What is June without Github and at least one diploma project, it would seem? A OT `!`

What is June without Github and at least one diploma project, it would seem? A OT `!`

I apologize in advance for the clickbait title, this is a quote from a Twitter thread, which will be discussed below with punctuation. The post turned out to be long, full of emotions and personal opinion and not at all technical, but that’s how it turned out. Sometimes words stick more cleanly than any hooks and hit harder than a fist.

When I had ice at Gaijin, I managed to work with many wonderful people and professionals in their field, including Zhenya K. and David F., and even after moving to another studio, we continue to keep in touch, throwing each other interesting news and congratulations Happy birthday. Actually, a few months ago, another hochmotred about the Junes came to me.don’t go there so as not to create a habrefect) I will list interesting quotes below. But at that moment it was very tight with the passage of time, another milestone, job search, interviews with new boyfriends, upheavals in the company, it was not particularly important for threads and articles, I would have crawled to bed without falling asleep on the way. Then he took a quick look, noted the strange presentation of the material and forgot.

And here I got to read more thoughtfully and combine it with the realities of recruiting not only juniors, but middles and seniors for the positions needed in the studio. The girl-developer Nataliya conducted a small experiment, made a fake job vacancy and collected reviews, looked at your worst resumes and collected statistics based on which she began to give recommendations on how to improve these same resumes in order to apply for this fake vacancy.


June without github is normal

The first thing that caught the eye in the description of the vacancy itself:

[дайте] a link to your best project on github and to the production version, and also briefly tell about it;

and then even more insistently:

What is June without Github and at least one diploma project, it would seem? And HERE! I don’t know how and why it is necessary to learn so that the github is empty, with a couple of green squares, but even such an empty git was only 69% of people with companions. And for github to be minimally decently designed, and there was at least something clear, this is a terrible rarity. Less than a quarter of guys with escorts!

Wait, you’re asking jun, right jun, for the best project on github? That is, maybe several of them? Does an active project imply any prod, commits? And they didn’t mention the green squares, maybe they should also be dark green? And you yourself will pass your filter on github (I will not provide a link to Nataliya’s git, it will be googled in a couple of minutes), something is not particularly true there, the tracker is turning green, and there is also a problem with projects in progress.

We’ve been trying to track down the renderer and the colleagues who brought him here, who do know that he has an active, but closed, repo where he shows some particularly cool work. But the main repository is almost empty (https://github.com/Raikiri/LegitEngine)

About courses

And here, I will repeat myself, we have before us Jun, who, judging by the comments in the thread about the quality of the resume, most likely came from courses or an institute, and from which courses he came is also not clear, maybe from normal private ones, or maybe from [xyz]andex’s. According to the quality of the current students from the institute, I could be wrong, because there was a very limited sample of 12 people, there everything was fine with the knowledge of Git, but only within the framework of issuing and launching commits, dying branches, but from words – you need to dissect commits in search of regression – a misunderstanding began.

If a person graduated from paid courses (yes, such were also leaked, but we tried not to take from the courses, we were called for an interview only if he passed the test on “ofigitelno”), because you can easily come across a gpt-programmer (unfortunately, the realities of modern life), there is no guarantee that the repository will be open. My opinion is that if the courses are paid, the repo should be closed, and only the course and I will be available, otherwise everyone will write off each other – once, who needs to index this course and pass it off as their own – that’s two.

There is also an opinion among my acquaintances and colleagues that in public raps it is not the code of the students, but the code of their teachers, which the students for some reason repeat to themselves in the projects, so maybe it is not the students who should be called after the analysis of the repository? And what, excuse me, as a result, the author of the thread wants to see in this code? Find stars who understand why they are doing it, were able to pass the teacher’s review and convince him that their option is better, optimal and faster? Then I have bad news, they are sorted out even before hh, usually in the third course, and after leaving the university they are already with work and projects in the background. And they definitely don’t go to such courses because they were bored at university.

Who else did we have from the newcomers? Self-taught, like that guy from github, he won’t get into your selection either, he has 500+ stars there and probably we weren’t the only ones who came to him. Or there was a Jun from another company who has already gained experience there and is looking away. But tell me honestly, how many companies will allow a developer to store the code of their know-how in a public repository so that it can be shown and shared? A million questions, of course.

Builds are not an indicator of knowledge

Yes, and the project should still be in production. are you serious What do we want to know about the candidate from the presence of images? The ability to press a couple of buttons on github or make a config from ready-made examples? So this pipeline 99.9999% will not be in the office where he will get a job and will have to study and learn especially. Or do we learn that the candidate knows how to copy code so that he successfully earns and does it “easy and simple” through ready-made actions on GitHub Actions?

Prod is the presence of a deployment, which is done in a few mouse clicks.

Of course, I’m not a welder at all, and I don’t weld on the Internet, I hope you’re already launching rockets into space at lkm. The game developer also has its own assembly plans, when a build for a couple of tens of gigabytes is collected from a thousand resources, resources are checked, autotests are run, the game plays itself for a couple of hours and then a working build/bundle appears, which is sent by the QA team. And I don’t go there, because there is a build engineer who is responsible for this whole zoo. So prod is configured and understandable processes, and not a mindless rollout of ready-made actions in a couple of clicks.

In addition, this apparent simplicity hides another sublevel from Jun. the world knowledge that exists only in documentation and old fables, and is obtained through personal experience. Obscured by its simplicity, the understanding of the assembly and delivery process (and the bundle must still be sent to Steam or the PS store) to the end user of the bundle in the style of “press a couple of buttons” will not lead to anything good. A newcomer in a normal office will be admitted to the building farm, if only after a couple of months – just to look and breathe next to him, because this is not his field of activity at all, but that of building engineers.

About conferences

Do you know what, in my opinion, is the main problem of people who have had a bad time? Not that they are bad or lazy (although there were, of course, inattentive slobs).
And the fact that they are as far as possible from any large communities and trends. Their IT world is tiny and narrow. These guys clearly show that the wider your professional world, the more people you see in it, the better your software and hardware, by the way, too. The 7 guys who did the cool thing definitely didn’t sit in isolation. They are clearly involved in the community, from there they also learned good practices for resumes, got pet ideas, and learned about cool ways to design profiles on Git. Communities gave them a chic competitive edge

I had a colleague who quite often asked for a day off to download memes. He almost always went to a meeting, it is important to self-teach, and for some reason the bugs were scattered throughout the department, i.e. he himself went to the meetings, and the whole department fixed the bugs.

Moreover, I see this attitude not only in the author of the thread, it has generally become fashionable to go to meetups a lot, and to sit in IT threads on weekdays, passing it off as self-study. But when you return a bug that has not been fixed or point out obvious mistakes, they are offended.

From there, the conclusion is that the 7 people who passed the test are hanging out in the IT party and participate “in the big community”? And what, we have banned books and lectures, without social media, now the knowledge of new standards cannot be pumped?

Meetings and conferences are important and necessary. They provide an opportunity to communicate not only with colleagues and show where you can grow. And you can come up with interesting topics yourself, and it’s a really good experience. Perhaps metaps give a certain percentage of transitions between companies that will simply know about your existence, and you will know about theirs. But he did not meet such among his acquaintances.

About the community

The same community repeats the mantra of how to solve problems with litcode. Seriously – it turned out to be the most unnecessary idea to check the knowledge of litcode tasks on social media. At first everything was fine, they took two mediums, one solved two medium tasks close to the algorithm from Litkod, the other wrote worse, but he was also a worker. And in fact, during the probationary period, the litkodev was filled with one bug after another, the knowledge seems to be there, but the litkod didn’t teach how to apply it. The second one was also not without problems, but he fixed the bugs. Guess who had to break up with.

Non-working code and algorithms fly out of my head faster than uninteresting advertising, and to find a regular and justified application of ВД and DF algorithms – well, ok, once every six months something will be found where to apply it.

The same community repeats with a second mantra, do DRY, do KISS, do what else they have invented… But this DRY put in Jun’s head will be spoiled in the strangest way, and will spoil the life of everyone around when writing complex systems. And if this engineer grows up with a lack of understanding of such methods and carries good and eternal life? I will not envy his colleagues and students when they make a complex system.

Do it Smartly and Conveniently, I believe in this DRY, but these courses are often taught only by life itself, with painful bumps and sleepless nights hugging the debugger.

Riding cats against identical bridles

I strongly oppose SUCH studies, because Natalya runs a huge blog with almost 19k subscribers and actually influences the opinion of a large group of people who will accept her reasoning and conclusions, presented in the form of research, as genuine. Not only will they believe it themselves, but they will also carry it further, turning it from 19k to 190k. But real projects show that this approach does not work! That’s probably why, out of 30 positions for programmers in the studio, only 4 were closed in six months, because they were not asking for mitapes, LinkedIn profiles and projects on Github, but for real tasks.

The author expects his candidates, who have just started doing something with their hands, to participate in a large community – it’s so easy! You can subscribe to a couple of hundred publications just by scrolling through C++Russia and reading them instead of “pure code” before going to bed, and then not understanding why using std::vector in a loop is such an endeavor, or why it slows down so ungodly for(string a: strings), but everything according to feng shui, everything according to the latest practices of the large community.

And from such posts, the owners of the construction galley “Labean & Partners LLC”, having looked at the rutubo-visas of powerful research from Natasha, begin to mindlessly copy the race for rock stars, without understanding what they are doing when they raise the conditions on June positions, sorry, to copper level. A long time ago, these lifters themselves were Junes? And have all the projects on Github and soft-skills Mitapny and the dead ostrich been read? Maybe so, but then these are not junkies, they are people with good actual experience, and they can be called junkies as soon as there is no commercial experience.

June is just a young engineer

Jun (IMHO) should know the syntax of crosses (or the language in which the office writes), the basics of mathematics and logic, and preferably understand the basic stack on which he will have to work. Then the interview turns into a couple of questions about the language, an assessment of general adequacy and basic questions about the world (just to understand whether he really knows and is interested) and whether the person is able to program. Then we talked a little about life to understand whether a person would fit in the department… and that was it.

This is the stage from which all developers start, perhaps they are taken (and often it turns out) to solve specific tasks of the company, the solution of which will take time and resources to learn. We do not set a five-year-old child the task of building brick buildings according to GOST, we give her wooden cubes and watch how she learns from mistakes.

And such Natashas with such an attitude towards Junes, I’m sorry, it’s better to turn to palmistry. There the results are more honest.

Related posts