Slack & Chat

❌ "Hey, quick question. I was working on the auth feature
and noticed something interesting. I've been thinking about
it and I'm not sure what to do. What do you think?"

✅ "The auth token expires after 5 minutes but the docs say
30 minutes. Should I update the code or the docs?"

❌ "It's not working"

✅ "The login endpoint returns 401 after the recent auth
changes. I tested with a valid token. Error log: [link]"

High: Production down, security issue, blocking team
Medium: Code review needed today, clarification needed
Low: General questions, improvements, nice-to-haves

"[Context] I'm implementing OAuth for the mobile app.

[Problem] The refresh token flow fails with error 400.

[What I tried]
- Checked token isn't expired (it's not)
- Verified client_id matches
- Tested with Postman (works there)

[Question] Are there known issues with mobile OAuth?
Any suggestions on what else to check?

[Resources]
- Error log: [link]
- Code: [GitHub PR link]"

❌ "Anyone know OAuth?"
❌ "This is broken"
❌ "Can someone help?"
❌ "Quick question..." (then disappear)

✅ "I'm getting a race condition in our Redis cache.

Reproducible example:
1. User logs in (writes to cache)
2. User immediately updates profile (reads cache)
3. Profile update uses stale data

Expected: Fresh data from database
Actual: Stale cache data

Code snippet: [gist link]

Should we add cache invalidation here or is there a
better pattern?"

"🐛 Bug: Users can't upload files > 5MB

Environment: Production
Impact: ~50 users affected
Priority: Medium (workaround: use FTP)

Steps to reproduce:
1. Go to /upload
2. Select file > 5MB
3. Click Upload
4. See error: 'File too large'

Expected: Files up to 10MB should work (per docs)
Actual: Fails at 5MB

Logs: [link]
Video: [recording]

I checked nginx config - max size is 10MB there.
Might be in our Node middleware?"

"📝 Ready for review: Add Redis caching to user API

PR: [link]
Priority: Low (feature, not blocking)
Context: Users were seeing slow load times (2-3s).
This adds caching to reduce DB queries.

Changes:
- Added Redis client
- Cached user queries (15min TTL)
- Added cache invalidation on user update
- Tests for cache hit/miss

Performance: Load time reduced from 2.5s to 200ms

Please check:
- Is 15min TTL reasonable?
- Should we cache more endpoints?
- Any security concerns?

Take your time - I'm moving on to another task."

"📅 Status Update - Jan 15

✅ Yesterday:
- Finished auth refactor (#1234)
- Reviewed 3 PRs
- Fixed bug in email service

🔄 Today:
- Start on payment integration
- Review design doc for new feature
- Team retro at 2pm

🚫 Blockers:
- Need API keys for Stripe (pinged @alice)"

❌ "That's a terrible idea"
❌ "We can't do that"

✅ "I see the value, but I'm concerned about X. Have we
considered Y instead?"

✅ "I don't think I can take this on right now. I'm focused
on Z which is higher priority. Could someone else help, or
can this wait until next week?"

✅ "This would require refactoring the entire auth system.
That's ~2 weeks of work. Is this worth delaying the Q1
roadmap?"

❌ "You're wrong"
❌ "That won't work"

✅ "I have a different perspective. I think X because Y.
What am I missing?"

✅ "I see your point about A. My concern is B. How do we
balance both?"

✅ "From my experience, X tends to cause Y. Have you seen
that differently?"

✅ Reply in thread to keep discussion organised
✅ Summarise conclusion in main channel
✅ Use "Also send to #channel" for important updates
✅ Keep thread focused on one topic

❌ Start multiple side conversations in one thread
❌ Lose important decisions in threads
❌ Expect everyone to read all threads

"🐛 Production error: Users can't log in
Thread for debugging ⬇️"

Person A: "Seeing 'invalid token' errors in logs"
Person B: "I see 401s in Datadog"
Person A: "Checking Redis... tokens are there"
Person C: "Is Redis in different region?"
Person A: "Yes! That's it. Latency causing timeouts"
Person C: "Deploying fix now"

[10 more messages of debugging]

Person A in main channel:
"✅ Fixed! Was Redis cross-region latency. Moved to same
region. Details in thread."

❌ "What?"
❌ "I don't understand"
❌ Ignore the message

✅ "Just to clarify: are you asking about X or Y?"
✅ "Can you give an example? I want to make sure I understand"
✅ "I'm not sure I follow. Do you mean [your interpretation]?"

❌ Ignore
❌ Guess incorrectly
❌ Feel bad

✅ "I don't know, but @alice might know"
✅ "Not sure, but this doc might help: [link]"
✅ "I'm not familiar with that area. Might be worth asking
in #backend"
✅ "I can look into this tomorrow if no one else responds"

✅ "In a meeting now, will respond in an hour"
✅ "Deep in debugging. Can I get back to you in 2 hours?"
✅ "I'm heads-down on a deadline. Could you ask in #general?
Someone else might be able to help faster"

First ping: "Hey, quick question..."
You: [Busy, don't respond immediately]

Second ping (5 minutes later): "Did you see my message?"

Your response:
✅ "I saw it - I'm in a meeting. I'll respond in 30 minutes"

Not:
❌ [Ignore again]
❌ [Respond defensively]

DM them:
"Hey, I noticed you often need quick responses. I usually
check Slack every 2 hours. For urgent things, could you
call me or mark as urgent? Otherwise I'll respond when I
can. Does that work?"

@alice: Direct question to Alice
@engineering: Relevant to engineering team (use sparingly)
@here: People currently online (urgent only)
@channel: EVERYONE including offline (very rare)

✅ "Question about auth: @alice do you know if we rate-limit
by IP or user?"

✅ "@here Production is down, investigating"

❌ "@channel Anyone know how to use Git?" (use docs or DM)
❌ "@everyone Happy Friday!" (unnecessary noise)

Without emoji:
"I'll take a look" (neutral, could be annoyed?)

With emoji:
"I'll take a look! 👍" (friendly, helpful)
"I'll take a look 😊" (definitely friendly)

❌ "Hey! 👋 I was just wondering 🤔 if maybe 🤷 you could
possibly 🙏 help me 😅 with this thing? 💭"

✅ "Could you help me with this? 🙏"

Can sound cold:
"That works"

Warmer:
"That works! Thanks for figuring that out 👍"

Warmer still:
"Perfect! That's exactly what we needed. Thanks for
figuring that out 🎉"

Too direct:
"Review this"

Better:
"Could you review this when you have a chance?"

Better:
"Would you mind reviewing this when you have time? No rush!"

Harsh:
"This won't work. We need to use X instead"

Better:
"Good start! One concern: performance might be an issue with
large datasets. Have we considered using X instead?"

Someone writes: "This code has problems"

Don't assume: "They think I'm a bad developer"
More likely: "They want to help improve the code"

If unsure:
"Just want to make sure I understand your tone - are you
saying we need to change this urgently, or is it more of a
suggestion for future improvement?"

❌ "Let's meet at 3pm"
✅ "Let's meet at 3pm EST / 12pm PST / 8pm UTC"

Use:
- Specific time zones
- Tools like World Time Buddy
- Calendar invites (auto-converts)

In your profile:
"🇧🇷 Brazil (UTC-3)
Working hours: 9am-5pm BRT (8am-4pm EST)"

In status:
"Out of office - back tomorrow at 9am BRT"

❌ "Can you look at that thing?"

✅ "Can you review the authentication PR when you have time?
Link: [url]. Main concern is the token expiry logic. No
rush - I'm working on another feature."

❌ "Should I use Redis or Postgres for caching?" 
[Waits 2 hours for response, does nothing]

✅ "Should I use Redis or Postgres for caching? I'm leaning
toward Redis for performance. I'll start with that and we
can adjust if you have concerns."

Morning:
"Starting on payment integration"

Evening:
"Payment integration is 50% done. Stripe is integrated,
working on error handling tomorrow. On track for Friday."

"/review-ready"
"Ready for review: [PR link]
Context: [brief description]
Please check: [specific concerns]
Priority: [low/medium/high]"

"/standup"
"Yesterday: [work]
Today: [plans]
Blockers: [blockers or 'none']"

"/not-urgent"
"Thanks for reaching out! I'm focused on a deadline right
now. I'll get back to you in [timeframe]. If it's urgent,
please mark as such."

Instead of: Responding every 5 minutes all day

Try:
- Check at 10am, respond to all
- Deep work 10am-12pm (DND)
- Check at 12pm, respond
- Lunch
- Check at 2pm, respond
- Deep work 2pm-4pm (DND)
- Check at 4pm, respond
- Check before end of day

from:@alice authentication
in:#backend redis
after:2024-01-01 deployment
has:link architecture

Not just:
"The code is done"

Also:
"The code is done! Now to enjoy my weekend 😊 Any plans?"

"🎉 Congrats on shipping the feature @alice!"
"Great debugging @bob, that was a tricky one!"
"Thanks for the quick review @carol, really helpful!"

In #random:
"What's everyone working on this week?"
"Anyone else watching [show]?"
"TIL about [interesting thing]"

"Trying to debug while my cat walks across the keyboard 😅"
"First day back from vacation, catching up on 200 messages"
"Just got my desk setup upgraded! [photo]"

❌ "Just saying hi" (then nothing else)
❌ Passive-aggressive tone
❌ Vague complaints without solutions
❌ Arguing in public channels
❌ Ignoring messages completely
❌ Never using emoji (might seem cold)
❌ Overusing emoji (unprofessional)
❌ All CAPS (shouting)
❌ no punctuation or capitalization (sloppy)
❌ Writing essays when 2 lines would do
❌ @channel for non-urgent things
❌ Sending 10 separate messages instead of one

✅ Be responsive (even if just "I'll look at this later")
✅ Be concise but friendly
✅ Give context
✅ Make it easy to help you
✅ Say thank you
✅ Admit when you're wrong
✅ Celebrate others
✅ Ask questions
✅ Share knowledge
✅ Be patient with async

❌ 
Person A: "Hi"
Person B: "Hi!"
Person A: "How are you?"
Person B: "Good, you?"
Person A: "Good! Quick question about the API..."

✅
Person A: "Hi! Quick question about the API: [actual question]"

❌ "The thing isn't working"

✅ "The OAuth login endpoint is returning 401. I checked the
token and it's valid. Could it be a CORS issue? Error log:
[link]"

❌ "Can you review this? [link]"

✅ "Can you review this Redis caching PR? It should fix the
slow queries we discussed yesterday. Main thing to check is
if 15-min TTL seems right. Link: [url]"

❌ "So I was working on the feature and I thought about using
approach A but then I considered B and actually there's also
C which might work and I read this article about D and I'm
not sure if we should use A or B or maybe C but D seems
interesting though it's complex and I tried implementing A
but ran into issues so then I tried B and that also had
problems and..."

✅ "I'm implementing the cache feature. Option A (Redis) is
faster but costs more. Option B (in-memory) is cheaper but
less reliable. I'm leaning toward A. Thoughts?"