The 40s News Express

  • Kuaishou announced that from July, it will cancel “big and small weeks” : it will pay twice as much for working overtime on weekends
  • Bytedance Intern Deletes All Machine Learning Models Below GB: Suspected to have been deleted by mistake
  • The blogger who leaked Mi 10 Exclusive Edition in advance will be paid 1 million RMB
  • Tencent has launched a new “small contract” function to quickly generate legally binding cloud contracts
  • Xiaomi 11 has released a “cool down” patch that can limit the Snapdragon 888’s fever. It finally stops burning!
  • Chinese Academy of Sciences releases domestic open source RISC-V processor Xiangshan
  • Sources say Huawei will build its first chip factory in Wuhan in 2022
  • Microsoft Explains Why Windows 11 Mandates TPM 2.0 Hardware Support
  • Huawei P50 Standard Edition Specification Exposition: or equipped with Snapdragon 888 4G
  • Nintendo Switch Pro processor: NV Orin S, code name Black Knight
  • Google’s new address book is designed to look a lot like Google Plus
  • V language written operating system: Vinix
  • Spring Boot 2.4.8 is released
  • Rocky Linux 8.4 released
  • Linux 5.14 will support the OpenPower Microwatt Soft CPU kernel
  • IntelliJ IDEA 2021.2 EAP 4 was released

Industry information

Kuaishou announced that from July, it will cancel “big and small weeks” : it will pay twice as much for working overtime on weekends

Kuaishou has just announced that starting July 1, it will cancel all working weeks and pay employees overtime wages according to relevant regulations. Internally, Kuaishou referred to the weeks as “Focus Days” and began trial operation in January this year, but decided to cancel it after half a year of trial operation. Weeks are one week off, then two weeks off the next, and so on. Employees will be paid twice as much for working on weekends and three times as much for working on national holidays.

Bytedance Intern Deletes All Machine Learning Models Below GB: Suspected to have been deleted by mistake

Recently, a netizen reported that an intern at Bytedance deleted all the lightweight Lite machine learning models of the company. The owner of Bytedance said that the Lite model is almost all the machine learning models under the size of GB in the company, and all of them have been deleted. What the intern directly deleted was the parent directory and added Skip Trash (temporarily disabling recovery when deleting files) operation, resulting in that the deleted model could not be restored. The whole company announced that night, and was directly listed in the P0 accident grade (serious accident). But it was said that the intern would not be fired “as long as it is not intentional”. Some netizens said that the intern should not have the authority to operate such an important thing, and that 80% of the problem should be handed over to the management. Instead, the intern exposed the problem, and from another point of view, the intern has done a good job.

The blogger who leaked Mi 10 Exclusive Edition in advance will be paid 1 million RMB

Digital blogger Beautiful Technology has issued an apology on Weibo, saying that it violated its confidentiality agreement with Xiaomi by Posting the unbox video of the Mi 10 Extra Commemorative Edition on August 11, 2020. After arbitration by the China International Economic and Trade Arbitration Commission, the company was ordered to pay 1 million RMB to Xiaomi.

Tencent has launched a new “small contract” function to quickly generate legally binding cloud contracts

June 24, Tencent electronic sign announced the “small contract” function on the line. Using the function of small contract, we can quickly generate a standard rental contract. After both parties have completed their personal identity verification and have no objection to the contract information, they can sign this legally effective contract at any time. This will effectively solve the rights and interests disputes that may arise between the two parties. In addition, the landlord will be able to issue a legally valid receipt for the tenant during the tenancy period through a “small receipt” function.

Xiaomi 11 has released a “cool down” patch that can limit the Snapdragon 888’s fever. It finally stops burning!

The Mi 11 is the world’s first model to be powered by the Snapdragon 888, and it’s because the Snapdragon 888 chip itself doesn’t do a very good job of suppressing the fever issue that’s where the jokes come in. As a matter of fact, the Mi 11 does have a certain fever problem. Although the sales of this phone has been considerable since its release, the fever problem has still been ridiculed by many netizens. But the Mi 11 fever problem has finally been solved. Based on the feedback from the Xiaomi community, it seems that the patch significantly limits the performance release of the Snapdragon 888 chip, with daily experience experiencing frame-drop lags and even games that can only use 60Hz refresh rate.

Chinese Academy of Sciences releases domestic open source RISC-V processor Xiangshan

At the first RISC-V China Summit held in Shanghai University of Science and Technology this week, Bao Yungang, a professor at the Chinese Academy of Sciences and a researcher at the Institute of Computing Technology of the Chinese Academy of Sciences, unveiled Xiangshan, a domestic open source high-performance RISC-V processor core. He says there is currently no open-source mainline in the CPU space like Linux, so the team decided that the industry needed an open source high-performance RISC-V core that could be widely used by industry while supporting academic experimentation with innovative ideas, with the goal of surviving for at least 30 years like Linux. Xiangshan is under the support of the Institute of Computing Technology of the Chinese Academy of Sciences and Pengcheng Laboratory, through the China Open Directive Ecological (RISC-V) Alliance, the industry enterprises together to develop an open source high-performance RISC-V processor core, the first mass-production version “Yanqi Lake” is scheduled to be released in July this year, using TSMC 28nm process.

Sources say Huawei will build its first chip factory in Wuhan in 2022

By 2022, Huawei will build its first chip factory in Wuhan. With the change of external environment, Huawei has to make many changes, among which the most important one is to produce its own chips. Back in April, we learned that the company was looking for talent for the project. In May, we learned that its subsidiary, HiSilicon, would continue to exist, but had instead begun developing its first processor under an open RISC-V architecture because it was not allowed to use the ARM architecture.

At that time, Huawei provided a development board HI3861 to HongMeng developers, which caused concern, although Heis did not disclose the architecture of the main chip used in this development board. However, there is mention of using GCCriscv32, so industry insiders speculate that it should be a development board based on RISC-V architecture.

Unfortunately, the information available to us at this time is very scarce. Sources from Digitimes in Taiwan have reported only a few words. All we know is that the foundry will be built in Wuhan, Hubei Province of China. So we don’t know what the wafer specifications and process were when we started.

If it can build its own factory and manufacture silicon wafers, and if open architecture-based chip development goes well, Huawei will be able to bypass the restrictions and re-launch its HiSilicon chip, and could potentially re-enter the mobile devices, desktop and server, 5G and other business markets.

Microsoft Explains Why Windows 11 Mandates TPM 2.0 Hardware Support

Many early adopters of Microsoft’s Windows 11 operating system this week saw a mandatory prompt for the Trusted Security Module (TPM) when they ran the System Update Inspector. This means that Microsoft is focusing more on the security improvements that Windows 11 can bring, in addition to a new user interface (UI), Android application support (WSA subsystem), and stylus interaction. So support for TPM 2.0 is not negotiable.

TPM is a chip that can store encryption keys and account credentials. It is usually associated with a PC motherboard or processor, but not all system users in recent years have chosen to enable TPM.

Another big bone of contention is Microsoft’s vague description of “Windows 11 compatibility.” That was until David Weston, the company’s director of enterprise and operating system security, posted a new post on the Windows Security Blog:

Future PCs will need this root-of-trust of modern hardware to help defend against common and sophisticated attacks, such as ransomware and sophisticated attacks by heavyweight hacker groups.

The TPM 2.0 requirement for Windows 11 will also help improve hardware security standards by enforcing built-in roots of trust.

Based on this, by default, any Windows 11-certified system will include the TPM 2.0 chip.

It will be part of a more comprehensive set of security tools, including the ability to use Windows Hello and support for hardware-level enforced stack protection on Intel/AMD systems that can be used right out of the box.

David Weston added that the new Windows operating system’s security requirements for new hardware are designed to build a more robust and resilient foundation for attacks on authenticated devices.

We know how effective this approach can be — PCs with a secure core are twice as resistant to malware infection.

What is known is that OEMs, including Acer, Asustek, Dell, HP, Lenovo, Panasonic and others, will provide solid support for TPM 2.0 for the new system. It remains to be seen how the consumer PC DIY market will evolve.

Huawei P50 Standard Edition Specification Exposition: or equipped with Snapdragon 888 4G

As usual, Huawei was due to launch its new flagship phone, the P50 series, in April. However, due to some uncontrollable reasons, the event has been repeatedly delayed. According to relevant Revelations, Huawei P50 series will be officially unveiled at the end of July, as the press conference is approaching, the recent Revelations about the aircraft is also emerging in an endless stream, this afternoon the blogger @Bald Ran Panda brought the latest news.

It is said that Huawei P50 Standard Edition will adopt the center perforated face screen scheme, which is also the most popular screen design in the market. The screen size is 6.3 inches, the body circumference is 156.7*74*8.3mm, and the weight is only 187G. It is a very thin flagship phone.

It is worth noting that Huawei P50 and P50 Pro will mainly promote 4G version, and the stock of 5G version may be relatively small. As for the top-equipped version of Huawei P50 Pro+, the release may be delayed, and it will be separately launched after the P50/P50 Pro two models are first launched.

The 4G version of Huawei’s P50 is likely to be powered by the Snapdragon 888 4G version. Previously, it was reported that Qualcomm has restored the supply of chips to Huawei, but only for 4G, while 5G chips have not yet been approved.

As for the 4G version of the Snapdragon 888, it has not been officially launched to the market yet. The source said that the chip is mainly aimed at countries and regions that are lagging behind in 5G development, but Huawei will also become a major customer of the Snapdragon 4G chip due to the 5G limitations of Huawei phones.

Nintendo Switch Pro processor: NV Orin S, code name Black Knight

Nintendo didn’t announce the Switch Pro at its E3 conference on Tuesday, but it’s likely that the device is making its final push. The latest update from source Kopite7kimi suggests that the new processor (SoC) chip on the Switch Pro is code-named Black Knight.

He cited an earlier tweet about Switch Pro using the T239 version of the Orin chip, which seems to be further confirmation.

Nvidia Orin was released as a car self-driving chip, which supports L2 level assisted driving of the most entry version power consumption of 15 watts, it is speculated that the high probability of Switch Pro based on this magic change.

NBC reports that Nintendo uses the Orin S chip. It’s not clear whether the 12-core Cortex-A78e CPU will be fully powered, but it will definitely be more powerful than the existing Tegra X1+ (4X Cortex-A57 + 4X Cortex-A53). Moreover, the integration is the AMPERE architecture GPU unit.

Of course, until all is said and done, take this revelation with a grain of salt, as Nintendo has never been about piling up hardware to attract players.

Google’s new address book is designed to look a lot like Google Plus

As Microsoft launches its next operating system, Google also launches a new version of its address book. According to the company, the new Google contact book experience, which includes additional information in this circling style, will help Google Workspace users get to know their colleagues better. While the personal version of the Google Plus social platform has been abandoned, Google seems to want to improve the experience for enterprise customers.

As shown in the figure, the new Google Address Book displays a chart of your organization, as well as a history of contacts with specific people (including emails, meetings, etc.), as part of the Address Book UI.

With a clean, organized look, the new Google address book is designed to help you “easily get to know your colleagues and other stakeholders”. But people who have used Google Plus may recall the bad experience of the early 2010s.

At the time, Google forced Google+ to sort and provide information to everyone, but spam incidents like “emails from Google+” left a big negative impression.

Thankfully, the Google address book update is unlikely to be pushed to individual users in the coming weeks. Instead, it will only be available to users of Google Workspace and G Suite productivity suites at various subscription levels.

Latest Technology Trends

Language written for the operating system: Vinix

V language development team to release (https://github.com/vlang/vinix), the first operating system written in V – Vinix, and said that the operating system can now run mlibc and bash.

Vinix’s source code is hosted on GitHub under the GPLv2 open source license, and its repo shows that it is written in V and aims to be a modern, fast, and useful operating system.

Vinix’s tentative goals are as follows:

  • Make sure your code is as straightforward as possible
  • Try to write it in V
  • Make a usable operating system that runs on real hardware, not just an emulator
  • Oriented to modern 64-bit architectures and CPU features
  • Maintains good source-level compatibility with Linux for porting programs

The reasons given by the development team for creating Vinix:

  • Explore V’s ability to program in bare metal
  • Improve the compiler by providing feedback for the unusual requirements of bare-metal programming
  • For the fun of it

According to Vinix’s README, you currently have to have Docker installed and running in order to build Vinix, which means there is no support for installing Vinix directly to your computer.

V is a static language that combines Go’s simplicity with Rust’s security features. The authors say that V is very similar to Go, and that if you know Go, you already know about 80% of V.

V Improvements on GO:

https://vlang.io/compare#go

V Main characteristics

  • Simple (author claims to be able to learn V in less than an hour)
  • Fast compilation (the compiler is only 400KB and has no third party dependencies)
  • Easy to develop: V compiles in less than a second
  • Security: No NULL, no global variables, no undefined values, boundary detection, default Immutable structure
  • Support C/C++ conversion
  • Easy to use cross – compile
  • Provide a cross-platform UI library
  • Built-in graphics library
  • Built-in ORM
  • Built-in Web Framework
  • …

Spring Boot 2.4.8 is released

Spring Boot 2.4.8 is now available. This release includes 33 bug fixes, documentation improvements, and dependency upgrades.

Major Updates

  • EnvironmentPostProcessors correct this not used for instantiation
  • Bind the name of a nonexistent Class to Class
    attribute failed
  • Configuration file-specific parsing should still occur when processing the “spring.config.import” attribute
  • If the kotlin.version extra attribute is set, the version of the Kotlin plug-in should not be read
  • Disable automatic timing MetricsClientHttpRequestInterceptor of memory leaks
  • In depends on the org. Springframework. The boot: spring – the boot – configuration – processor used in the project configuration cache, Gradle build failures and display “does not support when executed call ‘Task. The project’
  • When you use Flyway or LiquiBase with spring.flyway.url or spring.liquiBase. url and no spring-JDBC on your classpath, the application will not start
  • BeanNotOfRequiredTypeFailureAnalyzer report type of misleading information
  • DevTools’s Livereload supports HTTP header handling that is case sensitive
  • Dependence on repackaging cannot be ruled out
  • The spring.datasource. Name property is not correctly described
  • RandomValuePropertySource uuid support of the document
  • Update the reference documentation to reflect @SpringBootApplication’s use of @SpringBootConfiguration
  • Make document versioning more obvious

Rocky Linux 8.4 released

Rocky Linux 8.4 is now available. Rocky Linux is a community version of the Enterprise operating system designed to be 100% bug-for-bug compatible with Red Hat Enterprise Linux 8.4. Officially, since this is the first release of Rocky Linux, the release notes only reflect changes in upstream functionality from one release to another. Migration from Rocky Linux 8.3RC1, Rocky Linux 8.4RC1, or any other candidate to Rocky Linux 8.4 is not supported.

The Rocky Linux team provides the Migrate2Rocky tool to help users migrate from other enterprise Linux systems to Rocky Linux 8.4. This tool has been tested and works, but you use it at your own risk.

A new module

The new Module Streams in Rocky Linux 8.4 includes the following:

  • Python 3.9
  • SWIG 4.0
  • Subversion 1.14
  • Redis 6
  • PostgreSQL 13
  • Mar

Major changes

The main changes to Rocky Linux 8.4 are in the areas of security, networking, kernel and high availability, and clustering.

security

  • IPSec VPN provided by Libreswan now supports TCP encapsulation and security tags for IKEV2
  • SCAP-Security-Guide has been updated to 0.1.54 and OpenSCAP has been updated to 1.3.4. These updates provide substantial improvements, including optimized memory management
  • The FEAPOLICYD framework now provides integrity checking, and the RPM plug-in registry is updated by either the YUM package manager or the RPM package manager

network

  • Fully supported Nmstate (the host’s network API), these Nmstate packages provide a library and nmstatectl command line to declaratively manage the host’s network Settings
  • Support for MPLS (Multi-Protocol Label Switching)
  • Iproute2 introduces three new flow control (TC) operations: MAC \_push, push\_eth, and pop_eth, and adds MPLS tags

The kernel

  • Active Compression: Periodically start memory compression before an allocation request is made. As a result, latency for specific memory allocation requests is reduced.
  • A tablet memory controller for control group technology is provided. The Slab memory controller optimizes memory utilization and is able to move memory accounting from the page level to the object level. As a result, you can observe a significant decrease in total kernel memory usage and an improvement in memory fragmentation.
  • Temporal namespace functionality: This functionality is useful for changing the date and time within the Linux container. It is now also possible to adjust the clock in the container after the checkpoint is restored.
  • Support error detection and correction (eDAC) kernel modules set in 8th and 9th generation Intel Core processors.

High availability and clustering

  • Maintaining persistence of state data: The Pacemaker resource agent can detect faults asynchronously and inject faults into Pacemaker immediately, without waiting for the next monitoring interval. Persistent resource brokers can also speed up cluster response times for services with high state overhead, because maintaining state data reduces the state overhead of cluster operations (such as start, stop, and monitor) by not invoking state separately for each operation.

Compilers and development tools

The following compiler toolsets have been updated:

  • GCC Toolset 10
  • LLVM Toolset 11.0.0
  • Rust Toolset 1.49.0
  • Go Toolset 1.15.7

Identity management

  • Rocky Linux 8.4 provides Ansible modules for automated management of role-based access control (RBAC) in identity management (IDM), an Ansible Role for backing up and restoring IDM servers, and an Ansible module for location management.

Download Address:

https://rockylinux.org/download/

Linux 5.14 will support the OpenPower Microwatt Soft CPU kernel

According to the latest submissions, the upcoming Linux 5.14 release will add support for Microwatt.

Microwatt ISA small Open Power ISA soft core based on FPGA written in VHDL 2008, launched in 2019. It conforms to the Power ISA 3.0 instruction set, can run on a variety of FPGA hardware, and was the first processor to be written from scratch using the open Power ISA 3.0 specification, and became one of the reference designs for that specification. Last month, the kernel was also selected by OpenPower for the Efabless Open MPW Shuttle program.

Currently, Microwatt platform support, Microwatt device tree, and various other patches for enabling OpenPower Microwatt support have been submitted in the PowerPC-Next tree of the Linux kernel. This support is sufficient to handle an FPGA implementation of Microwatt. However, the patches also emphasize that the platform is still in its early days and does not yet have any of the features of SMP, VMX, VSX, transactional memory, or other mature POWER hardware.

If nothing unexpected, Linux 5.14 will add support for Microwatt.

IntelliJ IDEA 2021.2 EAP 4 was released

IntelliJ IDEA 2021.2 EAP 4 is now available. This release brings a number of feature updates and bug fixes, including a new Prot-Wide Analysis, as well as some nice new Java and Kotlin checks and intent actions.

It is important to note that starting with the EAP version, you must have a JetBrains account and log in to participate in IntelliJ IDEA Ultimate. If you are already logged in with your credentials, then you have automatic access to the EAP build. If you’re not already logged in, the IDE will redirect you to Account.JetBrains.com, where you can enter your login and password or create a new JetBrains account.

Project-wide analysis for Java projects

New features in IntelliJ IDEA help you track all errors throughout your Java project. Project-Wide Analysis can be performed in the Problems tool window. Simply click the new icon in the upper left corner of the window. One advantage of doing this is that you can start using Project-Wide Analysis before you even compile the project.

When you click on the icon, IntelliJ IDEA will start examining your project and then show you any problems that have occurred, including those that would normally only be detectable by scanning the entire project. The first examination may take longer than subsequent examinations.

IntelliJ IDEA shows a progress bar at the bottom, but you can click to expand it into a dialog box.

If you need to stop the analysis, simply click the icon that starts the process again. After the IDE completes the inspection, the results are displayed in the Problems tool window.

This feature is only suitable for small and medium-sized projects.

Java

Improved user experience during inspection review by adding relatively simple but effective checks and quick fixes.

Floating-point range tracking

Data flow analysis allows you to track the range of float and double values. IntelliJ IDEA can now use doubles and float. isNan statistics to validate conditions.

Update the collection size trace for the method

Another useful update to data flow analysis is that it now tracks the size of a collection after it has been updated by the following methods: Add, AddAll, Put, PutAll, Remove, RemoveAll, and RetainAll. If any errors are detected, the check highlights them.

Write object checks only

IntelliJ IDEA provides several checks that warn you when you modify an object but never query it, for example, a query and update mismatch for a collection or a read and write mismatch check for an array. Write-only object checks perform the same checks on your project and on some of the custom classes defined in the library. For example, it supports AtomicReference.

Quick fixes that address compilation errors like this

Added a new quick fix that you can apply to all the same compile errors in files with one click.

Improved overlay mapping, collection, or array element checking

Checking for handling overwritten keys or elements now supports local variables. If there is an overlay due to a switch fault, IntelliJ IDEA will detect it.

The same checks now apply to overridden array elements and highlight duplicate array element allocations.

Unused Assignment Inspection is also extended to detect overwritten fields.

Kotlin

Check to simplify the chain of calls for collection types

Now, when IntelliJ IDEA calls a method in a chain within a collection, it is possible to simplify the syntax and combine multiple calls into one.

The intended action for the buildString transform

Other checks and intention actions

There are also new checks and intent actions that now apply to Kotlin and help resolve specific cases: [KTIJ-6994], [KTIJ-8888], [KTIJ-7615], [KTIJ-1045], [KTIJ-1045].

.NET Maui Preview5 released

.NET Maui Preview 5 has been released. In this release, the development team enabled animation and View Transformation, completed the migration of multiple UI components, and improved a single project template. In addition, the first batch of preview documents covering the introduction and basics of.NET Maui have been released:

https://docs.microsoft.com/

animation

Currently,.NET MAU provides several ways to perform animations, the simplest of which is using view extension methods, such as fadeTo, RotateTo, scaleTo, translateTo, and so on. In the following example, get a reference to each view that is bound to the layout by using the new HandlerAttached event:

<DataTemplate x:Key="FavouriteTemplate">
    <Frame
        AttachedHandler="OnAttached"
        Opacity="0">
        ...
    </Frame>
</DataTemplate>
<FlexLayout
    BindableLayout.ItemTemplate="{StaticResource FavouriteTemplate}"
    BindableLayout.ItemsSource="{Binding Favorites}"
    >
    ...
</FlexLayout>

As the page appears, the view is animated in a slightly staggered fashion to create a beautiful cascading effect.

public partial class FavoritesPage : ContentPage { List<Frame> tiles = new List<Frame>(); void OnAttached(object sender, EventArgs e) { Frame f = (Frame)sender; tiles.Add(f); } protected override async void OnAppearing() { base.OnAppearing(); await Task.Delay(300); TransitionIn(); } async void TransitionIn() { foreach (var item in tiles) { item.FadeTo(1, 800); await Task.Delay(50); }}}

UI components

In this release, all the properties and events of the multiple controls have been migrated to the Xamarin.Forms renderer architecture’s handlers, including the ActivityIndicator, Checkbox, Image, and Stepper. In previous previews, developers had to check whether controls had been migrated and register renderers from compatibility packages for renderers that were not available. In.NET Maui Preview 5, this is made easier by updating the UsemauiApp extension to connect all the controls for developers, whether they are handler based or renderer based.

Another new feature in Preview 5 is the introduction of the Shell for the first time, which is an application container that provides URI navigation and a quick way to implement pop-up menus and tabs.

Individual project template updates

The development team made progress in this release by merging multiple WinUI projects into one. Now, when creating a project (DotNet New Maui), the developer will see two projects: the multi-objective.NET Maui project and the WinUI project.