'll say "I have no knowledge of how you are choosing to solve this problem, but get it done this way" Just make sue you're documenting this stuff! Get it in writing somehow, even if it's just you sending him an email saying "per our earlier discussion, i will implement x as follows..." It may never be useful, but just think of the incredible schadenfreude if you could point the finger squarely at him for some HIPAA violation.
|
|
#
?
Jun 5, 2018 08:05
|
|
|
bvj191jgl7bBsqF5m posted:I'm also in charge of updating everybody's JIRA tickets now, which I found out last week after the CEO got mad at me for not making sure all of the other developers' JIRA tickets were updated. So I guess I'm a product manager now or something? Project manager 
|
|
#
?
Jun 5, 2018 12:55
|
|
|
baquerd posted:Sometimes your boss seems more terribly insecure that they have no idea how to manage people or write software than they are exhibiting outright maliciousness. This seems a large part, but the power trips make him feel good about himself and his fragile ego, at the expense of others and that makes him evil.
|
|
#
?
Jun 5, 2018 14:16
|
|
|
bvj191jgl7bBsqF5m posted:One where people say "Javascripts" We just started using JIRA about two months ago, but I've only really started to use it in the past few weeks. I cannot overstate how much I loving hate using it, but I also can't decide if I just need time to get familiar with it or if it's genuinely terrible or if we're just using it wrong. It just seems like there are endless Sprints with Stories and Tasks and Subtasks all layered together and no clear way to tell what's actually mine and what isn't. I also can't decide if I should speak up to my boss about how much I don't like it, or if the company is at a point where they won't back down after we've gone this far into integrating it. Either way I want to  every time we have our newly implemented Weekly Standup meeting to discuss JIRA, or when an email comes in reminding us all to update JIRA.
|
|
#
?
Jun 5, 2018 15:51
|
|
|
It can work, but with any sort of central system there needs to be clear guidelines on how it's to be used. Otherwise you're going to be stuck with a quagmire of per-team or even per-layer implementations.
|
|
#
?
Jun 5, 2018 16:10
|
|
|
any tool is there to support a workflow, maybe just not yours any tool mandated by management will be used wrong, sometimes for extended periods of time, sometimes on purpose any tool that is unfamiliar to the user will seem poo poo Jira is presented as a Very Good Solution where it is often Mediocre But Still Better Than the Competition.
|
|
#
?
Jun 5, 2018 16:17
|
|
|
When we switched to SAFe (which I'm still convinced is just waterfall disguised as agile), we had a big company-wide meeting on how to input tickets into TFS and how to progress the different types through the statuses. It actually worked, with the exception of the people who are employed at a software development company using a Microsoft stack who still somehow can't operate a computer, or Skype for Business, or Office. But those people are doomed to fail anyhow. We implemented some pretty strict policies, though, so the workflow mostly Just Works. You can't do stupid poo poo for the sake of stupid poo poo. You have to enter dev notes into bugs, user stories must have acceptance criteria and must be vetted by a PO before entering any team's backlog, et cetera. It cut down on the "thing is broken, plz help" sort of bugs in our primary support queue. Also, TFS and Git is actually a really nice pairing. We use TFS just for the Team Services, not TFS version control.
|
|
#
?
Jun 5, 2018 16:48
|
|
|
Keetron posted:any tool mandated by management will be used wrong, sometimes for extended periods of time, sometimes on purpose My former company's switch to fully Agile using Scrum. The initial presentation was fine and closely pinned to how it should have been implemented, it just wasn't a good fit for our team that was split between three products, with one member being extremely part-time and another working on the other side of the world in Scotland (we were using Kanban previously due to being mostly code maintenance before a department shakeup). And then management decided to use the effort point system and velocity as methods to determine performance of individual members in a fashion that would also pit them against each other. A couple months later they doubled down and stated that all points would be determined by the time it would take the average joe developer to complete the task, presumably because there was more cross-team work occurring and metrics were not coming out the way they expected them to. It was a shitshow. Sage Grimm fucked around with this message at 17:28 on Jun 5, 2018 |
|
#
?
Jun 5, 2018 16:53
|
|
|
Protocol7 posted:When we switched to SAFe (which I'm still convinced is just waterfall disguised as agile), I think it depends on your PO and Scrum Master. If they are good at bridging the Business/Dev divide and ensure that they only demand enough info to determine external dependencies and level of effort enough to have a rough roadmap to show dependent teams during PI planning that is understood to be subject to change during the period to react to business needs, it works well. If they are planning the minutiae of the entire next three months to the second, and inflexibly sticking to it, then yeah, it's dumb. My SAFe team's was really fortunate that the scrum people were good and the organization knew the limitations of the process and was willing to work around it. Also, I loved PI planning because I could see the context and effects of my work, interact with users and fellow consultants on other teams, and get loads of free gourmet quality food. Honestly, that goddamn grocery chain had a better digital operation than a lot of actual digital and engineering companies I worked for.
|
|
#
?
Jun 5, 2018 17:24
|
|
|
BurntCornMuffin posted:Also, I loved PI planning because I could see the context and effects of my work, interact with users and fellow consultants on other teams, and get loads of free gourmet quality food. PI Planning is really great, and it's good to get everyone together and discuss. I think SAFe has provided a lot of value in that regard. But we still have fixed release dates and predefined scope for those releases, so aside from doing DSU meetings and other scrum board things, it's still not very agile. A limitation of my POs and SMs, I agree. It's just unfortunate that they try to call it agile when the releases and everything around that process are still very waterfall-y. I would love if we switched to a true agile methodology.
|
|
#
?
Jun 5, 2018 19:44
|
|
|
|
| # ? Jun 6, 2024 21:00 |
|
|
Protocol7 posted:fixed release dates and predefined scope for those releases Found our disconnect, my org didn't have this.
|
|
#
?
Jun 5, 2018 21:00
|
|