In this article, I will delve into whether artificial intelligence (AI), particularly the emergence of tools like generative AI (GenAI), will replace classical software outsourcing as we know it.

AI vs software outsourcing: An opportunity or a threat?

computer security, software security and more. This cohort-based program trains computational specialists who can construct models, develop algorithms, and write software using state-of-the-art graduate-level knowledge and techniques. The program has been ...

Master's in Professional Computer Science

What computer do I need for the best VFX software? VFX software, like even the best free video editing software, can be very resource-intensive. So, for top performance, it's best to use a ...

Best VFX software of 2023

If you’re on a computer and you get a text on your phone, wouldn’t it be great if you could reply without picking it up? Yes, it would, and there are great apps and services that enable just ...

How to send a text message from a computer

Graduate students are a vital part of the extensive research conducted in the Faculty of Computer Science. Many of our faculty members are also engaged in industry-related, funded research at the national and international level. Research here is extensive ...

Master's program

House Republicans were on the verge of open revolt Wednesday after the ideologically fractious conference failed to coalesce around a speaker nominee, leaving the chamber rudderless and leaderless ...

Republicans fail to coalesce around speaker choice, leaving House in limbo

Ryan on Tuesday night asked Romney to pinpoint how soon donors would have to coalesce around one candidate — essentially forcing lower-tier candidates from the race — to avoid the “train ...

Mitt Romney urges his donors to coalesce behind one challenger to Trump

Computer vision at the edge is unlocking actionable ... Dell Validated Designs connects multiple ISVs (independent software vendors) and partners to optimize application-based solutions that ...

Delivering 5 Key Outcomes With Computer Vision At The Edge

Get the most recent info and news about Coalesce on HackerNoon, where 10k+ technologists publish stories for 4M+ monthly readers.

Coalesce is full-stack product shop that solves problems for businesses, brands, and brave souls with great ideas.

We believe great products are built by great teams. With backgrounds ranging from art to ergonomics, our approach is often unorthodox and insight always unique. From early-stage strategy and product roadmaps to architecture, visual design, and code-agnostic development, we can fill in your product gaps in a comprehensive, low-risk way. Whether you’re struggling with startup-specific challenges, agency stresses, or software scalability, we have a suite of cost-effective services that deliver real value, so you can meet your deadlines head on. 

Ready to build something for your business but not sure how to explain the blueprints? That’s our speciality. We can't wait to work with you. 




Coalesce

Authors:
(1) Yelena Mejova, ISI Foundation, Italy;
(2) Arthur Capozzi, Università degli Studi di Torino, Italy;
(3) Corrado Monti, CENTAI, Italy;
(4) Gianmarco De Francisci Morales, CENTAI, Italy. Table of Links Introduction Background and Related Work Methods Results Discussion and References 3 METHODS 3.1 Data collection To examine the use of memetic engagement by Ukraine during the first 8 months of the 2022 conflict, we begin with a large collection of tweets matching war-related keywords (see Appendix A) spanning from 27 February 2022 to 12 October 2022. By examining this collection, we identify three prominent accounts: (1) @Ukraine: self-described as “the official Twitter account of Ukraine”, this verified account is labeled as an “Ukraine government organization” by Twitter. It posts war-related and political updates about Ukraine. (2) @uamemesforces: self-described as the “Source of the best Ukrainian memes”, this unverified account states as its goal ”to convey the truth about what is happening now in Ukraine with the help of memes”.[8] (3) @DefenceU: or “Defense of Ukraine”, is the “Official page of the Ministry of Defense of Ukraine”. Like @Ukraine, this verified account is described by Twitter as a “Ukraine government organization”. Its posts concern battlefield updates and news. We select the accounts @Ukraine and @DefenceU because of their large following (at the end of the collection period), because they are official accounts not associated with individuals or specific media outlets, and because of the focus on the conflict (many popular Ukrainian accounts are by non-relevant celebrities). Additionally, we select @uamemesforces, the most popular Ukrainian unofficial account that posts memes about the conflict. Note that we are limited in the number of accounts we can study due to the labor-intensive labeling process. We retrieve all of the posts for these accounts via Twitter Timeline API. We treat each post with media content as a potential meme, no matter the format. To ensure preserving content posted by these accounts, we also download the images associated with the posts and the users who retweeted or liked the post (see Table 1 for summary statistics). We associate each user with a country in the GeoNames database via their location field (a free-text field optionally filled by the user). Considering only tweets with media, out of 586 042 users, 44% are successfully mapped to a country. We retain only countries with at least 1000 geolocated retweeting users for geographic analysis. In order to explore the relationship between Ukrainian memes and each country’s actions towards the conflict, we use the Ukraine Support Tracker data from the Kiel Institute for the World Economy (downloaded on 20 June 2023), which “quantifies military, financial, and humanitarian aid transferred by governments to Ukraine since the end of diplomatic relations between Russia and Ukraine on January 24, 2022”.[9] The dataset provides a quantitative measure of a country’s aid to Ukraine. Finally, to examine the attitude of each country’s population towards Ukraine, we use the Eurobarometer STD98 survey which took place in winter 2022-2023 in 30 European countries [22]. Specifically, we use use QE2, answer 6: “To what extent do you agree or disagree with: providing financial support to Ukraine.” 3.2 Annotation To understand the visual and narrative attributes of the captured data, all authors perform a manual annotation of a random selection of the tweets. We use all of @Ukraine tweets and a random sample of roughly 40% of those from the other two accounts are selected for analysis, for a total of 1063 tweets (last column of Table 1). We remove self-replies (an account replying to its own post) which have no retweets of their own. Guided by previous literature, we compose a codebook to capture visual and narrative aspects of the content. First, we adopt the significant features of the content characterization framework proposed by Ling et al. [41], which is used to model the virality of political memes. These attributes include: (1) Number of panels: multiple or single; (2) Type of image: photo, screenshot, or illustration; (3) Scale: close-up, medium shot, long shot, other; (4) Type of subject: object, character, scene, creature, text, other; (5) Main attribute of the subject (if character): facial expression, posture, poster (6) Character emotion (if character): positive, negative, or neutral; (7) Contains words: whether image itself (not tweet caption) contains words. Second, we employ the narrative framework by Karpman [35] later extended by de Saint Laurent et al. [18]. It defines two key dimensions to characterize a narrative: moral quality (benevolent or malevolent) and power (strong or weak), thus resulting in four character archetypes: heroes (benevolent, strong), victims (benevolent, weak), villains (malevolent, strong), and fools (malevolent, weak). We also label the actors mentioned in these narratives (as an open category, guided by an initial set of main actors in the conflict, such as Zelensky and Putin). We further contextualize the narratives by their emotional appeal [8], in particular humor, pride, fear, outrage, and compassion (the label was originally open, but the labelers coalesced on this set). Finally, we record whether there is a specific intent in terms of a call to action or information sharing. Note that, as much as possible, we try to label the narratives as intended from the point of view of the posting account (although some biases may creep in, as we discuss in Section 5). Appendix B reports the full codebook. Using this literature-driven codebook, all authors of this manuscript perform a deductive coding of the selected posts (for more on such process see Linneberg and Korsgaard [42]). We rely on Google Translate to translate non-English language content. All coders first annotated a selection of 50 posts, discussed insights and disagreements, and came up with standard labels for emotional appeal, intent, and actor (though the label remained open). Subsequently, the rest of the posts are annotated by one author each, with uncertain cases discussed collectively. Appendix C reports the inter-annotator agreement on a selection of 48 posts (sampled from each account). On average, the agreement is 0.750 for @Ukraine, 0.604 for @uamemesforces, and 0.667 for @DefenceU (with a Krippendorff’s alpha of around 0.65 for all accounts and tasks), which is high considering the subjective nature of the tasks, and that some features were open and could have multiple labels. The annotated dataset is available at https://github.com/corradomonti/ukranian-tweets. 3.3 Popularity model Using the data collected and annotated as per the previous sections, we construct models to find whether there are trends in (𝑖) what kind of content becomes more popular, and (𝑖𝑖) who it reaches. Specifically, we use linear regression to model the log-transformed number of retweets with the content variables outlined above as independent variables. As the three accounts have posted a vastly different number of tweets, we weigh the data points (tweets) inverseproportionally to their representation in the dataset, so that each account weighs equally on the outcomes. We build several versions of the model to ascertain the benefit of each class of variables: visual [41], actor, emotion, intent, and narrative. A baseline model simply uses the identity of the author (i.e., group fixed effects), the number of followers of that account (log transformed) as a proxy for the potential audience, and the number of days from the invasion to account for changing interest in the topic over time. We then choose the best model by the Bayesian Information Criterion (BIC)—a model selection metric that captures both unexplained variation in the dependent variable and the number of explanatory variables. The coefficients of this model reveal the associations between these variables and the virality of the content. Next, we group retweets by the country of the user (those who can be geolocated) to understand the preferences of users who spread these messages. We enrich our characterization of the relationships between each country and Ukraine by considering the amount of financial assistance the country provides to Ukraine (normalized by the country’s GDP) or by their opinion on this topic (via the Eurobarometer survey). As mentioned above, we consider these measures as different proxies of support for Ukraine’s military efforts. First, we examine the retweeting rate (normalized by population). Then, we compare the narrative preferences to the amount of support for Ukraine: are there narratives associated with increased support for Ukraine? We operationalize this preference as the log odds of users from a country retweeting content having a specific narrative. Finally, we examine the geographical distribution of the retweeting rate and narrative preferences across Europe by plotting these on choropleth maps. [8] https://www.patreon.com/uamemesforces [9] https://www.ifw-kiel.de/topics/war-against-ukraine/ukraine-support-tracker This paper is available on arxiv under CC 4.0 license. Authors: (1) Yelena Mejova, ISI Foundation, Italy; (2) Arthur Capozzi, Università degli Studi di Torino, Italy; (3) Corrado Monti, CENTAI, Italy; (4) Gianmarco De Francisci Morales, CENTAI, Italy. Authors: (1) Yelena Mejova, ISI Foundation, Italy; (2) Arthur Capozzi, Università degli Studi di Torino, Italy; (3) Corrado Monti, CENTAI, Italy; (4) Gianmarco De Francisci Morales, CENTAI, Italy. Table of Links Introduction Introduction Background and Related Work Background and Related Work Methods Methods Results Results Discussion and References Discussion and References 3 METHODS 3.1 Data collection To examine the use of memetic engagement by Ukraine during the first 8 months of the 2022 conflict, we begin with a large collection of tweets matching war-related keywords (see Appendix A) spanning from 27 February 2022 to 12 October 2022. By examining this collection, we identify three prominent accounts: (1) @Ukraine: self-described as “the official Twitter account of Ukraine”, this verified account is labeled as an “Ukraine government organization” by Twitter. It posts war-related and political updates about Ukraine. @Ukraine: (2) @uamemesforces: self-described as the “Source of the best Ukrainian memes”, this unverified account states as its goal ”to convey the truth about what is happening now in Ukraine with the help of memes”.[8] @uamemesforces: (3) @DefenceU: or “Defense of Ukraine”, is the “Official page of the Ministry of Defense of Ukraine”. Like @Ukraine, this verified account is described by Twitter as a “Ukraine government organization”. Its posts concern battlefield updates and news. @DefenceU: We select the accounts @Ukraine and @DefenceU because of their large following (at the end of the collection period), because they are official accounts not associated with individuals or specific media outlets, and because of the focus on the conflict (many popular Ukrainian accounts are by non-relevant celebrities). Additionally, we select @uamemesforces, the most popular Ukrainian unofficial account that posts memes about the conflict. Note that we are limited in the number of accounts we can study due to the labor-intensive labeling process. We retrieve all of the posts for these accounts via Twitter Timeline API. We treat each post with media content as a potential meme, no matter the format. To ensure preserving content posted by these accounts, we also download the images associated with the posts and the users who retweeted or liked the post (see Table 1 for summary statistics). We associate each user with a country in the GeoNames database via their location field (a free-text field optionally filled by the user). Considering only tweets with media, out of 586 042 users, 44% are successfully mapped to a country. We retain only countries with at least 1000 geolocated retweeting users for geographic analysis. In order to explore the relationship between Ukrainian memes and each country’s actions towards the conflict, we use the Ukraine Support Tracker data from the Kiel Institute for the World Economy (downloaded on 20 June 2023), which “quantifies military, financial, and humanitarian aid transferred by governments to Ukraine since the end of diplomatic relations between Russia and Ukraine on January 24, 2022”.[9] The dataset provides a quantitative measure of a country’s aid to Ukraine. Finally, to examine the attitude of each country’s population towards Ukraine, we use the Eurobarometer STD98 survey which took place in winter 2022-2023 in 30 European countries [22]. Specifically, we use use QE2, answer 6: “To what extent do you agree or disagree with: providing financial support to Ukraine.” 3.2 Annotation To understand the visual and narrative attributes of the captured data, all authors perform a manual annotation of a random selection of the tweets. We use all of @Ukraine tweets and a random sample of roughly 40% of those from the other two accounts are selected for analysis, for a total of 1063 tweets (last column of Table 1). We remove self-replies (an account replying to its own post) which have no retweets of their own. Guided by previous literature, we compose a codebook to capture visual and narrative aspects of the content. First, we adopt the significant features of the content characterization framework proposed by Ling et al. [41], which is used to model the virality of political memes. These attributes include: (1) Number of panels: multiple or single; (2) Type of image: photo, screenshot, or illustration; (3) Scale: close-up, medium shot, long shot, other; (4) Type of subject: object, character, scene, creature, text, other; (5) Main attribute of the subject (if character): facial expression, posture, poster (6) Character emotion (if character): positive, negative, or neutral; (7) Contains words: whether image itself (not tweet caption) contains words. Second, we employ the narrative framework by Karpman [35] later extended by de Saint Laurent et al. [18]. It defines two key dimensions to characterize a narrative: moral quality (benevolent or malevolent) and power (strong or weak), thus resulting in four character archetypes: heroes (benevolent, strong), victims (benevolent, weak), villains (malevolent, strong), and fools (malevolent, weak). We also label the actors mentioned in these narratives (as an open category, guided by an initial set of main actors in the conflict, such as Zelensky and Putin). We further contextualize the narratives by their emotional appeal [8], in particular humor, pride, fear, outrage, and compassion (the label was originally open, but the labelers coalesced on this set). Finally, we record whether there is a specific intent in terms of a call to action or information sharing. Note that, as much as possible, we try to label the narratives as intended from the point of view of the posting account (although some biases may creep in, as we discuss in Section 5). Appendix B reports the full codebook. Using this literature-driven codebook, all authors of this manuscript perform a deductive coding of the selected posts (for more on such process see Linneberg and Korsgaard [42]). We rely on Google Translate to translate non-English language content. All coders first annotated a selection of 50 posts, discussed insights and disagreements, and came up with standard labels for emotional appeal, intent, and actor (though the label remained open). Subsequently, the rest of the posts are annotated by one author each, with uncertain cases discussed collectively. Appendix C reports the inter-annotator agreement on a selection of 48 posts (sampled from each account). On average, the agreement is 0.750 for @Ukraine, 0.604 for @uamemesforces, and 0.667 for @DefenceU (with a Krippendorff’s alpha of around 0.65 for all accounts and tasks), which is high considering the subjective nature of the tasks, and that some features were open and could have multiple labels. The annotated dataset is available at https://github.com/corradomonti/ukranian-tweets. 3.3 Popularity model Using the data collected and annotated as per the previous sections, we construct models to find whether there are trends in (𝑖) what kind of content becomes more popular, and (𝑖𝑖) who it reaches. Specifically, we use linear regression to model the log-transformed number of retweets with the content variables outlined above as independent variables. As the three accounts have posted a vastly different number of tweets, we weigh the data points (tweets) inverseproportionally to their representation in the dataset, so that each account weighs equally on the outcomes. We build several versions of the model to ascertain the benefit of each class of variables: visual [41], actor, emotion, intent, and narrative. A baseline model simply uses the identity of the author (i.e., group fixed effects), the number of followers of that account (log transformed) as a proxy for the potential audience, and the number of days from the invasion to account for changing interest in the topic over time. We then choose the best model by the Bayesian Information Criterion (BIC)—a model selection metric that captures both unexplained variation in the dependent variable and the number of explanatory variables. The coefficients of this model reveal the associations between these variables and the virality of the content. Next, we group retweets by the country of the user (those who can be geolocated) to understand the preferences of users who spread these messages. We enrich our characterization of the relationships between each country and Ukraine by considering the amount of financial assistance the country provides to Ukraine (normalized by the country’s GDP) or by their opinion on this topic (via the Eurobarometer survey). As mentioned above, we consider these measures as different proxies of support for Ukraine’s military efforts. First, we examine the retweeting rate (normalized by population). Then, we compare the narrative preferences to the amount of support for Ukraine: are there narratives associated with increased support for Ukraine? We operationalize this preference as the log odds of users from a country retweeting content having a specific narrative. Finally, we examine the geographical distribution of the retweeting rate and narrative preferences across Europe by plotting these on choropleth maps. [8] https://www.patreon.com/uamemesforces [9] https://www.ifw-kiel.de/topics/war-against-ukraine/ukraine-support-tracker This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Understanding Memetic Warfare Tactics: Social Media Tactics in the Russo-Ukrainian War

Memes and Narrative Strategies in the Russo-Ukrainian War: A Comprehensive Study

রুশো-ইউক্রেনীয় যুদ্ধে মেমস এবং বর্ণনামূলক কৌশল: একটি ব্যাপক অধ্যয়ন

Meme und Erzählstrategien im Russisch-Ukrainischen Krieg: Eine umfassende Studie

Memes y estrategias narrativas en la guerra ruso-ucraniana: un estudio exhaustivo

Mèmes et stratégies narratives dans la guerre russo-ukrainienne : une étude approfondie

रूस-यूक्रेनी युद्ध में मीम्स और कथात्मक रणनीतियाँ: एक व्यापक अध्ययन

ロシア・ウクライナ戦争におけるミームと物語戦略：包括的研究

러시아-우크라이나 전쟁의 밈과 서사 전략: 종합 연구

Memes e estratégias narrativas na guerra russo-ucraniana: um estudo abrangente

Мемы и нарративные стратегии в российско-украинской войне: комплексное исследование

Rus-Ukrayna Savaşında Memler ve Anlatı Stratejileri: Kapsamlı Bir Çalışma

Memes và chiến lược kể chuyện trong Chiến tranh Nga-Ukraina: Một nghiên cứu toàn diện

俄乌战争中的模因与叙事策略：综合研究

Authors:
(1) Vinod S. Khandkar and Manjesh K. Hanawal, Industrial Engineering and Operations Research Indian Institute of Technology Bombay, Mumbai, India and {vinod.khandkar, mhanawal}@iitb.ac.in. Table of Links Abstract & Introduction Related Work and Background Challenges in TD Detection Measurement Setup Development Case Study : Wehe - TD Detection Tool for Mobile Environment Shortcoming of Wehe on HTTPS Traffic TD Detection of HTTPS Traffic Conclusion & References Abstract We consider the problem of detecting deliberate traffic discrimination on the Internet. Given the complex nature of the Internet, detection of deliberate discrimination is not easy to detect, and tools developed so far suffer from various limitations. We study challenges in detecting the violations (focusing on the HTTPS traffic) and discuss possible mitigation approaches. We focus on ‘Wehe,’ the most recent tool developed to detect net-neutrality violations. Wehe hosts traffic from all services of interest in a common server and replays them to mimic the behavior of the traffic from original servers. Despite Wehe’s vast utility and possible influences over policy decisions, its mechanisms are not yet validated by others. In this work, we highlight critical weaknesses in Wehe where its replay traffic is not being correctly classified as intended services by the network middleboxes. We validate this observation using a commercial traffic shaper. We propose a new method in which the SNI parameter is set appropriately in the initial TLS handshake to overcome this weakness. Using commercial traffic shapers, we validate that SNI makes the replay traffic gets correctly classified as the intended traffic by the middleboxes. Our new approach thus provides a more realistic method for detecting neutrality violations of HTTPS traffic. Index Terms—Net neutrality, traffic differentiation, detection tools I. INTRODUCTION Net neutrality is a guiding principle promoting the “equal” treatment of all packets over the Internet. However, for economic benefits, ISPs may apply traffic differentiation to a specific service, user, content provider, or any other traffic group on the Internet without making any public declaration. It gives rise to a need to have tools that can detect such malicious activities over the Internet. Traffic differentiation (TD) detection involves the coalescence of many elements such as end-systems (user-client and server), probing traffic generation, expected network responses, and TD detection algorithms. These are interdependent components or operations. Hence, in developing TD detection tools, one faces challenges such as crafting internet traffic and conditioning measured network response that suits their detection algorithm. Moreover, as HTTPS traffic becomes prevalent, it is unclear what policies the middle-boxes apply for discrimination as payload provides no signatures. Our first goal is to study various challenges in designing a reliable TD detection mechanism for HTTPS traffic Several methods have been proposed in the literature to detect Network neutrality violations as documented in recent surveys [1], [2]. The literature is rich with various forms of discrimination and its detection, like discrimination of content providers [3], end-users [4], specific services like BitTorrent [5]. Our interest in this work is discrimination of services, specifically on streaming services (both audio and video) which are potential candidates from description due to their commercial values and high bandwidth requirements. Recently several tools are developed to detect discrimination of services like, NANO [6] ChkDiff [7], Netpolice [8], Shaperprobe [9], Packsen [10], Glasnost [11], Bonafide [12], and Wehe [13]. NANO is based on passive measurements, and all others are based on either active or differential probes. Wehe is the latest tool that overcomes many of the drawbacks of the other tools. However, due to the Internet’s complexity, many issues persist that prevent its use as a reliable tool. Wehe [13] follows a client-service architecture in which content of service of interests (Youtube, Netflix, etc.) are copied (or sniffed) from their respective original servers and hosted on a common server (referred to as replay server). To check if a particular service is discriminated, the client accesses the content of that service from the replay server. The server transfers the requested data with timing relations that mimic the original service’s data transfer characteristics over the Internet. The client accesses content from the replay server with and without VPN and compares them to ascertain if there was any difference in the quality. The basic idea of Wehe is that connection over VPN is encapsulated and will not be subjected to any discrimination as the middleboxes cannot classify the traffic correctly. Whereas content accessed without VPN can be classified correctly due to the signatures induced by the timing relations and can be subjected to discrimination policies. Despite the Wehe tool’s vast utility and possible influences over policy decisions, its mechanisms are not yet fully validated by other researchers. We investigate Wehe’s differentiation detection’s performance on traffic using HTTPS protocol, which is predominantly used by all services for security reasons and to avoid detection. We noticed that Wehe uses port 80 (HTTP) for all communications. Middleboxes process all the traffic on this port as non-encrypted traffic and may not classify them correctly as they will not see any signatures (recall that replayed traffic is HTTPS). On the other hand, when Wehe access traffic over port 443 (HTTP), traffic is not classified as intended traffic as we demonstrate it using a commercial traffic shaper. As middleboxes do not see traffic from the Wehe’s replay server as that coming from the original server, they may not apply intended traffic discrimination policies. Thus Wehe may not detect any deliberate discrimination resulting in false negatives. in false negatives. To overcome the limitations of Wehe for HTTPS traffic, we propose a new method in which traffic is accessed on port 443, but with a modification in the TLS handshake. In the TLS handshake, we explicitly send the Server Name Indication (SNI) corresponding to the actual service. We demonstrate that the HTTPS traffic get correctly classified by the middleboxes with these modifications. Our method ensures the middleboxes classify traffic from replay servers in the same way they do from the original servers for each service and apply the same discriminate policies (if any) on traffic from both replay and original servers. Thus, our method makes it possible to detect discrimination of HTTPS traffic in a realistic setting. Our contributions can be summarized as: We study various challenges associated with detection of discrimination of HTTPS traffic. We present the categorization of these challenges based on their source, e.g., as protocol and operational environment.


We demonstrate limitations of Wehe in detection of discrimination of HTTPS traffic. Specifically, we demonstrate that HTTPS traffic will not be correctly classified in the Wehe setup using a commercial traffic shaper.


We propose a mechanism in which traffic of services accessed from the replay server is treated as if it originated from the actual servers. Thus traffic from the replay server also gets subjected to the same discriminated policies applied on the actual service. We


We validate that replay traffic of our mechanism is treated in the same way originating from actual servers using a commercial traffic shaper. The paper is organizaed as follows. Sec. II provides the necessary background and related work. Sec. III describes all identified challenges in measurement setup for TD detection. Sec. IV describes the Wehe tool and its mechanisms in the context of identified challenges, Sec. V provides the validation results. Our proposed method is described in Sec. VI and Sec. VII gives the conclusions. This paper is available on arxiv under CC 4.0 license. Authors: (1) Vinod S. Khandkar and Manjesh K. Hanawal, Industrial Engineering and Operations Research Indian Institute of Technology Bombay, Mumbai, India and {vinod.khandkar, mhanawal}@iitb.ac.in. Authors: Authors: (1) Vinod S. Khandkar and Manjesh K. Hanawal, Industrial Engineering and Operations Research Indian Institute of Technology Bombay, Mumbai, India and {vinod.khandkar, mhanawal}@iitb.ac.in. Table of Links Abstract & Introduction Abstract & Introduction Related Work and Background Related Work and Background Challenges in TD Detection Measurement Setup Development Challenges in TD Detection Measurement Setup Development Case Study : Wehe - TD Detection Tool for Mobile Environment Case Study : Wehe - TD Detection Tool for Mobile Environment Shortcoming of Wehe on HTTPS Traffic Shortcoming of Wehe on HTTPS Traffic TD Detection of HTTPS Traffic TD Detection of HTTPS Traffic Conclusion & References Conclusion & References Abstract We consider the problem of detecting deliberate traffic discrimination on the Internet. Given the complex nature of the Internet, detection of deliberate discrimination is not easy to detect, and tools developed so far suffer from various limitations. We study challenges in detecting the violations (focusing on the HTTPS traffic) and discuss possible mitigation approaches. We focus on ‘Wehe,’ the most recent tool developed to detect net-neutrality violations. Wehe hosts traffic from all services of interest in a common server and replays them to mimic the behavior of the traffic from original servers. Despite Wehe’s vast utility and possible influences over policy decisions, its mechanisms are not yet validated by others. In this work, we highlight critical weaknesses in Wehe where its replay traffic is not being correctly classified as intended services by the network middleboxes. We validate this observation using a commercial traffic shaper. We propose a new method in which the SNI parameter is set appropriately in the initial TLS handshake to overcome this weakness. Using commercial traffic shapers, we validate that SNI makes the replay traffic gets correctly classified as the intended traffic by the middleboxes. Our new approach thus provides a more realistic method for detecting neutrality violations of HTTPS traffic. Index Terms —Net neutrality, traffic differentiation, detection tools Index Terms I. INTRODUCTION Net neutrality is a guiding principle promoting the “equal” treatment of all packets over the Internet. However, for economic benefits, ISPs may apply traffic differentiation to a specific service, user, content provider, or any other traffic group on the Internet without making any public declaration. It gives rise to a need to have tools that can detect such malicious activities over the Internet. Traffic differentiation (TD) detection involves the coalescence of many elements such as end-systems (user-client and server), probing traffic generation, expected network responses, and TD detection algorithms. These are interdependent components or operations. Hence, in developing TD detection tools, one faces challenges such as crafting internet traffic and conditioning measured network response that suits their detection algorithm. Moreover, as HTTPS traffic becomes prevalent, it is unclear what policies the middle-boxes apply for discrimination as payload provides no signatures. Our first goal is to study various challenges in designing a reliable TD detection mechanism for HTTPS traffic Several methods have been proposed in the literature to detect Network neutrality violations as documented in recent surveys [1], [2]. The literature is rich with various forms of discrimination and its detection, like discrimination of content providers [3], end-users [4], specific services like BitTorrent [5]. Our interest in this work is discrimination of services, specifically on streaming services (both audio and video) which are potential candidates from description due to their commercial values and high bandwidth requirements. Recently several tools are developed to detect discrimination of services like, NANO [6] ChkDiff [7], Netpolice [8], Shaperprobe [9], Packsen [10], Glasnost [11], Bonafide [12], and Wehe [13]. NANO is based on passive measurements, and all others are based on either active or differential probes. Wehe is the latest tool that overcomes many of the drawbacks of the other tools. However, due to the Internet’s complexity, many issues persist that prevent its use as a reliable tool. Wehe [13] follows a client-service architecture in which content of service of interests (Youtube, Netflix, etc.) are copied (or sniffed) from their respective original servers and hosted on a common server (referred to as replay server). To check if a particular service is discriminated, the client accesses the content of that service from the replay server. The server transfers the requested data with timing relations that mimic the original service’s data transfer characteristics over the Internet. The client accesses content from the replay server with and without VPN and compares them to ascertain if there was any difference in the quality. The basic idea of Wehe is that connection over VPN is encapsulated and will not be subjected to any discrimination as the middleboxes cannot classify the traffic correctly. Whereas content accessed without VPN can be classified correctly due to the signatures induced by the timing relations and can be subjected to discrimination policies. Despite the Wehe tool’s vast utility and possible influences over policy decisions, its mechanisms are not yet fully validated by other researchers. We investigate Wehe’s differentiation detection’s performance on traffic using HTTPS protocol, which is predominantly used by all services for security reasons and to avoid detection. We noticed that Wehe uses port 80 (HTTP) for all communications. Middleboxes process all the traffic on this port as non-encrypted traffic and may not classify them correctly as they will not see any signatures (recall that replayed traffic is HTTPS). On the other hand, when Wehe access traffic over port 443 (HTTP), traffic is not classified as intended traffic as we demonstrate it using a commercial traffic shaper. As middleboxes do not see traffic from the Wehe’s replay server as that coming from the original server, they may not apply intended traffic discrimination policies. Thus Wehe may not detect any deliberate discrimination resulting in false negatives. in false negatives. To overcome the limitations of Wehe for HTTPS traffic, we propose a new method in which traffic is accessed on port 443, but with a modification in the TLS handshake. In the TLS handshake, we explicitly send the Server Name Indication (SNI) corresponding to the actual service. We demonstrate that the HTTPS traffic get correctly classified by the middleboxes with these modifications. Our method ensures the middleboxes classify traffic from replay servers in the same way they do from the original servers for each service and apply the same discriminate policies (if any) on traffic from both replay and original servers. Thus, our method makes it possible to detect discrimination of HTTPS traffic in a realistic setting. Our contributions can be summarized as: We study various challenges associated with detection of discrimination of HTTPS traffic. We present the categorization of these challenges based on their source, e.g., as protocol and operational environment. We demonstrate limitations of Wehe in detection of discrimination of HTTPS traffic. Specifically, we demonstrate that HTTPS traffic will not be correctly classified in the Wehe setup using a commercial traffic shaper. We propose a mechanism in which traffic of services accessed from the replay server is treated as if it originated from the actual servers. Thus traffic from the replay server also gets subjected to the same discriminated policies applied on the actual service. We We validate that replay traffic of our mechanism is treated in the same way originating from actual servers using a commercial traffic shaper. We study various challenges associated with detection of discrimination of HTTPS traffic. We present the categorization of these challenges based on their source, e.g., as protocol and operational environment. We study various challenges associated with detection of discrimination of HTTPS traffic. We present the categorization of these challenges based on their source, e.g., as protocol and operational environment. We demonstrate limitations of Wehe in detection of discrimination of HTTPS traffic. Specifically, we demonstrate that HTTPS traffic will not be correctly classified in the Wehe setup using a commercial traffic shaper. We demonstrate limitations of Wehe in detection of discrimination of HTTPS traffic. Specifically, we demonstrate that HTTPS traffic will not be correctly classified in the Wehe setup using a commercial traffic shaper. We propose a mechanism in which traffic of services accessed from the replay server is treated as if it originated from the actual servers. Thus traffic from the replay server also gets subjected to the same discriminated policies applied on the actual service. We We propose a mechanism in which traffic of services accessed from the replay server is treated as if it originated from the actual servers. Thus traffic from the replay server also gets subjected to the same discriminated policies applied on the actual service. We We validate that replay traffic of our mechanism is treated in the same way originating from actual servers using a commercial traffic shaper. We validate that replay traffic of our mechanism is treated in the same way originating from actual servers using a commercial traffic shaper. The paper is organizaed as follows. Sec. II provides the necessary background and related work. Sec. III describes all identified challenges in measurement setup for TD detection. Sec. IV describes the Wehe tool and its mechanisms in the context of identified challenges, Sec. V provides the validation results. Our proposed method is described in Sec. VI and Sec. VII gives the conclusions. This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Challenges in Net Neutrality Violation Detection: A Case Study of Wehe Tool and Improvements

This paper is available on arxiv under CC 4.0 license. Authors: (1) Qin Wang, CSIRO Data61, Australia; (2) Shiping Chen, CSIRO Data61, Australia. Table of Links Abstract & Introduction Ethereum Accounts Account Abstraction Security Analysis Further Discussions Conclusion & References III. ACCOUNT ABSTRACTION A. Previous Attempts A series of continuous efforts have been made to make existing accounts extensible with more functions. We list a series related token standards in Tab.II. In Ethereum’s early phases, attempts were made to distinguish between EOA and CA by introducing various new transaction types. EIP-86/208 leverage the differentiation between two accounts to design customizable and collisionresistant contract addresses. This, in turn, led to the implementation of EIP-1014 and EIP-2470. EIP-859 introduced transaction-initiated contract deployment. It allowed for onthe-spot deployment of contract addresses if none existed, forming the basis for the fundamental functionality of EIP4337. EIP-2718 ushered in compatibility for future Ethereum iterations with any newly suggested transaction types. EIP293 systematically cataloged various advantages of contract accounts, encompassing recovery, key rotation, customized identity verification algorithms, and meta-transactions, which collectively entrenched the role of contract accounts. However, these efforts resulted in unwieldy complexity. Altering transaction types requires concurrent modifications to the underlying signature verification algorithms. This encompasses considerations like miner acceptance of new types, ensuring incentives are not lower than regular transactions to encourage verification, and addressing concerns about account address management and conflicts. Unfortunately, this approach lacked both backward and forward compatibility. Ultimately, these attempts have coalesced around two primary directions as stated in our Sec.I. B. Account Abstraction Account abstraction introduces a suite of functions designed to personalize fundamental elements that enable smart contract functionalities. These customizations encompass various aspects, including user operations, fee payment methods, and transaction packing mechanisms. . The workflow is summarized as follows: ① Users initiate interactions with the frontend abstraction (also referred to as application) layer, where their actions are translated into underlying transactions (via UserOperation). ② Bundlers aggregate multiple UserOperation instances, consolidating them into a single transaction that is then transmitted to the EntryPoint contract. ③ Within the EntryPioint contract, user signatures are verified and transactions initiated by the abstraction layer are processed. ④ The entries logged in UserOperation prompt the activation of relevant smart contracts, initiating state transitions tailored to their specific prerequisites. ⑤ Optionally, the user’s operations can be aggregated and authenticated using BLS signatures. Meanwhile, transaction fees for user actions are managed by the Paymaster contract. Consequently, on-chain applications interact with user actions in a manner similar to their interaction with standard external accounts. Overview UserOperation serves as a transaction-like entity, representing a user’s request achieved by transaction events. Within UserOperation, multiple requests and supplementary data can be encapsulated, facilitating the execution of transactions that a CA can perform. While sharing similarities with conventional transactions, UserOperations exhibit several distinctive characteristics. ✑ User operation. UserOperation includes new fields in the transaction structure, e.g., EntryPoint, Bundler, Paymaster and Aggregator. • Additional fields. UserOperation is sent to a separate mempool, where bundlers can package them into transactions which get included in a block. • Alternate mempool. Today, transaction inputs are specific, e.g., swap 2k USDC for 1.2ETH. In contrast, UserOperation can be decorated with additional metadata to be more outcome-focused, e.g., I want to trade 2k USDC for the most amount of ETH possible. • Intent based. Bundlers constitute a pivotal component of the infrastructure necessary for the realization of EIP-4337. These entities are a specific type of Ethereum node dedicated to facilitating the processing of UserOperations. These UserOperations are directed to a network of bundlers, which diligently monitor the mempool. Bundlers efficiently consolidate multiple UserOperations into a singular transaction, expertly packaging and submitting them to the blockchain on behalf of users. In exchange for this service, bundlers receive compensation for their efforts in executing these tasks. ✑ Bundler. It is a singleton smart contract that undertakes the verification and execution of UserOperations. ✑ EntryPoint. The process checks the account balance. It involves assessing whether the wallet possesses sufficient funds to cover the maximum potential gas expenditure (derived from the gas field in the UserOp). Transactions lacking adequate funds will be declined. • Verification loop. Upon verification, the transaction will be executed. Correspondingly, an amount from CA is deducted to reimburse the bundler. This reimbursement is required to cover the gas expenses. • Execution loop. In scenarios where multiple messages are signed using distinct keys, an aggregator plays a role in producing an authenticated signature. The signature inherently validates the authenticity of all individual signatures within the collection. For accounts associated with particular signature types that enable aggregation (such as BLS), the verification process of the account’s signature is deferred to an external contract. This external contract, in turn, is responsible for verifying a solitary signature across an entire bundle, thus streamlining the validation process. ✑ Aeggregator (optional). Through the consolidation of multiple signatures into a singular entity, aggregators contribute to the efficient management of data availability. This allows for the validation of numerous bundled UserOperations in a single consolidated process. The Paymaster takes charge of implementing gas payment policies, offering developers the capability to offer their end users gas-free interactions through sponsored or ERC-20 gas policies. These policies introduce flexibility in how gas is paid, such as the choice of currency (e.g., native ETH or ERC-20 tokens), and by whom it is paid. This effectively eliminates the need for users to possess native blockchain tokens to engage with the blockchain. ✑ Paymaster (optional). C. AA Benefits Based on its design, AA brings a series of direct benefits that previous CA and EOA do not have. Capitalizing on the capabilities of the EntryPoint contract, the consolidation of multiple transactions can guard against frontrunning attacks and potential MEV threats. • Transaction batching. . Departing from reliance solely on a single signer (EOA, private key holder), AA introduces the capability of multi-signatures sourced from a spectrum of entities. • Multisig ERC-20 tokens are accepted as valid payment for these fees, unshackling the restrictions previously confined to ether (ETH). • Gas abstraction. AA is engineered to encompass a mechanism for recovery through a social network of trusted associates. This mechanism offers an alternative route to regain access to the wallet. • Social recovery. Users are empowered with the ability to craft bespoke modules catering to specific functions. • Custom module. D. AA Surroundings EIP-2771 is designed to enable gasless transactions for users by introducing the concept of metatransactions. It permits third parties to cover a user’s gas expenses without necessitating modifications to Ethereum. The standards include the following key steps: The transaction signer signs and dispatches transactions to a gas relay. The Gas Relay, operating off-chain, receives these signed requests and subsidizes the gas costs, transforming them into valid transactions that are then routed through a trusted forwarder. This trusted forwarder is essentially a contract entrusted by recipients to accurately verify signatures and nonces. Supportive standards. EIP-3047 is designed to enable users to delegate control of their EOA to a contract, a parallel approach discussed in Sec.I. This feature allows any EOA to function as a smart contract wallet without requiring the deployment of a separate contract. As part of this standard, two new opcodes are introduced to the Ethereum Virtual Machine (EVM): AUTH and AUTHCALL. These opcodes facilitate proper invocation from smart contracts. Although this standard are incorporated into future updates of Ethereum’s account infrastructure, it still provides valuable insights. EIP-6900 introduces a standard aimed at synchronizing the implementation efforts of CA developers and plugin developers. It outlines the creation of a modular CA with the capability to accommodate all plugins adhering to standard protocols. In this standard, CAs are rendered as fully programmable accounts capable of housing assets within on-chain smart contracts. In parallel, account plugins serve as interfaces for smart contracts that facilitate the integration of compositional logic within CAs. This approach enhances data portability for users and alleviates the need for plugin developers to target specific wallet implementations. EIP-6551 introduces a token standard primarily focused on NFTs [14]. This standard enables the binding of one or multiple Ethereum accounts to an NFT. By doing so, NFTs gain the capability to possess assets and engage with applications, all without necessitating modifications to current smart contracts or infrastructure. These accounts, which are associated with tokens, seamlessly integrate with existing on-chain asset standards and have the potential to encompass forthcoming asset standards. Ethereum’s abstract account model offers an enhanced approach to implementing this standard, exemplified by the Web3 game project Sapienz [15]. In light of the forthcoming updates to Ethereum, a diverse range of competitive blockchain platforms are also stepping up their game. Notable contenders such as BNB Chain, Avalanche, Optimism, Polygon, StarkNet, and zkSync have all unveiled their respective initiatives for implementing crucial updates. Supportive platforms. Complementing the infrastructural advancements, an equally pivotal component lies within the realm of wallets. Serving as the hub where users manage, utilize, and safeguard their accounts, wallets play a pivotal role. Multiple wallet service providers, including MetaMash, Argent, Beam, Safe, Trust Wallet, Braavos, and OKX Wallet, have expressed their commitment to seamlessly incorporate these anticipated updates into their products. Supportive wallet. E. Real Case Adoption . As demonstrated in [16], Visa is actively exploring a solution that leverages the Paymaster smart contract. This solution aims to abstract the core interactions of blockchain technology and enhance the user payment experience by introducing a self-custodial smart contract wallet. The primary goal is to streamline the process for users conducting transactions within their wallets. One of the intriguing aspects is that users can now use any token to pay for gas fees, and these gas costs will be covered by Paymaster. In more detail, users have the capability to make payments to the Paymaster contract using USDC. Subsequently, the Paymaster contract converts the USDC into ETH and facilitates the transaction on the blockchain network. After a period of on-chain processing, the recipient of the transaction can receive an equivalent amount converted back to USDC. As of now, this particular use case is still in the conceptual verification stage. Visa By seamlessly integrating AA into existing EOA wallets, the accounts are elevated to the status of smart contract wallets, enriched with programmable logic and functionalities. For instance, Gnosis Safe [17] brings forth a smultisignature approach, necessitating multiple authorized entities to provide signatures for the same account, rather than relying solely on individual private keys. Argent [18] introduces the concept of social recovery, enabling users to recover lost or forgotten private keys. It allows users to utilize email addresses and phone numbers for offline recovery, introducing a familiar two-factor authentication mechanism. Users in Braavos [19] can access their wallets using the biometric features of their smartphones, such as facial or fingerprint recognition, adding an extra layer of security. Web3 wallet.

Account Abstraction, Analysed: Account Abstraction

As a developer, how many times each day do you look something up online? I'm not talking about a simple piece of syntax, I'm talking about the things that are a bit harder to find. For example: Is there a good way to compile C to WebAssembly and run it in a browser without needing to modify the code itself? Can the Stripe UI export promotion code redemptions as a CSV? Is there a tool for hot module reloading in Go? good Is it socially acceptable to go out in public in South Dakota with blue hair? While working as a team lead a few years ago, I remember asking Jeffy, a developer on my team: "Is there a way to get to use pointers for nullable columns instead of the sql.Null types?" sqlc Jeffy went heads down for... much longer than I would have expected. After more than a full day of research, he came back with: "No, there isn't a way. I found something that seemed promising, but it doesn't work. I think we should fork the project and add it ourselves." I've got some questions. Hol' up. First, you're telling me that you found what we want in the API of a well-known project, but it's broken and knows that it's broken but you? You are confident there aren't any skill issues involved? open-source no one else Second, you're telling me that your recommended course of action is for our team of four to fork a project with many thousands of lines of code and maintain our own version ad infinitum? All for a little quality of life improvement? No, Jeffy. No. What was Jeffy's problem? Counterintuitively, I believe Jeffy's problem was that he didn't know when to give up. I don't mean on the task as a whole, I mean on each individual path he was taking to find the answer. Here's a chart. Jeffy was too attached to his initial searching methods. Maybe he was too attached to one page from the docs, to his initial search query, or to Google as a search mechanism altogether. Jeffy found a and became stuck. He should have bailed and tried looking elsewhere. local maximum so much sooner If I don't find good evidence that the answer I'm looking for exists in the place I'm searching, I'm bailing pronto. That doesn't mean I'm 100% certain it doesn't exist, but you usually don't have to read an entire Stack Overflow thread to know whether or not it relates to your problem. I'm honestly amazed at how some developers take so long to look stuff up. Now, in their defense, even if you're an efficient searcher, you can still get stuck in a "quagmire": a local maximum. If you're not careful, you can find yourself reading on and on about a topic that seems like what you're looking for, only for it to skim over the details that you actually need. Bummer. exactly I'm a big fan of using AI as a replacement for some of my would-be Google searches, but Chat GPT is the quagmire, i.e its the ultimate local maximum. It's the worst kind of quagmire because at every step it will confidently reassure you that it does, in fact, know the answer. When you insist that it's wrong, it confidently continues to provide information that sounds correct but isn't. ultimate If you use Chat GPT for programming assistance, you to develop a strong sense of when ChatGPT is making stuff up. In short, need you need to know when to bail. RTFM Sometimes we talk about "Googling things" as a skill unto itself, and it kind of is. But similar to ChatGPT, it's less about formulating the most pristine search query, and it's more about knowing when to give up on a goose chase. Does the SERP indicate that your original search query was a good one? No? Bail. Does the first page you visit have a directly relevant title or first paragraph? No? Bail. After spending another minute on the page are you still convinced the answer is somewhere below? No? Bail. Is it likely the answer can be found more easily in the docs, the user manual, or, , the code itself? Yes? Bail. gasp Reading the manual, perusing the official docs, or digging through the source code usually takes longer than skimming an on-topic article or tutorial, but SEO-spammed articles often coalesce into an entire sea of local maxima. They tease you by scratching the surface of what you're looking for, but if your issue isn't the most common one, you can waste hours trudging through the garbage. Sometimes, taking the into the topic is the fastest (only?) way to find answers to the more esoteric questions. time to go deep Back to Jeffy I don't want to leave you hanging on this story. As it turned out, the documentation for SQLC was complete, but it was tricky to understand how to use pointers for nullable columns if you landed on the wrong page. There are a lot of configuration options for SQLC, and it's easy to get lost. I just cloned the repo, grepped for "pointer", and found the code that parsed the configuration option. It was clear after just a few minutes of reading the source that we just needed to do three things: Specify the we want (e.g. "string" instead of ) go_type sql.NullString Set nullable: true Set pointer: true Here's the code: overrides:
  - db_type: "text"
    nullable: true
    go_type:
      type: "string"
      pointer: true Sometimes you gotta rtfm. Sometimes you gotta suck it up and grep the code. Also published here. By the way , is awesome and if you like Go you should be aware of it. sqlc

Every Developer Should Learn to Give Up Sooner

This paper is available on arxiv under CC 4.0 license. Authors: (1) Amit Kumar Chatterjee, Yukawa Institute for Theoretical Physics, Kyoto University & Department of Physics, Ramakrishna Mission Vidyamandira; (2) Satoshi Takada, Department of Mechanical Systems Engineering and Institute of Engineering; (3) Hisao Hayakawa, Yukawa Institute for Theoretical Physics, Kyoto University. Table of Links Introduction Model and Protocol Different Regimes in d˜− Γ˜ Plain and Multiple QMPE QMPE in Region (d): Second-order Ep QMPE in Region (a1): Oscillations Summary Appendix A: Eigenvectors of L in region (d) Appendix B: Eigenvectors of L in region (a1) Appendix C: QMPE in region (b): purely exponential relaxation Appendix D: QMPE in region (e): 3rd order EP References Appendix D: QMPE in region (e): 3rd order EP The region (e) in the (d,˜ Γ) plane of Hatano’s model is composed of a single point It is a third order EP where all the non-zero eigenvalues become equal and their eigenvectors coalesce. Since this is a single point with fixed values of ˜d and Γ, the right eigenvectors are explicitly given by Of course |r1i and hℓ1| have same expressions both in Eqs. (3) and (D2). The left eigenvectors are ) The non-zero eigenvalue of the Lindbladian [Eq. (2)] in region (e) is −iλ = −i4√ 3. Importantly, as a consequence of the 3rd order EP, we can only achieve an approximately diagonal form i.e. Jordan normal form LbJn leading to the equation Due to the off-diagonal nonzero defects in Eq. (D5), the time evolution of density matrix elements and other observables pick up additional algebraic terms in time. 1. Density matrix elements In region (e), the density matrix elements have the expressions below In Fig. 15(a), we present existence of t˜+ and t˜− as function of the control parameters ˜dI and ˜dII using the fixed dissipation protocol. The solutions’ surfaces in Fig. 15(a) reveal that there are regions where only one of the solutions t˜+ or t˜− exist leading to single QMPE, and in other regions both the solutions exist resulting in double QMPE. Particularly, lower values of ˜dI and ˜dII correspond to single intersection time solution and therefore single QMPE. To understand the behaviors of t˜+,− more clearly, we fix ˜dI and display t˜+,− as functions of ˜dII in Fig. 15(b). The Fig. 15(b) shows that the two branches t˜+ and t˜− interchange as ˜dI < ˜dII. More specifically, for ˜dI > ˜dII we have t˜+ > t˜− and for ˜dI < ˜dII we have t˜+ < t˜−. The underlying reason is that the discriminant p γ 2 1 − 4γ2γ0 in Eq. (D8) is proportional to sign of ( ˜dI − ˜dII). Thus the two intersection times interchange their roles by changing the sign accompanying the discriminant as we go from ˜dI > ˜dII to ˜dI < ˜dII. The Fig. 16(a) demonstrates an explicit example of double QMPE in ground state probability in region (e). Like other regions, here also the second intersection (shown in the inset of Fig. 16(a)) is much feeble in comparison to the first intersection. 2. Average energy For the same parameters used to obtain QMPE for ρgg(t) in Fig. 16(a), we would like to examine the possibility of QMPE in average energy E(t). Indeed, we observe in Fig. 16(b) that energy also exhibits double QMPE, marked by the twice intersection of zero by the energy difference ∆E(t) between the two copies. The insets of Fig. 16(b) shows separately the two intersections in the time evolution of the energies for the two copies. The second intersection is much weaker in magnitude compared to the first intersection just like the case of ground state probability. Note that the first intersection between the energies of I and II occur after they overshoot their steady state value which they reach during some early time of the relaxation process. 3. von Neumann entropy Here we investigate the von Neumann entropy SvN(t), starting from two copies I and II. We use same parameters that have produced double QMPE both in ground state probability [Fig. 16(a)] and average energy [Fig. 16(b)]. In fact, the Fig. 17(a) shows that the entropy also exhibits double QMPE indicated by twice zero intersections (the second one shown in inset of Fig. 17(a)) of the entropy difference ∆SvN(t) between the two copies. 4. Temperature In this section we study the thermal QMPE in region (e). We observe thermal QMPE in region (e) by using variable dissipation protocol in Fig. 17(b). There we observe single intersection and reversal of identities (i.e. hotter becoming colder and vice versa) occur after the negative temperatures have happened. So, such temperatures can be considered as well-defined and therefore this intersection in Fig. 17(b) corresponds to single QMPE in temperature in region (e).

Multiple Quantum Mpemba Effect: Exceptional Points and Oscillations: Appendix D: QMPE in Region (e)

Authors: (1) Xiyu Zhou, School of Computer Science, Wuhan University, Wuhan, China; (2) Peng Liang, School of Computer Science, Wuhan University, Wuhan, China; (3) Zengyang Li, School of Computer Science, Central China Normal University, Wuhan, China; (4) Aakash Ahmad, School of Computing and Communications, Lancaster University Leipzig, Leipzig, Germany; (4) Mojtaba Shahin, School of Computing Technologies, RMIT University, Melbourne, Australia; (4) Muhammad Waseem, Faculty of Information Technology, University of Jyväskylä, Jyväskylä, Finland. II. METHODOLOGY The goal of this study is to systematically identity the issues that developers encountered in the practical application of GitHub Copilot, as well as their underlying causes and potential solutions. We formulated three RQs to direct subsequent phases of the methodology as shown in Fig. 1. The RQs and their rationales are detailed in Section II-A. A. Research Questions Rationale: Copilot is a relatively new product, and little is known about the specific challenges and issues users face while using it in the software development practice. By identifying these issues, this study can help us gain a better understanding of the obstacles that arise when using AI code generation tools like Copilot in practical situations from the perspective of software developers. RQ1: What are the issues faced by users while using Copilot in the software development practice? Rationale: Understanding the underlying causes of the issues identified in RQ1 is essential for developing effective solutions to address the issues. By identifying these causes, the study can provide insights into how to improve the design and functionality of Copilot. RQ2: What are the underlying causes of these issues? Rationale: RQ3 aims to identify potential solutions to the issues identified in RQ1, which can ultimately enhance the usability of Copilot for users. By exploring the existing practices and methods utilized by developers to tackle the issues, the study can gain insights into potential solutions that enhance the functionality and usability of Copilot. RQ3: What are the potential solutions to address these issues? B. Data Collection We collected data from three sources: GitHub Issues[1] , GitHub Discussions[2] , and SO posts[3] . GitHub Issue is a commonly used feature on GitHub for tracking bugs, feature requests, and other issues related to software development projects, which allow us to capture the specific problems and difficulties that users have encountered when coding with Copilot. GitHub Discussion, on the other hand, is a newer feature for more open-ended discussions among project contributors and community members, which also offers a central hub for project-related discussions and knowledge sharing among community members. Discussion topics can cover a wide range from technical questions and proposals to broader topics related to Copilot. Stack Overflow is a popular technology community that provides a Q&A platform covering a wide range of programming, development, and technical questions, including Copilot-related questions. Considering that Copilot was announced and started its technical preview on June 29, 2021, we chose to collect the data that were created after that date. The data collection was conducted on June 18, 2023. To assist us in addressing RQ2 and RQ3, we chose to collect closed GitHub issues and answered GitHub discussions, which contain known causes and solutions. For SO posts, we found that the amount of Copilot-related posts was relatively low. Hence, we chose to retrieve all potentially relevant posts, including unanswered ones, to obtain a more comprehensive dataset. For GitHub Issues, we used “Copilot” as the keyword to search closed Copilot-related issues in the entire GitHub, and a total of 4,057 issues were retrieved. We also employed “Copilot” as the keyword to search at SO, resulting in 679 retrieved posts. Different from GitHub Issues and SO posts, GitHub Discussions related to certain product are grouped under a specific subcategory, and “Copilot” was included as a subcategory under the “Product” category. Given the high relevance of these discussions to Copilot, we collected all the discussions that were answered under the “Copilot” subcategory as part of the data source for our study. In total, 925 discussions were obtained. C. Data Filtering To ensure that the data can be used to answer our RQs, we set the filtering criterion: the issue, discussion, or post should contain specific information related to the use of Copilot. We conducted the data labelling on the collected data to filter out the data which cannot be used for this study by following the criterion. Pilot Data labelling: To minimize personal bias in the formal labelling process, the first and third authors conducted a pilot data labelling. For GitHub issues and discussions, we randomly selected 100 and 25 respectively, which constitute 2.5% of the total count. Due to the smaller quantity of SO posts, we randomly selected 35, which constitutes 5% of the total posts. Selecting a certain proportion of data from different platforms respectively is to verify whether the criteria of the two authors are consistent across various platforms. The consistency of the labelled issues was measured by the Cohen’s Kappa coefficient [14], resulting in values of 0.824, 0.834, and 0.806. The results indicate that a reasonable level of agreement between the two authors. For the labelling results with differences of opinion, the two authors will engage in discussions with the second author to reach a consensus. 2) Data labelling: The first and third authors then conducted the formal data labelling. During this process, we excluded a large number of data not related to our research. For instance, in some issues, Copilot refers to other meanings, such as the co-pilot of an aircraft. Additionally, Copilot might be mentioned in a straightforward manner without additional information, like someone said, “You can try using Copilot, which is amazing”. We also excluded such issues since they could not provide useful information about the use of Copilot. During the labelling process, any result on which the two authors disagreed was subject to discussion with the second author until an agreement was reached. Ultimately, the two authors collected 476 GitHub issues, 706 GitHub discussions, and 184 SO posts for data extraction. D. Data Extraction and Analysis Extract Data: To answer the three RQs mentioned in Section II-A, we established a set of data items for data extraction, as presented in Table I. Data items D1-D3 intend to extract the information of issues, underlying causes, and possible solutions from the filtered data, to answer RQ1-RQ3, respectively. The first author conducted a pilot data extraction with the third author on 20 randomly selected GitHub issues, 20 discussions, and 20 SO posts, and in case of any discrepancies, the second author was involved to reach a consensus. Based on the observation, we established the following standards for data extraction: (1) If the same issue was identified by multiple users, we recorded it only once. (2) If multiple problems were identified within the same issue, discussion, or posts, we recorded each one separately. (3) For an issue that has multiple causes, we only recorded the confirmed cause by the contributor of the issue or the Copilot team as the root cause. (4) For an issue that has multiple solutions suggested, we only recorded the solutions that were confirmed by the contributor of the issue or the Copilot team to actually solve the issue. The first and third authors conducted data extraction from the filtered issues, discussions, and posts based on the data items, and then discussed and reached a consensus with the second author on inconsistencies to ensure that the data extraction process adhered to the predetermined criteria. Each extracted data item was reviewed multiple times by the three authors to ensure accuracy. The final data extraction results were compiled and recorded in MS Excel [13]. 2) Analyze Data: To answer the three RQs in Section II, we conducted data analysis using the Constant Comparison method [12]. The specific steps are as follows: 1) The first author meticulously reviewed the data obtained during the data extraction phase and assigned codes to each data. These codes constituted descriptive summarizations of the data, aimed at capturing the underlying themes. For instance, the issue in Discussion #10598 was coded as “Stopped Giving Inline Suggestions”. 2) The first author compared different codes to identify patterns, commonalities, and distinctions among them. Through this iterative comparison process, similar codes coalesced into higher-level types and categories. For example, the coding of Discussion #10598, along with other akin codes, fell under the same type of FUNCTIONALITY USAGE ISSUE, which further belonged to the category of Usage Issue. Ultimately, the first author engaged in discussions with the second and third authors to achieve a consensus on the taxonomies of issues, causes, and solutions, which are presented in Section III. [1] https://docs.github.com/en/issues [2] https://github.com/orgs/community/discussions/categories/copilot [3] https://stackoverflow.com/ This paper is under CC 4.0 license. available on arxiv

GitHub Copilot in Practice: Empirical Insights into User Experiences and Practical Challenges

In an engaging conversation with Dimitrios, CEO of the groundbreaking startup Ithaca, we explore his fascinating transition from the high echelons of Goldman Sachs to the dynamic world of cryptocurrency and blockchain technology. Dimitrios shares invaluable insights into how his extensive background in finance and his entrepreneurial ventures, especially with Nomisma, have shaped his strategic vision for Ithaca. Exclusive Insight: How Ithaca Stands Apart in the Crowded Crypto Space Ishan Pandey: Hi Dimitrios, I am delighted to welcome you to our "Behind the Startup" series. Can you share insights into your background, particularly how your experiences at Goldman Sachs and your involvement in prior ventures such as Nomisma have shaped your perspective and approach in your role as the CEO of Ithaca? Dimitrios : I feel blessed to have experienced markets at the highest level from all perspectives: from performing academic research with Nobel prize winners at the University of Chicago, to rising to partner through the ranks of Goldman Sachs, and finally through my own ventures. Goldman has an entrepreneurial culture, and I was fortunate to thrive in that environment. Whether I was working with trailblazing market teams in Europe in the 2000s, navigating through the GFC, or building and leading businesses in Asia Pacific, those experiences prepared me for what was to come. I went on to work in commodities, emerging markets and crypto, and built and led a multi-family office business. Along the way I learned a great deal, including hard lessons about asset allocation and risk management. I started dabbling in crypto in 2013 and was too early in building a regulated digital asset derivative business in Nomisma. Now having spent the last bull market and its denouement at a firm that was deeply enmeshed in the workings of crypto markets I feel as though my long career has been preparing me to build Ithaca all along. Derivatives are the most exciting segment in a truly revolutionary market and I am lucky to have the background to seize this opportunity. Ishan Pandey: How does Ithaca's options trading protocol differ from existing solutions in both centralized and decentralised venues? Dimitrios : I will take it one step further. We differ from anything that has been presented across Tradfi, CeFi and DeFi across enough dimensions that you could deem our approach to be revolutionary, both in terms of liquidity creation for the benefit of end users, but also in terms of potential disruption to existing markets. Our major innovation is the ability of our matching engine to directly incorporate a variety of rules that are usually internalized by market makers, bringing these to bear at the user level to improve market liquidity. We can do that because we utilize auctions instead of continuous limit order books. While there is some debate on the relative merits of auctions in the context of spot / delta one markets, they have not been implemented in the context of derivatives where they are most needed. The principles of replication and portfolio dominance in the context of full composability of all statically replicable payoffs - including, by the way,  digital options - allow for synergistic liquidity across all orders for all products to coalesce around the one auction. One gets a much bigger buck for a given liquidity buck. Ishan Pandey: How does Ithaca address the challenges related to on-chain governance? Dimitrios : We will present an internally consistent, incentive aligned governance framework in the context of our point system. I am confident that our community engagement, our investor support, and our background in mechanism design and corporate finance will not only allow us to address usual concerns, but also to promote the healthy growth of our ecosystem for all stakeholders. Ishan Pandey: Could you provide insights into the decision to use Off-chain Java backend pre-match processing and Arbitrum post-match settlement in Ithaca's architecture? Dimitrios: We like to describe our system as an off-chain oracle that manages matching and collateral optimization as well as on-chain settlement. The approach is well established among other DeFi protocols. In any case, we would like the community’s focus to be on our financial engineering innovation. The computational intensity of the matching engine does not allow for on-chain calculations, but there are both trivial and nontrivial solutions towards more trustlessness and decentralization. This is a road we will travel on post launch. Ishan Pandey: Can you explain Ithaca's revenue streams, including the key offerings like trading fees, Position Builder, Collateral Optimization, and Pre-packaged Strategies? Dimitrios: Our fees will be competitive with any other DeFi or DeFi options offering. Considering the fact we will be offering liquid digital options, our associated fees by transparency will be market-leading. Further considering our composability competitive edge, our fee structure will incentivize experimentation with custom conditional orders for experienced users, and pre-defined strategies and ‘stories‘ for users looking for more convenience or in need of a bit more hand-holding. Our state of the art collateral optimization engine will be free at the point of use. Ishan Pandey: As the CEO, how do you ensure compliance and regulatory engagement for Ithaca's operations? Dimitrios: There is a clear consensus emerging on how aspiring DeFi protocols protect themselves and their users from land mines: we follow best practices in terms of restricted jurisdictions and the protocol presents thorough and strict terms and conditions of use. Ishan Pandey: How do you see the future of Web3 derivatives trading evolving, and what role does Ithaca play in shaping that future? Dimitrios : Options are punching below their equivalent weight in crypto compared to any other asset class. Ithaca will not only address all the pain points, but it is also uniquely designed to provide a unified framework towards short-cutting the creation of liquid derivatives markets on new asset classes. As RWAs flock on chain and streams of social activity are ‘financialized’, Ithaca is uniquely placed to be the vehicle of their derivatization: the opportunity is mind boggling. Don’t forget to like and share the story! This author is an independent contributor publishing via our brand-as-author program. Be it through direct compensation, media partnerships, or networking, the author has a vested interest in the company/ies mentioned in this story. HackerNoon has reviewed the report for quality, but the claims herein belong to the author. #DYOR Vested Interest Disclosure:

COALESCE

coalesce.nyc

11-50 employees

Since 2005

COMPANY RANKING

COALESCE

EVERGREEN INDEX #3339

Representing this company? Click here to claim and customize this page!

Coalesce WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

COALESCE

coalesce.nyc

11-50 employees

Since 2005

COMPANY RANKING

COALESCE

EVERGREEN INDEX #3339

RELATED COMPANIESRANK

STORIES ABOUT COALESCE

Representing this company? Click here to claim and customize this page!

Coalesce WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!