Home Blog Page 3882

How AI Can Enhance Community Safety


Community managers are about so as to add a robust new safety expertise to their toolbox: synthetic intelligence.

AI guarantees the power to research huge volumes of information with out having a human round, says Shing-hon Lau, senior AI safety researcher at Carnegie Mellon College’s Software program Engineering Institute, in an e-mail interview. “AI might be focused towards automating routine processes, releasing people to research advanced threats, or by making an attempt to detect novel threats by quickly analyzing massive portions of information.”

The first differentiator AI brings to the desk is the power to research massive quantities of information and establish occasion patterns and correlations of that merely aren’t possible for a human, says Marc Herren, a director at expertise advisory and analysis agency ISG, through e-mail. “Over time, AI can acknowledge the beginning of a sample and supply predictive alerts permitting community and safety directors to take corrective actions earlier than an incident occurs.”

Why AI is an effective match

Resulting from the truth that community safety entails numerous information, AI is especially well-suited for important infrastructure safety. “When you could have massive quantities of information, you possibly can have a excessive price of false positives,” says Sushila Nair, vice chairman and head of the North American cybersecurity follow at enterprise advisory agency Capgemini, in an e-mail interview.

AI will also be utilized in community segmentation, isolating vital networks and lowering an assault’s the general blast radius. “This strategy is a crucial step in a zero-trust framework,” Nair says. AI-powered micro-segmentation offers organizations with segmentation suggestions based mostly on community visitors. It accomplishes this process by detecting overly permissive guidelines in order that firms can phase based mostly on software utilization and asset criticality, which helps in limiting an assault’s scope.

AI on the edge

AI can analyze community visitors utilizing deep packet inspection at a velocity that human analysts merely cannot match. “AI can crunch by means of the volumes of information quickly to establish each exterior and inner threats,” Nair says. “Since community information tends to be particularly voluminous, AI is pivotal for this use case.”

Understanding community visitors helps community managers detect threats. “It means that you can set up governance over the kind of visitors that is going by means of your community,” Nair says. “A cloud entry safety dealer (CASB), for example, means that you can perceive if there’s visitors going to high-risk SaaS purposes or if there’s information exfiltration.” In the meantime, AI-powered micro-segmentation instruments permit community adjustments to be made simply and rapidly.

A rising variety of safety and community expertise distributors at the moment are incorporating AI into their merchandise to supply anomaly and risk detection capabilities. Enterprises ought to reap the benefits of such options, Herren says. “For a holistic resolution, organizations must also think about impartial AI instruments that may combine throughout a number of vendor options and community layers and segments.”

An built-in resolution

AI-powered community detection and response (NDR) options are notably helpful since they’ll detect malware, hidden assaults, and anomalies throughout networks, Nair says. Mixed with safety info and occasion administration (SIEM) and safety orchestration, automation, and response (SOAR), an NDR might be a part of an built-in resolution powered by AI that not solely detects assaults however can allow automated responses.

Dynamic microsegmentation is vital for holding cyberattacks by stopping attackers from transferring laterally, Nair says. “CASBs can use network-driven AI to allow visibility and management over information in addition to customers in cloud purposes,” she explains. “You possibly can, subsequently, achieve an understanding of who’s utilizing what SaaS software and if that software is excessive danger or unsanctioned by the group.”

Getting began with AI community safety

One of the simplest ways to get began with AI community safety is by utilizing the expertise to complement current instruments and processes. “AI takes time and many information to turn into efficient, and engineers have to refine and validate AI fashions earlier than a corporation might be assured within the accuracy of the analytics,” Herren says.

Perceive your key belongings and your dangers, Nair advises. “Set the objectives for what you need to obtain with AI-driven community safety so that you perceive the way it will add worth to the enterprise,” she recommends. “Is it getting used for quicker detection occasions, quicker response occasions, or is there one other enterprise driver?”

Nair suggests starting with a pilot challenge. “Combine it together with your cybersecurity ecosystem,” she says. “Relying on the use case, you could want to combine it together with your SIEM and SOAR resolution or different cybersecurity instruments,” Nair suggests. AI-powered instruments typically require some fine-tuning. “Overview the outcomes towards your challenge objectives and plan for enlargement if the pilot outcomes achieved the said objectives.”

The human contact

AI is not foolproof, and human judgment will proceed to enhance AI expertise in community administration, Nair says. In its present state, AI nonetheless requires a human to observe and fine-tune operations. “The AI mannequin consumes massive quantities of information, a few of it with privateness implications, so guarantee you could have governance and assurance course of for introducing AI into your surroundings,” she recommends. “As your community grows, you could have to improve your AI system because the quantity of information requiring evaluation will enhance.”

A basic understanding of AI expertise is vital to assist dispel the hype and fanfare, Lau says. “AI is not magic—it is a instrument that is appropriate to be used in some circumstances,” he observes. “Determining whether or not your particular circumstance is one the place AI may help requires a minimum of a base stage of understanding about how AI works.”

Associated articles:



Turing-Full-RAG (TC-RAG): A Breakthrough Framework Enhancing Accuracy and Reliability in Medical LLMs By Dynamic State Administration and Adaptive Retrieval


The sector of huge language fashions (LLMs) has quickly developed, significantly in specialised domains like drugs, the place accuracy and reliability are essential. In healthcare, these fashions promise to considerably improve diagnostic accuracy, remedy planning, and the allocation of medical assets. Nonetheless, the challenges inherent in managing the system state and avoiding errors inside these fashions stay important. Addressing these points ensures that LLMs will be successfully and safely built-in into medical observe. As LLMs are tasked with processing more and more advanced queries, the necessity for mechanisms that may dynamically management and monitor the retrieval course of turns into much more obvious. This want is especially urgent in high-stakes medical situations, the place the results of errors will be extreme.

One of many major points going through medical LLMs is the necessity for extra correct and dependable efficiency when coping with extremely specialised queries. Regardless of developments, present fashions often wrestle with points reminiscent of hallucinations—the place the mannequin generates incorrect data—outdated data, and the buildup of faulty knowledge. These issues stem from missing strong mechanisms to regulate and monitor retrieval. With out such mechanisms, LLMs can produce unreliable conclusions, which is especially problematic within the medical subject, the place incorrect data can result in critical penalties. Furthermore, the problem is compounded by the dynamic nature of medical data, which requires techniques that may adapt and replace constantly.

Numerous strategies have been developed to deal with these challenges, with Retrieval-Augmented Era (RAG) being one of many extra promising approaches. RAG enhances LLM efficiency by integrating exterior data bases and offering the fashions with up-to-date and related data throughout content material era. Nonetheless, these strategies usually fall quick as a result of they should incorporate system state variables. These variables are important for adaptive management, making certain the retrieval course of converges on correct and dependable outcomes. A mechanism to handle these state variables is important to keep up the effectiveness of RAG, significantly within the medical area, the place choices usually require intricate, multi-step reasoning and the power to adapt dynamically to new data.

Researchers from Peking College, Zhongnan College of Economics and Legislation, College of Chinese language Academy of Science, and College of Digital Science and Expertise of China have launched a novel Turing-Full-RAG (TC-RAG) framework. This method is designed to deal with the shortcomings of conventional RAG strategies by incorporating a Turing Full strategy to handle state variables dynamically. This innovation permits the system to regulate and halt the retrieval course of successfully, stopping the buildup of faulty data. By leveraging a reminiscence stack system with adaptive retrieval and reasoning capabilities, TC-RAG ensures that the retrieval course of reliably converges on an optimum conclusion, even in advanced medical situations.

The TC-RAG system employs a complicated reminiscence stack that displays and manages the retrieval course of via actions like push and pop, that are integral to its adaptive retrieval and reasoning capabilities. This stack-based strategy permits the system to selectively take away irrelevant or dangerous data selectively, thereby avoiding the buildup of errors. By sustaining a dynamic and responsive reminiscence system, TC-RAG enhances the LLM’s capability to plan and purpose successfully, much like how medical professionals strategy advanced circumstances. The system’s capability to adapt to the evolving context of a question and make real-time choices primarily based on the present state of information marks a big enchancment over present strategies.

In rigorous evaluations of real-world medical datasets, TC-RAG demonstrated a notable enchancment in accuracy over conventional strategies. The system outperformed baseline fashions throughout varied metrics, together with Actual Match (EM) and BLEU-4 scores, exhibiting a median efficiency acquire of as much as 7.20%. As an example, on the MMCU-Medical dataset, TC-RAG achieved EM scores as excessive as 89.61%, and BLEU-4 scores reached 53.04%. These outcomes underscore the effectiveness of TC-RAG’s strategy to managing system state and reminiscence, making it a robust software for medical evaluation and decision-making. The system’s capability to dynamically handle and replace its data base ensures that it stays related and correct, at the same time as medical data evolves.

In conclusion, the TC-RAG framework addresses key challenges reminiscent of retrieval accuracy, system state administration, and the avoidance of faulty data; TC-RAG gives a sturdy answer for enhancing the reliability and effectiveness of medical LLMs. The system’s revolutionary use of a Turing Full strategy to handle state variables dynamically and its capability to adapt to advanced medical queries set it aside from present strategies. As demonstrated by its superior efficiency in rigorous evaluations, TC-RAG has the potential to change into a useful software within the healthcare trade, offering correct and dependable help for medical professionals in making important choices.


Take a look at the Paper. All credit score for this analysis goes to the researchers of this mission. Additionally, don’t neglect to observe us on Twitter and be part of our Telegram Channel and LinkedIn Group. In case you like our work, you’ll love our publication..

Don’t Overlook to hitch our 49k+ ML SubReddit

Discover Upcoming AI Webinars right here


Asif Razzaq is the CEO of Marktechpost Media Inc.. As a visionary entrepreneur and engineer, Asif is dedicated to harnessing the potential of Synthetic Intelligence for social good. His most up-to-date endeavor is the launch of an Synthetic Intelligence Media Platform, Marktechpost, which stands out for its in-depth protection of machine studying and deep studying information that’s each technically sound and simply comprehensible by a large viewers. The platform boasts of over 2 million month-to-month views, illustrating its recognition amongst audiences.



South Africa Celebrates 150 Days With out Loadshedding. New Utility-Scale Photo voltaic PV Public sale Oversubscribed


Join every day information updates from CleanTechnica on electronic mail. Or observe us on Google Information!


South Africans had been going through rising electrical energy rationing (loadshedding) over the previous couple of years. The consequences of loadshedding on houses and companies have been brutal, and the South African Reserve Financial institution says that throughout the greater phases of load shedding, the place shoppers can expertise 12 hours of load shedding per day, South Africa was shedding as much as R900 million ($50 million) per day. There’s some superb information for South Africans, although. At the moment, Eskom, South Africa’s nationwide electrical energy firm, introduced a big achievement of 150 consecutive days (equal to 5 months) with out loadshedding.

This exceptional milestone displays the profitable implementation of Eskom’s Technology Restoration Plan, which started in March 2023. Eskom says the restoration plan has not solely made Eskom’s energy stations extra dependable, however has additionally demonstrated enhanced effectivity in its operations. Consequently, there was a exceptional discount in diesel expenditure, roughly R10.60 billion, which is 75% much less in comparison with the identical time final yr. Eskom maintained a mean Power Availability Issue (EAF) of 68% over the previous seven days, with the best-performing stations—Kusile, Kendal, Majuba, Lethabo, and the Peaking stations—recording an EAF of 70%. Moreover, three extra energy stations have achieved an EAF above 60%. This enchancment is because of the continued advantages of accelerating and executing deliberate upkeep, partnering with the unique gear producers (OEMs), and the dedication of energy station managers and their groups.

Eskom says its operational effectivity continues to surpass its winter expectations, with present unplanned outages averaging between 9,800 MW and 10,900 MW over the previous seven days. At the moment’s determine is 10,482 MW, which is 5,018 MW decrease than the winter 2024 forecast. This ends in an accessible technology capability of 31,963MW, whereas yesterday night’s peak forecast (23 August 2024) was 26,420MW.

Abstract of bids courtesy of South Africa’s Division of Minerals and Power.

The intensive loadshedding skilled by South Africans led to a rush by houses and enterprise to put in rooftop photo voltaic, backed by battery storage generally. South African houses and companies have added 3,526 MW of rooftop photo voltaic in simply two years! It’s simply fantastic to see how briskly electrical energy technology capability may be added from rooftop photo voltaic. Based on information from South Africa’s nationwide electrical energy utility firm, Eskom, there have been about 2,264.5 MW of rooftop photo voltaic PV put in in South Africa as of July 2022. Based on Eskom’s newest replace, South African houses and companies have now put in 5,790.5 MW of photo voltaic PV. Including the contribution from rooftops and different distributed photo voltaic vegetation, the full put in renewable capability in South Africa now stands at 12,221 MW! This progress in distributed photo voltaic PV, plus the two.3 GW of utility-scale PV, is now contributing to South Africa’s personal duck curve.

There’s extra excellent news. For South Africa’s seventh Bid Window Beneath The Renewable Power Impartial Energy Producer Procurement Programme (REIPPPP BW7), a complete of 8,526 MW of bids for photo voltaic PV have been obtained in opposition to a goal of 1,800 MW. For wind, 1,692 MW of bids have been obtained in opposition to goal of three,200 MW. It might be tremendous cool if all of the bids obtained for photo voltaic and wind might be accommodated throughout the adjudication and funding for all organized if potential. That might supercharge South Africa’s wind and photo voltaic put in capability to 22.4 GW! That’s excluding all the extra rooftop photo voltaic from houses and that might be deployed between now and finalisation of the REIPPPP BW7. A number of components would have to be thought-about and brought care of to accommodate all that photo voltaic and wind obtained in REIPPPP BW7, resembling availability of grid capability for interconnection in addition to funding of tasks outdoors the targets. Could be tremendous cool although if it occurred.


Have a tip for CleanTechnica? Need to promote? Need to recommend a visitor for our CleanTech Discuss podcast? Contact us right here.


Newest CleanTechnica.TV Movies

Commercial



 


CleanTechnica makes use of affiliate hyperlinks. See our coverage right here.

CleanTechnica’s Remark Coverage




ios – Utilizing Auto Structure in UITableView for dynamic cell layouts & variable row heights


TL;DR: Don’t love studying? Soar straight to the pattern tasks on GitHub:

The primary 2 steps under are relevant no matter which iOS variations you’re creating for.

1. Set Up & Add Constraints

In your UITableViewCell subclass, add constraints in order that the subviews of the cell have their edges pinned to the sides of the cell’s contentView (most significantly to the highest AND backside edges). NOTE: do not pin subviews to the cell itself; solely to the cell’s contentView! Let the intrinsic content material dimension of those subviews drive the peak of the desk view cell’s content material view by ensuring the content material compression resistance and content material hugging constraints within the vertical dimension for every subview will not be being overridden by higher-priority constraints you’ve gotten added. (Huh? Click on right here.)

Keep in mind, the thought is to have the cell’s subviews related vertically to the cell’s content material view in order that they’ll “exert strain” and make the content material view develop to suit them. Utilizing an instance cell with a number of subviews, here’s a visible illustration of what some (not all!) of your constraints would wish to appear to be:

Example illustration of constraints on a table view cell.

You’ll be able to think about that as extra textual content is added to the multi-line physique label within the instance cell above, it might want to develop vertically to suit the textual content, which can successfully drive the cell to develop in top. (After all, you could get the constraints proper to ensure that this to work accurately!)

Getting your constraints proper is unquestionably the hardest and most necessary half of getting dynamic cell heights working with Auto Structure. When you make a mistake right here, it may stop every little thing else from working — so take your time! I like to recommend organising your constraints in code as a result of you already know precisely which constraints are being added the place, and it is lots simpler to debug when issues go mistaken. Including constraints in code might be simply as simple as and considerably extra highly effective than Interface Builder utilizing format anchors, or one of many implausible open supply APIs accessible on GitHub.

  • When you’re including constraints in code, you need to do that as soon as from inside the updateConstraints technique of your UITableViewCell subclass. Word that updateConstraints could also be known as greater than as soon as, so to keep away from including the identical constraints greater than as soon as, be certain to wrap your constraint-adding code inside updateConstraints in a verify for a boolean property similar to didSetupConstraints (which you set to YES after you run your constraint-adding code as soon as). However, you probably have code that updates present constraints (similar to adjusting the fixed property on some constraints), place this in updateConstraints however exterior of the verify for didSetupConstraints so it may run each time the tactic is known as.

2. Decide Distinctive Desk View Cell Reuse Identifiers

For each distinctive set of constraints within the cell, use a singular cell reuse identifier. In different phrases, in case your cells have a couple of distinctive format, every distinctive format ought to obtain its personal reuse identifier. (An excellent trace that you could use a brand new reuse identifier is when your cell variant has a special variety of subviews, or the subviews are organized in a definite trend.)

For instance, in case you have been displaying an e mail message in every cell, you might need 4 distinctive layouts: messages with only a topic, messages with a topic and a physique, messages with a topic and a photograph attachment, and messages with a topic, physique, and picture attachment. Every format has utterly completely different constraints required to attain it, so as soon as the cell is initialized and the constraints are added for one in all these cell sorts, the cell ought to get a singular reuse identifier particular to that cell kind. This implies while you dequeue a cell for reuse, the constraints have already been added and are able to go for that cell kind.

Word that on account of variations in intrinsic content material dimension, cells with the identical constraints (kind) should still have various heights! Do not confuse basically completely different layouts (completely different constraints) with completely different calculated view frames (solved from similar constraints) on account of completely different sizes of content material.

  • Don’t add cells with utterly completely different units of constraints to the identical reuse pool (i.e. use the identical reuse identifier) after which try and take away the outdated constraints and arrange new constraints from scratch after every dequeue. The inner Auto Structure engine is just not designed to deal with massive scale modifications in constraints, and you will notice huge efficiency points.

For iOS 8 – Self-Sizing Cells

3. Allow Row Peak Estimation

To allow self-sizing desk view cells, you could set the desk view’s
rowHeight property to UITableViewAutomaticDimension. You will need to additionally
assign a worth to the estimatedRowHeight property. As quickly as each of
these properties are set, the system makes use of Auto Structure to calculate the
row’s precise top

Apple: Working with Self-Sizing Desk View Cells

With iOS 8, Apple has internalized a lot of the work that beforehand needed to be carried out by you previous to iOS 8. As a way to permit the self-sizing cell mechanism to work, you could first set the rowHeight property on the desk view to the fixed UITableView.automaticDimension. Then, you merely must allow row top estimation by setting the desk view’s estimatedRowHeight property to a nonzero worth, for instance:

self.tableView.rowHeight = UITableView.automaticDimension;
self.tableView.estimatedRowHeight = 44.0; // set to no matter your "common" cell top is

What this does is present the desk view with a brief estimate/placeholder for the row heights of cells that aren’t but onscreen. Then, when these cells are about to scroll on display, the precise row top will likely be calculated. To find out the precise top for every row, the desk view mechanically asks every cell what top its contentView must be based mostly on the recognized fastened width of the content material view (which relies on the desk view’s width, minus any extra issues like a piece index or accent view) and the auto format constraints you’ve gotten added to the cell’s content material view and subviews. As soon as this particular cell top has been decided, the outdated estimated top for the row is up to date with the brand new precise top (and any changes to the desk view’s contentSize/contentOffset are made as wanted for you).

Typically talking, the estimate you present would not need to be very correct — it’s only used to accurately dimension the scroll indicator within the desk view, and the desk view does a superb job of adjusting the scroll indicator for incorrect estimates as you scroll cells onscreen. You need to set the estimatedRowHeight property on the desk view (in viewDidLoad or comparable) to a relentless worth that’s the “common” row top. Provided that your row heights have excessive variability (e.g. differ by an order of magnitude) and also you discover the scroll indicator “leaping” as you scroll do you have to trouble implementing tableView:estimatedHeightForRowAtIndexPath: to do the minimal calculation required to return a extra correct estimate for every row.

For iOS 7 assist (implementing auto cell sizing your self)

3. Do a Structure Cross & Get The Cell Peak

First, instantiate an offscreen occasion of a desk view cell, one occasion for every reuse identifier, that’s used strictly for top calculations. (Offscreen which means the cell reference is saved in a property/ivar on the view controller and by no means returned from tableView:cellForRowAtIndexPath: for the desk view to truly render onscreen.) Subsequent, the cell have to be configured with the precise content material (e.g. textual content, pictures, and so on) that it might maintain if it have been to be displayed within the desk view.

Then, drive the cell to right away format its subviews, after which use the systemLayoutSizeFittingSize: technique on the UITableViewCell‘s contentView to seek out out what the required top of the cell is. Use UILayoutFittingCompressedSize to get the smallest dimension required to suit all of the contents of the cell. The peak can then be returned from the tableView:heightForRowAtIndexPath: delegate technique.

4. Use Estimated Row Heights

In case your desk view has greater than a pair dozen rows in it, you can find that doing the Auto Structure constraint fixing can rapidly bathroom down the primary thread when first loading the desk view, as tableView:heightForRowAtIndexPath: is known as on each row upon first load (so as to calculate the scale of the scroll indicator).

As of iOS 7, you’ll be able to (and completely ought to) use the estimatedRowHeight property on the desk view. What this does is present the desk view with a brief estimate/placeholder for the row heights of cells that aren’t but onscreen. Then, when these cells are about to scroll on display, the precise row top will likely be calculated (by calling tableView:heightForRowAtIndexPath:), and the estimated top up to date with the precise one.

Typically talking, the estimate you present would not need to be very correct — it’s only used to accurately dimension the scroll indicator within the desk view, and the desk view does a superb job of adjusting the scroll indicator for incorrect estimates as you scroll cells onscreen. You need to set the estimatedRowHeight property on the desk view (in viewDidLoad or comparable) to a relentless worth that’s the “common” row top. Provided that your row heights have excessive variability (e.g. differ by an order of magnitude) and also you discover the scroll indicator “leaping” as you scroll do you have to trouble implementing tableView:estimatedHeightForRowAtIndexPath: to do the minimal calculation required to return a extra correct estimate for every row.

5. (If Wanted) Add Row Peak Caching

When you’ve finished all of the above and are nonetheless discovering that efficiency is unacceptably gradual when doing the constraint fixing in tableView:heightForRowAtIndexPath:, you will sadly must implement some caching for cell heights. (That is the method prompt by Apple’s engineers.) The overall concept is to let the Autolayout engine remedy the constraints the primary time, then cache the calculated top for that cell and use the cached worth for all future requests for that cell’s top. The trick in fact is to be sure you clear the cached top for a cell when something occurs that might trigger the cell’s top to alter — primarily, this may be when that cell’s content material modifications or when different necessary occasions happen (just like the consumer adjusting the Dynamic Kind textual content dimension slider).

iOS 7 Generic Pattern Code (with a lot of juicy feedback)

- (UITableViewCell *)tableView:(UITableView *)tableView cellForRowAtIndexPath:(NSIndexPath *)indexPath
{
    // Decide which reuse identifier ought to be used for the cell at this 
    // index path, relying on the actual format required (you will have
    // only one, or might have many).
    NSString *reuseIdentifier = ...;

    // Dequeue a cell for the reuse identifier.
    // Word that this technique will init and return a brand new cell if there is not
    // one accessible within the reuse pool, so both method after this line of 
    // code you'll have a cell with the right constraints able to go.
    UITableViewCell *cell = [tableView dequeueReusableCellWithIdentifier:reuseIdentifier];
         
    // Configure the cell with content material for the given indexPath, for instance:
    // cell.textLabel.textual content = someTextForThisCell;
    // ...
    
    // Be sure the constraints have been arrange for this cell, because it 
    // might have simply been created from scratch. Use the next strains, 
    // assuming you're organising constraints from inside the cell's 
    // updateConstraints technique:
    [cell setNeedsUpdateConstraints];
    [cell updateConstraintsIfNeeded];

    // In case you are utilizing multi-line UILabels, do not forget that the 
    // preferredMaxLayoutWidth must be set accurately. Do it at this 
    // level in case you are NOT doing it inside the UITableViewCell subclass 
    // -[layoutSubviews] technique. For instance: 
    // cell.multiLineLabel.preferredMaxLayoutWidth = CGRectGetWidth(tableView.bounds);
    
    return cell;
}

- (CGFloat)tableView:(UITableView *)tableView heightForRowAtIndexPath:(NSIndexPath *)indexPath
{
    // Decide which reuse identifier ought to be used for the cell at this 
    // index path.
    NSString *reuseIdentifier = ...;

    // Use a dictionary of offscreen cells to get a cell for the reuse 
    // identifier, making a cell and storing it within the dictionary if one 
    // hasn't already been added for the reuse identifier. WARNING: Do not 
    // name the desk view's dequeueReusableCellWithIdentifier: technique right here 
    // as a result of this can end in a reminiscence leak because the cell is created however 
    // by no means returned from the tableView:cellForRowAtIndexPath: technique!
    UITableViewCell *cell = [self.offscreenCells objectForKey:reuseIdentifier];
    if (!cell) {
        cell = [[YourTableViewCellClass alloc] init];
        [self.offscreenCells setObject:cell forKey:reuseIdentifier];
    }
    
    // Configure the cell with content material for the given indexPath, for instance:
    // cell.textLabel.textual content = someTextForThisCell;
    // ...
    
    // Be sure the constraints have been arrange for this cell, because it 
    // might have simply been created from scratch. Use the next strains, 
    // assuming you're organising constraints from inside the cell's 
    // updateConstraints technique:
    [cell setNeedsUpdateConstraints];
    [cell updateConstraintsIfNeeded];

    // Set the width of the cell to match the width of the desk view. This
    // is necessary in order that we'll get the right cell top for various
    // desk view widths if the cell's top is determined by its width (on account of 
    // multi-line UILabels phrase wrapping, and so on). We do not want to do that 
    // above in -[tableView:cellForRowAtIndexPath] as a result of it occurs 
    // mechanically when the cell is used within the desk view. Additionally notice, 
    // the ultimate width of the cell is probably not the width of the desk view in
    // some circumstances, for instance when a piece index is displayed alongside 
    // the correct facet of the desk view. You will need to account for the decreased 
    // cell width.
    cell.bounds = CGRectMake(0.0, 0.0, CGRectGetWidth(tableView.bounds), CGRectGetHeight(cell.bounds));

    // Do the format move on the cell, which can calculate the frames for 
    // all of the views based mostly on the constraints. (Word that you could set the 
    // preferredMaxLayoutWidth on multiline UILabels contained in the 
    // -[layoutSubviews] technique of the UITableViewCell subclass, or do it 
    // manually at this level earlier than the under 2 strains!)
    [cell setNeedsLayout];
    [cell layoutIfNeeded];

    // Get the precise top required for the cell's contentView
    CGFloat top = [cell.contentView systemLayoutSizeFittingSize:UILayoutFittingCompressedSize].top;

    // Add an additional level to the peak to account for the cell separator, 
    // which is added between the underside of the cell's contentView and the 
    // backside of the desk view cell.
    top += 1.0;

    return top;
}

// NOTE: Set the desk view's estimatedRowHeight property as a substitute of 
// implementing the under technique, UNLESS you've gotten excessive variability in 
// your row heights and also you discover the scroll indicator "leaping" 
// as you scroll.
- (CGFloat)tableView:(UITableView *)tableView estimatedHeightForRowAtIndexPath:(NSIndexPath *)indexPath
{
    // Do the minimal calculations required to have the ability to return an 
    // estimated row top that is inside an order of magnitude of the 
    // precise top. For instance:
    if ([self isTallCellAtIndexPath:indexPath]) {
        return 350.0;
    } else {
        return 40.0;
    }
}

These tasks are totally working examples of desk views with variable row heights on account of desk view cells containing dynamic content material in UILabels.

Xamarin (C#/.NET)

When you’re utilizing Xamarin, try this pattern venture put collectively by @KentBoogaart.

Russian Hackers Utilizing Pretend Model Websites to Unfold DanaBot and StealC Malware

0


Aug 16, 2024Ravie LakshmananMalware / Information Theft

Russian Hackers Utilizing Pretend Model Websites to Unfold DanaBot and StealC Malware

Cybersecurity researchers have make clear a classy info stealer marketing campaign that impersonates official manufacturers to distribute malware like DanaBot and StealC.

The exercise cluster, orchestrated by Russian-speaking cybercriminals and collectively codenamed Tusk, is claimed to embody a number of sub-campaigns, leveraging the status of the platforms to trick customers into downloading the malware utilizing bogus websites and social media accounts.

“All of the lively sub-campaigns host the preliminary downloader on Dropbox,” Kaspersky researchers Elsayed Elrefaei and AbdulRhman Alfaifi mentioned. “This downloader is chargeable for delivering extra malware samples to the sufferer’s machine, that are principally info-stealers (DanaBot and StealC) and clippers.”

Cybersecurity

Of the 19 sub-campaigns recognized so far, three are mentioned to be presently lively. The identify “Tusk” is a reference to the phrase “Mammoth” utilized by the menace actors in log messages related to the preliminary downloader. It is price noting that mammoth is a slang time period typically utilized by Russian e-crime teams to consult with victims.

The campaigns are additionally notable for using phishing ways to deceive victims into parting with their private and monetary info, which is then offered on the darkish net or used to realize unauthorized entry to their gaming accounts and cryptocurrency wallets.

The primary of the three sub-campaigns, referred to as TidyMe, mimics peerme[.]io with a lookalike web site hosted on tidyme[.]io (in addition to tidymeapp[.]io and tidyme[.]app) that solicits a click on to obtain a bug for each Home windows and macOS methods. The executable is served from Dropbox.

The downloader is an Electron software that, when launched, prompts the sufferer to enter the CAPTCHA displayed, after which the principle software interface is displayed, whereas two extra malicious recordsdata are covertly fetched and executed within the background.

Each the payloads noticed within the marketing campaign are Hijack Loader artifacts, which in the end launch a variant of the StealC stealer malware with capabilities to reap a variety of knowledge.

DanaBot and StealC Malware

RuneOnlineWorld (“runeonlineworld[.]io”), the second sub-campaign, includes using a bogus web site simulating a massively multiplayer on-line (MMO) recreation named Rise On-line World to distribute the same downloader that paves the best way for DanaBot and StealC on compromised hosts.

Additionally distributed by way of Hijack Loader on this marketing campaign is a Go-based clipper malware that is designed to observe clipboard content material and substitute pockets addresses copied by the sufferer with an attacker-controlled Bitcoin pockets to carry out fraudulent transactions.

Rounding off the lively campaigns is Voico, which impersonates an AI translator challenge referred to as YOUS (yous[.]ai) with a malicious counterpart dubbed voico[.]io so as to disseminate an preliminary downloader that, upon set up, asks the sufferer to fill out a registration type containing their credentials after which logs the data on the console.

Cybersecurity

The ultimate payloads exhibit related conduct as that of the second sub-campaign, the one distinction being the StealC malware used on this case communicates with a unique command-and-control (C2) server.

“The campaigns […] reveal the persistent and evolving menace posed by cybercriminals who’re adept at mimicking official tasks to deceive victims,” the researchers mentioned. “The reliance on social engineering strategies akin to phishing, coupled with multistage malware supply mechanisms, highlights the superior capabilities of the menace actors concerned.”

“By exploiting the belief customers place in well-known platforms, these attackers successfully deploy a spread of malware designed to steal delicate info, compromise methods, and in the end obtain monetary acquire.”

Discovered this text fascinating? Observe us on Twitter and LinkedIn to learn extra unique content material we submit.