Vaya

vaya

Being quite as cunning as Ali himself

The French commander Nicole, surnamed the “Pilgrim,”

Ali Pacha by Alexandre Dumas, is part of the HackerNoon Books Series. You can jump to any chapter in this book here . Chapter V CHAPTER V After taking possession of Argyro-Castron, which he had long coveted, Ali led his victorious army against the town of Kardiki, whose inhabitants had formerly joined with those of Kormovo in the outrage inflicted on his mother and sister. The besieged, knowing they had no mercy to hope for, defended themselves bravely, but were obliged to yield to famine. After a month’s blockade, the common people, having no food for themselves or their cattle, began to cry for mercy in the open streets, and their chiefs, intimidated by the general misery and unable to stand alone, consented to capitulate. Ali, whose intentions as to the fate of this unhappy town were irrevocably decided, agreed to all that they asked. A treaty was signed by both parties, and solemnly sworn to on the Koran, in virtue of which seventy-two beys, heads of the principal Albanian families, were to go to Janina as free men, and fully armed. They were to be received with the honours due to their rank as free tenants of the sultan, their lives and their families were to be spared, and also their possessions. The other inhabitants of Kardiki, being Mohammedans, and therefore brothers of Ali, were to be treated as friends and retain their lives and property. On these conditions a quarter of the town; was to be occupied by the victorious troops. One of the principal chiefs, Saleh Bey, and his wife, foreseeing the fate which awaited their friends, committed suicide at the moment when, in pursuance of the treaty, Ali’s soldiers took possession of the quarter assigned to them. Ali received the seventy-two beys with all marks of friendship when they arrived at Janina. He lodged them in a palace on the lake, and treated them magnificently for some days. But soon, having contrived on some pretext to disarm them, he had them conveyed, loaded with chains, to a Greek convent on an island in the lake, which was converted into a prison. The day of vengeance not having fully arrived, he explained this breach of faith by declaring that the hostages had attempted to escape. The popular credulity was satisfied by this explanation, and no one doubted the good faith of the pacha when he announced that he was going to Kardiki to establish a police and fulfil the promises he had made to the inhabitants. Even the number of soldiers he took excited no surprise, as Ali was accustomed to travel with a very numerous suite. After three days’ journey, he stopped at Libokhovo, where his sister had resided since the death of Aden Bey, her second son, cut off recently by wickness. What passed in the long interview they had no one knew, but it was observed that Chainitza’s tears, which till then had flowed incessantly, stopped as if by magic, and her women, who were wearing mourning, received an order to attire themselves as for a festival. Feasting and dancing, begun in Ali’s honour, did not cease after his departure. He spent the night at Chenderia, a castle built on a rock, whence the town of Kardiki was plainly visible. Next day at daybreak Ali despatched an usher to summon all the male inhabitants of Kardiki to appear before Chenderia, in order to receive assurances of the pacha’s pardon and friendship. The Kardikiotes at once divined that this injunction was the precursor of a terrible vengeance: the whole town echoed with cries and groans, the mosques were filled with people praying for deliverance. The appointed time arrived, they embraced each other as if parting for ever, and then the men, unarmed, in number six hundred and seventy, started for Chenderia. At the gate of the town they encountered a troop of Albanians, who followed as if to escort them, and which increased in number as they proceeded. Soon they arrived in the dread presence of Ali Pacha. Grouped in formidable masses around him stood several thousand of his fierce soldiery. The unhappy Kardikiotes realised their utter helplessness, and saw that they, their wives an children, were completely at the mercy of their implacable enemy. They fell prostrate before the pacha, and with all the fervour which the utmost terror could inspire, implored him to grant them a generous pardon. Ali for some time silently enjoyed the pleasure of seeing his ancient enemies lying before him prostrate in the dust. He then desired them to rise, reassured them, called them brothers, sons, friends of his heart. Distinguishing some of his old acquaintances, he called them to him, spoke familiarly of the days of their youth, of their games, their early friendships, and pointing to the young men, said, with tears in his eyes. “The discord which has divided us for so many years has allowed children not born at the time of our dissension to grow into men. I have lost the pleasure of watching the development of the off-spring of my neighbours and the early friends of my youth, and of bestowing benefits on them, but I hope shortly to repair the natural results of our melancholy divisions.” He then made them splendid promises, and ordered them to assemble in a neighbouring caravanserai, where he wished to give them a banquet in proof of reconciliation. Passing from the depths of despair to transports of joy, the Kardikiotes repaired gaily to the caravanserai, heaping blessings on the pacha, and blaming each other for having ever doubted his good faith. Ali was carried down from Chenderia in a litter, attended by his courtiers, who celebrated his clemency in pompous speeches, to which he replied with gracious smiles. At the foot of the steep descent he mounted his horse, and, followed by his troops, rode towards the caravanserai. Alone, and in silence, he rode twice round it, then, returning to the gate, which had just been closed by his order, he pulled up his horse, and, signing to his own bodyguard to attack the building, “Slay them!” he cried in a voice of thunder. The guards remained motionless in surprise and horror, then as the pacha, with a roar, repeated his order, they indignantly flung down their arms. In vain he harangued, flattered, or threatened them; some preserved a sullen silence, others ventured to demand mercy. Then he ordered them away, and, calling on the Christian Mirdites who served under his banner. “To you, brave Latins,” he cried, “I will now entrust the duty of exterminating the foes of my race. Avenge me, and I will reward you magnificently.” A confused murmur rose from the ranks. Ali imagined they were consulting as to what recompense should be required as the price of such deed. “Speak,” said he; “I am ready to listen to your demands and to satisfy them.” Then the Mirdite leader came forward and threw back the hood of his black cloak. “O Pacha!” said he, looking Ali boldly in the face, “thy words are an insult; the Mirdites do not slaughter unarmed prisoners in cold blood. Release the Kardikiotes, give them arms, and we will fight them to the death; but we serve thee as soldiers and not as executioners.” At these words; which the black-cloaked battalion received with applause, Ali thought himself betrayed, and looked around with doubt and mistrust. Fear was nearly taking the place of mercy, words of pardon were on his lips, when a certain Athanasius Vaya, a Greek schismatic, and a favourite of the pacha’s, whose illegitimate son he was supposed to be, advanced at the head of the scum of the army, and offered to carry out the death sentence. Ali applauded his zeal, gave him full authority to act, and spurred his horse to the top of a neighbouring hill, the better to enjoy the spectacle. The Christian Mirdites and the Mohammedan guards knelt together to pray for the miserable Kardikiotes, whose last hour had come. The caravanserai where they were shut in was square enclosure, open to the sky, and intended to shelter herds of buffaloes. The prisoners having heard nothing of what passed outside, were astonished to behold Athanasius Vaya and his troop appearing on the top of the wall. They did not long remain in doubt. Ali gave the signal by a pistol-shot, and a general fusillade followed. Terrible cries echoed from the court; the prisoners, terrified, wounded, crowded one upon another for shelter. Some ran frantically hither and thither in this enclosure with no shelter and no exit, until they fell, struck down by bullets. Some tried to climb the walls, in hope of either escape or vengeance, only to be flung back by either scimitars or muskets. It was a terrible scene of despair and death. After an hour of firing, a gloomy silence descended on the place, now occupied solely by a heap of corpses. Ali forbade any burial rites on pain of death, and placed over the gate an inscription in letters of gold, informing posterity that six hundred Kardikiotes had there been sacrificed to the memory of his mother Kamco. When the shrieks of death ceased in the enclosure, they began to be heard in the town. The assassins spread themselves through it, and having violated the women and children, gathered them into a crowd to be driven to Libokovo. At every halt in this frightful journey fresh marauders fell on the wretched victims, claiming their share in cruelty and debauchery. At length they arrived at their destination, where the triumphant and implacable Chainitza awaited them. As after the taking of Kormovo, she compelled the women to cut off their hair and to stuff with it a mattress on which she lay. She then stripped them, and joyfully narrated to them the massacre of their husbands, fathers, brothers and sons, and when she had sufficiently enjoyed their misery they were again handed over to the insults of the soldiery. Chainitza finally published an edict forbidding either clothes, shelter, or food to be given to the women and children of Kardiki, who were then driven forth into the woods either to die of hunger or to be devoured by wild beasts. As to the seventy-two hostages, Ali put them all to death when he returned to Janina. His vengeance was indeed complete. But as, filled with a horrible satisfaction, the pacha was enjoying the repose of a satiated tiger, an indignant and threatening voice reached him even in the recesses of his palace. The Sheik Yussuf, governor of the castle of Janina, venerated as a saint by the Mohammedans on account of his piety, and universally beloved and respected for his many virtues, entered Ali’s sumptuous dwelling for the first time. The guards on beholding him remained stupefied and motionless, then the most devout prostrated themselves, while others went to inform the pacha; but no one dared hinder the venerable man, who walked calmly and solemnly through the astonished attendants. For him there existed no antechamber, no delay; disdaining the ordinary forms of etiquette, he paced slowly through the various apartments, until, with no usher to announce him, he reached that of Ali. The latter, whose impiety by no means saved him from superstitious terrors, rose hastily from the divan and advanced to meet the holy sheik, who was followed by a crowd of silent courtiers. Ali addressed him with the utmost respect, and endeavoured even to kiss his right hand. Yussuf hastily withdrew it, covered it with his mantle, and signed to the pacha to seat himself. Ali mechanically obeyed, and waited in solemn silence to hear the reason of this unexpected visit. Yussuf desired him to listen with all attention, and then reproached him for his injustice and rapine, his treachery and cruelty, with such vivid eloquence that his hearers dissolved in tears. Ali, though much dejected, alone preserved his equanimity, until at length the sheik accused him of having caused the death of Emineh. He then grew pale, and rising, cried with terror: “Alas! my father, whose name do you now pronounce? Pray for me, or at least do not sink me to Gehenna with your curses!” “There is no need to curse thee,” answered Yussuf. “Thine own crimes bear witness against thee. Allah has heard their cry. He will summon thee, judge thee, and punish thee eternally. Tremble, for the time is at hand! Thine hour is coming—is coming—is coming!” Casting a terrible glance at the pacha, the holy man turned his back on him, and stalked out of the apartment without another word. Ali, in terror, demanded a thousand pieces of gold, put them in a white satin purse, and himself hastened with them to overtake the sheik, imploring him to recall his threats. But Yussuf deigned no answer, and arrived at the threshold of the palace, shook off the dust of his feet against it. Ali returned to his apartment sad and downcast, and many days elapsed before he could shake off the depression caused by this scene. But soon he felt more ashamed of his inaction than of the reproaches which had caused it, and on the first opportunity resumed his usual mode of life. The occasion was the marriage of Moustai, Pacha of Scodra, with the eldest daughter of Veli Pacha, called the Princess of Aulis, because she had for dowry whole villages in that district. Immediately after the announcement of this marriage Ali set on foot a sort of saturnalia, about the details of which there seemed to be as much mystery as if he had been preparing an assassination. All at once, as if by a sudden inundation, the very scum of the earth appeared to spread over Janina. The populace, as if trying to drown their misery, plunged into a drunkenness which simulated pleasure. Disorderly bands of mountebanks from the depths of Roumelia traversed the streets, the bazaars and public places; flocks and herds, with fleeces dyed scarlet, and gilded horns, were seen on all the roads driven to the court by peasants under the guidance of their priests. Bishops, abbots, ecclesiastics generally, were compelled to drink, and to take part in ridiculous and indecent dances, Ali apparently thinking to raise himself by degrading his more respectable subjects. Day and night these spectacles succeeded each other with increasing rapidity, the air resounded with firing, songs, cries, music, and the roaring of wild beasts in shows. Enormous spits, loaded with meat, smoked before huge braziers, and wine ran in floods at tables prepared in the palace courts. Troops of brutal soldiers drove workmen from their labour with whips, and compelled them to join in the entertainments; dirty and impudent jugglers invaded private houses, and pretending that they had orders from the pacha to display their skill, carried boldly off whatever they could lay their hands upon. Ali saw the general demoralization with pleasure, especially as it tended to the gratification of his avarice, Every guest was expected to bring to the palace gate a gift in proportion to his means, and foot officers watched to see that no one forgot this obligation. At length, on the nineteenth day, Ali resolved to crown the feast by an orgy worthy of himself. He caused the galleries and halls of his castle by the lake to be decorated with unheard-of splendour, and fifteen hundred guests assembled for a solemn banquet. The pacha appeared in all his glory, surrounded by his noble attendants and courtiers, and seating himself on a dais raised above this base crowd which trembled at his glance, gave the signal to begin. At his voice, vice plunged into its most shameless diversions, and the wine-steeped wings of debauchery outspread themselves over the feast. All tongues were at their freest, all imaginations ran wild, all evil passions were at their height, when suddenly the noise ceased, and the guests clung together in terror. A man stood at the entrance of the hall, pale, disordered, and wild-eyed, clothed in torn and blood-stained garments. As everyone made way at his approach, he easily reached the pacha, and prostrating himself at his feet, presented a letter. Ali opened and rapidly perused it; his lips trembled, his eyebrows met in a terrible frown, the muscles of his forehead contracted alarmingly. He vainly endeavoured to smile and to look as if nothing had happened, his agitation betrayed him, and he was obliged to retire, after desiring a herald to announce that he wished the banquet to continue. Now for the subject of the message, and the cause of the dismay it produced. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Alexandre Dumas (2006). Ali Pacha. Urbana, Illinois: Project Gutenberg. Retrieved https://www.gutenberg.org/cache/epub/2753/pg2753-images.html This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at www.gutenberg.org , located at https://www.gutenberg.org/policy/license.html .

Ali led his victorious army against the town of Kardiki

Procesando Datos Para Deep Learning: Datasets, visualizaciones y DataLoaders en PyTorch

Para extraer datos de websites, puede usar las herramientas de extracción de datos como . Estas herramientas pueden extraer datos de website automáticamente y guardarlos en muchos formatos, como Excel, JSON, CSV, HTML o en su propia base de datos a través de . Solo toma unos minutos puede extraer miles de líneas de datos, la mejor es que no se necesita codificación en este proceso. Octoparse API Tabla de contenidos Método #1: Extracción Sencilla con Las Plantillas de Web Scraping Método #2: Extracción Personalizada con Modo Avanzado Tome Google Search como ejemplo. Si estamos interesados en información sobre el "smoothie" y queremos extraer todos los títulos, descripciones y URLs de los resultados de búsqueda. Para extraer datos de Google Search, puede usar una . plantilla de web scraping La plantilla es un crawler preformateado que listo para usar sin ninguna configuración. Hay más de 50 plantillas para elegir. Verá todas las plantillas desde sitios web de comercio electrónico como Amazon y eBay hasta canales de redes sociales como Facebook, Twitter e Instagram. Octoparse también ofrece plantillas personalizadas. Método #1: Extracción Sencilla con Las Plantillas de Web Scraping Primero: Elija una plantilla de web scraping Para usar la plantilla, necesita instalar Octoparse en la computadora. Seleccione el modo "Task Template". Vaya a la plantilla de web scraping de Google Search en la categoría "search engine", Segundo: lee las instrucciones de la plantilla Abre la plantilla. Consulte las instrucciones y la salida de muestra para asegurarse de que esta plantilla le proporcionará los datos que necesita. Puede pasar el cursor sobre los campos de datos para ver qué elementos de los sitios web se extraerán. Consulte los parámetros para comprender mejor lo que necesita ingresar. Los parámetros variarían en diferentes plantillas, porque necesitan diferentes términos de búsqueda para continuar. Puede ser una URL, una palabra clave, una lista de URL/palabras clave , la cantidad de páginas que desea obtener, etc. En este caso, debemos ingresar el término de búsqueda "smoothie" Tercero: Usa la plantilla y comienza datos extracción Sigue haciendo clic en "use template", luego ingrese "smoothie" y presione "save and run". Si es una-vez proyecto, simplemente puede ejecutar el crawler en su computadora local. Mientras que, si está manejando un proyecto en curso, puede extraer datos en la  plataforma de la nube de Octoparse. Después de la extracción, puede exportarla a muchos formatos, como Excel, CSV y txt. Acabamos de presentar cómo usar una plantilla de web scraping para extraer web datos de Google Search. También puede crear su propio crawler utilizando el "Modo avanzado". Es posible que necesita algunas configuraciones, pero es muy flexible en términos de extracción de datos. Método #2: Extracción Personalizada con Modo Avanzado Paso 1: Ingrese la URL de destino para crear un crawler Si está tratando de , puede ingresar una lista de hasta 10,000 URL en el cuadro. En este caso, como solo estamos scraping solo un sitio web, peguemos la URL de destino en el cuadro y haga clic en "save URL" para continuar. extraer datos a gran escala Paso 2: Crea un bucle de paginación Cambie el navegador a Firefox 45. Ahora Octoparse ha cargado la página web en el navegador incorporado con éxito. Luego, debemos crear una paginación haciendo clic en el botón de página "Next" y seleccionando "Loop click next page" en el panel de Consejos de acción. Verá el bucle de paginación que acabamos de crear en el área de flujo de trabajo. Paso 3: Extraer los datos e iniciar la extracción Ahora podemos extraer los datos. Haga clic en el título de un resultado de búsqueda y haga clic en "select all". Después de seleccionar todos los títulos, se resaltarán en verde. Haga clic en "extract text of the selected element" para extraer todos los títulos. Hagamos una pausa por un momento para ver el flujo de trabajo. Como veis, acabamos de construir un bucle extracción dentro del ciclo de paginación. Todo el proceso de extracción funcionará de esta manera: el bot primero abrirá la página web, extraerá los títulos en la primera página uno por uno, y luego pasará a la página siguiente para repetir la extracción hasta que la extracción se detenga o se complete. Además de Google, las herramientas de extracción de datos pueden extraer datos de muchos otros sitios web, y se utilizan ampliamente en todas las industrias. Por ejemplo, las compañías pueden extraer mapas de Yellowpages, Yelp y Google para generar oportunidades de ventas. Puede consultar otros usos y aplicaciones de extracción de datos . Previously published at http://www.octoparse.es/blog/scrape-websites-sin-ser-bloqueado

Extraer Datos del Website a Excel Automáticamente

So, what does it take to create a startup that succeeds? Is it knowing the art of building successful apps? Or it is all about technical skills? For many startups where non-technical founders have created award-winning tech products, nothing could be further from the truth. For those who liken creating apps without tech knowledge to cooking in the absence of the right recipe, here’s news. Founders and CEOs with non-technical skills are shaping tech startups worldwide. Let’s start with the startup that revolutionized the concept of a home away from home: Airbnb. Brian Chesky Brian Chesky is the co-founder of Airbnb. He was also a designer to start with. Founding the peer-to-peer lodging service and becoming an internet billionaire was simply a matter of succeeding where great minds have failed. Brian was also named among the 100 most influential people by Time’s Magazine. Brian worked as an industrial designer at an LA firm right at the start of his career post a Bachelor of Fine Arts in Industrial Design way back in 2004. The Rhode Island School of Design alumnus is now brushing shoulders with the likes of Facebook’s techie founder Mark Zuckerberg. Airbnb has emerged as a market leader in recent times in the online hospitality industry, showing there’s always room for a great idea. In January 2017, Airbnb invested in Resy, a restaurant reservation app along with Gary Vaynerchuk and Mike Montero. The company first became profitable in 2016. It’s 2017 acquisitions ranged across Canada’s Luxury Retreat, social payment startup Tilt and travel accessibility startup Accomable. In 2018, Airbnb plus was offered along with luxury rentals by Beyond Airbnb. In March 2019, Airbnb completed an acquisition of over USD 400 million by purchasing hotel booking website HotelTonight. Melody McCloskey The founder of StyleSeat, an online appointment management system for stylists and the largest marketplace in the beauty services industry is a non-tech innovator. Melody McCloskey and co-founder Dan Levine work incredibly well together, sharing a common passion for making solutions that appeal to the beauty industry. More than USD 300 million worth of appointments has been booked through StyleSeat. Melody’s company offers an impressive listing of 200K stylists in 15,000 cities. After completing a degree in French and International Relations at UC Davis, the last thing on Melody’s mind was starting a tech company. This was about to change when she was at the receiving end of not one, but three styling disasters priced at USD 300 each. She worked in PR on the product side and made history when she quit the job to start StyleSeat. Giving women a chance to find the dream stylist at the single click of a mouse or a swipe of a smartphone proved to be a revolutionary idea. Innovating for women in tech, Melody’s venture has provided clear proof that tech skills can only take you so far, especially while building a tech startup. At the end of the day, speed and decision making capabilities count as much. Rather than worrying about writing code, bring excellent product management skills to the board and you’ll do well, too. From fundraising to monetizing the concept, Melody took beauty with a purpose to a whole new level. Tim Westergren When it comes to tech behemoths and giants, tech titans like Bill Gates and Mark Zuckerberg come to mind. But tech startups and products need not always have a leader with a technical background to succeed. Pandora Radio CEO and founder Tim Westergren is the case in point. He is a sound engineer and musician with a strong understanding of what drives a musical piece and the tech industry. Tim donned many hats, working as a music composer, recorder and even a nanny to make ends meet, before setting up Pandora Radio. Named as one of Time’s Most Influential 100 people in 2010, Tim has led the tech startup to establish itself as a key player in the music industry. From having USD 2 million in back wages to closing a 9 million funding deal, Tim has seen it all as far as Pandora Radio is concerned. Stepping down as CEO in 2017, Tim continues to inspire and lead the team at Pandora. Alex Turnbull Slow and steady wins the race, and as a non-tech guy who started a premier tech company, Alex Turnbull of Groove HQ knows this better than anyone else. He started with sketching to come up with wireframes. This led to innovation when he founded Groove, a helpdesk software company. In his blog, Alex explains how he used graphics editing software Photoshop to create mockups of the app and bridge the gap between co-founders and himself in terms of understanding. An avid surfer and a passionate software enthusiast without any technical skills, it took courage and grit to set up the software. But Alex was a visionary in 2011, introducing live chat when no one else even thought of it. Because of his amazing grasp of the tech industry, a lot of iterations and money were saved. A serial entrepreneur, Alex’s company earns millions per month in recurring revenue. Groove was a culmination of a lot of hard work and persistence. And from the stats coming in as Groove draws in investors and advisors like David Hauser, noted entrepreneur, speaker and angel investor. Perri Gorman Chase Non-tech co-founder Perri Gorman Chase thought of an email unsubscribe service and with the help of her tech friends, Unroll.me was born. Here’s what she learned in her own words. While prototypes look good at the front end, visual design made it easier for others to see what Perri wanted to accomplish. She succeeded in communicating with engineers and developers who translated her vision into reality. The focus is on understanding if the engineering team can speak one’s language and vice versa. Perri spoke of how one needs to know the landscape, databases, programming and API integrations or (in other words) how things fit together. Getting the basics right and controlling business outcomes are linked. Always maintain a database and an understanding of engineering processes and like Perri, you’ll get it right! Walker Williams Founder and CEO of Teespring, Walker Williams worked as a cartoonist and writer before he set out to form a tech company, with zero knowledge of coding or programming. Right from the start, Walker pursued diverse interests. At 16, he made up his mind to be an entrepreneur. And then secured a BA degree in Arts and History from Brown University. Fresh out of university, he collaborated with co-founder Evans Stites Clayton and created a platform for custom merchandise. Unique custom designs, setting sales goals and even prices for items is what distinguishes Teespring from others in the business. In 2016, Teespring was valued at USD 30 million alone. Walker was listed as the leading entrepreneur because of his revolutionary social commerce company. But it all began when the two friends Evans and Walker while still at University, designed and sold a t-shirt to commemorate the closing of a local hangout. Now, Teespring attracts 100K orders per month and growing! It has even been featured in Y Combinator’s blog. AJ Forsythe Founder and CEO of iCracked, AJ Forsyth actually holds a degree in science and biology/psychology! He’s tried his hands at everything from the winery to beekeeping before settling for the on-demand smartphone repair service through iCracked. While Forsythe graduated from California Polytechnic State University in 2011, he founded iCracked way before that. From a college dorm room in 2010 to the world’s largest, most efficient repair service for tablets and smartphones, Forsythe has come a long way as has his venture iCracked. The company has over 4K certified iTech professionals and operates in 11 countries. AJ also made it to Forbes 30 under 30 for his achievement. With the next move towards smart homes for all, iCracked is an early adopter with a difference. Stephanie Tilenius Founder and CEO of Vida Health Inc., Stephanie Tilenius is a Harvard and Brandeis University alumni with degrees in Economics and International Finance. So what motivated her to found a tech despite her non-tech background? Stephanie sought to provide expert, on-demand, personalized health coaching along with programs from experienced healthcare providers and leading medical centers. health company Close to 133 million individuals in the US live alone and face chronic health conditions. Her father had 4 chronic health conditions and as she struggled to build a health infrastructure to support him, Stephanie realized the need for having a scalable continuous mobile care platform and Vida Health was formed. In Stephanie’s words, Vida Health is the first “horizontal digital therapeutic” services provider. She also supports women entrepreneurs through Rivet Ventures, a fund. Sean Rad Founder and Chairman of Tinder, Sean Rad was a college dropout until he founded the dating app and became a CEO! Rad founded Tinder in 2012 with two other co-founders. Tinder managed a million matches within just 2 months of its launch. Further, Tinder had over 3.7 million paid subscribers, up by 81% over the same quarter in 2017. Presently, Tinder is valued at USD 3 billion. It is one of the highest grossing apps in the app store. Jamie Wong Founder and CEO of Vayable, Jamie Wong is a filmmaker par excellence and founder of tech company Vayable that uses technologies to promote travel. Jamie has a background in communication, advocacy, and disruptive media. She has pursued Art, History, and Politics, going onto a BA in History at Wesleyan University and finally the Columbia University’s Graduate School of Journalism. She founded Vayable in 2010, driven by the need to share collective and unique travel experiences, and it’s been quite a journey so far. Jamie’s tech company has a presence in 240 cities and was also featured in NYT, WSJ, and CNN. Evan Sharp Co-founder and Chief Creative Officer at Pinterest, Evan Sharp was definitely not a techie when he came up with the idea. Evan studied Architecture at Columbia University. Prior to that, he pursued a BA in History from the University of Chicago. While enrolled at the architecture course, he met Ben Silbermann, co-founder, and CEO of Pinterest. The image sharing site was launched in 2010. It is one of the most creative social media channels out there. In October 2018, Pinterest spanned a user base of 250 million monthly active individuals and 175 billion Pinterest Pins. Conclusion Pinterest, Vayable or any of the other tech products created by non-tech founders are doing so well. This proves you need innovation, not technical knowledge if you want to thrive in the tech startup world. So don’t let the absence of a formal degree deter you. Remember, knowledge grows with experience and expertise is not about coding alone!

11 Non-Technical Startup Founders Who Built Great Tech Products

11 fundadores de empresas emergentes no técnicas que crearon excelentes productos tecnológicos

Go is a programming language often used for applications in which performance matters. Optimizing your code based on assumptions is not a best practice of course. You need to have insights about your code performance and bottlenecks to be able to optimize it efficiently. What is a profiler? Profiler is a dynamic performance analysis tool that provides critical execution insights in various dimensions which enable resolving performance issues, locating memory leaks, thread contention and more. Go comes with built-in support for profiling! What kinds of profiles can I get? Go has several built in profiles: stack traces of all current Goroutines Goroutine: stack traces of CPU returned by runtime CPU: a sampling of memory allocations of live objects Heap: a sampling of all past memory allocations Allocation: stack traces that led to the creation of new OS threads Thread: stack traces that led to blocking on synchronization primitives Block: stack traces of holders of contended mutexes Mutex: How to get a profile? There are a few ways to create a profile. 1. Using “go test” to generate profile Support for profiling the standard testing package. As an example, the following command runs all benchmarks and writes CPU and memory profile to and : built into cpu.prof mem.prof go -cpuprofile cpu.prof -memprofile mem.prof -bench . test 2. Download live profile using HTTP If you need to have a live profile in a long-running service, package is your ultimate solution. Solely import the package in your code and you are good to go: net/http/pprof _ import "net/http/pprof" The initialization of this package is the reason you need to import it solely. {
	http.HandleFunc( , Index)
	http.HandleFunc( , Cmdline)
	http.HandleFunc( , Profile)
	http.HandleFunc( , Symbol)
	http.HandleFunc( , Trace)
} func init () "/debug/pprof/" "/debug/pprof/cmdline" "/debug/pprof/profile" "/debug/pprof/symbol" "/debug/pprof/trace" As you can see the package initialization registers handlers for the given URL in . These handlers write the profile into a which you can download. Now the only thing you need to do is running a HTTP server with a handler: DefaultServeMux http.ResponseWriter nil { {
		http.ListenAndServe( , )
	}()
} func init () go func () ":1234" nil Note that if you already run your http server with , you don’t need to run another server and profiles are available on your server. DefaultServeMux Now, you can download live profile data from the URL. /debug/pprof/ 3. Profiling in code Using , You can also profile directly within the code. For example you can start a CPU profile using and then stop it by . Now you have the CPU profile written to the given . runtime/pprof pprof.StartCPUProfile(io.Writer) pprof.StopCPUProfile() io.Writer But wait, there is an easier option. has developed a package named which can make profiling easy as pie. First you have to get the package using following command (or using Go modules): Dave Cheney profile go get github.com/pkg/profile Then you can use it like this: { profile.Start(profile.ProfilePath( )).Stop() } func main () defer "." // do something Above code will write CPU profile into in working directory. The default has been set to CPU profiling but you can pass options to to be able to get other profiles as well. cpu.pprof profile.Start() profile.Start(profile.CPUProfile).Stop() profile.Start(profile.GoroutineProfile).Stop() profile.Start(profile.BlockProfile).Stop() profile.Start(profile.ThreadcreationProfile).Stop() profile.Start(profile.MemProfileHeap).Stop() profile.Start(profile.MemProfileAllocs).Stop() profile.Start(profile.MutexProfile).Stop() // CPUProfile enables cpu profiling. Note: Default is CPU defer // GoroutineProfile enables goroutine profiling. // It returns all Goroutines alive when defer occurs. defer // BlockProfile enables block (contention) profiling. defer // ThreadcreationProfile enables thread creation profiling. defer // MemProfileHeap changes which type of memory profiling to // profile the heap. defer // MemProfileAllocs changes which type of memory to profile // allocations. defer // MutexProfile enables mutex profiling. defer How to use a profile? So we have collected the profile data, what now? is a tool for visualization and analysis of profiling data. It reads a collection of profiling samples in format and generates reports to visualize and help analyze the data. It can generate both text and graphical reports. pprof profile.proto All of the ways that mentioned above are using under the hood and this package writes runtime profiling data in the format expected by the pprof visualization tool. runtime/pprof Write A Simple Code To Profile For a better understanding, I’ve written the below code to profile. It is a simple HTTP server which handles requests on URL. Every request comes in, the function will declare a and create a child Goroutine at line 16. The child will decode the given JSON log within the and if it was successful it does a time consuming (~400 millisecond) write and then sends to . Meanwhile the function blocks on select statement (line 27) until either when the child sends the result to (line 24) or when a timeout happens (line 30). /log logHandler chan int Request.Body http.StatusOK ch logHandler ch ( _ ) {
	http.HandleFunc( , logHandler)
	http.ListenAndServe( , )
} {
	ch := ( ) {
		obj := ( [ ] ) err := json.NewDecoder(r.Body).Decode(&obj); err != {
			ch <- http.StatusBadRequest } time.Sleep(time.Duration(rand.Intn( )) * time.Millisecond)
		ch <- http.StatusOK
	}() { status := <-ch:
		w.WriteHeader(status) <-time.After( * time.Millisecond):
		w.WriteHeader(http.StatusRequestTimeout)
	}
} import "encoding/json" "math/rand" "net/http" "net/http/pprof" "time" func main () "/log" ":8080" nil func logHandler (w http.ResponseWriter, r *http.Request) make chan int go func () make map string float64 if nil return // simulation of a time consuming process like writing logs into db 400 select case case 200 I’ve compiled and run the code and have sent some loads to it using . Here is the output: hey Summary:
  Total:        0.8292 secs
  Slowest:      0.2248 secs
  Fastest:      0.0039 secs
  Average:      0.1574 secs
  Requests/sec: 241.1882Latency 

distribution:
  10% 0.0409 secs
  25% 0.1143 secs
  50% 0.2007 secs
  75% 0.2013 secs
  90% 0.2093 secs
  95% 0.2187 secs
  99% 0.2245 secsStatus code 

distribution:
  [200] 87 responses
  [408] 113 responses in in in in in in in pprof Basics After writing, running, and sending load to our service, it is time to run the pprof tool on the profile of the service to see which functions were the most consuming ones during the load. Install pprof Tool You can use command or install to use it directly. go tool pprof pprof go get -u github.com/google/pprof Run pprof On Interactive Mode Now let’s run pprof on allocated objects using the following command: pprof -alloc_objects http://:8080/debug/pprof/allocs This will open a simple interactive shell that takes pprof commands to generate reports. Now enter the command , the output will be the top memory consumers. top Dropped Nodes As you can see in the output, 11 nodes has been dropped (line 3). A node is an object entry and dropping a node can help to reduce the noise, but sometimes it may hide the root cause of the issue. Adding the flag or assigning it directly inside the attractive shell will include all data of the profile in the result. -nodefraction=0 Difference Between Flat And Cumulative In the above output, you can see two columns, and . Flat means , while cum (cumulative) means For better understanding, let’s assume the following function, the flat time of function is 4s and the cum is 11s. flat cum only by this function by this function and functions called down the stack. A {
   B() DO STH DIRECTLY C() } func A () // takes 1s // takes 4s // takes 6s Difference Between Allocations And Heap Profile An important note to consider is that the difference between the allocations and heap profile is the way the pprof tool reads them at start time. Allocations profile will start pprof in a mode which displays the total number of allocated objects since the program began which as well, but heap profile will start in a mode that displays number of live allocated objects which . includes garbage collected ones does not include garbage collected bytes That means in the pprof tool you can change the mode regardless of the profile type (heap or allocations). It can be done using the following values. You can either set them as a pprof flag or assign them to inside the pprof interactive shell: sample_index amount of memory allocated and not released yet (heap) inuse_space : amount of objects allocated and not released yet (heap) inuse_objects : total amount of memory allocated (regardless of released) alloc_space : total amount of objects allocated (regardless of released) alloc_objects : Get An SVG Graph From The Profile You can generate a graph of your profile in an SVG format, and open it with a web browser. The following command will request for a 5s CPU profile and will launch a browser with an SVG file. pprof -web http://:8080/debug/pprof/profile?seconds=5 Run pprof Via A Web Interface One of the useful pprof features is the web interface. If the flag is specified, pprof starts a web server at the specified host:por that provides an interactive web-based interface to pprof. -http pprof -http :8080 http://:8080/debug/pprof/goroutine The preceding command should automatically open your web browser at the right page; if not, you can manually visit the specified port in your web browser. Monitoring Goroutines From Command Line As you can see in the Goroutine profile, 115 Goroutines are alive which is not normal for idle service. It should be a Goroutine leak. So let’s check Goroutines status using the tool. grmon is a command line monitoring for Goroutines. Let’s install it and run it on host that is serving the profile data: grmon go get -u github.com/bcicen/grmon
grmon -host :8080 The output is: It seems the function blocks on a action. After reviewing the code I found the bug. It has occurred due to sending into an unguaranteed channel. I’ve fixed that by turn the to a buffered one channel (line 15). It was the very same mistake that I have mentioned . logHandler chan send ch this post How does a profiler work? Although stack traces generated by profilers are often intelligible, it is sometimes necessary to understand how profilers work under the hood to get even more details from the generated profiles. CPU Profiler The Go CPU profiler uses a signal to record code execution statistics. Once the signal got registered, it will deliver every specified time interval. This timer unlike typical timers will increment when the CPU is executing the process. The signal will interrupt code execution and make it possible to see which code was interrupted. SIGPROF When the function is called, the signal handler will be registered to call every 10 milliseconds interval by default (100 Hz). On Unix, it uses a syscall to set the signal timer. pprof.StartCPUProfile SIGPROF setitimer(2) On every invocation signal, the handler will trace back the execution by unwinding it from the current value. This will generate a stack trace and increment its hit count. The whole process will result in a list of stack traces grouped by the number of times each stack trace is seen. PC After the profiler is stopped, the stack traces will be converted to pprof-supported stack frames with source file names, line numbers and function names. Memory Profiler The memory profiler samples allocations. It will show function calls allocations. Recording all allocation and unwinding thestack trace would be expensive, therefore a sampling technique is used. heap The sampling process relies on a pseudo random number generator based on to sample only a fraction of allocations. The generated numbers define the distance between samples in terms of allocated memory size. This means that only allocations that cross the next random sampling point will be sampled. exponential distribution The sampling rate defines the mean of exponential distribution . The default value is 512 KB which is specified by the . That means to tune the size of the fraction to be sampled, you have to change value. Setting the value to 1, will include every allocated block in the profile. Obviously this will slow down your application. runtime.MemProfileRate runtime.MemProfileRate Stack traces with corresponding sampled allocations information is available any time, since the memory allocation profiler is always active. Setting the to 0, will turn off the memory sampling entirely. runtime.MemProfileRate Once the stack traces got ready, those will be converted to pprof-supported stack frames. Also published here .

Go: The Complete Guide to Profiling Your Code

Go: la guía completa para perfilar su código

 Wow, pnpm, You're Really Fast: How to Use It on Heroku

 Vaja, pnpm, sou molt ràpid: com utilitzar-lo a Heroku

Vaya, pnpm, eres muy rápido: cómo usarlo en Heroku

 وای، pnpm، شما واقعاً سریع هستید: نحوه استفاده از آن در Heroku

 וואו, pnpm, אתה ממש מהיר: איך להשתמש בו ב-Heroku

すごい、pnpm、本当に速いですね: Heroku での使い方

 Wah, pnpm, Anda Sungguh Pantas: Cara Menggunakannya pada Heroku

 Uau, pnpm, ești cu adevărat rapid: cum să-l folosești pe Heroku

 Wow, pnpm, Ju jeni vërtet i shpejtë: Si ta përdorni atë në Heroku

ว้าว pnpm คุณเร็วจริงๆ: วิธีใช้งานบน Heroku

 Wow, pnpm, U Hatlisa Hakunene: Ndlela Yo Yi Tirhisa Eka Heroku

 Voy, pnpm, siz juda tezsiz: uni Heroku-da qanday ishlatish kerak

Wow, pnpm, Bạn thực sự nhanh: Cách sử dụng nó trên Heroku

Most often, is defined as clear execution under sound judgement and discipline. rationality When investing, most tend to complement their thesis within a rational decision-making framework, as if that made their analysis "fundamental". Under the spectrum of business and economic valuation techniques, that might be correct. However, the markets are never wrong, and the world is not always rational. In the cryptoverse, over and under valuations have significant elements that make them prone to more subjectivity than traditional stocks. Premia and discounts, for instance, are only temporarily perceived, due to high volatility conditions. Supply curves, energy shortages, structural changes... none of that seems to fit into the narrative presented in the latest quarterly report or newsletter of a trending crypto fund, so one might think that timeliness is what makes the difference. It can, indeed. Those funds have bought a lot cheaper than any of its followers on Twitter. Let's take Solana, which despite having massively dropped along the rest of the crypto market, it is still at a 40x relative to its price 16 months ago. Markets always change, but do investors change? Does human nature change? Greed is still greed, and fear is still fear. Let's put some leverage into the mix and let's get ready to test rationality under the stress conditions of the bear market. How about finding talent? Not necessarily for building, but also for investing. Talk to any CEO and they'll say that recruiting high-quality people is one of their top priorities. In fact, spotting talent early is the key for most ventures in emergent markets. This is by no means an easy task, but whenever the opportunity comes, whether in bull or bear mode, one should be powered with both cash and confidence. Best way to spot this? Find out what the talent does during downtime. Also, keep in mind that “maximalists” are also born either out of irrationality or out of quantitative analysis. We are all familiar with the digital gold maxis and the toxicity that can potentially abound from some bold unfounded claims (miners have to pay bills, economic security is a thing…). However, if you are in crypto, you should know that VCs are part of the game too, and that behind their thesis, there is a founding team. Long-term sustainable business are not built out of irrational acts such as reacting to a vampire-attack, promising unsustainable yields... Instead, the yield comes from high velocity markets where the teams who are willing to adapt and learn are the ones who ultimately thrive. The difference between a senior and a freshman team should not be ignored. One important thing to note here, specially when assessing senior teams, it is that domain expertise teaches you about "you can't do", not about what "you could do". Venture capital is hard, it is not only about financial, but also about mental constraints. For example, when you are running low on capital with the combination of caring for people—many people will get crushed by the brutal situation of money running out. is the name of the game. If you can convince retail that a project is safer and has less selling pressure than it actually does, you can hit big. Asymmetric information Ultimately, investing is more than just assessing financial statements and valuations. It is also about making calculated bets on the future, which often may not have a quantifiable basis. Same for financial decisions, it is more than just whether it is fairly-valued today, but whether it has a compelling and achievable long-term vision – an end-game that it can works towards and bring into reality. The holy-grail is ultimately a survival game, where only the paranoid survive. How many VC funds actually sit down and examine their own death? In my opinion, this exercise is worth exploring: take the worst case scenario, and exacerbate its negative outcome by 3x or 5x. In other words, evaluate your own cause of death: the failure to survive. Exploration over hesitation is the key at this point. Honesty is the key: there are no guarantees of future success unless changes are made, specially while assessing highly innovative endeavors. Only the strongest survive, and those will be the ones who are willing to justify their own existence by adapting quickly, making smart moves, and not resting on their laurels. " I fear living a life where I could have accomplished something and didn't." Neil deGrasse Tyson It is the knowledge that I'm going to die that creates the focus that I bring to being alive. The urgency of accomplishment. When investing, there is no such thing as protection. Anything can certainly die, so the investors must justify their own existence one way or another. When will something die? Probably when it runs out of cash. The cause of death is unknown and, because of that,  all shapes and sizes must be carefully assessed under its proper risk-management framework. There will always be fear of the unknown. This is often neglected and leads many to under-water territories. Starting a company nowadays is not as expensive as it used to be, so many are tricked into believing that venture is not as risky as it was in the past. However, none of that removes the risk of failure. The risk can certainly be lower because of the existence of more established playbooks. And this is what makes most ventures be afraid of embarking into uncharted categories of investment, where big risks could cause them to miss out on the next generation of venture returns. No big risks also means deterioration: less experimentation, innovation, and creativity. With such big funds and such fast deployment cycles it is much more difficult to take big bets on riskier technologies. The bear market has come, and VCs just started dreaming too small, and building out their product offerings more thoughtfully. Otherwise, they risk destroying their reputation. Spray and pray no longer works, and reputation has become the new currency in the Venture Capital Universe. Embrace reality, accept the changing environment and reevaluate with intention rather than regret. Forget about the V-shape recoveries and prepare accordingly. In times of austerity, capital is expensive. Take the lead and adapt quickly, now it is about survival of the fittest: “it is not the strongest of the species that survive, nor the most intelligent, but the one most responsive to change.” Moving fast means to anticipate events. Proactivity will make you accept the pain of discipline before it is too late, which will be a lot better than suffering the pain of regret. Confronting the reality of a crash is the hardest step, but the reality is that we suffer more in our minds than we do in reality.

Spray, Pray, and Go Away: Investing is an Art

Rocíe, ore y váyase: Invertir es un arte

Vaporisez, priez et partez : investir est un art

स्प्रे, प्रार्थना, और चले जाओ: निवेश एक कला है

スプレーして、祈って、立ち去ってください: 投資は芸術です

Borrife, Ore e Vá embora: Investir é uma Arte

Xịt, Cầu nguyện và Đi: Đầu tư là một nghệ thuật

喷洒、祈祷和走开：投资是一门艺术

In previous weeks, I've analyzed several libraries and frameworks that augment the client with AJAX capabilities. Vue.js
Alpine.js
HTMX
Vaadin In this post, I'll compare them across several axes. Analysis Frontend skills:
Remember that I started this series from the point of view of a backend developer. In this section, I grade how much you need to know about client technologies to complete the job.


Team organization:
In the introduction, I hinted that the decoupling of frontend and backend teams profoundly impacted projects. Each team is fast on its own, and they can parallelize their work, but integrating the two can double the initial development time. Here, I grade how easy it is to integrate frontend and backend.


Ease of setup:
How easy it is to set up the tool from the backend.


Ease of styling:
Backend developers are not designers by default. Does the tool offer a default, at least average-looking style? How hard is it to create one? For all intents and purposes, Vue.js and Alpine.js are similar; I'll refer to them as JavaScript frameworks. JavaScript frameworks

HTMX

Vaadin



Frontend skills

Need the full HTML, JavaScript (it's in the name), and CSS tryptic

Only need HTML and CSS, HTMX takes care of the JavaScript burden

No frontend skills needed, Vaadin takes care of everything



Team organization

Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI

Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI

Each developer can develop their use case from the database to the UI



Ease of setup

* Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page

* Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page

Everything is in the POM. As Vaadin generates the whole frontend, you don't need additional setup



Ease of styling

No default; one needs to use an existing library, e.g., Bootstrap or create their own

No default; one needs to use an existing library, e.g., Bootstrap or create their own

Vaadin comes bundled with the Lumo theme. Other themes are available in the Vaadin Add-ons Directory, such as the Parity Theme. Applying a theme is as easy as setting it as a dependency and adding an annotation.Creating a custom theme is no small potatoes, though. You can ease the task by starting from an existing one and changing it bit by bit. Vaadin, the company, also provides custom themes for a fee. Time to choose If you are still unsure how to proceed, here are my recommendations. If you're working on a regular business app, e.g., forms, choose Vaadin. Business apps are Vaadin's primary use case and will shine there, immensely increasing productivity. If your app requires good-looking visualization widgets, choose Vaadin as well. For example, its Vaadin Charts component is truly amazing. Note that it's commercially licensed, though. If you want to offer an API from the start, choose Vue or Alpine. While it's possible to use HTMX or Vaadin, it doesn't make sense in this context. I also emphasize "from the start": everybody plans to offer an API at some point, but most never do. The possible productivity potential you plan to have in the future is not worth the guaranteed productivity in the next months. The same goes for distributing your app over several channels - from the start (bis repetita placent). If you're in none of these situations, it's time to go into more detail. Are your developers skilled in frontend technologies? Are they willing to learn to close the gap? Will you need these skills in the near future? These are a couple of questions that can help you decide which way to go. This post concludes my series on AJAX and SSR. I hope you had as much fun reading it as I did writing it. The complete source code for this post can be found on GitHub: https://github.com/ajavageek/compare-frontends?embedable=true In previous weeks, I've analyzed several libraries and frameworks that augment the client with AJAX capabilities. Vue.js Alpine.js HTMX Vaadin Vue.js Vue.js Alpine.js Alpine.js HTMX HTMX Vaadin Vaadin In this post, I'll compare them across several axes. Analysis Frontend skills:
Remember that I started this series from the point of view of a backend developer. In this section, I grade how much you need to know about client technologies to complete the job. Team organization:
In the introduction, I hinted that the decoupling of frontend and backend teams profoundly impacted projects. Each team is fast on its own, and they can parallelize their work, but integrating the two can double the initial development time. Here, I grade how easy it is to integrate frontend and backend. Ease of setup:
How easy it is to set up the tool from the backend. Ease of styling:
Backend developers are not designers by default. Does the tool offer a default, at least average-looking style? How hard is it to create one? Frontend skills: Remember that I started this series from the point of view of a backend developer. In this section, I grade how much you need to know about client technologies to complete the job. Frontend skills: Remember that I started this series from the point of view of a backend developer. In this section, I grade how much you need to know about client technologies to complete the job. Team organization: In the introduction, I hinted that the decoupling of frontend and backend teams profoundly impacted projects. Each team is fast on its own, and they can parallelize their work, but integrating the two can double the initial development time. Here, I grade how easy it is to integrate frontend and backend. Team organization: In the introduction, I hinted that the decoupling of frontend and backend teams profoundly impacted projects. Each team is fast on its own, and they can parallelize their work, but integrating the two can double the initial development time. Here, I grade how easy it is to integrate frontend and backend. Ease of setup: How easy it is to set up the tool from the backend. Ease of setup: How easy it is to set up the tool from the backend. Ease of styling: Backend developers are not designers by default. Does the tool offer a default, at least average-looking style? How hard is it to create one? Ease of styling: Backend developers are not designers by default. Does the tool offer a default, at least average-looking style? How hard is it to create one? For all intents and purposes, Vue.js and Alpine.js are similar; I'll refer to them as JavaScript frameworks. JavaScript frameworks

HTMX

Vaadin



Frontend skills

Need the full HTML, JavaScript (it's in the name), and CSS tryptic

Only need HTML and CSS, HTMX takes care of the JavaScript burden

No frontend skills needed, Vaadin takes care of everything



Team organization

Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI

Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI

Each developer can develop their use case from the database to the UI



Ease of setup

* Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page

* Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page

Everything is in the POM. As Vaadin generates the whole frontend, you don't need additional setup



Ease of styling

No default; one needs to use an existing library, e.g., Bootstrap or create their own

No default; one needs to use an existing library, e.g., Bootstrap or create their own

Vaadin comes bundled with the Lumo theme. Other themes are available in the Vaadin Add-ons Directory, such as the Parity Theme. Applying a theme is as easy as setting it as a dependency and adding an annotation.Creating a custom theme is no small potatoes, though. You can ease the task by starting from an existing one and changing it bit by bit. Vaadin, the company, also provides custom themes for a fee. JavaScript frameworks

HTMX

Vaadin Frontend skills

Need the full HTML, JavaScript (it's in the name), and CSS tryptic

Only need HTML and CSS, HTMX takes care of the JavaScript burden

No frontend skills needed, Vaadin takes care of everything Team organization

Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI

Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI

Each developer can develop their use case from the database to the UI Ease of setup

* Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page

* Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page

Everything is in the POM. As Vaadin generates the whole frontend, you don't need additional setup Ease of styling

No default; one needs to use an existing library, e.g., Bootstrap or create their own

No default; one needs to use an existing library, e.g., Bootstrap or create their own

Vaadin comes bundled with the Lumo theme. Other themes are available in the Vaadin Add-ons Directory, such as the Parity Theme. Applying a theme is as easy as setting it as a dependency and adding an annotation.Creating a custom theme is no small potatoes, though. You can ease the task by starting from an existing one and changing it bit by bit. Vaadin, the company, also provides custom themes for a fee. JavaScript frameworks HTMX Vaadin JavaScript frameworks JavaScript frameworks HTMX HTMX Vaadin Vaadin Frontend skills Need the full HTML, JavaScript (it's in the name), and CSS tryptic Only need HTML and CSS, HTMX takes care of the JavaScript burden No frontend skills needed, Vaadin takes care of everything Frontend skills Frontend skills Need the full HTML, JavaScript (it's in the name), and CSS tryptic Need the full HTML, JavaScript (it's in the name), and CSS tryptic Only need HTML and CSS, HTMX takes care of the JavaScript burden Only need HTML and CSS, HTMX takes care of the JavaScript burden No frontend skills needed, Vaadin takes care of everything No frontend skills needed, Vaadin takes care of everything Team organization Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI Each developer can develop their use case from the database to the UI Team organization Team organization Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI Depends on each developer's skills: * Either separation between frontend and backend development* Or they can develop their use case from the database to the UI Depends on each developer's skills:* Either separation between frontend and backend development* Or they can develop their use case from the database to the UI Depends on each developer's skills: * Either separation between frontend and backend development* Or they can develop their use case from the database to the UI Each developer can develop their use case from the database to the UI Each developer can develop their use case from the database to the UI Ease of setup * Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page * Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page Everything is in the POM. As Vaadin generates the whole frontend, you don't need additional setup Ease of setup Ease of setup * Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page * Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page * Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page * Thanks to WebJars, you can manage dependencies in the POM* WebJars Locator allows not specifying the version number in the HTML* You still need to reference each library in the HTML page Everything is in the POM. As Vaadin generates the whole frontend, you don't need additional setup Everything is in the POM. As Vaadin generates the whole frontend, you don't need additional setup Ease of styling No default; one needs to use an existing library, e.g., Bootstrap or create their own No default; one needs to use an existing library, e.g., Bootstrap or create their own Vaadin comes bundled with the Lumo theme. Other themes are available in the Vaadin Add-ons Directory, such as the Parity Theme. Applying a theme is as easy as setting it as a dependency and adding an annotation.Creating a custom theme is no small potatoes, though. You can ease the task by starting from an existing one and changing it bit by bit. Vaadin, the company, also provides custom themes for a fee. Ease of styling Ease of styling No default; one needs to use an existing library, e.g., Bootstrap or create their own No default; one needs to use an existing library, e.g., Bootstrap or create their own No default; one needs to use an existing library, e.g., Bootstrap or create their own No default; one needs to use an existing library, e.g., Bootstrap or create their own Vaadin comes bundled with the Lumo theme. Other themes are available in the Vaadin Add-ons Directory, such as the Parity Theme. Applying a theme is as easy as setting it as a dependency and adding an annotation.Creating a custom theme is no small potatoes, though. You can ease the task by starting from an existing one and changing it bit by bit. Vaadin, the company, also provides custom themes for a fee. Vaadin comes bundled with the Lumo theme. Other themes are available in the Vaadin Add-ons Directory , such as the Parity Theme . Applying a theme is as easy as setting it as a dependency and adding an annotation. Creating a custom theme is no small potatoes, though. You can ease the task by starting from an existing one and changing it bit by bit. Vaadin, the company, also provides custom themes for a fee. Vaadin Add-ons Directory Parity Theme Time to choose If you are still unsure how to proceed, here are my recommendations. If you're working on a regular business app, e.g. , forms, choose Vaadin. Business apps are Vaadin's primary use case and will shine there, immensely increasing productivity. e.g. If your app requires good-looking visualization widgets, choose Vaadin as well. For example, its Vaadin Charts component is truly amazing. Note that it's commercially licensed, though. Vaadin Charts If you want to offer an API from the start , choose Vue or Alpine. While it's possible to use HTMX or Vaadin, it doesn't make sense in this context. I also emphasize "from the start": everybody plans to offer an API at some point, but most never do. The possible productivity potential you plan to have in the future is not worth the guaranteed productivity in the next months. from the start The same goes for distributing your app over several channels - from the start ( bis repetita placent ). from the start bis repetita placent If you're in none of these situations, it's time to go into more detail. Are your developers skilled in frontend technologies? Are they willing to learn to close the gap? Will you need these skills in the near future? These are a couple of questions that can help you decide which way to go. This post concludes my series on AJAX and SSR. I hope you had as much fun reading it as I did writing it. The complete source code for this post can be found on GitHub: https://github.com/ajavageek/compare-frontends?embedable=true https://github.com/ajavageek/compare-frontends?embedable=true

Comparing Vue.js, Alpine.js, HTMX, and Vaadin: A Backend Developer's Guide to AJAX and SSR

Stories, News & Mentions about VAYA

Vaya’s Stories on HackerNoon