Phew, ይሄ ነገር በጣም አስቀያሚ ስራ ነበር ሚካኤልን cuz እኔ ያለሁበት ሰፈር ብዙ አሉ እና ስራ ፈተው ሲያሳዝኑ like ለምግብ እስኪቸገሩ ልጅ ያለው ምናምን ነበር እና ምንም irresponsible የሆነ ውሳኔ ነበር good thing ትንሽ ይዘገያሉ እንጂ ያስባሉ I think😁

I can't believe it is on V4 and i never been to any of them jesos😅😅😅

Go and ask ChatGpt :

Fill this room with everything you know about me please

Here is my result 😮‍💨

You may think የማይገባ bug or break up or being broke is hurting huh you thinh they are painfull



Have you tried being Man united fan 😭😭 have you treid watching this 11 piece off shits play football 😭😭😭😭

Source Control Needs to Evolve for the AI Coding Era

yesterday i submitted a PR with 27K+ changes across 200+ files , it wasn't an overnight change, it was an over 2 weeks worth of task and it was about 60 or 70 commit containing continuous back and forth changes.....
What i did wrong was doing it all on a single branch and managing it under a single PR, and ofc this has cursor made changes, wdym 😁.... but all are properly monitored and supervised but it still feels overwhelming and this made me think

we need to update our Source Control logic

like in this coding agent supported coding culture , we do know the changes are reflected with large amount diff(A diff is a list of lines that changed) and Reviewing it means reading those lines. so the source control management should be changed in a way that it can work with this AI enhancement and coding style so just like all the IDE's and frameworks SaaS platforms evolved to adopt the AI era even the way we write and debug code has changed, so do our Source Controlling standard should be evolved

This is my take Source control-our Git Logic has to evolve to adapt the current ai era or coding. I’ll be honest, I genuinely don’t have a full solution or a concrete proposal. But I know the current model has to change, from a syntax-based( In an AI-assisted workflow, that’s the wrong level of abstraction) and tracking which lines and characters changed to see is purpose-based or action-based source control, one that tracks what we were trying to do rather than only which lines moved. and go to line based from that level to the LOC(Line of Code) based diff . with this the intent becomes the unit of review, and the line-level diff is just supporting evidence.That would make long large , agent assisted changes reviewable by checking the purpose , the scope, and a few representative samples, rather than scrolling through thousands of lines and hoping nothing slips through.
so unless we upgrade our source controll ,we’ll keep ending up with PRs that feel overwhelming even when the work behind them was careful and controlled.


I’d like to hear how you guys are handling this. How are you reviewing large agent assisted changes today?

I am so flattered right now 😂😂
I got asked by a realtor if I wanted to buy an apartment 😅


I wonder what made them think I could buy an apartment 😂😂

Someone need to join the channel the sub count is......😂😂😂😂

Thank you 'Fariha'😅😅

It was a pleasant day 😮‍💨🤌🏼

here is the game it is so simple just downlaod and have fun if you are some one with the memory 😅