Betrayed by github activity graph
-
Let's keep everything aside for a moment. Fuck the stupid chart and the weekend thingy. Why did you have to go and tell this to the HR. If anything, you're the one who betrayed the programming community by giving more leverage to HR. You gotta undo this somehow.
-
@Jaloopa said in Betrayed by github activity graph:
Because nothing changed in those 16 years,
You're right, I still punch cards all day
-
@blakeyrat said in Betrayed by github activity graph:
@Gąska said in Betrayed by github activity graph:
You know, there are people in the world that aren't like you.
Degenerates.
With that attitude, no wonder you have a job you hate.
@Gąska said in Betrayed by github activity graph:
Some programmers actually enjoy programming,
I enjoy creating software. (I even do it as a hobby.) I don't enjoy programming. Programming is by far the worst, most boring, most un-fun part of creating software.
To you. I, for one, like coding. (Sometimes. Under specific circumstances.) I also enjoy shipping high quality products, but that's another matter. I definitely don't enjoy making documentation, testing, talking with POs, and all the other supporting tasks - but hey, everyone's different, and I don't see it as wrong if you enjoy that.
@Gąska said in Betrayed by github activity graph:
and like all hobbyists, they can see differences between different software projects that normal people wouldn't ever notice, and these differences determine how much fun they're having working on a particular project.
So they get all giddy because the code uses one implementation-detail technology over another?
Sometimes, yes.
Who gives a shit. It's all implementation-detail.
I do. I think you might not know what the word "hobbyist" means.
The only thing that matters is the parts the user sees.
And why can't I take joy in things that don't matter? Do you give the same speech to people who like sightseeing?
@Gąska said in Betrayed by github activity graph:
as well as solving interesting problems without obvious solution.
99.9% of the time this just means the developer:
- Doesn't have enough experience, doesn't read enough, etc. (In other words, they think the problem has no obvious solution when there's like 30.)
Who cares? I'm talking about how much fun the developer has developing. Of course less experienced developer will find things more interesting. But there are projects that even experienced developers struggle with, for example graphics programming for blockbuster AAA games.
- Their solution sucks ass.
Again, I'm talking about fun, nothing else. In another conversation, I'd agree with you - but in this one, it doesn't matter.
The other 0.1% of the time the code could have been avoided entirely if better design decisions had been made up-front.
Design is part of solution. To me, the most interesting part. In fact, I like designing (architecture, not UI) more than programming itself.
-
@blakeyrat said in Betrayed by github activity graph:
@TimeBandit said in Betrayed by github activity graph:
I believe you made the wrong career choice
When I got into the industry in the late 90s it was a more interesting industry. That was back when it was all about usability and user experience and coming up with new and innovative development tools, before it was taken-over by the cabal of Unix developers screaming "DO EVERYTHING LIKE WE DID IT IN 1986 FOREVER!"
Hell back then, Filemaker and Microsoft Access had better GUI-builders than the most popular open source-y languages now do. Apple had a macro recorder that worked system-wide and didn't require one touch of a CLI. AOL was inventing "buddy lists" and had a well-designed IM program that read your messages using text-to-speech.
It's not me who changed, it's the industry.
And yes I hate everything about my career, thanks for reminding me.
-
@Gąska said in Betrayed by github activity graph:
With that attitude, no wonder you have a job you hate.
...and it's completely impossible I was making a joke!
-
@blakeyrat said in Betrayed by github activity graph:
@Gąska said in Betrayed by github activity graph:
You know, there are people in the world that aren't like you.
Degenerates.
@Gąska said in Betrayed by github activity graph:
Some programmers actually enjoy programming,
I enjoy creating software. (I even do it as a hobby.) I don't enjoy programming. Programming is by far the worst, most boring, most un-fun part of creating software.
@Gąska said in Betrayed by github activity graph:
and like all hobbyists, they can see differences between different software projects that normal people wouldn't ever notice, and these differences determine how much fun they're having working on a particular project.
So they get all giddy because the code uses one implementation-detail technology over another? Who gives a shit. It's all implementation-detail.
The only thing that matters is the parts the user sees.
@Gąska said in Betrayed by github activity graph:
as well as solving interesting problems without obvious solution.
99.9% of the time this just means the developer:
- Doesn't have enough experience, doesn't read enough, etc. (In other words, they think the problem has no obvious solution when there's like 30.)
- Their solution sucks ass.
The other 0.1% of the time the code could have been avoided entirely if better design decisions had been made up-front.
-
@blakeyrat said in Betrayed by github activity graph:
@Gąska said in Betrayed by github activity graph:
One of them isn't software development work, two others totally involve check-ins.
Well I suppose it depends on how your company's run. If you don't have a good solution for document sharing, you might check-in your documentation.
Sounds like you think checking in documentation into source control is a bad way of doing things. If that's indeed what you think (it's not your exact words, and because it's you, I'm extra careful about making assumptions), would you care to elaborate on that?
I don't know why you'd check-in a quick and dirty prototype.
If it takes more than two days, or is to be worked on by more than one person, it's better if it's stored somewhere it can be retrieved from when the developer's laptop goes missing. Note that the post I replied to didn't contain words "quick and dirty" - and in some companies, it's not uncommon to have prototype software projects going on for half a year or more, with over a dozen developers assigned.
Maybe in a different repo if you think it'll be around longer than a couple days.
I never said "checking in to the same repo as the main product". I just said "checking in".
-
@blakeyrat said in Betrayed by github activity graph:
BTW how does this dumb feature treat pull requests? If I build up a pull request every day but my asshole boss only approves all of them on Fridays, does that count as a check-in every day or just one big one on Friday?
It goes by commit dates. Commit, not PR. And it counts each commit separately.
-
@Gąska said in Betrayed by github activity graph:
Sounds like you think checking in documentation into source control is a bad way of doing things.
I wouldn't say it's a bad way of doing it. (It's certainly better than having it only one one computer and not backed-up anywhere.) But it's not the best way of doing it.
As we've discussed before, Git (and other SCMs) are absolutely ass at doing diffs on anything that isn't a plain text file. Word (or Google Docs, or whatever) however is actually quite good at it, so if you're not using that, you're being an idiot. Plus Word and competitors can let two people edit the file simultaneously.
@Gąska said in Betrayed by github activity graph:
If it takes more than two days, or is to be worked on by more than one person, it's better if it's stored somewhere it can be retrieved from when the developer's laptop goes missing.
Yeah I keep mine in OneDrive. Most companies have something like that set up for backing up files, right?
@Gąska said in Betrayed by github activity graph:
It goes by commit dates. Commit, not PR. And it counts each commit separately.
What if you have the jerk on your team who thinks it's a good idea to merge the PR as a single commit?
-
@marczellm I'd say that the moment in spelunking when you realize "it's a miracle the system doesn't just collapse, for instance right now", is pretty exciting.
-
@cartman82 said in Betrayed by github activity graph:
@Gąska said in Betrayed by github activity graph:
I am just saying, if this image was the only thing someone had to go on when deciding whether to try and recruit me, they might wonder.
Maybe you should stop coming up with impossible hypothetical situations.
In the real world, you have to make some kind of determination.
Do you not have this kind of useful info:
1.) How long candidate has been with current/prior employers (if he truly was slacking off on weekends, it'd probably not last him long at that company)
2.) References (not of current employer, since unless their collapse is imminent, they don't know the employee is looking for a new job, but any employer prior to that could at least verify the employee worked as long as he said he did)
3.) The interview itself. Yes, they can bullshit themselves in, but if you ask the right questions you can get some hints that they're bullshitting.
Barring that, if you come across "the only thing you can go on" and it sucks as a metric, then I assert that going by that is worse than going by nothing, because it's likely to give you a false positive/negative. It's like if I had to blindly choose a car's reliability and the only thing I had to go on was the seat had a small tear in it. I might as well cut my losses and admit that I have nothing to go on and everything is by chance.
But, again, with the three things above, I really don't. There are more reliable ways to judge how well a candidate will perform. They aren't perfect indicators. Nothing is. But they're better than snooping on one's commit frequency without knowing anything about how their current employer does business.
-
@blakeyrat said in Betrayed by github activity graph:
@Gąska said in Betrayed by github activity graph:
With that attitude, no wonder you have a job you hate.
...and it's completely impossible I was making a joke!
It's certainly impossible to miss the irony of what you've just said in this post.
-
@blakeyrat said in Betrayed by github activity graph:
As we've discussed before, Git (and other SCMs) are absolutely ass at doing diffs on anything that isn't a plain text file. Word (or Google Docs, or whatever) however is actually quite good at it, so if you're not using that, you're being an idiot. Plus Word and competitors can let two people edit the file simultaneously.
What do you think about using a plain text based format for documentation? Markdown, reStructuredText etc.?
-
@marczellm Blakeyrant incoming about how implementation details are for stupid idiot morons.
-
@blakeyrat said in Betrayed by github activity graph:
As we've discussed before, Git (and other SCMs) are absolutely ass at doing diffs on anything that isn't a plain text file.
Git's slightly better at this than most other SCMs (SVN, CVS, whatever garbage you use), because it is so stupid.
[ed. was too lazy to extend quote]
There's value in keeping the version history of the documentation aligned with that of the codebase, and it's a lot of value. This means collaborative edit needs to be able to be committed from. I'm not cool with "lol the documentation has no versions".
-
@marczellm said in Betrayed by github activity graph:
What do you think about using a plain text based format for documentation? Markdown, reStructuredText etc.?
Don't.
-
@blakeyrat said in Betrayed by github activity graph:
But probably not. He probably just spent weekends with his family, the slacker.
The point is there's nothing in that chart that supports either conclusion. Maybe he did spent all weekend in a drunken bender and would have worked if he could have managed to stumble to his computer. Chart won't tell you that's why; it won't tell you that's not why.
Well, however he spent his weekends, we know he wasn't writing and committing code for the job. It'd be a better signal if he had. Even if just in that he might be more open to being poached.
-
@cartman82 said in Betrayed by github activity graph:
Well, however he spent his weekends, we know he wasn't writing and committing code for the job. It'd be a better signal if he had.
I think you're all alone on that one.
-
@blakeyrat What would you think about getting a screwdriver that wasn't physically attached to the toolbox it came with via a welded-on steel tether?
-
@stillwater said in Betrayed by github activity graph:
Let's keep everything aside for a moment. Fuck the stupid chart and the weekend thingy. Why did you have to go and tell this to the HR. If anything, you're the one who betrayed the programming community by giving more leverage to HR. You gotta undo this somehow.
It's too late now. You can't put the genie back in the bottle.
Even if I killed the HR lady and burned down the office, I bet she'd already spread this new tactic to all her HR friends/allies. Pretty soon, all HR recruiting will be based solely on github activity chart and the world will be hell.
And it will all be my fault.
-
@The_Quiet_One said in Betrayed by github activity graph:
But, again, with the three things above, I really don't. There are more reliable ways to judge how well a candidate will perform. They aren't perfect indicators. Nothing is. But they're better than snooping on one's commit frequency without knowing anything about how their current employer does business.
Well, they also involve a lengthy, multi-day process of contacting the candidate, trying to get them to apply, analyzing their actual CV and portfolio and then conducting multiple phases of on-site interviews.
Not very helpful when all I need to make is a determination whether to contact the candidate at all.
-
@cartman82 I don't even use GitHub. I'm screwed.
-
@blakeyrat said in Betrayed by github activity graph:
Well, however he spent his weekends, we know he wasn't writing and committing code for the job. It'd be a better signal if he had.
I think you're all alone on that one.
I am a visionary ahead of my time. Like the Galileo of analyzing git commit history.
-
@cartman82 said in Betrayed by github activity graph:
we know he wasn't writing and committing code for the job
No you don't. You only know he wasn't committing any.
-
@cartman82 At least you'd have tried, though. If that office is just HR it wouldn't even be like it was something that didn't need doing.
You will need to visit the Dealy Lama for spiritual deodorizing.
-
@The_Quiet_One said in Betrayed by github activity graph:
@cartman82 I don't even use GitHub. I'm screwed.
Well nothing's stopping you from putting all your portfolio code on an FTP server and putting a link in CV...
-
@cartman82 said in Betrayed by github activity graph:
@stillwater said in Betrayed by github activity graph:
Let's keep everything aside for a moment. Fuck the stupid chart and the weekend thingy. Why did you have to go and tell this to the HR. If anything, you're the one who betrayed the programming community by giving more leverage to HR. You gotta undo this somehow.
It's too late now. You can't put the genie back in the bottle.
Even if I killed the HR lady and burned down the office, I bet she'd already spread this new tactic to all her HR friends/allies. Pretty soon, all HR recruiting will be based solely on github activity chart and the world will be hell.
And it will all be my fault.
If you ever want to break into management you can at least put it on your CV as "Invented Github Activity-based profiling". Turn your faults into strengths!
-
@cartman82 said in Betrayed by github activity graph:
Even if I killed the HR lady and burned down the office
What do you mean by "if"?
-
@blakeyrat said in Betrayed by github activity graph:
@Gąska said in Betrayed by github activity graph:
Sounds like you think checking in documentation into source control is a bad way of doing things.
I wouldn't say it's a bad way of doing it. (It's certainly better than having it only one one computer and not backed-up anywhere.) But it's not the best way of doing it.
That I agree with. Though "good enough" is in most cases good enough. And plain text or HTML is often good enough.
As we've discussed before, Git (and other SCMs) are absolutely ass at doing diffs on anything that isn't a plain text file. Word (or Google Docs, or whatever) however is actually quite good at it, so if you're not using that, you're being an idiot.
Or maybe I don't have Google Docs or Word 365. Okay, I admit that argument doesn't work anymore - everyone has Word 365 nowadays.
One advantage of keeping docs in project repo is that it's easier to find and less likely to get lost.
@Gąska said in Betrayed by github activity graph:
If it takes more than two days, or is to be worked on by more than one person, it's better if it's stored somewhere it can be retrieved from when the developer's laptop goes missing.
Yeah I keep mine in OneDrive. Most companies have something like that set up for backing up files, right?
Probably, never used it so I don't know. It certainly sounds like a suboptimal way to store code, especially when you're cooperating.
@Gąska said in Betrayed by github activity graph:
It goes by commit dates. Commit, not PR. And it counts each commit separately.
What if you have the jerk on your team who thinks it's a good idea to merge the PR as a single commit?
Then his stats are shit.
-
@dcon said in Betrayed by github activity graph:
we know he wasn't writing and committing code for the job
No you don't. You only know he wasn't committing any.
If you write 2 days of code and don't commit, that's even worse than not writing at all.
-
@cartman82 I suggest that you recuse yourself from all hiring decisions for a period of not less than 7 years.
-
@Gąska said in Betrayed by github activity graph:
One advantage of keeping docs in project repo is that it's easier to find and less likely to get lost.
Well you can shove 'em in there periodically I guess.
Maybe when you peel off your release-1.0 branch you can shove the docs as of release-1.0 in there. I dunno.
-
@cartman82 said in Betrayed by github activity graph:
If you write 2 days of code and don't commit, that's even worse than not writing at all.
Or you might be doing something investigative, trying to figure out how to do something complex, and it turns out that the idea you had doesn't work and can't possibly be made to work. It's not great when that happens, but sometimes it does.
I'm not sure if the github contribution history includes anything that hasn't made it into a project's main branch yet.
-
@izzion said in Betrayed by github activity graph:
Yeah, sure, I can totally see that from the code quality and commit dates.
sigh. yet another kid playing games with his commit dates. do they think I learned Git yesterday?You must have had more industrious professors than I had. With mine, I'd be shocked if they looked at the code, understood the code, understood how to get the commit log (if learned Git at all), or bothered to do any of the above. As with executives, professors want screenshots, they want a write-up, and they want an excellent presentation. The code quality doesn't matter.
-
@Gąska said in Betrayed by github activity graph:
Sounds like you think checking in documentation into source control is a bad way of doing things. If that's indeed what you think (it's not your exact words, and because it's you, I'm extra careful about making assumptions), would you care to elaborate on that?
Confluence.
At least that's this year's attempt at a documentation repository at Vandelay Industries. Previously we've tried Sharepoint, Trac, source code (via doxygen) and I forget which others...
The fact that we cannot seem to standardise on our doc repository for any decent length of time because every time we get a new manger they have a 'better way' of doing things, or some C*O has had his ear bent by some saleman so the whole company must use FootPrints or Jira, or Discourse doesn't really help matters.
-
@cartman82 said in Betrayed by github activity graph:
@The_Quiet_One said in Betrayed by github activity graph:
@cartman82 I don't even use GitHub. I'm screwed.
Well nothing's stopping you from putting all your portfolio code on an FTP server and putting a link in CV...
I've got NDAs that would say otherwise.
-
@The_Quiet_One the set
covered by NDA
andportfolio code
, these sets they are disjoint sets.
-
@cartman82 said in Betrayed by github activity graph:
@dcon said in Betrayed by github activity graph:
we know he wasn't writing and committing code for the job
No you don't. You only know he wasn't committing any.
If you write 2 days of code and don't commit, that's even worse than not writing at all.
Nonsense. I don't know if you're being hyperbolic or not, but while you should commit as often as possible, how could you say delaying your commits by a few days while actually working on shit is worse than not actually working on shit, especially if you're on your own long-term feature branch by yourself? That doesn't make any sense.
-
@heterodox said in Betrayed by github activity graph:
@izzion said in Betrayed by github activity graph:
Yeah, sure, I can totally see that from the code quality and commit dates.
sigh. yet another kid playing games with his commit dates. do they think I learned Git yesterday?You must have had more industrious professors than I had. With mine, I'd be shocked if they looked at the code, understood the code, understood how to get the commit log (if learned Git at all), or bothered to do any of the above. As with executives, professors want screenshots, they want a write-up, and they want an excellent presentation. The code quality doesn't matter.
My chemistry final project in high school could be switched for creating a program to do something in chemistry. So I opted to make a stoichiometry calculator, cursing the stupidity of the Eclipse GUI builder all the while. The teacher, for some goddamn reason, wanted the code printed out and stapled into a packet along with the title page.
-
@Gribnit said in Betrayed by github activity graph:
@The_Quiet_One the set
covered by NDA
andportfolio code
, these sets they are disjoint sets.No. Portfolios represent your professional history. Yes, you could include hobby and personal projects, but if you aren't doing that kind of stuff in your free time, it shouldn't negatively affect your job status if you are good at your day job.
-
@heterodox
I mean, I went to school for a Business degree. So none of my professors ever read my code either.
-
My professors did not even know how to code. #Indianthings
-
@stillwater said in Betrayed by github activity graph:
@cartman82 said in Betrayed by github activity graph:
But if I thought of this, other HR consultants/interviewers will too
No they don't.
In all fairness, they probably have or will. You just don't want to work for those companies.
-
@blakeyrat said in Betrayed by github activity graph:
Apple had a macro recorder that worked system-wide and didn't require one touch of a CLI.
Ohhhhhhhh, I am so close to blakeyrat Bingo. He just has to say the word.
-
@stillwater said in Betrayed by github activity graph:
Let's keep everything aside for a moment. Fuck the stupid chart and the weekend thingy. Why did you have to go and tell this to the HR. If anything, you're the one who betrayed the programming community by giving more leverage to HR. You gotta undo this somehow.
He obviously has to kill her before the next HR convention.
Never mind. At HR conferences they have Forrest Gump and Geena Davis as guest speakers. No HR actually happens at HR conferences.
That was a so big that my wife sent it to me.
-
@blakeyrat said in Betrayed by github activity graph:
@Gąska said in Betrayed by github activity graph:
With that attitude, no wonder you have a job you hate.
...and it's completely impossible I was making a joke!
If we based it on prior experience, yes.
-
@cartman82 said in Betrayed by github activity graph:
@stillwater said in Betrayed by github activity graph:
Let's keep everything aside for a moment. Fuck the stupid chart and the weekend thingy. Why did you have to go and tell this to the HR. If anything, you're the one who betrayed the programming community by giving more leverage to HR. You gotta undo this somehow.
It's too late now. You can't put the genie back in the bottle.
Even if I killed the HR lady and burned down the office, I bet she'd already spread this new tactic to all her HR friends/allies. Pretty soon, all HR recruiting will be based solely on github activity chart and the world will be hell.
And it will all be my fault.
You're going to need a lot of Quik-Lime.
-
@cartman82 said in Betrayed by github activity graph:
@blakeyrat said in Betrayed by github activity graph:
Well, however he spent his weekends, we know he wasn't writing and committing code for the job. It'd be a better signal if he had.
I think you're all alone on that one.
I am a visionary ahead of my time. Like the
GalileoTheodore Kaczynski of analyzing git commit history.FTFY
-
@Polygeekery said in Betrayed by github activity graph:
At HR conferences they have Forrest Gump and Geena Davis as guest speakers.
For anyone who wants to see for themselves that this is really a thing:
https://insights.ceridian.com/program/keynotes/
Forrest Gump is a fucking keynote speaker at a huge HR conference in Vegas. There should be an HR version of WTDWTF, but that would violate basically all the privacy laws and agreements.
-
@PJH said in Betrayed by github activity graph:
The fact that we cannot seem to standardise on our doc repository for any decent length of time because every time we get a new manger they have a 'better way' of doing things, or some C*O has had his ear bent by some saleman so the whole company must use FootPrints or Jira, or Discourse doesn't really help matters.
I hope that your documentation won't need too many posts for updates if it becomes the latter...