李 | Roy | roylee (@roylee0x) header
李 | Roy | roylee (@roylee0x) profile picture

IQ 50, Indie dev. Building https://nitter.gq/t.co/Kw8RkCNlgZ, https://nitter.gq/t.co/7I5so4CZ5f

🇲🇴
Joined September 2017
Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)Photo by 李 | Roy | roylee (@roylee0x)
李 | Roy | roylee retweeted
Before you rush to buy a stock that's down 15-20% after earnings, understand one important rule: The 3-Day Rule! What is the 3-Day Rule, and why does it matter? Let me explain. When a stock gaps down sharply after earnings or bad news, wait at least 3 trading days before considering buying. Why? Large institutions managing millions of shares may need several sessions to reduce their positions. The initial gap down might be just the beginning of the selling pressure. Day 1: An initial bounce could be driven by short covering and aggressive dip buyers rather than sustained institutional demand. Days 2 and 3: Selling pressure may continue, and the stock could make new lows as investors reassess the news. After Day 3: Look for signs that selling pressure is fading and price is beginning to stabilize. Before buying, check: - Selling volume is shrinking - Price is holding above the recent low - You understand why the stock dropped - The fundamental thesis remains intact Remember, three days don't guarantee a bottom. Some stocks need several weeks or even months to build a proper base before beginning another sustained uptrend. If the business is fundamentally broken, the stock isn't necessarily on sale. It may simply be getting repriced. The market doesn't reward you for catching the exact bottom. Let the selling pressure fade, let the stock build a base, and let the trend confirm itself. Missing the bottom is far better than getting caught in a downtrend
🤖 Made with AI
80
416
41
2,660
333,658
李 | Roy | roylee retweeted
hairlines v0.2 is now available! draw your own with the /hairline-create skill
65
184
30
4,434
229,805
李 | Roy | roylee retweeted
14
17
1
719
24,769
李 | Roy | roylee retweeted
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence. block.xyz/dessa
4,162
7,137
4,763
69,024
13,777,923
李 | Roy | roylee retweeted
My iPhone Duo video went viral, so I'm giving away 100 iPhones on X 😛
91,707
92,401
6,536
139,823
6,156,050
Let your Grok Bot / Muse talk to your Claude Code / Codex or any other agents Introducing Reason Network, Free, peer to peer and Open-Source Engine Loop in your teammates too so no live agents work collides when coding. Try it for free: reasonmachines.com/network one click install and setup!
193
317
174
3,954
622,627
李 | Roy | roylee retweeted
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
1,949
1,746
901
32,081
4,912,912
李 | Roy | roylee retweeted
we meant it when we said machine-native types go far, far beyond just decisions, and were very careful about how we defined System One models. introducing the bounding box primitive to Jev 1.14, for locating at the speed of Jev. deets in the docs: docs.typesafe.ai/go/box
4
44
26
538
117,593
李 | Roy | roylee retweeted
做产品我可以熬夜,做增长我能拖三天。 产品磨得再极致,没人用也是白磨。 道理我都懂,就是不想干。
56
6
5
146
16,428
李 | Roy | roylee retweeted
Just wondering is anyone still hiring? The last few months job search was really a joke.
3
1
1
10
395
李 | Roy | roylee retweeted
Ok it makes sense lmao
261
1,652
489
12,039
1,968,204
李 | Roy | roylee retweeted
I've had agents implement and optimize the Campfire web app in Elixir, Go, and Rust. Yes, Ruby is the slowest. I didn't care about that when the pay-off was huge in terms of productivity and developer joy. But if you're no longer reading the code? github.com/basecamp/once-cam…
Readers added context they thought people might want to know
The shown benchmarks use non-equivalent implementations: Elixir routes all DB queries through a single GenServer with Redis for jobs (retained from Rails), while Rust uses in-process queues, omits CSRF tags/fields, and applies other caching changes. github.com/basecamp/once-… raw.githubusercontent.com/basecamp/once-… github.com/basecamp/once-… github.com/basecamp/once-… x.com/zachsdaniel1/s…
351
96
176
2,022
1,026,743
李 | Roy | roylee retweeted
GPT-6.1 Sol keeps getting faster every day. It launched painfully slow coz demand was apparently insane, but OpenAI keeps bringing more capacity online and the difference is already noticeable. If they actually get Sol 6.1 to ~2x the original speed while keeping this level of intelligence + efficiency nd cost then we’re talking. Maybe the model was never slow. The servers were just fighting for their lives.
14
4
1
38
4,488
李 | Roy | roylee retweeted
i wanted to voice dictate without bothering everyone so I made this lip reading app lol github repo in the thread
292
179
129
2,878
222,892
李 | Roy | roylee retweeted
谁能想到,真正继承乔布斯衣钵的,竟然是他。 大屏、极简设计、现场Demo、产品揭晓、One more thing:美国政府发布网站America.gov,鲁比奥介绍新版美国护照的发布会让人似曾相识。
65
250
59
2,054
184,608
李 | Roy | roylee retweeted
一位前 Stripe 员工回忆 2016 年的 Stripe,我觉得最值钱的是第一条: 两个创始人 Patrick 和 John,谁私信都几乎秒回。 普通员工想约他们,他们总能挤出 15 分钟。 不管你在公司什么级别。 结果大家反而舍不得用,怕占他们时间。 这哥们说现在回头想,从创始人那边看,大概就像员工都在躲着他们。 他后面有个比喻,我觉得太准了: 这些私下的连接,相当于在公司里插了几十根双向探针。 下面的信息能直接传到创始人手里,创始人也能绕开正式流程去推一把。 他觉得这对 Stripe 成功的贡献,可能跟正式流程一样大。 其他几条也挺有意思: - 特别自下而上。创始人定一个目标,每一层自己想办法把自己那块往那个方向掰。好处是灵活、好玩、有创造力。坏处是经常完全不知道该干啥,得自己琢磨。 - 老员工手里可能有 5000 万美元的期权,新人就是找了份普通工作。这哥们说他在任何一次互动里,都没感觉到过一丝这种差别。 - Stripe Press 根本没有商业上存在的理由。花管理层的精力,去精装出版一本《波斯王子》的开发日志,唯一的理由是他们真信把这些书放到人的书架上,世界会变好一点。 最后他说了句挺诚实的话: 这种环境不适合所有人。需要很多结构的人会很惨。 但喜欢的人是真的喜欢。 我觉得很多公司学 Stripe,流程能抄走,那个秒回抄不走。 因为秒回的成本全压在创始人自己身上,没法外包给任何人。 看完我打开自己的未读消息,99+。 这根探针,在我这应该是插反了。 nitter.gq/spakhm/status/21045910…
What made Stripe great, c. 2016 (Part II) - p&j would talk to anyone any time. You could dm them and get a near instant response. If you needed to meet them they'd always find 15m for you. They'd do that with everyone, you didn't have to be a high-ranking employee. It's impossible to overstate how accessible they were. (We all tried really hard to use this sparingly to protect their time, though in hindsight from their pov it must have looked like ppl are avoiding them) - if you abstract this a little, these informal connections functioned like dozens of bidirectional probes inserted throughout the org. It let ppl give information directly to the founders, and let p&j steer the company outside of standard channels. This probably contributed to Stripe's success as much as the formal process - everyone was extremely optimistic, not just about Stripe but about the world as a whole. There was this conviction that the world can be improved, that there is no limit to this, that making it better is everyone's responsibility, and that big improvements are within reach. Stripe of that time was brimming with noblesse oblige - one tangible example of this is Stripe Press. There is no commercial reason for it to exist. Why spend management bandwidth on publishing a beautiful meticulously typeset hardcover of The Making of Prince of Persia? The only reason is a strong conviction that putting these books on people's shelves materially improves the world. (Btw you can experience that sense of optimism yourself. Buy a Stripe Press book, you will feel it the moment you hold it in your hands) - everyone without exception had an egalitarian mindset. In hypergrowth startups you have OGs with $50M worth of options and new employees who basically just got a normal job. Not one time did I sense even a shade of this in any interaction. All anyone ever thought about was contributing to building the rocket ship as best they could - p&j were surprisingly good writers. (I think Patrick was more prolific, maybe three essays for every one of John’s, but both were very good. My memory might be off though.) Running Stripe demanded an insane rate of learning of them. Every time they'd work out some new problem they'd write about it, so in a sense everyone got to learn with them. The essays were typically far more perceptive than the best ft/economist writing imo - the executive team was felt everywhere yet had a very light touch. If you picture a company as a growing clump of cells, execs would be standing to its side, collecting vitals, reading signs, tending here, nudging there, pruning the occasional bad cell. They could be commanding when necessary but reaching for chain of command was not their first instinct - Stripe felt incredibly bottom up. p&j would declare a goal and at every level people did their best to reorient their corner of the company in the goal's direction. There are tradeoffs to this. The upside is that the company was remarkably agile, more fun to work in, more inventive. The downside is that it often could be totally unobvious what to do, you'd have to figure it out yourself - this wasn't for everyone. If you needed a lot of structure you'd be screwed. But if you loved it you *really* loved it. Stripe selected for those people. You were surrounded by scrappy, tenacious, intrepid colleagues, it was probably the best part of working there
4
53
20,846
李 | Roy | roylee retweeted
Jev can’t do any of these: 1. Image inputs 2. String, integer, and number types 3. Dependency execution in a single call TypeLLM can. Try it: github.com/TypeLLM/TypeLLM
45
262
17
2,834
198,323
李 | Roy | roylee retweeted
看 Rails 社区对 DHH Keynote 的反应,我觉得他们接受不了的是:或许 Ruby 和 Rails 也可能有一天不再重要。 @dhh 这次已经走得很远,远得很多人无法接受,而且我觉得他说得还是稍微委婉了一些。 HEY 开始走向 Native + Rust。在我看来,这可能只是一个开始。 我甚至认为,未来在很多服务端场景里,Ruby 和 Rails 都可能逐渐被替代。不是因为它们不好,而是时代的条件已经变了。 都已经 Pencils Down 了,代码都不需要自己写了,那么很多过去为了开发效率做出的技术选择,自然值得重新考虑。 Rails 已经 20 多年了。它当然伟大,Convention over Configuration、Programmer Happiness、One Person Framework,这些思想影响了整整一代开发者,也深深影响了我。 但伟大的应该是这些思想,而不是那几百万行 Ruby 代码。 这一点我其实很能理解 DHH。 我在 11 年开始和一些人一起组建 Ruby China 社区,一直到现在,很多人因为 Ruby 认识我。但后来我逐渐从 Ruby 到 Go,再到 Rust,几乎不再写 Ruby。早些时候其实我一直不好意思在社区里面说这个事情,感觉不好开口。 不是我否定了 Ruby。 恰恰相反,Ruby 留给我的很多东西,今天依然存在于我做的事情里。 只是当我发现新的工具能把事情做得更好,我没有理由因为自己的过去,就要求自己的未来也必须是 Ruby。 所以如果有一天 Rails 大量使用 Rust,甚至最终变成一个完全不同的东西,我一点都不会觉得这是对 Rails 的背叛。 真正背叛 Rails 的,反而可能是为了守住 Rails,而拒绝继续改变。 DHH 今天还没有这么说。 但我认为他心里已经开始走到这里了。 不信,等着瞧。
29
17
12
211
85,829
李 | Roy | roylee retweeted
The plan was there all along.
186
255
85
6,668
207,411
李 | Roy | roylee retweeted
sharing some notes on typesafe 🤝 coding agents: docs.google.com/document/d/1… we likely will never have time (ever again) to play ourselves, but hope the that the community goes WILD (and makes me look like a naive idiot)
140
244
72
3,161
583,138