BITESIZE

bitesize

Let’s say you have a job interview in a few days. How should you prepare for it so that you can make an informed decision about joining the company, as well as make sure that your interests are taken care of? Prepare your questions An interview is a two-way street: it’s not only the company investigating you as a potential employee; in addition, you are investigating the company as a potential employer. Hopefully, you have some other options available—such as staying at your current place or trying out other companies interested in you. In the best case, you’ll have the flexibility to decide whether you want to follow through with the company or not. To learn about the potential employer, you need to ask questions. The question will depend on what’s important for you at the current stage of your career. Some things I would like to know before deciding to join a new company are as follows: What is their development process? If they use some standardized methodology such as Scrum, I would like to hear how they implement it. Especially whether they have Scrum Master and Product Owner roles covered—and whether those people are easily reachable. I would also ask if they do retrospective meetings. How do they do quality automatization: things such as unit tests, e2e, CI & code-style enforcing? For things that they don't have, I would ask if they would like to see it in the project. I would ask about their product and how they do product design/discovery. Life–work balance questions inspired by : How do they deal with emergencies? How often is there a need to receive a call or respond to messages outside working hours? Healthy Software Developer’s video how long do employees usually work? Salary negotiation The point of going to the interview is to get to the salary negotiation. The result of the negotiation will have a great impact on your life for the whole time you spend at the company—it’s a smart move to prepare well. The earlier you start, the better—as you will have more time to investigate, ask questions, and plan your strategy. The best resource for preparing for an interview in IT is . You can buy the complete e-book, read the chapters online, or get the key advice from the bite-size articles on the website. Fearless Salary Negotiation Research compensation Before you start negotiation, figure out the current situation in the market. A little bit of time spent on systematic investigation can give you a much more precise picture than anecdotes heard here or there. Check out this . chapter on market value estimation Decide your minimum Great advice is to decide on the minimum pay that you’ll accept you start speaking with the company. It could happen that during the various rounds of interviews, they will manage to the company very well or that the effort you have already invested into the company will make you more flexible than you should be. In those cases, having the minimum salary decided ahead of time can be very helpful. . before sell Relevant chapter from the book here Think openly about your options. As you negotiate with the company, it’s good to be flexible about the options you consider. If you are asking for more money than they have budgeted for, you can propose reduced hours with the same hourly rate that you are happy with. Similarly, commuting to the office can easily take 1–2 hours of your day, so a lower salary could be worth it if it comes with guaranteed remote work. Practice interview Try role-playing an interview with someone. If you can handle it, you can ask them to be tough on you—stricter than what you would expect from a real interview. Ideally, you’ll have a tough, low-stakes interview that would make you relaxed during the real thing. Investigate the company Learn about the company—all you can do before the meeting, that is. Check their website, their LinkedIn page, and their key employees. Play around with their product if you can. In short, spend an hour or two trying to learn as much as you can about the company. It will help you ask important questions and make a good impression. Summary A bit of preparation can go a long way. As developers, we often think about the technical part of the interview, but learning about the company and preparing for salary negotiation can help us even more. We can filter out companies we shouldn’t join, and we can make sure that we get as good a deal as possible. Also published . here

Nailing Your Job Interview: Preparation, Research, and Looking Out For Your Best Interests

When we think of debugging, we think of breakpoints in IDEs, stepping over, inspecting variables, etc. However, there are instances where stepping outside the conventional confines of an IDE becomes essential to track and resolve complex issues. This is where tools like DTrace come into play, offering a more nuanced and powerful approach to debugging than traditional methods. This blog post delves into the intricacies of DTrace, an innovative tool that has reshaped the landscape of and system analysis. debugging https://youtu.be/3M0AhZnVoUk?embedable=true As a side note, if you like the content of this and the other posts in this series, check out my that covers this subject. If you have friends who are learning to code, I'd appreciate a reference to my . If you want to get back to Java after a while, check out my Debugging book Java Basics book Java 8 to 21 book . DTrace Overview DTrace was first introduced by Sun Microsystems in 2004. DTrace quickly garnered attention for its groundbreaking approach to dynamic system tracing. Originally developed for Solaris, it has since been ported to various platforms, including MacOS, Windows, and Linux. DTrace stands out as a dynamic tracing framework that enables deep inspection of live systems – from operating systems to running applications. Its capacity to provide real-time insights into system and application behavior without significant performance degradation marks it as a revolutionary tool in the domain of system diagnostics and debugging. Understanding DTrace’s Capabilities DTrace, short for Dynamic Tracing, is a comprehensive toolkit for real-time system monitoring and debugging, offering an array of capabilities that span across different levels of system operation. Its versatility lies in its ability to provide insights into both high-level system performance and detailed process-level activities. System Monitoring and Analysis At its core, DTrace excels in monitoring various system-level operations. It can trace system calls, file system activities, and network operations. This enables developers and system administrators to observe the interactions between the operating system and the applications running on it. For instance, DTrace can identify which files a process accesses, monitor network requests, and even trace system calls to provide a detailed view of what's happening within the system. Process and Performance Analysis Beyond system-level monitoring, DTrace is particularly adept at dissecting individual processes. It can provide detailed information about process execution, including CPU and memory usage, helping to pinpoint performance bottlenecks or memory leaks. This granular level of detail is invaluable for performance tuning and debugging complex software issues. Customizability and Flexibility One of the most powerful aspects of DTrace is its customizability. With a scripting language based on C syntax, DTrace allows the creation of customized scripts to probe specific aspects of system behavior. This flexibility means that it can be adapted to a wide range of debugging scenarios, making it a versatile tool in a developer’s arsenal. Real-World Applications In practical terms, DTrace can be used to diagnose elusive performance issues, track down resource leaks, or understand complex interactions between different system components. For example, it can be used to determine the cause of a slow file operation, analyze the reasons behind a process crash, or understand the system impact of a new software deployment. Performance and Compatibility of DTrace A standout feature of DTrace is its ability to operate with remarkable efficiency. Despite its deep system integration, DTrace is designed to have minimal impact on overall system performance. This efficiency makes it a feasible tool for use in live production environments, where maintaining system stability and performance is crucial. Its non-intrusive nature allows developers and system administrators to conduct thorough debugging and performance analysis without the worry of significantly slowing down or disrupting the normal operation of the system. Cross-Platform Compatibility Originally developed for Solaris, DTrace has evolved into a cross-platform tool, with adaptations available for MacOS, Windows, and various . Each platform presents its own set of features and limitations. For instance, while DTrace is a native component in Solaris and MacOS, its implementation in Linux often requires a specialized build due to kernel support and licensing considerations. Linux distributions Compatibility Challenges on MacOS On MacOS, DTrace's functionality intersects with System Integrity Protection (SIP), a security feature designed to prevent potentially harmful actions. To utilize DTrace effectively, users may need to disable SIP, which should be done with caution. This process involves booting into recovery mode and executing specific commands, a step that highlights the need for a careful approach when working with such powerful system-level tools. We can disable SIP using the command: csrutil disable We can optionally use a more refined approach of enabling SIP without dtrace using the following command: csrutil enable --without dtrace Be extra careful when issuing these commands and when working on machines where dtrace is enabled. Back up your data properly! Customizability and Flexibility of DTrace A key feature that sets DTrace apart in the realm of system monitoring tools is its highly customizable nature. DTrace employs a scripting language that bears similarity to C syntax, offering users the ability to craft detailed and specific diagnostic scripts. This scripting capability allows for the creation of custom probes that can be fine-tuned to target particular aspects of system behavior, providing precise and relevant data. Adaptability to Various Scenarios The flexibility of DTrace's scripting language means it can adapt to a multitude of debugging scenarios. Whether it's tracking down memory leaks, analyzing CPU usage, or monitoring I/O operations, DTrace can be configured to provide insights tailored to the specific needs of the task. This adaptability makes it an invaluable tool for both developers and system administrators who require a dynamic approach to problem-solving. Examples of Customizable Probes Users can define probes to monitor specific system events, track the behavior of certain processes, or gather data on system resource usage. This level of customization ensures that DTrace can be an effective tool in a variety of contexts, from routine maintenance to complex troubleshooting tasks. Following in a simple hello world dtrace probe: sudo dtrace -qn 'syscall::write:entry, syscall::sendto:entry /pid == $target/ { printf("(%d) %s %s", pid, probefunc, copyinstr(arg1)); }' -p 9999 The kernel is instrumented with hooks that match various callbacks. dtrace connects to these hooks and can perform interesting tasks when these hooks are triggered. They have a naming convention, specially: . In this case the provider is a system call in both cases. We have no module, so we can leave that part blank between the colon ( ) symbols. We grab a write operation and entries. When an application writes or tries to send a packet, the following code event will trigger. provider:module:function:name : sendto These things happen frequently, which is why we restrict the process ID to the specific target with . This means the code will only trigger for the PID passed to us in the command line. The rest of the code should be simple for anyone with basic C experience; it's a printf that would list the processes and the data passed. pid == $target Real-World Applications of DTrace DTrace's diverse capabilities extend far beyond theoretical use, playing a pivotal role in resolving real-world system complexities. Its ability to provide deep insights into system operations makes it an indispensable tool in a variety of practical applications. To get a sense of how dtrace can be used, we can use the command whose output on my Mac is below: man -k dtrace bitesize.d(1m)           - analyse disk I/O size by process. Uses DTrace
cpuwalk.d(1m)            - Measure which CPUs a process runs on. Uses DTrace
creatbyproc.d(1m)        - snoop creat()s by process name. Uses DTrace
dappprof(1m)             - profile user and lib function usage. Uses DTrace
dapptrace(1m)            - trace user and library function usage. Uses DTrace
dispqlen.d(1m)           - dispatcher queue length by CPU. Uses DTrace
dtrace(1)                - dynamic tracing compiler and tracing utility
dtruss(1m)               - process syscall details. Uses DTrace
errinfo(1m)              - print errno for syscall fails. Uses DTrace
execsnoop(1m)            - snoop new process execution. Uses DTrace
fddist(1m)               - file descriptor usage distributions. Uses DTrace
filebyproc.d(1m)         - snoop opens by process name. Uses DTrace
hotspot.d(1m)            - print disk event by location. Uses DTrace
iofile.d(1m)             - I/O wait time by file and process. Uses DTrace
iofileb.d(1m)            - I/O bytes by file and process. Uses DTrace
iopattern(1m)            - print disk I/O pattern. Uses DTrace
iopending(1m)            - plot number of pending disk events. Uses DTrace
iosnoop(1m)              - snoop I/O events as they occur. Uses DTrace
iotop(1m)                - display top disk I/O events by process. Uses DTrace
kill.d(1m)               - snoop process signals as they occur. Uses DTrace
lastwords(1m)            - print syscalls before exit. Uses DTrace
loads.d(1m)              - print load averages. Uses DTrace
newproc.d(1m)            - snoop new processes. Uses DTrace
opensnoop(1m)            - snoop file opens as they occur. Uses DTrace
pathopens.d(1m)          - full pathnames opened ok count. Uses DTrace
perldtrace(1)            - Perl's support for DTrace
pidpersec.d(1m)          - print new PIDs per sec. Uses DTrace
plockstat(1)             - front-end to DTrace to print statistics about POSIX mutexes and read/write locks
priclass.d(1m)           - priority distribution by scheduling class. Uses DTrace
pridist.d(1m)            - process priority distribution. Uses DTrace
procsystime(1m)          - analyse system call times. Uses DTrace
rwbypid.d(1m)            - read/write calls by PID. Uses DTrace
rwbytype.d(1m)           - read/write bytes by vnode type. Uses DTrace
rwsnoop(1m)              - snoop read/write events. Uses DTrace
sampleproc(1m)           - sample processes on the CPUs. Uses DTrace
seeksize.d(1m)           - print disk event seek report. Uses DTrace
setuids.d(1m)            - snoop setuid calls as they occur. Uses DTrace
sigdist.d(1m)            - signal distribution by process. Uses DTrace
syscallbypid.d(1m)       - syscalls by process ID. Uses DTrace
syscallbyproc.d(1m)      - syscalls by process name. Uses DTrace
syscallbysysc.d(1m)      - syscalls by syscall. Uses DTrace
topsyscall(1m)           - top syscalls by syscall name. Uses DTrace
topsysproc(1m)           - top syscalls by process name. Uses DTrace
Tcl_CommandTraceInfo(3tcl), Tcl_TraceCommand(3tcl), Tcl_UntraceCommand(3tcl) - monitor renames and deletes of a command
bitesize.d(1m)           - analyse disk I/O size by process. Uses DTrace
cpuwalk.d(1m)            - Measure which CPUs a process runs on. Uses DTrace
creatbyproc.d(1m)        - snoop creat()s by process name. Uses DTrace
dappprof(1m)             - profile user and lib function usage. Uses DTrace
dapptrace(1m)            - trace user and library function usage. Uses DTrace
dispqlen.d(1m)           - dispatcher queue length by CPU. Uses DTrace
dtrace(1)                - dynamic tracing compiler and tracing utility
dtruss(1m)               - process syscall details. Uses DTrace
errinfo(1m)              - print errno for syscall fails. Uses DTrace
execsnoop(1m)            - snoop new process execution. Uses DTrace
fddist(1m)               - file descriptor usage distributions. Uses DTrace
filebyproc.d(1m)         - snoop opens by process name. Uses DTrace
hotspot.d(1m)            - print disk event by location. Uses DTrace
iofile.d(1m)             - I/O wait time by file and process. Uses DTrace
iofileb.d(1m)            - I/O bytes by file and process. Uses DTrace
iopattern(1m)            - print disk I/O pattern. Uses DTrace
iopending(1m)            - plot number of pending disk events. Uses DTrace
iosnoop(1m)              - snoop I/O events as they occur. Uses DTrace
iotop(1m)                - display top disk I/O events by process. Uses DTrace
kill.d(1m)               - snoop process signals as they occur. Uses DTrace
lastwords(1m)            - print syscalls before exit. Uses DTrace
loads.d(1m)              - print load averages. Uses DTrace
newproc.d(1m)            - snoop new processes. Uses DTrace
opensnoop(1m)            - snoop file opens as they occur. Uses DTrace
pathopens.d(1m)          - full pathnames opened ok count. Uses DTrace
perldtrace(1)            - Perl's support for DTrace
pidpersec.d(1m)          - print new PIDs per sec. Uses DTrace
plockstat(1)             - front-end to DTrace to print statistics about POSIX mutexes and read/write locks
priclass.d(1m)           - priority distribution by scheduling class. Uses DTrace
pridist.d(1m)            - process priority distribution. Uses DTrace
procsystime(1m)          - analyse system call times. Uses DTrace
rwbypid.d(1m)            - read/write calls by PID. Uses DTrace
rwbytype.d(1m)           - read/write bytes by vnode type. Uses DTrace
rwsnoop(1m)              - snoop read/write events. Uses DTrace
sampleproc(1m)           - sample processes on the CPUs. Uses DTrace
seeksize.d(1m)           - print disk event seek report. Uses DTrace
setuids.d(1m)            - snoop setuid calls as they occur. Uses DTrace
sigdist.d(1m)            - signal distribution by process. Uses DTrace
syscallbypid.d(1m)       - syscalls by process ID. Uses DTrace
syscallbyproc.d(1m)      - syscalls by process name. Uses DTrace
syscallbysysc.d(1m)      - syscalls by syscall. Uses DTrace
topsyscall(1m)           - top syscalls by syscall name. Uses DTrace
topsysproc(1m)           - top syscalls by process name. Uses DTrace There's a lot here; we don't need to read everything. The point is that when you run into a problem, you can just search through this list and find a tool dedicated to debugging that problem. Let’s say you're facing elevated disk write issues that are causing the performance of your application to degrade... But is it your app at fault or some other app? rwbypid.d can help you with that. It can generate a list of processes and the number of calls they have for read/write based on the process ID, as seen in the following screenshot: We can use this information to better understand IO issues in our code or even in 3rd party applications/libraries. is another tool that helps us track IO operations but with more details: iosnoop In diagnosing elusive system issues, DTrace shines by enabling detailed observation of system calls, file operations, and network activities. For instance, it can be used to uncover the root cause of unexpected system behaviors or to trace the origin of security breaches, offering a level of detail that is often unattainable with other debugging tools. Performance optimization is the main area where DTrace demonstrates its strengths. It allows administrators and developers to pinpoint performance bottlenecks, whether they lie in application code, system calls, or hardware interactions. By providing real-time data on resource usage, DTrace helps in fine-tuning systems for optimal performance. Final Words In conclusion, DTrace stands as a powerful and versatile tool in the realm of system monitoring and debugging. We've explored its broad capabilities, from in-depth system analysis to individual process tracing and its remarkable performance efficiency that allows for its use in live environments. Its cross-platform compatibility, coupled with the challenges and solutions specific to MacOS, highlights its widespread applicability. The customizability through scripting provides unmatched flexibility, adapting to a myriad of diagnostic needs. Real-world applications of DTrace in diagnosing system issues and optimizing performance underscore its practical value. DTrace's comprehensive toolkit offers an unparalleled window into the inner workings of systems, making it an invaluable asset for system administrators and developers alike. Whether it's for routine troubleshooting or complex performance tuning, DTrace provides insights and solutions that are essential in the modern computing landscape. Also published . here

Advanced Debugging Techniques: Revisiting DTrace for Troubleshooting

Twitter(now X) indubitably has the largest crypto community on the internet. Picture this: millions of users and a vast market for crypto investors to tap into. But as Elon Musk once admitted, is currently plagued by a bot problem that needs a quick fix. crypto Twitter A found that many of the biggest crypto profiles have 100,000s of fake ‘bot’ followers inflating their communities. These same fake profiles are also used to promote projects with artificial and low-quality engagement on social posts. Elon Musk’s attempts to prevent this since he acquired the company have had limited results, and what is more, Twitter/X is predicted to lose users over the next two years - almost 10% of its active user base. study earlier this year 30 million a new social-fi company, is amping up efforts to improve the level of social engagement on Twitter posts and help fight against the countless fake profiles across the platform. I had the opportunity to sit down with 8an, co-founder of /Reach, for a discussion about /Reach’s mission to help fix Crypto Twitter. /Reach, Q1: Can you tell us about yourself and your route to web3? My background started in consumer goods. I started my career at Rocket Internet as a Global Venture Developer, where we would launch day-one revenue-generating businesses in Europe and Emerging markets within 90 days from conception. After having gone the entrepreneurial route, namely as a founding team member of Numan (which grew to be the largest digital men’s health clinic in Europe), I founded Particle, a real-world assets on-chain fractionalization protocol focused on fine art. We famously tokenised Banksy’s $13M Love is in the Air in January 2022. I’ve been passively investing in Bitcoin since 2014, but I became more active in 2017 and then more ingrained in the community in 2021, like a lot of people. Q2: When Elon Musk bought Twitter a year ago, he promised to fix its bot problem once and for all. Do you think it is a herculean task to beat Twitter bots? It’s definitely not easy, but as with everything, recognising that there is a problem is a big first step. I think it's very much doable, especially with Musk's approach, which is to essentially make it economically deterrent for bots to exist - a similar approach to what /Reach is doing to those seeking to abuse or game the system. The goal, in the long run, is to make it unviable for a bot or exploiter ecosystem to exist. Q3: You have rightly pointed out that Crypto Twitter has a massive bot problem. Can you tell us how serious this problem is? Reportedly, up to 10% of X’s user base are bots. How deep this problem goes, we’ll only know once Musk reveals the extent of the issue as he has promised. What I can say is that within Crypto Twitter, it definitely is something everyone--from project owners, community leaders, artists, and individuals--sees as detrimental to their efforts. Q4: What can you tell us about the founding of /Reach, and how do you intend to tackle this problem? Attention and reach are the greatest challenges for brands, community builders, artists, and projects to crack. The X algorithm is broken, and Musk’s Ads Revenue Sharing program has fallen short of properly remunerating accounts for their reach. The “pay influencers” model is broken too - pseudo/low-value engagement doesn't help bring true visibility to comms that need to be promoted – with unfair and opaque pricing structures. Lastly, in an effort to maintain integrity on their profiles, even a brand’s greatest champions are reluctant to show support. Many aren’t even aware of which pieces of content would be most impactful for them to share. To elaborate further: Modern online tribes and communities, from gaming servers to Web3 projects, have become advertising-resistant, disregarding messages even from Key Opinion Leaders (KOLs). Traditional advertising struggles to infiltrate these tight-knit digital circles authentically. Reputable Web2 brands, despite having considerable media budgets, have yet to effectively access these communities and tribes in a manner that feels genuine and is welcomed. The outsider perception of brands is a significant barrier. Acceptance within these communities hinges on genuine endorsement from active, core tribe members. We approach this problem by harnessing the very ethos of Web3 in having the participants in a network equally participate in its success. The strategies we employ include the promotion of content, products, and services within active, fervent communities across diverse platforms – Web3, gaming, sports, and gambling, among others – organically and unobtrusively; overcoming the ‘Outsider Stigma’: Navigating through the walls of advertising resistance by enabling promotions through trusted, native voices within each tribe or community; and delivering a win-win scenario, allowing communities to monetize their user base while brands authentically connect with fresh, engaged audiences. Q5: What role does Twitter play in keeping the crypto community active? Why is Twitter the platform with the most crypto users? Twitter is our open forum and our playground, it is extremely important in the crypto community. It’s supported, of course, by Discord, Telegram, and WeChat, but Twitter’s model of short, bite-size news-esque delivered content is perfect for the fast-paced nature of Crypto. Q6: What are some of the strategies employed by /Reach in combating Crypto’s Twitter bot problem? We won’t be able to fully solve X’s bot problem alone, but we will be able to make sure it does not affect the spend creators put through the /Reach protocol. The /Reach protocol incentivizes both quality engagement and volume of engagement –both necessary in the success of visibility and fostering high caliber conversations and reach. We’ve implemented purpose-built, AI-powered, dynamic algorithms to power this effort. It is this quality engagement involving real people with authentic opinions that is our greatest weapon against bots on Twitter. Q7: From the information I have, you seem to have collected together a group of very impressive backers. Can you tell us a little bit about them and how you work with them? All of them are users of the /Reach protocol throughout the beta phase. The product and its results spoke for itself and they wanted to get involved. Many of the big names we speak to regularly to get their advice and perspective within the realm of their expertise. Q8: What are the next steps for /Reach after your launch? Grow the /Reach audience. We have already signed two deals with major communities to start using the /Reach protocol and have a pipeline of over 20 conversations at varying degrees of the sales cycle. The goal is to grow to an audience of more than 1,000,000 users within three months of launch.

Fixing Crypto Twitter: Exclusive Interview With /Reach Co-founder 8an

Ever chatted with an AI and gotten a response so out there, it felt like it came from a parallel universe? LLMs (like all of us) occasionally have day-dreams, also known as "hallucinations". Stick with me, and I'll show you a quick and easy fix to prevent this from happening to the product you're working on. Different Types of Hallucinations: The Obvious Untruths: An AI just told you about a president who never existed. The LLM has made up events, people, and places that are as real as unicorns, . entertaining, but not exactly helpful when you're after the facts The Oversharer: In contrast to making things up, some LLM responses go overboard with details. Before you know it, a user gets access to information it shouldn't have. . It raises concerns about privacy and the ethical use of data The Gibberish Generator: And sometimes, it's like they just smash the keyboard and hope for the best, similar to a baby with alphabet soup. . Not what you want if this output is fed to another function It is the fastest growing open source LLMOps platform, and the only one built for full stack developers with first-class TypeScript Support. Just a quick word about Pezzo: Like our mission? . We're building for developers, by developers 🧙‍♂️ Check Pezzo out and give us a star Why Do Hallucinations Happen? The Desire to Please: LLMs hate letting us down. When faced with gaps in data, it may invent details to fulfill the user's request, prioritizing continuity over accuracy. Outdated Information: An LLM's knowledge can be outdated. In fact, it is only as current as its last update. This can lead to hiccups in their responses. Preventing Hallucinations With "Sanity Checks" Sanity Checks are our way of teaching AIs to pause and ponder before blurting out whatever comes to mind. It's like giving them an internal filter to weed out the wacky before it hits our screens. This is how you do it: Add to your prompt an example of a thinking process where you stop after every response and perform a check about that response. The subject doesn't matter, the LLM will follow the reasoning structure. You can copy and paste this example to your prompt: Question: Who lived longer, Albert Einstein or Isaac Newton?

Answer:
Are follow up questions needed here: Yes.
Follow up: How old was Albert Einstein when he died?
Intermediate answer: Albert Einstein was 76 years old when he died.
Follow up: How old was Isaac Newton when he died?
Intermediate answer: Isaac Newton was 84 years old when he died.
So the final answer is: Isaac Newton. The Evolution of Prompting This prompting technique is a natural evolution from Direct Prompting and Chain-Of-Thought Prompting (see pic below). Here's the cool part: Sanity Checks Prompting reveals the LLM's thought process, letting us peek into the "black box" as it breaks down complex queries into bite-size follow-ups. The LLM knows it's reached the destination when it can seamlessly transition from intermediate answers to a final, confident conclusion. are a fixable glitch, not a dead end, and you're now ready to tackle them head-on. AI hallucinations Take these tools, push the AI to its limits, and make it work for you. If there's a specific bug you can't iron out with your AI, let's hear it below. Your challenge might just be the star of our next troubleshooting guide. Go on, give it a whirl, and report back 🚀 This article is inspired by the paper Measuring and Narrowing the Compositionality Gap in Language Models

A Secret Technique To Sidestepping LLM Hallucinations 

To say that artificial intelligence (AI) has been disrupting industries left and right is a mere understatement. From automation, and generative AI, to Natural Language Processing (NLP), these tech breakthroughs are continuously revolutionizing a myriad of organizations and domains over the last several years. With its lightning speed of global adoption and utilization, AI shows no signs of slowing down anytime soon. For legacy industries with traditional business models, embracing AI can still be a challenge. For one, there are some organizational facets that are simply irreplaceable by machines. While processes and operations are exponentially optimized, certain business aspects still call for critical thinking, out-of-the-box creativity, natural discernment, and human connection. The Learning & Development (L&D) space is no exception to this disruption. In fact, progressive business leaders and HR teams welcome these advancements and are looking forward to the vast opportunities brought about by AI. While of employees still prefer learning from an expert, L&D professionals are optimistic about the benefits of strategically incorporating AI into their initiatives and training programs. 87% Now, the question is: Can all L&D activities be led by AI? This article will dive into the L&D categories that can seamlessly adopt and utilize AI technologies, the segments that are irreplaceable by machines, and how to find the right balance in between. To decipher how AI can be implemented in the L&D domain, it’s paramount to understand the two main categories in the L&D space: and . on-demand/self-paced expert-led programs On-Demand/Self-Paced L&D Programs On-demand training empowers trainees to learn at their own pace via training platforms such as LinkedIn Learning. These programs grant them the flexibility to choose when and which courses to take, aligning with their needs and convenience. On-demand courses consist of diverse content elements, such as videos, class recordings, PDFs, and presentations, which each learner can navigate through at their individual speed. What’s more, the training platform effortlessly saves their progress, allowing them to resume from where they left off in their own time. This stands in contrast to traditional training methods, such as classroom settings with instructors who dictate the course's tempo and content in each session. Self-paced training provides learners with significantly more independence, enabling them to access pertinent materials when required and fostering greater engagement as they dictate their learning speed and priorities. Can it be led by AI? . Definitely Given its structure and nature, AI will be revolutionary in this segment – from pinpointing your specific learning objectives, discovering tailored content, and saving valuable time by avoiding irrelevant material. This technology promises a more interactive and adaptive learning experience, catering to your individual learning style and providing precisely targeted exercises. Moreover, AI is poised to enhance the on-demand, self-paced learning experience by introducing follow-up support and ensuring the practical implementation of acquired knowledge. Expert-Led L&D Programs Expert-Led L&D Programs are facilitated by industry experts, whether it's in an online or classroom setting. This model can take various forms, such as lectures, interactive workshops for hands-on skill practice, group/team coaching or 1:1 coaching/mentoring. Regardless of the delivery method, it serves as an excellent choice for organizations that require specialized training on intricate subjects, prioritize interactive and practical learning, value face-to-face engagement, and can allocate time for learners to fully immerse themselves in extended training sessions. Can it be led by AI? . Limitedly Yes, AI can streamline some of processes involved in expert led-learning, and automate some technical elements. Use cases include leveraging generative AI for better assessment of the learning targets, better matching of an expert to a cohort or individual, tailoring bite size learning to support the expert-led sessions, or automated follow-ups to support continuous learning. In doing so, it can uncover valuable insights and offer personalized recommendations – all with increased scalability and efficiency. Not in the foreseeable future. Expert-led L&D programs bank on direct person-to-person interactions and nuanced human experiences that machines currently don’t provide. Think of the 2013 movie “Her” – it’s the utopia of AI where an OS and a human are almost indistinguishable. However, it’s still science fiction, and we still have a long way to go before that becomes a seamless reality. But will machines completely replace human experts? At the end of the day, AI still lacks the genuine qualities of a human expert in your field of interest, and trusting a machine for learning and skill development can be dubious. In the short term, this trust may not yield significant benefits; over the long term, it could potentially show promise. But There Are Things AI Can’t Replace While AI can definitely optimize and automate low-quality L&D elements, there are things it can’t replace – primarily, the human connection. The unique quality of a human connection is the ability to foster trust and build rapport. When we interact with another person, we engage not only in words and actions but also in the exchange of emotions, empathy, and shared experiences. This multifaceted interaction creates a profound sense of trust that forms the foundation of many collaborations, including learning. On the other hand, AI operates within the confines of algorithms and data analysis. It can process vast amounts of information and make predictions or recommendations based on patterns and data trends. However, it lacks the emotional intelligence and intuition that humans naturally possess. AI (still) cannot read between the lines, pick up on subtle nuances, or understand the intricate web of human emotions that influence our decisions and actions. These factors are critical in learning and development – from motivating individuals to learn, and encouraging employees to advance their careers, to building a sound and healthy working relationship. The Key is Finding the Balance We are at a stage when no one can totally disregard AI because of its limitations, or completely replace all human-led L&D activities with machines. At the end of the day, it’s all about navigating the spot where AI can fill the void, bridge the gap, and streamline traditional training methods. Finding the right balance between human and machine is the key ingredient for you to thrive in today’s competitive world. Consider the challenge of identifying what aspects of one's work or skill set need enhancement. For some individuals, self-assessment may come naturally; they can readily discern their strengths and weaknesses. However, this self-awareness is not universal, and many people may struggle to pinpoint areas in which they need to develop to achieve desired outcomes. This is where AI steps in as a valuable ally. AI effortlessly analyzes vast datasets and identifies patterns, trends, and correlations that might be fallible to humans. In L&D activities, AI can serve as a guide, helping individuals recognize areas of improvement based on data-driven insights. Microsoft called their AI “co-pilot” for a reason. Think of AI as a co-pilot in an aircraft. The human pilot remains in control, making critical decisions and ensuring the overall success of the flight. However, the co-pilot provides support by managing side tasks, providing valuable information, and alerting the pilot to potential issues. This collaboration enhances the overall efficiency and safety of the journey. Conclusion The key takeaway is treating AI as a tool that can assist you; not necessarily replacing humans, but refreshing obsolete methods for efficiency and cost-effectiveness. Those organizations that will embrace, adapt, and utilize these inevitable advancements will find themselves earning an upper hand in today’s cutthroat digital age, in which those who still cling to old practices will be left behind. At the same time, an overreliance on machines is an equally dangerous territory that one has to avoid. Human intuition, empathy, and adaptability remain indispensable assets that cannot be replicated by AI. While incorporating AI into L&D strategies, it is essential to preserve the uniquely human qualities that contribute to innovation and resilience.

Can All L&D Activities Be Led By AI?

Presentations can be daunting to put together. I know what you are thinking, you've got a pile of information which all seems quite relevant which you need to simmer down into bite-size pieces to attract potential investors, right? Well, fear not, there are many different strategies that you can use to is not dull but rather engaging and timely so investors can go away with a clear understanding of the overall objective and message of the presentation. make sure your presentation Introducing to you the 10/20/30 Rule. A successful venture capitalist, , came up with the 10/20/30 Rule of PowerPoint. The suggested rule offers an excellent structure for presentations aiming to keep your audience's attention span from straying and is designed to keep your presentation compact, digestible and understandable. Guy Kawasaki Image Source: 180 Degrees Consulting The structure of the rule means your presentation should consist of no more than ten slides, shouldn't last longer than twenty minutes, and should contain a font size no smaller than thirty points. The great thing about the rule is that it is relevant to any business or type of presentation, be it an investor pitch or employee training, making it applicable and resourceful to all. It's quite easy to overfill the presentation with facts and text, which is why using visual data representation is a great way to present your findings. Keeping your presentation full of impressive visuals can help keep your potential investors engaged. The first part of the rule is to use a maximum of 10 slides per presentation. Make use of pre-made slides that have the structure already laid out to adhere to. The example below is based on a aimed at potential investors; however, the rule applies to all industries by customising the slides to meet your needs better. business presentation Image Source: poweredtemplate.com - in the first slide, identify the area of concern you will focus on in your presentation. The amount you’re seeking to raise should be discussed either during this slide or in your overview statement. Problem Focus on covering the following points: What’s the problem you’re solving? Explain the issue in detail, and why is it important to act now Why has no one solved it before? - once you have identified your problem, the key to any is to offer a solution to the problem. Ensure to speak facts and data. Explain why you’re the person/company to offer the solution, and how is your approach different from the others who’ve tried. Solution good presentation - now take a look at the business as a whole and how will you implement the solution considering the . Investors are primarily interested in making a profit (for some, it’s the one and only goal), hence, consider focusing on explaining how you’re going to return on their investments. Business model business strategy - discuss any new or existing software/hardware you will use to reach your goal. If you're a SaaS startup seeking a seed round, use this slide to show mockups of your technology/software. Technology - if the presentation is based on a product launch, for example, discuss how you will be getting the product to your audience. Reflect on the marketing budget and strategies you’re planning on implementing to drive traffic and awareness to the business (e.g. Paid Ads, SEO, Influencer Marketing, Content Marketing, etc). Marketing and Sales - outline what you're up against, who are your competitors? Explain how your model is different and how exactly are you going to beat the competition. Competition - gives everyone clear roles and collectively will help reach the end goal. Discuss the size of the team required, positions and hiring strategies. Team delegating tasks amongst a team - setting small goals to eventually lead up to the bigger target is a good way of approaching and managing the solution. Explain how you’re going to reach these steps. Projections and Milestones - timeline and forecast when tasks should be completed so that everyone is clear of the time frame to get the job done. Status and Timeline - talk about all the points you have discussed and what the next steps are to get the process started. Summary The second part of the rule is to stick to a 20-minute-long presentation. Keeping presentations short not only helps keeps your but also gives you more time at the end to answer questions or address any concerns your audience may have. With investor pitches, it’s definitely a good strategy to leave some time to respond to questions. audience engaged Running over time could potentially be dangerous, leading to an uninformed disengaged audience. The last part of the rule for a is to keep the font size at 30 or relatively close to it. It may seem like 30 points is big, but in retrospect, using smaller fonts can make your presentation harder to see - not something you’d want to happen during your pitch. winning presentation The use of smaller text can make the slide look text-heavy. The method of using a more prominent and clear font shows your knowledge and assurance in the information you are sharing. Most times, presenters squeeze information onto a single slide because they are not fully familiar with the topic. As you are the one who has to present, do the talking and persuade the investors to give you money, you have to know your stuff by heart, ensuring a flawless experience.

Winning Formulas: Obeying To The 10/20/30 Rule In Investor Pitching Presentations

Keeping up with the latest developments in Artificial Intelligence (AI) can be a challenging task given the rapid pace of innovation in this field. Newsletters can be an invaluable resource for staying up to date, providing a curated selection of the latest news, research, and insights directly to your inbox. Here are the top 20 AI newsletters you should consider subscribing to in 2023: : PromptPal is the platform where to share and discover the best AI prompts for ChatGPT, Midjourney, Bard, Bing, and more. Their newsletter is filled with the best free prompts, the best AI tools to use, the latest AI news, the best new Midjourney images and more. It’s our favourite AI newsletter. PromptPal : Provides a weekly collection of the latest in AI/ML research, tools, and projects. Ideal for professionals interested in a bite-size minimalistic format that covers the technical side of AI. Data Machina : Written by Jeremy Kahn of Fortune Magazine, this newsletter provides a long-form look into the AI industry and the implications of tools like ChatGPT. Fortune: Eye on AI : Provides a critical lens into the world of AI, covering the second-order implications of new artificial intelligence technologies and the ethical dilemmas that we face. The Road to AI We Can Trust : Perfect for a quick dose of AI news, The Neuron provides short reads and humorous updates in the AI world. The Neuron : A weekly newsletter written by Alex McFarland, AI Disruption highlights the most significant AI developments each week. AI Disruption : A daily newsletter that offers brief, punchy insights into new AI tools, how-to guides, and creative AI use cases. Perfect for anyone looking to understand the most interesting tweets and product releases in AI, regardless of technical skill level. Ben’s Bites : Provides AI news and commentary from the perspective of a builder in the space. This a fantastic resource if you are interested in the perspective of someone actively building AI tools and technologies. Machine Learnings : A weekly email that provides a deeper dive into the ethical and technical aspects of AI, perfect for those that are ready to dig further into the world of AI. AI Weekly Digest : Translates academic research in the field of AI, machine learning, and data science into useful insights for everyday people. One Useful Thing : A must-read monthly newsletter for anyone looking to gain a more comprehensive grasp of the AI industry, politics, and ethics. Guide to AI : Provides an in-depth look into cutting-edge AI research and explains why it matters in an easy-to-understand manner. Written by Jack Clark, the former Policy Director at OpenAI. Import AI : Offers short, organized, and very visual weekly roundups of AI news. Ideal for anyone with only a few minutes each week to read about AI news. Its emphasis is on providing quick and succinct updates without personal takes, summaries, or endless scrolling. AI Tidbits : Weekly AI summaries written by graduate students in the field, providing recaps of the biggest developments of the week. The emails focus on research and business-related AI news, while also providing great commentary on public policy and technical implications. Last Week in AI : A new newsletter delivering the latest AI tools and resources for the startup and builder audience. It offers simple, weekly emails that can be read in less than 5 minutes. Meta Minded : Provides daily updates on what’s happening in AI and robotics, offering varying opinions from a wide array of people in the industry. It also features a curated news section on the latest in Artificial Intelligence. Inside AI Newsletter : More focused on insiders and established professionals in the field of AI, Data Elixir shares information relevant to industry experts on the latest developments in artificial intelligence and machine learning. Expect to find new tools useful for AI engineers and job postings here. Data Elixir : Authored by Daniel Bourke, this monthly newsletter covers the most interesting news over the last month in AI. Each newsletter takes about 10 to 15 minutes to read, making it a good option for readers who don’t want a weekly/daily update on the industry. Zero to Mastery Newsletter : A highly respected newsletter by the MIT Technology Review that covers the latest news and trends in AI. It includes updates on machine learning, neural networks, and data science in general. The Algorithm : Curated by Rohin Shah, this weekly newsletter provides updates on technical and philosophical progress related to AI alignment, a field of study that focuses on ensuring that AI systems will be beneficial to humans. AI Alignment Newsletter : This popular TikTok channel also sends out newsletters featuring easy-to-follow how-to guides on AI image and text generation, along with fun examples of AI use cases. Tech N Trendz (AI Tools) There you have it! These are the top 20 AI newsletters you should consider subscribing to in 2023. They offer a wide range of perspectives and cover a variety of topics, from the latest AI tools and research to ethical considerations and industry trends. Happy reading!

21 AI Newsletters You Should Subscribe to in 2023

21 boletines informativos de IA a los que debería suscribirse en 2023

21 newsletters sur l'IA auxquelles vous devriez vous abonner en 2023

21 AI न्यूज़लेटर्स आपको 2023 में सब्सक्राइब करने चाहिए

2023 年に購読すべき 21 の AI ニュースレター

21 boletins informativos de IA que você deve assinar em 2023

21 bản tin AI bạn nên đăng ký vào năm 2023

2023 年你应该订阅的 21 份 AI 时事通讯

Often when debugging, we need to step outside of the comforting embrace of the IDE to reproduce or track an issue. In this series, I’d like to go over some tools you might find useful for these cases. I’ll try to limit myself to tools that are 100% debugging tools and not those that are useful for development testing. E.g. tools like or are remarkably useful. You can/should use them while debugging. But you probably used them while building and testing the feature. So you would already be familiar with them and should have some sense of what they do. I want to focus on tools that you would mostly reach for when debugging. In that sense, tools like etc. also make no sense here. curl jq SDKMan I also want to avoid tools like database tools. They are remarkably useful when debugging, but again, you’d probably use them during development, too. They’re also very vendor specific, so it’s a vast subject that’s too broad to cover here. The tools I'll cover in this series encompass the following categories: System monitoring tools - as is the case with strace and dtrace, which we'll discuss today Network monitors - also fits in the above bracket but warrants a category of its own VM/Runtime monitoring - e.g. tools that let us inspect the JVM, etc. Profilers and memory monitors In this first post I'd like to discuss two heavyweight champions: dtrace and strace. If you're a Java developer or a Windows user, there's a good chance you never heard of these tools. You might have inadvertently used one of them since so many tools were built on top of them, but that's probably not the case. Both tools will let you debug anything without the source code. You could detect problems and gain a level of understanding you never imagined. Dtrace Back in 2004, I first heard about DTrace while working at Sun Microsystems. It became all the rage in the hallways as it was an innovation that Sun Microsystems was promoting. Dtrace was ported later into MacOS X (it originated on Solaris). Today there are ports on Windows and Linux as well. Dtrace is a powerful low-level dynamic tracing framework. But that's just another superlative and if you never used such a tool and don't have a background in systems programming, you might be feeling a bit confused: what the hell does it do? It lets you "see" everything. Want to know which files a process has opened? OK Want to know who invokes a kernel API and get a stack trace to the caller? OK Want to know why a process dies? OK Want to know how much CPU Time is spent on an operation? OK You might think that dtrace is one of those tools that will destroy your CPU completely... But here's the killer feature: it's fast enough to run in production with minimal to no performance impact! It was revolutionary when it was launched almost two decades ago and it's still the case to this day! Running Dtrace Before we begin, a word of warning. Save your data! This tool can easily crash your machine and enabling it requires disabling important security facilities on MacOS. This is a risky "low level" system service and should be treated as such... On a Mac dtrace conflicts with "System Integrity Protection" which is a security feature that blocks some interactions between processes (among other things). Under normal circumstances, this would be great. But if you want to run dtrace this would be a problem. The solution is booting to recovery mode on Intel Macs; this means holding the keys on boot. On an ARM mac just long-press the power button. Command-R Then, in a recovery mode terminal, issue the command: . csrutil disable Then upon reboot, dtrace should work as expected. Basic Usage As mentioned before, dtrace is a very powerful tool. There are whole books written about it. It has its own programming language based on C syntax that you can use to build elaborate logic. E.g. the following command will log some information from the given callbacks: sudo dtrace -qn 'syscall::write:entry, syscall::sendto:entry /pid == $target/ { printf("(%d) %s %s", pid, probefunc, copyinstr(arg1)); }' -p [PID] The code snippet passed to the dtrace command listens to the sendto callback on the target process ID. Then prints out the information to the console, e.g.: (pid) text If this seems like a bit too much and too hard to get started with... You're 100% right. It's a powerful tool when you need it. But for most of our day-to-day usage, it's just too powerful. What we want is to know a bit of basic stuff and this is just too much! Simple Usage As luck would have it, we have a simple solution: man -k dtrace This prints out a list of tools that's worth reading just to get a sense of how extensive this thing is. Here are a few interesting lines of output from that command: bitesize.d(1m)           - analyse disk I/O size by process. Uses DTrace
dapptrace(1m)            - trace user and library function usage. Uses DTrace
errinfo(1m)              - print errno for syscall fails. Uses DTrace
iotop(1m)                - display top disk I/O events by process. Uses DTrace
plockstat(1)             - front-end to DTrace to print statistics about POSIX mutexes and read/write locks It's worth your time going over this list to realize what you can truly do here. Examples You're facing elevated disk write issues that are causing the performance of your application to degrade... But is it your app at fault or some other app? Just run: sudo rwbypid.d It will print out the reads/writes to the disk: PID CMD                       DIR    COUNT
  2957 wordexp-helper              W        1
  2959 wc                          W        1
  2961 grep                        W        1

... snipped for clarity ...

   637 firefox                     R     6937
   637 firefox                     W    15325
   343 sentineld                   W   100287 The security software is really holding performance down... You can also use to get more specific results on the number of bytes written/distribution. bitesize.d That's pretty high level though. What if you want specifics: file name, process name, etc? sudo iosnoop -a Prints out output that includes pretty much everything you would need: STRTIME              DEVICE  MAJ MIN   UID   PID D      BLOCK     SIZE                     PATHNAME ARGS
2022 Jun 30 12:16:56 ??        1  17   501  1111 W  150777072     4096 ??/idb/3166453069wcaw.sqlite-wal firefox\0
2022 Jun 30 12:16:56 ??        1  17   501   661 W  150777175   487424  ??/index-dir/the-real-index Slack Helper\0
2022 Jun 30 12:16:57 ??        1  17   499   342 W  150777294     4096 ??/persistent/.dat.nosync0156.ztvXap sentineld\0 I can see the process id and how many bytes it wrote to the specific file! Let's say your program spans processes and you want to see what's going on. E.g. I run a source code build in a server I built: sudo errinfo Lets me detect the error returned from system calls to and the command that originally triggered that: EXEC          SYSCALL  ERR  DESC
    WindowServer workq_kernreturn   -2 
    WindowServer workq_kernreturn   -2 
   SentinelAgent workq_kernreturn   -2 
   SentinelAgent workq_kernreturn   -2 
          Signal           Helper    0 
          Google           Chrome    0 
           Brave          Browser    0 
          Google           Chrome    0 These are just the tip of the iceberg. I suggest checking out this old from Oracle or . Disclaimer: I didn't read the book... dtrace tutorial the book strace Amusingly enough, the strace tool also originated at Sun Microsystems in this case in the 90s. This isn't surprising though as the list of technologies originating from Sun Microsystems is absolutely mind-numbing. Strace is much simpler than dtrace in both usage and capabilities. For better and for worse. Since dtrace requires deep OS support, it never became an official feature of common Linux distributions and as a result, on Linux people use strace instead of dtrace. They aren't exactly interchangeable, though. Strace is enabled thanks to the kernel feature known as ptrace. Since ptrace is already in Linux, we don't need to add additional kernel code or modules. Typically, dtrace requires deeper kernel support, to get around licensing issues on Linux it's in a separate loadable module but that still presents some challenges. Working with strace is akin to printing a log entry every time we make a kernel call. This creates very verbose logging for every command you execute. As a result, you can follow what's going on under the hood of a running process. really Running Strace Nowadays, strace is commonly used in Linux, it's my favorite system diagnostic tool on that platform. It's remarkably convenient to work with since we can run it with no special privileges. Notice that, unlike dtrace, you should keep strace away from production environments (unless the code is segregated). It carries a significant performance overhead and would bring a production system down. The most basic usage of strace is just passing the command line to it: strace java -classpath. PrimeMain The output of strace for this is pretty long, lets go over a few of the lines: execve("/home/ec2-user/jdk1.8.0_45/bin/java", ["java", "-classpath.", "PrimeMain"], 0x7fffd689ec20 /* 23 vars */) = 0
brk(NULL)                               = 0xb85000
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f0294272000
readlink("/proc/self/exe", "/home/ec2-user/jdk1.8.0_45/bin/j"..., 4096) = 35
access("/etc/ld.so.preload", R_OK)      = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/tls/x86_64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/tls/x86_64", 0x7fff37af09a0) = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/tls/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/tls", 0x7fff37af09a0) = -1 ENOENT (No such file or directory) Every one of these lines is a Linux system call. We can google each one of them to get a sense of what's going on. Here's a simple example: open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/tls/x86_64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) Java tries to load the library from the directory using a system open call to load the file. The exit code of the system call is , which means that the file isn't there. Under normal circumstances, we should get back a file descriptor value from this API. Looking in the directory, it seems the directory is missing. I'm guessing that this is because of a missing JCE installation. This is probably OK but might have been interesting in some cases. pthread tls -1 tls Obviously, the amount of output is overwhelming sometimes. We usually just want to see things like "which file was opened" and "what's going on with our network calls". We can easily accomplish that by only looking at specific system calls using the argument. -e strace -e open java -classpath . PrimeMain Will only show the open system calls: open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/tls/x86_64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/tls/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/x86_64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/tls/x86_64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/tls/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/x86_64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
open("/etc/ld.so.cache", O_RDONLY|O_CLOEXEC) = 3
open("/lib64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = 3
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/libjli.so", O_RDONLY|O_CLOEXEC) = 3
open("/home/ec2-user/jdk1.8.0_45/bin/../lib/amd64/jli/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) There are many system calls you can learn and use to track many behaviors such as this: connect, write, etc. This is only the tip of the iceberg of what you can do using strace. Julia Evans wrote some of the most . If you want to learn more about it, there's probably no better place (also check out her other stuff... Amazing resources!). exhaustive and entertaining posts on strace Strace and Java As you saw before, strace works great with the JVM. Since strace predates Java and is a very low-level tool, it has no awareness of the JVM. The JVM works like most other platforms and invokes system calls which you can use to debug its behavior. However, because of its unique approach to some problems, some aspects might not be as visible with strace. A good example is allocations. System tools use malloc, which maps to kernel allocation logic, but Java takes a different route. It manages its own memory for efficiency and easier garbage collection logic. As a result, some aspects of memory allocation will be hidden from strace output. This can be a blessing in disguise, as the output can be overwhelming sometimes. At the time of this writing, threading works well with strace. But this might not be the case moving forward as might change the one-to-one mapping between Java threads and system threads. This might make strace output harder to pinpoint in heavily threaded applications. project Loom Finally There's an alphabet soup of "*trace" utilities in various forms that keep borrowing ideas from one another. It's a significant challenge to keep up with all of that noise. There are just too many great tools to cover, I would like to discuss btrace in a future installment though. It's very similar to dtrace but also very JVM specific so probably worth a different post. The tools I discussed today take different approaches to solve similar problems: how do we understand what a binary application "really" does? Security researchers and hackers use these tools to understand your program. They don't need the code and don't need a disassembly to see what you're actually doing. You can also use these tools to understand the impact of your actions. We often invoke an API and let things end there. But the devil is in the details and those details can carry a heavy toll. As a Java developer, I rarely think about signal delivery, process management or such low-level staples. But I do spend time looking at these things since they ultimately impact the stability and performance of my application. Also Published Here

Using External Debugging Tools dtrace and strace

We all know that tech is a pretty good career money-wise. Still, you can do yourself a disservice by sticking to lower pay than a company might offer. Even a tiny difference in pay becomes an impressive amount when you multiply it by 12 months in a year and all the years you’ll be working. Let’s ensure that you don’t miss out on financial opportunities. 1. Talk about pay with your peers The way talking about money is perceived in the culture. I grew up in Poland, and talking about my income felt like boasting to me for a long time. You could see this being the case if a well-paid specialist were to talk about their income to people who are stuck in minimum wage jobs. On the other hand, for people who work in similar jobs, the possible outcomes are: There is a minimal difference in compensation. Someone—maybe you—might realize that they are underpaid. There are a few reasons this could happen: you lack some technical skill that has a significant impact on pay you lack negotiating skills the company you work for has no budget to pay better And the people who learn that they are paid better? They lose nothing. Besides the cultural taboo, the only reason not to talk about money would be if you are more invested in the company’s bottom line than in your and your colleagues’ welfare. For example, this could be the case if your share of company ownership is more significant than your part of the total payroll, or you believe fully in your employer’s mission. 2. Evaluate the broader market regularly If you are like me, your focus on your career fluctuates a lot over the years. Sometimes I spend a lot of time learning new stuff engaging with tech communities, and sometimes I have months when I do nothing IT-related besides my day job. There are always changes in the job market: new technologies rising in popularity, new companies in the local market offering better conditions, or far away companies opening up for remote candidates. It’s a good idea to make an appointment with yourself to regularly check up on those changes. At a minimum, setting aside one afternoon a year for searching for job offers should make you aware of how things are going outside your current company. 3. Don’t count on the company’s initiative The company you work for is interested in maintaining their operations undisturbed. Maybe your manager’s evaluation depends on how many resign from their team, but this gives them only a small incentive to keep your salary in line with the market. This incentive will be even smaller if you convince them that you like your work and you are not searching for anything else at the moment. If the company has a policy of regular pay raises, it will most likely just match inflation, not the IT job market changes. If you are promoted to a different salary bracket that comes with a raise, it looks like an improvement, but it may be that you are placed at the bottom of this new bracket. 4. Prepare yourself for the negotiation In 2019 I bought myself as a birthday gift. It was the greatest gift I ever got, and it paid back for itself in no time. The book gives much practical advice—step-by-step guides for what to do before negotiating the salary. Dealing with a new company is a big part of it, but many things’ll be useful in a more general context, and one chapter is dedicated to promotions and pay raises. Fearless Salary Negotiation If you are on a budget, the book author generously offers it for . You can get his advice from his website, in the bite-size piece: . free how to write an email to your manager starting the conversation 5. Ask for it! Bringing up the topic of compensation with your manager can be stressful. The proper preparation should help you achieve your target and help you with the stress. You can look at it from another perspective—a reasonable manager should appreciate that you talk about your issue with the salary with them first, instead of going directly to the competition. And if they turn out to be unreasonable? It will mean that talking with other companies is a good idea. 6. Don’t be shy; let your colleagues know Once you secure your raise, you can let your colleagues know! The standard industry practice of jumping from job to job to get better compensation means that your experienced colleague could be replaced one day by somebody who knows nothing about your project, making you carry a more significant workload. Sharing your good news can inspire other team members to try doing the same, improving morale and making sure they stay longer in the project. In the end, this will make your job easier and more pleasant. Also published . here

Stories, News & Mentions about BITESIZE

Bitesize’s Stories on HackerNoon