In 2013 Canonical tried to crowdfund Ubuntu Edge smartphone. Its main feature could be the ability to use the smartphone as a full-fledged PС. Unfortunatly, the crowdfunding campaign did not accumulate enough money, so a dream of having a universal device remained to be the dream.

I've been searching for universality, too, on the software side, not the hardware one. Today I can confidently say I found the necessary combination: Git and JavaScript.

As you know, I have already described the benefits of browser applications (nCKOB static site generator) and the benefits of using Git instead of yet another back-end with API (GitBudget to track personal spendings). Once GitBudget was out, I spent the remaining 2020 to build a system allowing one to create browser applications right inside browsers. GitJS is the name of that system.

Home of this article: https://robotics.snowcron.com/coins/02_head_or_tail.htm

The global objective of these articles is to build a coin classifier, capable of scanning your pocket change and find rare / valuable coins. This is a second article in a series, so let me remind you what happened earlier (https://habr.com/ru/post/538958/).

During previous step we got a rather large dataset composed of pairs of images, loaded from an online coins site meshok.ru. Those images were uploaded to the Internet by people we do not know, and though they are supposed to contain coin's head in one image and tail in the other, we can not rule out a situation when we have two heads and no tail and vice versa. Also at the moment we have no idea which image contains head and which contains tail: this might be important when we feed data to our final classifier.

So let's write a program to distinguish heads from tails. It is a rather simple task, involving a convolutional neural network that is using transfer learning.

Same way as before, we are going to use Google Colab environment, taking the advantage of a free video card they grant us an access to. We will store data on a Google Drive, so first thing we need is to allow Colab to access the Drive:

Hey everyone! This post was born from a question asked by an IT forum member. The summary of the question looked as follows:

• There is a set of text files containing routing tables collected from various network devices.
• Each file represents one device.
• Device platforms and routing table formats may vary.
• It is required to analyze a routing path from any device to an arbitrary subnet or host on-demand.
• Resulting output should contain a list of routing table entries that are used for the routing to the given destination on each hop.

The one who asked a question worked as a TAC engineer. It is often that they collect or receive from the customers some text 'snapshots' of the network state for further offline analysis while troubleshooting the issues. Some automation could really save a lot of time.

I found this task interesting and also applicable to my own needs, so I decided to write a Proof-of-Concept implementation in Python 3 for Cisco IOS, IOS-XE, and ASA routing table format.

In this article, I’ll try to reconstruct the resulting script development process and my considerations behind each step.

Let’s get started.

AngelList is a social network designed to connect startups with investors and vice versa. The founders of it were dissatisfied with how opaque the VC world was, and found a way to increase the amount of available data. The project began in partnership with just 50 volunteer investors wishing to allocate $80 million in capital, and has grown to be the leading website of its kind. Over the past three years more than 75% of startups that received seed funding from American investors used AngelList to make it happen.

All Prometheus metrics are based on time series - streams of timestamped values belonging to the same metric. Each time series is uniquely identified by its metric name and optional key-value pairs called labels. The metric name specifies some characteristics of the measured system, such as http_requests_total - the total number of received HTTP requests. In practice, you often will be interested in some subset of the values of a metric, for example, in the number of requests received by a particular endpoint; and here is where the labels come in handy. We can partition a metric by adding endpoint label and see the statics for a particular endpoint: http_requests_total{endpoint="api/status"}. Every metric has two automatically created labels: job_name and instance. We see their roles in the next section.

Prometheus provides a functional query language called PromQL. The result of the query might be evaluated to one of four types:

Scalar (aka float)

String (currently unused)

Instant Vector - a set of time series that have exactly one value per timestamp.

Range Vector - a set of time series that have a range of values between two timestamps.

At first glance, Instant Vector might look like an array, and Range Vector as a matrix.

If that would be the case, then a Range Vector for a single time series "downgrades" to an Instant Vector. However, that's not the case:

What is distributed tracing? Distributed tracing is a method used to profile and monitor applications, especially those built using a microservices architecture. Distributed tracing helps pinpoint where failures occur and what causes poor performance.

Let’s have a look at a simple prototype. A user fetches information about a shipment from `logistic` service. logistic service does some computation and fetches the data from a database. logistic service doesn’t know the actual status of the shipment, so it has to fetch the updated status from another service `tracking`. `tracking` service also needs to fetch the data from a database and to do some computation.

In the screenshot below, we see a whole life cycle of the request issued to `logistics` service:

Hi, Habr. Here I have prepared for you a small guide about NTFS Reparse points (hereinafter RP). This article is for those who are just starting to dive into the Windows kernel drivers development. In the beginning, I will explain the theory with examples, then I will give an interesting task to solve.

This is the third part of a series covering the implementation of algorithms in Go. In this article, we discuss the Merge Interval algorithm. Usually, when you start learning algorithms you have to deal with some problems like finding the least common denominator or finding the next Fibonacci number. While these are indeed important problems, it is not something that we solve every day. What I like about the Merge Interval algorithm is that we apply it in our everyday life, usually without even noticing that we are solving an algorithmic problem.

Let's say that we need to organize a meeting for our team. We have three colleagues Jay, May, and Ray and their time schedule look as follows (a colored line represents an occupied timeslot):

As we enter a new decade, the IT industry has also encountered many loopholes that we must work on. Seeing this, hundreds of motivated young minds have decided to become entrepreneurs. But is it so easy to succeed in a world where competition is already so high?

Bluntly, the answer is NO. It is not easy to be successful, and the reason for these startups' failure is the lack of digital presence or inadequate technical support. Another major reason that has been seen is the wrong choice of the web development company.

Hiring dedicated web developers may seem like an easy task at first, but it is not. It is one of the most challenging tasks. And to prevent you from falling into the pits of failure, I have created a list of the 7 best web development companies in India and abroad.

This article is about security. I’ll focus on this in the context of web applications, but I’ll also touch on other types of applications. Before I describe approaches and frameworks, I want to tell you a story.

Background

Throughout my years working in the IT sphere, I’ve had the opportunity to work on projects in a variety of fields. Even though the process of authenticating requirements remained relatively consistent, methods of implementing the authorization mechanism tended to be quite different from project to project. Authorization had to be written practically from scratch for the specific goals of each project; we had to develop an architectural solution, then modify it with changing requirements, test it, etc. All this was considered a common process that developers could not avoid. Every time someone implemented a new architectural approach, we felt more and more that we should come up with a general approach that would cover the main authorization tasks and (most importantly) could be reused on other applications. This article takes a look at a generalized architectural approach to authorization based on an example of a developed framework.

Approaches to Creating a Framework

As usual, before developing something new, we need to decide what problems we’re trying to solve, how the framework will help us solve them, and whether or not there is already a solution to these issues. I’ll walk you through each step, starting with identifying issues and describing our desired solution.

We’re focusing on two styles of coding: imperative and declarative. Imperative style is about how to get a result; declarative is about what you want to get as a result.

See more at robotics.snowcron.comThis is the first article in a serie dedicated to coins classification.Having countless "dogs vs cats" or "find a pedestrian on the street" classifiers all over the Internet, coins classification doesn't look like a difficult task. At first. Unfortunately, it is degree of magnitude harder - a formidable challenge indeed. You can easily tell heads of tails? Great. Can you figure out if the number is 1 mm shifted to the left? See, from classifier's view it is still the same head... while it can make a difference between a common coin priced according to the number on it and a rare one, 1000 times more expensive.Of course, we can do what we usually do in image classification: provide 10,000 sample images... No, wait, we can not. Some types of coins are rare indeed - you need to sort through a BASKET (10 liters) of coins to find one. Easy arithmetics suggests that to get 10000 images of DIFFERENT coins you will need 10,000 baskets of coins to start with. Well, and unlimited time.So it is not that easy.Anyway, we are going to begin with getting large number of images and work from there. We will use Russian coins as an example, as Russia had money reform in 1994 and so the number of coins one can expect to find in the pocket is limited. Unlike USA with its 200 years of monetary history. And yes, we are ONLY going to focus on current coins: the ultimate goal of our work is to write a program for smartphone to classify coins you have received in a grocery store as a change.Which makes things even worse, as we can not count on good lighting and quality cameras anymore. But we'll still try.In addition to "only Russian coins, beginning from 1994", we are going to add an extra limitation: no special occasion coins. Those coins look distinctive, so anyone can figure that this coin is special. We focus on REGULAR coins. Which limits their number severely.Don't take me wrong: if we need to apply the same approach to a full list of coins... it will work. But I got 15 GB of images for that limited set, can you imagine how large the complete set will be?!To get images, I am going to scan one of the largest Russian coins site "meshok.ru".This site allows buyers and sellers to find each other; sellers can upload images... just what we need. Unfortunately, a business-oriented seller can easily upload his 1 rouble image to 1, 2, 5, 10 roubles topics, just to increase the exposure.

So we can not count on the topic name, we have to determine what coin is on the photo ourselves.To scan the site, a simple scanner was written, based on the Python's Beautiful Soup library. In just few hours I got over 50,000 photos. Not a lot by Machine Learning standards, but definitely a start.After we got the images, we have to - unfortunately - revisit them by hand, looking for images we do not want in our training set, or for images that should be edited somehow. For example, someone could have uploaded a photo of his cat. We don't need a cat in our dataset.First, we delete all images, that can not be split to head/hail.

If you feel like the New Year just came, and you missed the first half of January, then all this time you've been busy looking for tricky bugs in the code you maintain. It also means that our article is what you need. PVS-Studio has checked the ELKI open source project to show you errors that may occur in the code, how cunningly they can hide there, and how you can deal with them.

One of our readers recommended paying heed to the Espressif IoT Development Framework. He found an error in the project code and asked if the PVS-Studio static analyzer could find it. The analyzer can't detect this specific error so far, but it managed to spot many others. Based on this story and the errors found, we decided to write a classic article about checking an open source project. Enjoy exploring what IoT devices can do to shoot you in the foot.

This is a step by step instruction for creating Node.JS web server application with typescript and express. Creating of new web application on Node.JS is is not making often, so the steps how to do it can be forgotten. So this article is a kind of reminder for me and other developers. Also I added in article links to video clips for more clarification. Here I show how to create application on Node.JS without using any frameworks for it. Just Node.JS and packages.

There’s a lot of talk about machine learning nowadays. A big topic – but, for a lot of people, covered by this terrible layer of mystery. Like black magic – the chosen ones’ art, above the mere mortal for sure. One keeps hearing the words “numpy”, “pandas”, “scikit-learn” - and looking each up produces an equivalent of a three-tome work in documentation.

I’d like to shatter some of this mystery today. Let’s do some machine learning, find some patterns in our data – perhaps even make some predictions. With good old Python only – no 2-gigabyte library, and no arcane knowledge needed beforehand.

Interested? Come join us.

Voice calling plays a crucial role in personal and professional communication. Be it a friendly conversation among classmates or a business deal between companies – voice calling has been the most easy, convenient, and affordable way to communicate for decades. Thus, communication corporations and developers have brought in innovative ways to integrate their Application Programming Interfaces (APIs) and invent newer ways of voice communication.

It may seem that when you are a beginner, you'll do simple things only. No need to learn data structures and algorithms. No need to understand Big O notation, complexity and stuff like that. 

This couldn't be further away from the truth!

In 2008, when I just started learning to program, I spent a lot of time reading books on PHP and MySQL. Months later, when I felt confident, I took my first freelance project. It was a real estate website. A simple one. I used a custom-made ORM and everything worked just fine!

When I released it, the search feature quickly became sluggish and made the website unusable. 

I was wondering what the heck had happened. I figured out that database queries became very slow when there were over 200 real estate objects added to it. 

This is it. What worked fine during testing did not work in real life.

I was a self-taught developer. I did not know how to measure if my project scaled well. I didn't even know that I had to do it.

I thought algorithms mattered only for launching a spaceship.

If I had some basic understanding of algorithms, I would have known that the more the input, the longer it takes. 

I am not saying I would have come up with a robust solution as a junior, but I would have looked for a solution because I knew there would be a problem. 

Please, don't make the same mistake!

Of course, data structures and algorithms are much more than that and they apply differently depending on what you work on.

But a basic understanding of data structures and algorithms is a must for every software developer. 

I'm Ivan Kopenkov, a senior front-end developer at Mail.ru Cloud Solutions. In this article, I will tell you about the approaches we have used for the UI library components customization. You will also learn how to significantly decrease bundle size, cutting off all the unnecessary modules Ant Design takes there.

In our case, we are making wrappers for original Ant Design components inside the project, changing their appearance, and developing their logic. At the same time, we import both customized and original components right from the ant-design module. That saves tree shaking functionality and makes complex library components use our wrappers instead of original nested elements.

If you are already or about to use Ant Design, this article will provide you with a better and more effective way to do so. Even if you have chosen another UI library, you might be able to implement these ideas.

Neural networks (NN) and computer vision models in particular are known to perform well in specific tasks, but often fail to generalize to tasks they have not been trained on. A model that performs well on a food data may perform poorly on satellite images. 

A new model from OpenAI named CLIP claims to close this gap by a large margin. The paper Open AI wrote presenting CLIP demonstrates how the model may be used on a various classification datasets in a zero-shot manner. 

In this article, I will explain the key ideas of the model they proposed and show you the code to use it. 

Currently, there are hardly any high quality / modern / free / public voice activity detectors except for WebRTC Voice Activity Detector (link). WebRTC though starts to show its age and it suffers from many false positives.

Also in some cases it is crucial to be able to anonymize large-scale spoken corpora (i.e. remove personal data). Typically personal data is considered to be private / sensitive if it contains (i) a name (ii) some private ID. Name recognition is a highly subjective matter and it depends on locale and business case, but Voice Activity and Number Detection are quite general tasks.

Key features:

• Modern, portable;
• Low memory footprint;
• Superior metrics to WebRTC;
• Trained on huge spoken corpora and noise / sound libraries;
• Slower than WebRTC, but fast enough for IOT / edge / mobile applications;
• Unlike WebRTC (which mostly tells silence from voice), our VAD can tell voice from noise / music / silence;
• PyTorch (JIT) and ONNX checkpoints;

Typical use cases:

• Spoken corpora anonymization;
• Can be used together with WebRTC;
• Voice activity detection for IOT / edge / mobile use cases;
• Data cleaning and preparation, number and voice detection in general;
• PyTorch and ONNX can be used with a wide variety of deployment options and backends in mind;