Hemanshu

11K posts

Hemanshu

Hemanshu

@heman_

Building https://t.co/bdSCAtVp6M & https://t.co/P8o2qEjz6w

Katılım Mart 2009
2.2K Takip Edilen200 Takipçiler
Hemanshu
Hemanshu@heman_·
@Prakashplutus Just glanced at infy and tcs fii holding .. they still have 30% if infy and 10% of tcs only to sell ... These companies have no growth story now with AI
English
0
0
0
947
Investing @ Prakash
Investing @ Prakash@Prakashplutus·
FIIs are not stopping , now they are selling @ $1 billion run rate. Flight of capital ? Pressure on currency ? Better we admit a mistake & correct it. 📊 Participant-wise Derivatives Data @ 23400 . 📌 FII Activity 🔹 Index Futures Added + 6,211 Long Reduced - 5,560 Short Long: Short Ratio –12:88 🔹 Stock Futures Added + 22,645 Long Added + 16,878 Short Long: Short Ratio – 58:42 FII in index futures added longs & reduced shorts & Stock futures added longs and shorts. 💰 Cash Market Update –16th March 2026 (Includes Block Deals – Provisional Data) DII Net : + 12,593 Cr FII Net : - 9,365 Cr 🔗 Refer to the disclaimer plutusadvisors.com/disclosure-ra
English
15
15
293
74.1K
Hemanshu
Hemanshu@heman_·
@johncrickett It is 100% over-engineered. It's best version was around their 100 releases back. I cancelled my claude sub. Trying codex. Or back to our good ol ways of coding by hand.
English
0
0
1
31
John Crickett
John Crickett@johncrickett·
I spent the weekend actually reading the Claude Code docs. It's a rabbit hole. CLAUDE.md files. MCP configs. Skills. Subagents. Hooks. Plugins. Agent Teams. You could spend more time configuring Claude Code than building software. All of it is productivity theatre. The only thing that actually matters: think first, then give it focused, relevant context.
English
114
30
738
78.3K
Hemanshu retweetledi
Anuradha Tiwari
Anuradha Tiwari@talk2anuradha·
ANI - Silent Aaj Tak - Silent Zee News - Silent ABP - Silent Times Now - Silent NDTV - Silent The entire ecosystem is against General Category !
English
365
5.5K
16.5K
133.2K
Hemanshu retweetledi
THE SKIN DOCTOR
THE SKIN DOCTOR@theskindoctor13·
ANI is not covering the anti-UGC protest at Ramlila Ground or related news. It’s not that the protest is too small for ANI to cover because they report almost anything, even if it matters to no one. Just announce you’re doing a hawan for the India vs New Zealand final and their reporter will reach your house immediately, as if the country’s future depends on that sacred footage. Just a few days ago they even posted (and later quietly deleted, after criticism) an insensitive clip of their reporter interviewing the hearsay van driver who had come to take cricketer Rinku Singh’s father’s body from his home, as if that was needed. Clearly, an intentional blackout.
English
283
4.1K
13.4K
233.8K
Hemanshu
Hemanshu@heman_·
As you guys rightly identify that working with Claude code and in general LLMs is an ART. Let us play with it. Your over engineering is spoiling our canvas.
English
0
0
0
181
Hemanshu
Hemanshu@heman_·
Another eagerness of @claudeai recently to do everything via code versus actual reading is causing the performance to be deteriorated in unintended ways. I just think these guys @bcherny and team should stop over engineering claude code. (1/2)
English
1
0
0
172
Hemanshu
Hemanshu@heman_·
The eagerness of Claude Code to get into implementation mode when we are brainstorming approach is annoying and frustrating. @bcherny @trq212
English
0
0
0
24
Hemanshu
Hemanshu@heman_·
Codex is a refreshing change. I have hesitated to switch from claude code. But now its slow, seems overengineered, and usage limit even on a 100$ plan exhausts very fast. I guess will cancel my max plan and move to codex!
English
0
0
0
84
Hemanshu
Hemanshu@heman_·
@trq212 @trq212 : The usage is again filling up very fast. Much faster than it has been over last two or three months.
English
0
0
0
7
Thariq
Thariq@trq212·
We've reset rate limits for all Claude Code users. Yesterday we rolled out a bug with prompt caching that caused usage limits to be consumed faster than normal. This is hotfixed in 2.1.62. Make sure you upgrade to the latest and hope you enjoy using Claude Code this weekend!
English
728
449
11.1K
1M
Hemanshu
Hemanshu@heman_·
@trq212 No wonder, I have never hit my limits and this time I hit limits in a few prompts from both my claude code accounts
English
0
0
0
77
Hemanshu
Hemanshu@heman_·
@CTOAdvisor What are those skills which are resilient to age discrimination?
English
0
0
2
157
Keith Townsend
Keith Townsend@CTOAdvisor·
I’ve seen world class talent not find a job at their skill level completely due to market conditions. This happens to experienced engineers as they age. As you age, you don’t have the luxury to only work on the things you love. You must consider ageism and add skill that is resilient to age discrimination.
English
3
0
19
2.4K
Hemanshu
Hemanshu@heman_·
@bcherny @trq212 : whatever u guys did for in the recent revisions of claude code w.r.t each prompt referring to a new plan isnt working. Old plans provide a lot of conext for continuing task. Now each time CC is getting into investigation & i have to teach it again. That sucks!
English
0
0
0
34
Hemanshu retweetledi
Suhail
Suhail@Suhail·
Make the margins next to zero for all these AI models. It was trained on humanity's data, it should be gift to ourselves. Doing so will save us from a few in control of our species. Distill at industrial scale! Distill, I say!
English
83
125
1.5K
56.5K
Hemanshu
Hemanshu@heman_·
@unclebobmartin Core of AI led development is going to be relentless testing. No unwanted behaviour should be introduced
English
0
0
0
25
Hemanshu
Hemanshu@heman_·
@unclebobmartin Jira shud go because features are getting developed so fast that you don't need multi days and multi sprints tracking.
English
1
0
0
30
Uncle Bob Martin
Uncle Bob Martin@unclebobmartin·
I have an itchy feeling that AI development is going to be a boon for QA and Jira.
English
41
11
243
41.3K
Hemanshu
Hemanshu@heman_·
Hey @bcherny @trq212 /insights is crashing. Every prompt is leading to a new plan when old plan would have given it a lot of context. Speed seems to have slowed a lot with 4.6. and a lot of times 15 to 20 mins of run give no output. Frustrating.
English
0
0
0
12
Hemanshu
Hemanshu@heman_·
@unclebobmartin Key to working with claude code is hell lot of functional testing. Refactoring for elegant code when you have achieved your desired functionality and then again functional testing. That's all.
English
0
0
0
14
Hemanshu
Hemanshu@heman_·
@unclebobmartin With deep respect sir, most of the clean coding principles especially around having very small functions or doing TDD dont work well with claude. The larger the code blocks easier it is for claude code to get relevant context. Lesser unnecessary code better it is
English
1
0
0
41
Uncle Bob Martin
Uncle Bob Martin@unclebobmartin·
One thing is abundantly clear. Wielding a power tool like Claude requires a lot of skill.
English
48
29
557
27.7K