4 ms·
To make decisions quickly, I totally agree. If you want to understand the why, it is essential to have it in writing. If the problem is complex, then one guy
by qznc 7y ago
To make decisions quickly, I totally agree.
If you want to understand the why, it is essential to have it in writing.
If the problem is complex, then one guy writing for 30 minutes can provide more clarity than 5 discussing for an hour.
- growse 7y agoEmail is a bad tool for discussion. However, it can be a good tool for information dissemination. (There are better tools - I'd argue that sending information should be pull rather than push, but that's just my preference). So yes, have a guy write something down for 30 minutes, get people to read it, and then have an actual conversation / chat about it.
- lukethomas 7y agoWhen collaborating over ambiguous topics where the potential to be misunderstood is high, you need to use the "richest" communication channel possible. Body language, facial expression, and the fast feedback loop can help you quickly establish a sense of common ground. On the other hand, status updates or sharing of metrics/information can easily be done asynchronously, because the potential to be misunderstood is low. It's also a more scalable approach and persists over time.
- jrandm 7y ago> When collaborating over ambiguous topics where the potential to be misunderstood is high, you need to use the "richest" communication channel possible. Body language, facial expression I would dispute that body language and facial expression are somehow always correctly interpreted or unambiguous, like most forms of communication. Written word, or a series of characters in a string, as I think we in a technical forum with many programmers in particular can appreciate, can in fact remove quite a bit of ambiguity if everyone is careful about what they type. Github has many examples of this in-progress.
- msla 7y agoEmail is a bad tool for discussion in that it lasts too long, and too many people can save copies; it's far better to have ephemeral and limited-distribution discussions, so if something obnoxious or actionable is said, nobody is opened up to legal actions. The modern trend of people believing accusers is pushing against this wisdom, but it's still far better to not have evidence. That said, I'm sure Slack can be persuaded to destroy incriminating logs if the right pressure is applied. One of the benefits of people moving away from self-hosted solutions.
- closeparen 7y agoEvery enterprise-friendly messaging system offers configurable retention policies. Messages involving certain roles or keywords may need to be retained for years to comply with regulations, while others can be deleted after a few days to minimize risk.
- qznc 7y agoI completely agree with you. By writing I was thinking about RFCs not emails. Have a face-to-face discussion and have a shared written document. If I have to chose between the two, the I would favor the written document though.
- mlindner 7y agoI completely disagree. If the problem is complex you need to have an in person discussion on a white board rather than writing a long winded email that will mostly be filler information that the reader already knows. It probably won't even answer the issue the reader wanted to know either. And now you've wasted 30 minutes.
- jrandm 7y agoHave you ever learned about something complex from a book? That would imply you don't need a whiteboard and 10 minute meeting in order to transfer information to one another.
- closeparen 7y agoInteresting question. I've tried to self-study from books alone, and it's never worked. The book always seems to work best in concert with the professor or TA's whiteboarding.
- qznc 7y agoWasn't thinking about email but more about RFCs. I have been in such whiteboard discussions. Then I had the same discussion again one week later because another one needed the information. And again. And again. I have heard from such whiteboard decisions. Then I heard a very different conclusion from the same whiteboard decision. And another. And another.