imchuncai

要问就问，不问拉倒，婆婆妈妈，语无伦次

@inostarling 第一，我没看怎么知道不想看；第二，就算有套路，都套路完了，别说假话了，谁还跟你说话呢

你问他啊，在这问个什么劲，莫名其妙

菜就多练

只友链 linux.do ，那不成跪着要饭的了吗？

没有任何调查，张口就来，你眼里的专家级别在我这只是玩具级别

不知所云

自己都学不会，给别人做 app

现在变成我的软广了

Hi Imchuncai,

I'm Ivan, founder of XERJ, an open-source, AI-native search engine: it give=
s coding agents the exact passage of code or documentation they need instea=
d of letting them guess from memory. On top of it we run a public Corpus Hu=
b: an open index of great code solutions that agents consult before writing=
their own version, so they repeat what worked instead of repeating mistake=
s.

We reviewed your "Add cache protol STATS" in imchuncai/umem-cache and selec=
ted it for the hub. It is a substantial piece of work: 7 files and about 17=
0 lines across the project root, in C.

Once it is in the hub, a developer whose coding agent hits the same kind of=
problem gets your implementation back as a worked reference, with your nam=
e and the repo attached, so the agent builds on code that already works ins=
tead of improvising its own version and repeating the mistakes you already =
solved.

If you don't mind, we will include your work in the public index, under the=
repo's licence (GPL-2.0) and with attribution to you and to imchuncai/umem=
-cache. If you would rather keep it out, one line back is enough and I will=
leave it out.

And if you use AI coding agents yourself, XERJ is worth a try: it cuts the =
tokens an agent burns by retrieving the exact passage instead of guessing, =
from reference coding against real implementations to search across large d=
ocument sets. It is free and open source: https://github.com/xerj-org/xerj

Ivan
Founder, XERJ
https://github.com/xerj-org/xerj
--HjDVEMUL4QPBIOtifdmG5mNz66gs6JHfb11K9tfo
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable