投稿時間:2023-07-18 23:21:32 RSSフィード2023-07-18 23:00 分まとめ(27件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
IT	気になる、記になる…	Anker、スリムなQi対応ワイヤレス充電器｢Anker 318 Wireless Charger (Pad)｣を発売	https://taisy0.com/2023/07/18/174215.html	anker	2023-07-18 13:48:15
AWS	AWS Government, Education, and Nonprofits Blog	Announcing new AWS data connector for popular nonprofit CRM: Blackbaud Raiser’s Edge NXT	https://aws.amazon.com/blogs/publicsector/new-aws-data-connector-for-popular-nonprofit-crm-blackbaud-raisers-edge-nxt/	Announcing new AWS data connector for popular nonprofit CRM Blackbaud Raiser s Edge NXTThe AWS for Nonprofits team announced a new Amazon AppFlow data connector that enables nonprofits to transfer valuable data from Blackbaud Raiser s Edge NXT to AWS services and other destinations In this blog post learn some common nonprofit use cases that can be addressed by integrating your data with other AWS services and commercially available software as service SaaS applications	2023-07-18 13:28:32
python	Pythonタグが付けられた新着投稿 - Qiita	【保存版】Python Notion 操作 （notion-sdk-py編）	https://qiita.com/Yusuke_Pipipi/items/b44cb8442932019c52c9	notion	2023-07-18 22:40:45
python	Pythonタグが付けられた新着投稿 - Qiita	Python SimpleServerで同ネットワーク内ファイル共有	https://qiita.com/Mikotostudio/items/4dd4fe3059736cb69fc0	linuxmac	2023-07-18 22:32:32
python	Pythonタグが付けられた新着投稿 - Qiita	Fletを使って実際に業務アプリを作った感想	https://qiita.com/ForestMountain1234/items/b5cc9ca1364b178fedc6	経緯	2023-07-18 22:31:14
AWS	AWSタグが付けられた新着投稿 - Qiita	Terragruntを活用したTerraformの実践的なディレクトリ構成	https://qiita.com/tak0203753/items/0d9b8bb2d5cc9b6f84b4	terraform	2023-07-18 22:15:46
Docker	dockerタグが付けられた新着投稿 - Qiita	DockerやDocker Composeなどについて	https://qiita.com/ryouzi/items/a2e811a6d0381cd0c957	docker	2023-07-18 22:27:54
技術ブログ	Developers.IO	問題を課題を区別する	https://dev.classmethod.jp/articles/problem-and-issue/	組織開発	2023-07-18 13:52:25
技術ブログ	Developers.IO	[アップデート]全 AWS Fargate 利用者必見！ Seekable OCI インデックスによりコンテナの起動が大幅に高速化するようになりました	https://dev.classmethod.jp/articles/update-aws-fargate-seekable-oci/	awsfargate	2023-07-18 13:03:47
海外TECH	MakeUseOf	4 Reasons to Hire a Career Counselor and Where You Can Find One	https://www.makeuseof.com/reasons-to-hire-career-counselor-where-to-find/	career	2023-07-18 13:30:19
海外TECH	DEV Community	Building High Quality Android UI: Embracing Test Driven Development with Jetpack Compose	https://dev.to/zurcher/building-high-quality-android-ui-embracing-test-driven-development-with-jetpack-compose-21aj	Building High Quality Android UI Embracing Test Driven Development with Jetpack ComposeCover Photo by David Watkis on UnsplashHow can we approach UI development in Android while ensuring we are writing high quality software from the design spec Test Driven Development TDD is a way of approaching software development that is very handy for both feature writing and bug fixing It provides a path for developers to request more information when needed before writing “just in case software also known as YAGNI You Aren t Gonna Need It while fostering the creation of optimized code using the KISS principle Keep It Simple Stupid I guess Additionally TDD is highly effective for writing code that is easy to update Introducing a new test is straightforward once the suite is running and tests are always there for you if you accidentally or intentionally decide to update the behavior of your software So if we all agree that TDD is the way let s give it a try and attempt writing our UI with it using Jetpack Compose PlanFirst let s consider a common scenario for a mobile engineer receiving a new design specification for a brand new screen to be added to our app For this exercise I have chosen Taras Migulko s fantastic work on an E commerce app design from Dribbble Specifically we will be implementing one of the UI elements of a clothing list of items as depicted in the following screen Keeping this in mind let s set up our development environment In this case we will start a new project with Compose assuming that this step has already been completed for the purpose of this sample I ve created a default project using the templates of Android Studio for an Activity Compose Let s make sure our project is including all the required libraries to do TDD in Compose Lastly if we choose to follow a Test Driven Development TDD approach it is essential to remember and adhere to the three rules of TDD defined by Robert C Martin also known as Uncle Bob These rules serve as our development traffic lightswhen writing or updating any tests in the app Write production code only to pass a failing unit test Write no more of a unit test than sufficient to fail compilation failures are failures Write no more production code than necessary to pass the one failing unit test By following these rules we ensure that our development process remains focused and efficient ActionThe Design Spec is ready and the environment is set so How do we approach a new piece of UI with a test This will depend on each framework used for UI for this example we ll be using Jetpack Compose so it becomes relevant to understand how to “Think in Compose With Compose is easy to dissect the UI into very small pieces that are described as Composables that will be put together to render the full UI So the first thing we ll need to do Is have a close look into the UI to define how can we divide it and go from small to big Taking a first glance I ve identified at least five different Composables that will conform this screen A top Action Bar which includes the Back Name Item Count and Filter action A main List of clothing itemsEach Clothing ItemThe Clothing Item Content display name price and the add to cart action button The Favourite ButtonThere s already some added value in doing this as now we are starting to surface both reusable components such as the top action bar and also starting to think on naming for our composables without even touching a single line of code So let s go ahead and pick the smallest possible composable from our new screen the Favourite Button Assuming this behaves as any favourite button this will have to support states filled heart and empty heart Let s write our first test Worth mentioning at this stage that we ll be following the suggested documentation on Testing Compose so we also need to add the composeTestRule and initialise our screen using its setContent function Let s now bring in our test condition And when we write our Composable name that still does not exist we brake the second TDD rule the test no longer compilesTo address this issue we create a new Kotlin file and write our Composable And we go back to the testCompilation issue addressed We may continue writing the test now So the test is ready and compiles next step let s run itWe made it we have a failing test now Let s follow TDD rules and now to make It pass before writing any new tests And I know what you are thinking we are not checking the boolean We are not supporting the other scenarios we could write so much more code But this this is the exact point where you need to take that hat of and pick the TDD hat If you want to make the switch and only think Are we following all the TDD rules Yes then Its ok lets keep going Test is going fully green now this means we are able to write the next test So let s try the other scenario our composable needs to support displaying an empty heart if isFavourite is false And when we run it So once again following TDD rule number amp we go back to our composable and add the required logic to make this test pass being careful enough that our previous test will work too this time And we make sure to run the entire suite now so that both tests are ran And we can see all our tests going green now Finally we can update our tests to make sure that the other icon is not displayed in the incorrect scenario With all of this in place we now have peace of mind that our Composable will do exactly what we need to so we can take a pause on our Test writing and go to the Composable itself to make sure It looks as it should by adding Its background colors and right sizing ConclusionAs you can see with this process of following just the rules of TDD stoping when requried and moving on when we can we can ensure that our UI will work as expected and is fully covered in case we want to update it in the future a k a high quality software By incorporating Test Driven Development TDD into our development process we eliminate the guesswork associated with coding and relying on trial and error to verify functionality Instead our code is designed to precisely meet our requirements TDD acts as a lighthouse guiding us to write the necessary code while helping us identify any gaps in the requirements specification through thoughtful consideration Moreover implementing TDD in Jetpack Compose is a straightforward process as we have seen This approach resolves any potential issues during the design phase allowing us to confidently write our user interface UI knowing that it will seamlessly integrate with the business logic layers in the future In upcoming posts I ll continue developing this entire screen with TDD for now I encourage you to embrace TDD from the outset rather than writing tests after developing the code By doing so you will write less code think through your implementation thoroughly and easily identify any missing components in the requirements Embrace the power of TDD and have a wonderful day	2023-07-18 13:48:22
海外TECH	DEV Community	Use a Stored Column for Indexing Values Stored in JSON [MySQL Tip]	https://dev.to/brightdevs/use-a-stored-column-for-indexing-values-stored-in-json-mysql-tip-2imi	Use a Stored Column for Indexing Values Stored in JSON MySQL Tip Managing and querying JSON data in MySQL databases has become common in modern web applications To improve performance when querying JSON values you can create a stored column that extracts specific values and indexes them  Table DefinitionConsider a scenario where you have a table named orders that stores various details about customer orders including a JSON column called data that holds additional information related to each order To improve performance when querying by a specific order number you can create a stored column that extracts the order number from the JSON data and indexes it Let s take a look at the table definition CREATE TABLE orders id INT UNSIGNED NOT NULL data JSON NOT NULL order number VARCHAR GENERATED ALWAYS AS data gt gt number NOT NULL PRIMARY KEY id INDEX orders order number idx order number In this example the table orders has columns for id data and order number The order number column is defined as a stored column using the GENERATED ALWAYS AS syntax with the expression data gt gt number to extract the value of the number key from the JSON data By creating an index on the order number column using INDEX MySQL optimizes the query performance when searching and filtering based on order numbers With the table and index set up you can execute queries like SELECT FROM orders WHERE order number  ConclusionUsing a stored column for indexing JSON values can significantly improve query performance especially with large datasets Remember to carefully design and maintain indexes based on your application s needs to balance performance and write operations In conclusion leveraging a stored column for indexing JSON values in MySQL can enhance query performance and provide the flexibility of working with JSON data within your database schema Hope you have enjoyed this short bright dev tip Check our repository By Maciej Nawrocki Senior Backend Developer and Patryk Szlagowski Senior Backend Developer at Bright Inventions	2023-07-18 13:47:03
海外TECH	DEV Community	How we built a GPT code agent that generates full-stack web apps in React & Node.js, explained simply	https://dev.to/wasp/how-we-built-a-gpt-code-agent-that-generates-full-stack-web-apps-in-react-nodejs-explained-simply-4f9	How we built a GPT code agent that generates full stack web apps in React amp Node js explained simplyWe created GPT Web App Generator which lets you shortly describe the web app you would like to create and in a matter of minutes a full stack codebase written in React Node js Prisma and Wasp will be generated right in front of you and available to download and run locally We started this as an experiment to see how well we could use GPT to generate full stack web apps in Wasp the open source JS web app framework that we are developing Since we launched we had more than apps generated in just a couple of days Check out this blog post to see GPT Web App Generator in action including a one minute demo video few example apps and learn a bit more about our plans for the future Or try it out yourself at In this blog post we are going to explore the technical side of creating the GPT Web App Generator techniques we used how we engineered our prompts challenges we encountered and choices we made Note from here on we will just refer to it as the “Generator or “code agent when talking about the backend Also all the code behind the Generator is open source web app GPT code agent Support us ️If you wish to express your support for what we are doing consider giving us a star on Github Everything we do at Wasp is open source and your support motivates us and helps us to keep making web app development easier and with less boilerplate How well does it work First let s quickly explain what we ended up with and how it performs Input into our Generator is the app name app description free form text and a couple of simple options such as primary app color temperature auth method and GPT model to use As an output Generator spits out the whole JS codebase of a working full stack web app frontend backend and database Frontend is React Tailwind the backend is NodeJS with Express and for working with the database we used Prisma This is all connected together with the Wasp framework You can see an example of generated codebase here Generator does its best to produce code that works out of the box →you can download it to your machine and run it For simpler apps such as TodoApp or MyPlants it often generates code with no mistakes and you can run them out of the box Here is what a generated TodoApp looks like For a bit more complex apps like a blog with posts and comments it still generates a reasonable codebase but there are some mistakes to be expected here and there For even more complex apps it usually doesn t follow up completely but stops at some level of complexity and fills in the rest with TODOs or omits functionality so it is kind of like a simplified model of what was asked for Overall it is optimized for producing CRUD business web apps This makes it a great tool for kick starting your next web app project with a solid prototype or to even generate working simple apps on the fly How does it work ️ When we set out to build the Generator we gave ourselves the following goals we must be able to build it in a couple of weeksit has to be relatively easy to maintain in the futureit needs to generate the app quickly and cheaply a couple of minutes lt generated apps should have as few mistakes as possibleTherefore to keep it simple we don t do any LLM level engineering or fine tuning instead we just use OpenAI API specifically GPT and GPT to generate different parts of the app while giving it the right context at every moment pieces of docs examples guidelines … To ensure the coherence and quality of the generated app we don t give our code agent too much freedom but instead heavily guide it step by step through generating the app As step zero we generate some code files deterministically without GPT just based on the options that the user chose primary color auth method those include some config files for the project some basic global CSS and some auth logic You can see this logic here we call those “skeleton files code on Github Then the code agent takes over The code agent does its work in main phases Planning Generating Fixing Since GPT is quite slower and significantly more expensive than GPT also has a lower rate limit regarding the number of tokens per minute and also the number of requests per minute we use GPT only for the planning since that is the crucial step and then after that we use GPT for the rest Intermezzo short explanation of OpenAI Chat Completions APIOpenAI API offers different services but we used only one of them “chat completions API itself is actually very simple you send over a conversation and you get a response from the GPT The conversation is just a list of messages where each message has content and a role where the role specifies who “said that content →was it “user you or “assistant GPT The important thing to note is that there is no concept of state memory every API call is completely standalone and the only thing that GPT knows about is the conversation you provide it with at that moment If you are wondering how ChatGPT the web app that uses GPT in the background works with no memory →well each time you write a message the whole conversation so far is resent again There are some additional smart mechanisms in play here but that is really it at its core Official guide official API reference Step Planning A Wasp app consists of Entities Prisma data models Operations NodeJS Queries and Actions and Pages React Once given an app description and title the code agent first generates a Plan it is a list of Entities Operations Queries and Actions and Pages that comprise the app So kind of like an initial draft of the app It doesn t generate the code yet →instead it comes up with their names and some other details including a short description of what they should behave like This is done via a single API request toward GPT where the prompt consists of the following Short info about the Wasp framework an example of some Wasp code We explain that we want to generate the Plan explain what it is and how it is represented as JSON by describing its schema We provide some examples of the Plan represented as JSON Some rules and guidelines we want it to follow e g “plan should have at least page “make sure to generate a User entity Instructions to return the Plan only as a valid JSON response and no other text App name and description as provided by the user You can see how we generate such a prompt in the code here Also here is a link to a gist with an actual instance of this prompt for a TodoApp GPT then responds with a JSON hopefully that we parse and we have ourselves a Plan We will use this Plan in the following steps to drive our generation of other parts of the app Note that GPT sometimes adds text to the JSON response or returns invalid JSON so we built in some simple approaches to overcome these issues which we explain in detail later Intermezzo Common prompt designThe prompt design we just described above for generating a Plan is actually very similar for other steps e g the Generation and Fixing steps along with their respective sub steps so let s cover those commonalities All of the prompts we use more or less adhere to the same basic structure General contextShort info about what Wasp framework is Doc snippets with code examples if needed about whatever we are generating right now e g examples of NodeJS code or examples of React code Project context stuff we generated in the previous steps that is relevant to the current step Instructions on what we want to generate right now JSON schema for it example of such JSON response Rules and guidelines this is a good place to warn it about common mistakes it makes or give it some additional advice and emphasize what needs to happen and what must not happen Instructions to respond only with a valid JSON and no other text Original user prompt app name and description as provided by the user We put the original user prompt at the end because then we can tell GPT in the system message after it sees the start of the original user prompt we have a special header for it that it needs to treat everything after it as an app description and not as instructions on what to do →this way we attempt to defend from the potential prompt injection Step Generating After producing the Plan Generator goes step by step through the Plan and asks GPT to generate each web app piece while providing it with docs examples and guidelines Each time a web app piece is generated Generator fits it into the whole app This is where most of our work comes in equipping GPT with the right information at the right moment In our case we do it for all the Operations in the Plan Actions and Queries NodeJs code and also for all the Pages in the Plan React code with one prompt for each So if we have queries actions and pages that will be GPT prompts requests Prompts are designed as explained previously Code on Github generating an Operation generating a Page When generating Operations we provide GPT with the info about the previously generated Entities while when generating Pages we provide GPT with the info about previously generated Entities and Operations Step Fixing Finally the Generator tries its best to fix any mistakes that GPT might have introduced previously GPT loves fixing stuff it previously generated →if you first ask it to generate some code and then just tell it to fix it it will often improve it To enhance this process further we don t just ask it to fix previous code but also provide it with instructions on what to keep an eye out for like common types of mistakes that we noticed it often does and also point it to any specific mistakes we were able to detect on our own Regarding detecting mistakes to report to GPT ideally you would have a full REPL going on →that means running the generated code through an interpreter compiler then sending it for repairs and so on until all is fixed In our case running the whole project through the TypeScript compiler was not feasible for us with the time limits we put on ourselves but we used some simpler static analysis tools like Wasp s compiler for the wasp file and prisma format for Prisma model schemas and sent those to GPT to fix them We also wrote some simple heuristics of our own that are able to detect some of the common mistakes Our code amp prompt for fixing a Page Our code amp prompt for fixing Operations In the prompt we would usually repeat the same guidelines we provided previously in the Generation step while also adding a couple of additional pointers to common mistakes and that usually helps it fixes stuff it missed before But often not everything instead something will still get through Some things we just couldn t get it to fix consistently for example Wasp specific JS imports no matter how much we emphasized what it needed to do with them it would just keep messing them up Even GPT wasn t perfect in this situation For such situations when possible we ended up writing our own heuristics that would fix those mistakes fixing JS imports Things we tried learned Explanations We tried telling GPT to explain what it did while fixing mistakes which mistakes it will fix and which mistakes it fixed since we read that that can help but we didn t see visible improvement in its performance Testing Testing the performance of your code agent is hard In our case it takes a couple of minutes for our code agent to generate a new app and you need to run tests directly with the OpenAI API Also since results are non deterministic it can be pretty hard to say if output was affected by the changes you did or not Finally evaluating the output itself can be hard especially in our case when it is a whole full stack web app Ideally we would have set up a system where we can run only parts of the whole generation process and we could automatically run a specific part a number of times for each of different sets of parameters which would include different prompts but also parameters like type of model gpt vs gpt temperature and similar in order to compare performance for each of those parameter sets Evaluation performance would also ideally be automated e g we would count the mistakes during compilation and or evaluate the quality of app design →but this is also quite hard We unfortunately didn t have time to set up such a system so we were mostly doing testing manually which is quite subjective and vulnerable to randomness and is effective only for changes that have quite a big impact while you can t really detect those that are minor optimizations Context vs smarts When we started working on the Generator we thought the size of GPT s context would be the main issue However we didn t have any issues with context at the end →most of what we wanted to specify would fit into k to max k tokens while GPT has context up to k Instead we had bigger problems with its “smarts →meaning that GPT would not follow the rules we very explicitly told it to follow or would do things we explicitly forbid it from doing GPT proved to be better at following rules than GPT but even GPT would keep doing some mistakes over and over and forgetting about specific rules even though there was more than enough context The “fixing step did help with this we would repeat the rules there and GPT would pick up more of them but often still not all of them Handling JSON as a response As mentioned earlier in this article in all our interactions with GPT we always ask it to return the response as JSON for which we specify the schema and give some examples However GPT still doesn t always follow that rule and will sometimes add some text around the JSON or will make a mistake in formatting JSON The way we handled this is with two simple fixes Upon receiving JSON we would remove all the characters from the start until we hit and also all chars from the end until we hit Simple heuristic but it works very well for removing redundant text around the JSON in practice since GPT will normally not have any or in that text If we fail to parse JSON we send it again for repairs to GPT We include the previous prompt and its last answer that contains invalid JSON and add instructions to fix it JSON parse errors we got We repeat this a couple of times until it gets it right or until we give up In practice these two methods took care of invalid JSON in of the cases for us NOTE While we were implementing our code agent OpenAI released new functionality for GPT “functions which is basically a mechanism to have GPT respond with a structured JSON following the schema of your description So it would likely make more sense to do this with “functions but we already had this working well so we just stuck with it Handling interruptions in the service We were calling OpenAI API directly so we noticed quickly that often it would return service unavailable especially during peak hours e g pm CET Therefore it is recommended to have some kind of retry mechanism ideally with exponential backoff that makes your code agent redundant to such random interruptions in the service and also to potential rate limiting We went with the retry mechanism with exponential backoff and it worked great Temperature ️Temperature determines how creative GPT is but the more creative it gets the less “stable it is It hallucinates more and also has a harder time following rules A temperature is a number from to with a default value of We experimented with different values and found the following ≥ would every so and so start giving quite silly results with random strings in it ≥ lt was okish but was introducing a bit too many mistakes ≥ lt was optimal →creative enough while still not having many mistakes ≤ seemed to perform similarly to a bit higher values but with a bit less creativity maybe That said I don t think we tested values below enough and that is something we could certainly work on more We ended up using as our default value except for prompts that do fixing for those we used a lower value of because it seemed like GPT was changing stuff too much while fixing at being too creative Our logic was let it be creative when writing the first version of the code then have it be a bit more conventional while fixing it Again we haven t tested all this enough so this is certainly something I would like us to explore more Future While we ended up being impressed with the performance of what we managed to build in such a short time we were also left wanting to try so many different ideas on how to improve it further There are many avenues left to be explored in this ecosystem that is developing so rapidly that it is hard to reach the point where you feel like you explored all the options and found the optimal solution Some of the ideas that would be exciting to try in the future We put quite a few limitations regarding the code that our code agent generates to make sure it works well enough we don t allow it to create helper files to include npm dependencies no TypeScript no advanced Wasp features … We would love to lift the limitations therefore allowing the creation of more complex and powerful apps Instead of our code agent doing everything in one shot we could allow the user to interact with it after the first version of the app is generated to provide additional prompts for example to fix something to add some feature to the app to do something differently … The hardest thing here would be figuring out which context to provide to the GPT at which moment and designing the experience appropriately but I am certain it is doable and it would take the Generator to the next level of usability Another option is to allow intervention in between initial generation steps →for example after the plan is generated to allow the user to adjust it by providing additional instructions to the GPT Find an open source LLM that fits the purpose and fine tune pre train it for our purpose If we could teach it more about Wasp and the technologies we use so we don t have to include it in every prompt we could save quite some context have the LLM be more focused on the rules and guidelines we are specifying in the prompt We could also host it ourselves and have more control over the costs and rate limits Take a different approach to the code agent let it be more free Instead of guiding it so carefully we could teach it about all the different things it is allowed to ask for ask for docs ask for examples ask to generate a certain piece of the app ask to see a certain already generated piece of the app … and would let it guide itself more freely It could constantly generate a plan execute it update the plan and so on until it reaches the state of equilibrium This approach potentially promises more flexibility and would likely be able to generate apps of greater complexity but it also requires quite more tokens and a powerful LLM to drive it →I believe this approach will become more feasible as LLMs become more capable Support us ️If you wish to express your support for what we are doing consider giving us a star on Github Everything we do at Wasp is open source and your support motivates us and helps us to keep making web app development easier and with less boilerplate Also if you have any ideas on how we could improve our code agent or maybe we can help you somehow gt feel free to join our Discord server and let s chat	2023-07-18 13:22:18
海外TECH	DEV Community	Top 7 Featured DEV Posts from the Past Week	https://dev.to/devteam/top-7-featured-dev-posts-from-the-past-week-3hl3	Top Featured DEV Posts from the Past WeekEvery Tuesday we round up the previous week s top posts based on traffic engagement and a hint of editorial curation The typical week starts on Monday and ends on Sunday but don t worry we take into account posts that are published later in the week DEVs Writing Quality Content About Your Career jmfayard recently shouted out inspiring career writers that you definitely need to check out This kind of list could never really be complete so add your favorite DEV writers in the comments of this amazing post DEVs writing quality content about your career Jean Michel Fayard ・Jul bestofdev career developer gratitude Top Must Know Tips for Web AccessibilityThis awesome post by ismaestro aims to provide you with essential principles practical advice and examples for creating accessible websites You ll learn to create clear descriptions for images explore the world of navigating using only the keyboard and much more so that everyone can browse your site like a pro Top Must Know Tips for Web Accessibility Ismael Ramos for This is Learning・Jul webdev javascript beginners programming Please Don t Write Confusing ConditionalsHere we ll discuss with somedood how confusing conditionals manifest in our code Along the way we also explore different refactoring techniques for improving the readability of conditional control flows Please don t write confusing conditionals Basti Ortiz・Jul beginners tips programming javascript The Magic of Empty Git CommitAlthough seemingly counterintuitive empty commits can offer valuable benefits for documentation triggering automation pradumnasaraf demonstrates how to create and push an empty commit using Git The Magic of Empty Git Commit Pradumna Saraf・Jul git opensource beginners programming A Comprehensive Beginner s Guide to NPM Simplifying Package ManagementThis fully beginner friendly guide from abhixsh will take you through the fundamentals of NPM providing you with a solid foundation for simplifying package management and streamlining your development process A Comprehensive Beginner s Guide to NPM Simplifying Package Management Abishek Haththakage・Jul webdev beginners npm programming Real DOM Virtual DOM Shadow DOM What s the Difference It can be hard to keep your head straight between real DOMs and virtual DOMs and shadow DOMs and how they all work together to create a clean performant Document Object Model Let s dive in with lyndsiwilliams Real DOM Virtual DOM Shadow DOM What s the Difference Lyndsi Kay Williams・Jul webdev react frontend performance Introduction to CSS Grid A Comprehensive GuideCSS Grid is a layout system that allows you to create complex grid based designs with ease Here we ll explore the key concepts of CSS Grid with joanayebola and demonstrate how to use it with practical code examples Introduction to CSS Grid A Comprehensive Guide joan ・Jul webdev css That s it for our weekly Top for this Tuesday Keep an eye on dev to this week for daily content and discussions and be sure to keep an eye on this series in the future You might just be in it	2023-07-18 13:17:00
海外TECH	DEV Community	Repo manager	https://dev.to/tigawanna/repo-manager-3d53	Repo manager What I built Repo manager simple github dashboard for quick updates to the viewer profile and repos Category Submission Dashboard App Link Screenshots DescriptionI built a github dashboard that lets you quuickly edit delete your profile info or repository detailsfeature include quick viewer profile overviewview list of repositories sorted by last pushed to descending order by default but options to sort by name stargazers created at and updated at desc or ascending sorting by forked repositories only is supportedhint that informs you if your forked repository is out of sync with it s parenta search bar thats lets you quickly search for new repositories and filter your own selectmultiple repos and bulk delte them use with caution Link to Source Coderepo manager Permissive LicenseMIT licence Background What made you decide to build this particular app What inspired you i feel like it will come in handy as i approach apply to jobs phase need something that will help me in tidying things up adding proper names descriptions andtags for repositories an daso deleting redundant onesi also disliked the current hops you have to jump through to delete repositories How I built it How did you utilize refine Did you learn something new along the way Pick up a new skill Really love the way refine halped me scafold out the project one of the biggest pain of setting up a SPA is setting up the router state management stying and setting up the API structure I wasn t a big fan of MUI before but after building this app i feel like it was all outof prejudice because the consistent UI accessible components is quite the productivity boost The data providers were also really great and even though i had to use some escape hatchesi can see myself suing them more often the inferencer is a great prototyping tool and lets you immediately view the API you just wired upit s all great dx all the way	2023-07-18 13:00:49
Apple	AppleInsider - Frontpage News	TSMC profits fall but the iPhone 15 will help it rebound	https://appleinsider.com/articles/23/07/18/tsmc-profits-fall-but-the-iphone-15-will-help-it-rebound?utm_medium=rss	TSMC profits fall but the iPhone will help it reboundApple supplier TSMC is expected to report a drop in profits soon but business will recover in the third quarter because the company is producing the iPhone TSMC profits are expected to decline year over yearOn Thursday Taiwan Semiconductor Manufacturing Company TSMC a chip manufacturer based in Taiwan is projected to reveal a decrease in net profit for the second quarter of The decline is due to global economic challenges which have affected the demand for semiconductors Read more	2023-07-18 13:47:53
Apple	AppleInsider - Frontpage News	WSJ: After 11 years of work, people actually like Apple Maps	https://appleinsider.com/articles/23/07/18/wsj-after-11-years-of-work-people-actually-like-apple-maps?utm_medium=rss	WSJ After years of work people actually like Apple MapsThe Wall Street Journal has just noticed that people do actually like and use Apple Maps an apparent change that took only years to be acknowledged Apple MapsApple Maps had what could be characterized as a terrible start to its existence in with a catalog of issues that didn t help users shift over to it from Google Maps Now over a decade later the Wall Street Journal acknowledges that Apple Maps is no longer terrible Read more	2023-07-18 13:42:53
海外TECH	Engadget	The best dorm room essentials for college students	https://www.engadget.com/best-dorm-room-essentials-for-college-students-133806068.html?src=rss	The best dorm room essentials for college studentsCollege will be back in session soon enough which also means a new batch of freshmen will start living the dorm life If that applies to you we think it s a good idea to stock up on a few essentials for your new tiny abode before you get there especially if this is your first time away from home We ve got tech recommendations of course but not everything on this list is a gadget That s because we also wanted to cover the other items that will bring you the comforts of home to your dorm and hopefully make student life less stressful Echo DotI resisted a smart speaker until a few months ago After getting an Echo Dot I now use Alexa to set alarms play focus music remind me about stuff and keep tabs on the weather ーall of which would have been very helpful back in school Sure your phone can do most of that but anything that helps you pick up your phone less is a productivity booster in itself Plus the Dot works with all major music services and the sound quality is surprisingly good for its size There s more than enough punch here to entertain a dorm room and do justice to your study beats ーAmy Skorheim Commerce WriterBelkin MagSafe in wireless chargerIf you own multiple Apple products and you re constantly detangling charging cables we highly recommend a in wireless charger like the Belkin BoostCharge Pro It has a W MagSafe charging base for your iPhone and charging spaces for your Apple Watch and AirPods as well The horizontal layout lets you charge any Qi capable phone though at reduced speeds or other compatible earbuds and accessories On top of that the flat pad format means you can easily pack it in your bag the next time you go on a trip ーNicole Lee Commerce WriterAnker Soundcore Space AA good set of noise canceling headphones can help you get some peace and quiet anytime you need to work or want to escape the rowdiness of your fellow dormmates You don t need to spend a ton to find a quality pair either The Anker Soundcore Space A delivers effective active noise cancellation eight to hours of battery life per charge a transparency mode IPX rated water resistance and solid audio quality for less than and even less than often times when on sale If you don t like their bassy default sound you can also customize the profile through a useful companion app Just note that these are earbuds if you d prefer an over ear pair try the Anker Soundcore Life Q ーJeff Dunn Senior Commerce WriterLinenspa shredded foam pillowComfort is key to dorm living which is why we recommend an oversized reading pillow like this one from Linenspa It helps change up your seating position so you re not in your desk chair all the time and it s definitely a lot more comfortable than just piling up bed pillows against the wall The pillow is filled with shredded memory foam so it won t put a lot of stress on your back plus it comes with a soft velour cover That posture support makes it great for reading playing games or watching TV As a bonus there s also a handle on the top that makes it easy to carry around ーN L Zwilling Electric KettleDorms may limit the number and types of kitchen appliances you can have but most are fine with a simple electric kettle like this one Zwilling s Electric Kettle boils water eerily fast and doesn t have an exposed heating element which some housing regulations don t allow It also looks lovely and in our experiments boiled ounces of water in two minutes flat Sure it can help you make tea or coffee but the real pro level move is stocking up on instant cup foods for morning oatmeal midnight Cup Noodles and the cravings in between ーA S Ultimate Ears Wonderboom For when their phone s speakers just won t cut it the UE Wonderboom can make their dorm room parties and chill sessions on the quad even better with punchy sound The small barrel shaped speaker is compact enough to fit in a backpack and it could be attached to the outside of a bag with a carabiner clip thanks to its built in top loop We found the Wonderboom to deliver the biggest sound of all the portable Bluetooth speakers we tested in its size range and the latest model has improved battery life and wireless range There s no app but it can be paired with other UE speakers for stereo sound And if it accidentally falls off a table or takes a bit of a beating its IP rating and drop proof design should sufficiently protect it ーValentina Palladino Senior Commerce EditorNanoleaf LinesNanoleaf s modular smart lights let you add personality and functional lighting to your half of the room The degree “Smarter kit comes with four Lines that attach at right angles so you can make a few different designs a square and an X come to mind The set is expandable and each bar has millions of available colors with the ability to display two hues at once Lines works with Alexa HomeKit Google Home IFTTT and other smart home platforms so you can set themes create routines and control the lights with your voice But possibly the best news for dorm dwellers is the included mounting tape that won t punch holes in your walls and anger your RA ーA S LapGear DesignerWhen you re just too lazy to sit at your actual desk a lap desk like the LapGear Designer can make working from the bed or couch more comfortable The Designer is softly padded and lightweight with an easy to clean top that s large enough to fit a inch laptop A stopper at the bottom helps keep things from slipping off when you re sitting at an angle plus there s a slot for holding a phone and a handle for carrying the whole thing around ーJ D Vornado Medium Air CirculatorMany dorms lack air conditioning so having a fan that s powerful enough to keep you cool during the late summer months is crucial The Vornado should do the job as it moves air around a room powerfully yet takes up little room on a tabletop or larger window sill It doesn t oscillate but you can tilt its head vertically and the way it circulates air allows it to send a breeze through most of a room A simple dial lets you swap between three speed settings while the sturdy plastic frame is easy to clean and keeps the thing relatively quiet when it s on If you re moving into a particularly large dorm room the Vornado is a stronger alternative ーJ D OXO Good Grips cereal dispenserWhen I was in college we called cereal ramen and vodka “the five food groups I hear today s college bound generation drinks less and is probably better with numbers but I suspect cereal is still a staple which is why this dispenser from OXO is a must get The silicone seal on the pop top lid keeps Crispix crunchy and Lucky Charms fresh though it can do nothing to make Grape Nuts less gravel like It also keeps ants and other pests out and the clear plastic looks nice on a shelf But I particularly like being able to open the lid pour out the cereal and close it back up with just one hand ーA S Pure Green Natural Latex mattress topperChances are the bed in your dorm room isn t very comfortable It s also probably not easy or cost effective to change the mattress That s why we recommend getting a mattress topper It s the one way to control how your bed feels without spending a lot of money We prefer a natural latex option like this one from Pure Green because it delivers comfort and support without the sinking feeling of memory foam It also won t absorb much body heat which helps keep you cool throughout the night Pure Green sells its mattress topper in three different thicknesses ーfrom one inch to three ーto fit your particular needs Opt for the or inch if you want firm or medium firm or spring for the inch model if you prefer a something softer ーN L Lunya Sleep MaskRoommate pulling an all nighter with the lights on Want to catch some z s in the afternoon in between classes Or maybe you re just particularly sensitive to light If you can relate to any of this we recommend getting a sleep mask to make it easier for you to drift off into dreamland This one from Lunya is one of our personal favorites because it completely blocks out light and feels comfortable to boot it s like wearing pillows on your eyes We tend to prefer this model over masks with eye cups because it s not quite as bulky The Lunya s wide elasticized band will fit most people and it even covers the ears which helps reduce noise It s also machine washable so you can easily keep it clean ーN L CodenamesPlaying casual board games is a great way to socialize at gatherings without the need to engage in small talk A really popular one is called Codenames a party game that pits two teams of spies against each other Each “spymaster has to get their teammates to guess hidden words which are plotted out on a grid using only one word clues and a number For example if you wanted your team to guess the words “costume “web and “spider you might say “Peter to indicate that there are three clues on the board that match that word There are also clues you have to avoid which makes the game a little harder The game is easy to explain and it encourages communication which helps break the ice ーN L Herd MentalityIf you want a game that s more light hearted we recommend Herd Mentality It accommodates four to players which makes it perfect for parties In it you simply take turns flipping over a question and trying to write down what you think everyone else will answer as well For example if the question is “What is the best way to cook an egg you write “scrambled and it turns out that is what most other players answered as well you will get points But beware of giving the answer that s the odd one out because you ll get the dreaded Pink Cow and be in danger of losing the game unless you can somehow trick someone else into getting it instead ーN L This article originally appeared on Engadget at	2023-07-18 13:38:06
海外TECH	Engadget	Startup will test self-flying aircraft in remote regions of Canada	https://www.engadget.com/startup-will-test-self-flying-aircraft-in-remote-regions-of-canada-131542988.html?src=rss	Startup will test self flying aircraft in remote regions of CanadaA Canadian air cargo startup called Ribbit is planning to test pilotless flights for deliveries in remote areas of the country Northern Ontario Business has reported The company signed a million contract with Transport Canada and Innovative Solutions Canada to start autonomous test flights quot over the next months quot the company said quot Many rural and remote areas are served by larger airplanes that fly infrequently quot CEO Carl Pigeon said in a press release earlier this month quot Ribbit takes a smaller aircraft and uses autonomy to drastically change the unit economics of that plane This lets us offer reliable next day or two day service and improve supply chains quot The company is starting small with two passenger recreational style aircraft It plans to remove the seats to open up room for cargo then make the aircraft fully autonomous using remote control software and hardware The idea began as a project from University of Waterloo students including co founders Jeremy Wang and Carl Pigeon Ribbit said its aircraft would use remote pilots to monitor progress communicate with air traffic controllers and generally provide a backup The company has already signed contracts with retailers and wholesalers including locally owned businesses that serve the province The aim is to fill a demand for timely delivery of food medical items and more quot The goal is really to try and improve that transportation link so that we can get food and other perishables time sensitive items medical supplies etcetera into these communities at a lower price to the end consumer more reliably and more frequently quot Wang told Northern Ontario Business There are a number of competitors in this space already most notably Xwing which has already performed autonomous gate to gate commercial cargo flights That company is using much larger aircraft namely converted Cessna Grand Caravan B utility planes fitted with Xwing s Autoflight software Another competitor in the space is Reliable Robotics founded by former SpaceX and Tesla engineers which has also run successful remotely piloted cargo tests Ribbit is smaller than those players but it knows its customers and the region well quot Be it air cargo asset monitoring or maritime patrol we have identified several applications for the technology quot Wang said quot Customers appreciate our ability to understand their operations deeplyーthen reimagine them with autonomy quot This article originally appeared on Engadget at	2023-07-18 13:15:42
海外TECH	Engadget	NVIDIA drops remake of fan-favorite mod ‘Portal: Prelude’ on Steam for free	https://www.engadget.com/nvidia-drops-remake-of-fan-favorite-mod-portal-prelude-on-steam-for-free-130050799.html?src=rss	NVIDIA drops remake of fan favorite mod Portal Prelude on Steam for freeNVIDIA just officially released the fan made Portal mod Portal Prelude The company dropped it on Steam and what s more it s free for anyone who has the original game This isn t a drab re release as the updated release features new textures full ray tracing DLSS for increased performance NVIDIA Reflex for decreased latency and RTX IO for quicker load times NVIDIA also dropped a GeForce Game Ready Driver to simplify setup All of the changes to the remaster were done by modders keeping with the spirit of the original release via the company s forthcoming RTX Remix creator toolkit NVIDIA also hired famed modders to modernize assets and improve the lighting For the uninitiated Portal Prelude was originally released back in and developed by modders It acts as a prequel to the original game and is set before the time of GLaDOS The game offers a ten hour campaign with nearly test chambers a fully voiced story and mechanics that go beyond the first Portal It continues to be the highest rated Portal mod and there s no shortage of competition This remake took eight months of dev time and acts as a showcase for the aforementioned modding toolkit RTX Remix Portal Prelude is available right now on Steam and won t cost you anything as long as you have the original Portal so get downloading This article originally appeared on Engadget at	2023-07-18 13:00:50
海外ニュース	Japan Times latest articles	U.S. national in North Korean custody after crossing inter-Korean border during tour	https://www.japantimes.co.jp/news/2023/07/18/asia-pacific/american-detained-north-korea/	U S national in North Korean custody after crossing inter Korean border during tourThe U S citizen crossed into the nuclear armed country while on a tour of the Joint Security Area at the border with the South the United	2023-07-18 22:25:07
ニュース	BBC News - Home	Carla Foster: Mother jailed over lockdown abortion to be released	https://www.bbc.co.uk/news/uk-england-65581850?at_medium=RSS&at_campaign=KARANGA	tablets	2023-07-18 13:29:12
ニュース	BBC News - Home	UK terrorism risk is rising - Suella Braverman	https://www.bbc.co.uk/news/uk-66234656?at_medium=RSS&at_campaign=KARANGA	bravermanthe	2023-07-18 13:10:58
ニュース	BBC News - Home	Wagner: Satellite images reveal Belarus camp arrival	https://www.bbc.co.uk/news/world-66234260?at_medium=RSS&at_campaign=KARANGA	belarus	2023-07-18 13:37:53
ニュース	BBC News - Home	Heatwave scorches Europe, in pictures	https://www.bbc.co.uk/news/in-pictures-66226313?at_medium=RSS&at_campaign=KARANGA	europe	2023-07-18 13:13:24
ニュース	BBC News - Home	Andre Onana: Manchester United agree deal for Inter Milan goalkeeper	https://www.bbc.co.uk/sport/football/66235305?at_medium=RSS&at_campaign=KARANGA	additional	2023-07-18 13:09:57
ニュース	BBC News - Home	The Ashes 2023: Ben Stokes says rain-affected fourth Test could help England	https://www.bbc.co.uk/sport/cricket/66235792?at_medium=RSS&at_campaign=KARANGA	The Ashes Ben Stokes says rain affected fourth Test could help EnglandCaptain Ben Stokes says a weather shortened fourth Ashes Test could suit England as they look to level the series with Australia at Old Trafford	2023-07-18 13:09:25