๐Ÿš€ Software Engineer with +4 years working abroad ๐Ÿ‡บ๐Ÿ‡ธ โณ +13 years of experience with IT โšก๏ธ I write code in PHP, JS, TS and GO ๐Ÿ‘‡๐Ÿป

Sรฃo Paulo, Brasil
Joined November 2010
Lucas Sahdo retweeted
Need help from the builders and PHP community! ๐Ÿš€ elephc is the first real PHP-to-native compiler (no VM, no runtime). Weโ€™re currently at ~250 stars and Iโ€™d love to reach 500 as a milestone. If you find the project interesting, a star would mean a lot and help more people discover it. โ†’ github.com/illegalstudio/eleโ€ฆ What do you think, is native PHP compilation exciting?
14
29
1
137
156,249
Hรก algumas semanas publiquei um vรญdeo sobre DDD estratรฉgico e tudo o que vocรช nรฃo ouvio falar sobre Domain Driven Design by the book. Para meus seguidores dessa rede segue youtu.be/FKNrpyPqK48?si=NYkCโ€ฆ
38
Sรกbado dia 16/05 eu tive a oportunidade de palestrar presencialmente sobre um tema que gosto muito e uso todos os dias nos projetos que atuo no mercado. Domain Driven Design. Minha mensagem tem sido constante e previsรญvel: Pecisamos modelar domรญnio de forma mais estratรฉgica!
1
80
Changing an old #Laravel project here. Iโ€™m renaming the persistence layer from #Repository to #DAO, since repos should be aggregate centric, right? And this one clearly follows more of a Laravel way style. For now: โ€œActionโ€ pattern, anemic domain, and DAO baby, cheers! ๐Ÿบ
1
75
If your #Repository is just a hashtag#proxy to the #ORM, you're missing the point. Repository is a bridge between your hashtag#domain and your #infrastructure. The whole idea is that your business logic doesn't know how data is stored. New video is out.
1
1
121
Lot of work to do lately but I just published my take on the Repository Pattern. "Repositories are useless! They're just proxy layers for ORMs!"ย ๐Ÿคฏ if you've heard this, someone probably missed the point.
1
50
Lucas Sahdo retweeted
the lessons I've learned scaling my startups: > 90%+ of users pick google sign in. just make it the default. > strip all formatting from your emails. send from a real name, not "company team." watch your open rates jump. > add "how did you hear about us?" to onboarding. makes marketing 10x easier. > 99% of people sliding into your dms with "opportunities" are wasting your time. protect your calendar. > creator sponsorships cost less than ads but eat more time. pick based on what you have more of. > building a good product is just relentlessly asking "what does my user actually want" and doing that. > copycats show up the moment you get traction. they'll clone everything. they still won't catch you. > i couldn't build something good if i wasn't using it daily. > always check logs right after pushing updates. > bugs are fine. slow fixes aren't. > people notice good ui/ux design even if they can't explain why. > first 10 paying customers are harder than the next 100. > always refund people who ask. not worth the energy or the bad review. give them 2 months free too. > marketing is just running experiments fast. steal from what works for similar products. > cheap accountants cost more in the long run. > users will hop on calls way more than you'd expect. these conversations are cheat codes. > strong testimonials let customers sell for you. > you won't have a "pmf moment." only way to know is when people tell their friends about your product (word of mouth) > even when things are going well you'll have days where you think it's all falling apart. it's not. keep building.
87
155
13
2,654
135,514
And then: ๐Ÿฎ๐Ÿฌ๐Ÿญ๐Ÿฐ: ๐—™๐—ฎ๐—ฐ๐—ฒ๐—ฏ๐—ผ๐—ผ๐—ธ ๐—ฟ๐—ฒ๐—น๐—ฒ๐—ฎ๐˜€๐—ฒ๐˜€ ๐—›๐—ฎ๐—ฐ๐—ธ ๐Ÿฎ๐Ÿฌ๐Ÿญ๐Ÿฒ: ๐—ฆ๐—น๐—ฎ๐—ฐ๐—ธ ๐—บ๐—ถ๐—ด๐—ฟ๐—ฎ๐˜๐—ฒ๐˜€ ๐˜๐—ผ ๐—›๐—›๐—ฉ๐—  ๐Ÿฎ๐Ÿฌ๐Ÿญ๐Ÿด: ๐—ฅ๐—ผ๐—ฎ๐—ฑ๐—ฅ๐˜‚๐—ป๐—ป๐—ฒ๐—ฟ ๐—ฟ๐—ฒ๐—น๐—ฒ๐—ฎ๐˜€๐—ฒ๐—ฑ ๐Ÿฎ๐Ÿฌ๐Ÿฎ๐Ÿฎ: ๐—™๐—ฟ๐—ฎ๐—ป๐—ธ๐—ฒ๐—ป๐—ฃ๐—›๐—ฃ ๐—ฟ๐—ฒ๐—น๐—ฒ๐—ฎ๐˜€๐—ฒ๐—ฑ
1
1
78
Today, if you're starting fresh and need high-performance #PHP, you have great options: Swoole, FrankenPHP, RoadRunner. But back then? These companies had to build their own paths...
1
108
Why Facebook, Slack, and ManyChat didn't choose Swoole, FrankenPHP, or RoadRunner to solve their performance problems with #PHP back there? The way I see it, the simple answer: ๐—ง๐—ถ๐—บ๐—ถ๐—ป๐—ด.+
1
1
160
Why I think that? Here's when each solution became mature: ๐Ÿฎ๐Ÿฌ๐Ÿญ๐Ÿฌ: ๐—™๐—ฎ๐—ฐ๐—ฒ๐—ฏ๐—ผ๐—ผ๐—ธ ๐—ผ๐—ฝ๐—ฒ๐—ป-๐˜€๐—ผ๐˜‚๐—ฟ๐—ฐ๐—ฒ๐˜€ ๐—›๐—ถ๐—ฝ๐—›๐—ผ๐—ฝ ๐Ÿฎ๐Ÿฌ๐Ÿญ๐Ÿญ: ๐—™๐—ฎ๐—ฐ๐—ฒ๐—ฏ๐—ผ๐—ผ๐—ธ ๐—ฎ๐—ป๐—ป๐—ผ๐˜‚๐—ป๐—ฐ๐—ฒ๐˜€ ๐—›๐—›๐—ฉ๐—  ๐Ÿฎ๐Ÿฌ๐Ÿญ๐Ÿฎ: ๐—ฆ๐˜„๐—ผ๐—ผ๐—น๐—ฒ ๐—ฑ๐—ฒ๐˜ƒ๐—ฒ๐—น๐—ผ๐—ฝ๐—บ๐—ฒ๐—ป๐˜ ๐—ฏ๐—ฒ๐—ด๐—ถ๐—ป๐˜€ (๐—ถ๐—ป ๐—–๐—ต๐—ถ๐—ป๐—ฎ)
1
51