Subjective review of some of the English-language educational platforms. Part 1: Hyperskill.org

As I promised in the previous article , I continue my research. For those who have not read my first article about Russian free educational platforms, I would like to inform you that I came to the following conclusion: everything is pretty sad with studies there - there are a lot of useless materials, empty promises, inconveniences and outright hack.



Since there is no quarantine now, there is not so much free time for research, so I will publish a review of English-language platforms in parts. Now we will focus on Hyperskill.org (hereinafter HS) aka JetBrains Academy (hereinafter JB). I invite interested readers to discuss the article and their online learning experience in the comments. Suggestions on the choice of interesting English-language educational platforms will also be useful. 



Disclaimer. I am not associated with any of the educational platforms and have not been associated with any relationship, except for free education. The article is not intended to offend anyone, cause financial damage, condemn someone's choice, call for the choice of something, and also inflate the holivar. If someone decides that this is another "plaintive" article, then I suggest paying attention to the first word of the title.



Hyperskill.org







The HS platform is designed to provide  paid  educational services. At  https://www.jetbrains.com/academy/  in English, it is proposed to undergo training in 4 areas, of which 2 are not yet fully ready (marked as beta ) - see below. There is no Russian language and, it seems, is not planned.







For long-term registered users, training until January 1, 2021 is free. I registered a long time ago ...



It's nice that there are no lies about any guarantees of employment: “ We believethat learning at JetBrains Academy will definitely help you become more prepared for your first job interview. The more complex projects are inspired by test assignments for junior developer positions. We do not have any special agreements with IT companies and therefore can't give any guarantees " .



I would very much like to know what this belief is based on. Usually, this confidence is based on the impressive credentials of the content author / authors. I didn’t see anything like this on HS ... I’ll write more about the authors of HS just below in this article.



The HS home page mentions somewhere down below that you will need to install the corresponding JB product to complete the course. I think that very few people will read this - I, now, have not read it ... Further in the course (for example, in the topic about IntelliJ IDEA) will talk about the awesomeness of JB's products. But you won't be told what to do if another IDE is already installed. Is the JB product required to continue the course? Later it turns out that yes : without installation, it will not work to continue training.



I will continue the review.







At the first step, HS offers to choose the direction of study: Java, Python, Kotlin and, judging by the screenshot above, JavaScript. The training time is:



  • The full Java track takes 144 hours to complete on average.

  • The full Python track takes 93 hours to complete on average.

  • The full Kotlin ᵝ track takes 47 hours to complete on average.







With JavaScript, the situation on HS is incomprehensible - after clicking the "Choose your track" button, this item is missing. But there is still one JavaScript project at this address .



I decided to learn Java: Java Developer track , 29 projects, 144 hours, 371 topics. Further, all conclusions about HS are made from the experience of learning Java.



The question immediately arose about the reality of estimating the time spent on training: how, for example, the level of complexity and the average time of the solution are combined? Well, only HS can answer ...







After choosing a track, HS offers to choose a project for training one of four difficulty levels: from Easy to Challenging... I was not looking for easy ways. For each project, you can find out its average student grade, average duration, the number of topics studied in the project. Issuance of a certificate is not provided.







If you scroll through the list in full, you can see that even the most complex project affects less than 1/3 of all available topics: 110 out of 371. Therefore, after completing the first project, you can take on the next one or look at the so-called. Knowledge map and study the remaining topics of interest directly from there.







Each topic consists of theory and several questions and / or exercises. The theory is given in the form of text - no video. This is a plus for me. Both theory and questions can be discussed with other students. Sometimes posts of HS team members also appear there: mainly about fixing errors in the topic - fixed



Both questions and theory can be assessed and explained. I have already noted that HS tried to fix obvious errors in the content. But not all ...



The study of the knowledge map allows (highly likely) to assume that the project sets the sequence for the study of topics, and the material itself does not adapt to the project. It is noticeable that themes are created by different people - there is a difference in styles. Caught funny samples. For example, in the topic about IDE you can see the following phrase: " If you're thinking" wait, it sounds too good, what's the catch? " you are right ... ". This topic is common to all languages ​​in HS: Java, Kotlin, Python.



The principle of HS content creation makes it difficult to assimilate the material, because content authors do not fully consider what has already been learned by a particular student and what has not. This is somewhat strange, since there are dependencies of topics on the Knowledge Map , and in a large number. Most likely, traditionally, “ something went wrong ” and some dependencies were lost. For example, " Theory: Standard output " is given after the output tasks. Or, another example from the very beginning of the course.





Work the JVM.



In the correspondence with HS, we managed to find out that the content is created according to the following model: the interested person (most likely a student) writes a topic and several examples for it. For this he receives 2000 ... 3000 rubles. I tried to convince the HS representative that such a model obviously leads to a low rate and the presence of errors. I have not achieved success. However, as often happens, Habr helped: I came across an article about online schools . There I found a possible explanation for such a position of HS: "the third mistake is the high cost of the product - the authors want the course to be of the highest quality . " HS, it seems, does not want to make this mistake: with their approach, the cost of content will be close to the minimum, and " people hawala"so. Although not always - quite often in the discussion it is advised to read the topic on, for example, http://tutorials.jenkov.com/ - there is often not enough information on HS to solve the problem.



For my part, I can also recommend the famous CodeGym . I read their articles - I liked it ...



If someone wants to work at HS for the above price list (or maybe a higher one - as you agree), then contact https://vk.com/wall-118500697_392 .



An interesting feature of HS is that it tries to determine the level of the student before starting the course - he is asked several questions. Then, based on the results of the entrance testing, it is suggested to skip some topics in the track. This is called a personal curriculum.... But even after successfully passing the test, during training, strange tasks of the form come across: paste the code above into this field . See screenshot for example. This is considered medium difficulty! And in the test there were questions, for example, about inheritance ...



To make it clear: I answered all the test questions about inheritance. And on others too ...







When going through a personal plan, there are other oddities - they can offer a problem already solved during testing.



A big drawback of the HS platform is the testing system - task checks are very slow : this task, for example, is checked for about 5 seconds.







But this one is more than 30 seconds.







It seems that the HS problems are the same as those discussed in my previous Stepik.org review. Taking into account the common "parent", it can be assumed that tasks are checked on shared capacities. Since there are no deadlines, then, if you wish, you will pass the course even with such a slow check, but you may have to pay an extra month ...



Another supposed means of delaying the educational process is the so-called. Code quality - it is not clear what it affects; I personally did not notice such an influence at all. But warnings are tiring (and you can't turn them off) and encourages you to spend more time on the task. And $$$ are coming ... It looks like it's a conspiracy , yes ...



The screenshot shows an example of a "gross" stylistic error from the point of view of HS - one "missile" was taken away.







Another example of a “gross” stylistic error: “ int x2 = 0, x3 = 0, x4 = 0, x5 = 0; ". It is believed that it is impossible to write that way. And missing a space is also a "gross" mistake.



Or another example of a style check: correct, but can be improved .







Moreover, HS itself often uses such constructions in theory, for example, here .







And in the recommended solutions too:





Editor's choice!



By the way, after that I went through a dozen more topics, but they haven't told me anything about the error " The user-supplied array is stored directly ". Well, we know that ... But those who do not know, with such a "patchwork" training, can hang over this error.



I guess the content is mostly students.for a small price. Their writings seem to be verified in a similar way; if at all checked - wrote above about the posts of members of the HS team. Their main content: corrected , corrected ... Therefore, there are stupid mistakes that the students did not indicate. For example, in the already mentioned Java knowledge test, one of the questions came across two identical answers. Or the tasks in the test are formally different, but in fact, to solve them, you need to remember the keywords when declaring the main method



Similarly, when solving problems on a specific topic already in the track: often the tasks differ only in the name of the variable and its type. 



HS tasks can also be artificially cumbersome., for example, write several inheritors of the base class. To understand the essence, one heir would be enough, especially since the heirs are almost identical .







A huge drawback of HS is the difficulty of free movement in the track by topics and tasks in the topic. This is not a lack of content, but of the platform itself.







Suppose I am stuck on some task in the track. In other systems, I can continue to study the course further: if I gain my mind, I will return to this task. Or I won't come back ... And here - it is not clear how to do it. This is again, in my opinion, to the issue of $$$.



There are other errors in the interface of the HS platform itself: how many sections do I have until the end of the topic ?!





 

I think I've written enough about Hyperskill. I turn to the conclusions.



  1. HS .

  2. , 25$ 67 .

  3. , «»: , , , . , . 

  4. HS , – : «» , . , . .

  5. «» HS.

  6. , HS . , . , , . , . HS JavaScript, , , . 

  7. , , – . .

  8. . !

  9. : HS? 25$ , – 50$ . , .2 .




All Articles