2023. 3. 18. 07:09ㆍ미국박사유학
Put what you want to ask at the front.
Question I originally made: If you don't deploy multi-cluster services, what is the major reason?
Question B rephrased: What are the major reasons for your services that are not deployed multi-cluster? (You can choose more than one.)
The choices for one of the survey questions I made were too verbose.
Survey questions should be short, simple, straightforward, easy-to-read and clear.
Choices I created at the first place.
- Multi-cluster deployment looks complicated.
- We tried multi-cluster but failed.
- We deployed our service in multi-cluster but didn’t see any notable benefit, so no more multi-cluster (e.g., no performance improvement).
- We do not because we don’t see enough motivation to deploy multi-cluster. (why do you even need multi-cluster!?!)
- Other
Choices that were rephrased by B.
- Too complicated
- Little to no benefit for these services
- Not sure / have not evaluated multi-cluster deployments
- Other
My sentences were already short but he even made them evn more easy-to-understand and clearer.
Worded by me originally
- To place the cluster closer to the end user.
- To achieve better load distribution.
- To achieve isolation of workloads between clusters.
- To achieve better scalability. (It seems that scalability should be defined more clearly.)
- To comply some laws (e.g., GDPR).
- Other
Rephrased by B
- Reduce latency to end users
- Better load distribution
- Isolate workloads
- Scale out compute capacity
- Compliance or security policy
- Utilize multiple cloud providers
- Other.
Another small copyedit B commented.
- If you don't deploy multi-cluster service --> If you don't deploy any multi-cluster services
It was something hard to catch for me as a person who speaks English as a second language. B explicitly commented to add any. In other words, adding any or not is not trivial difference.
Negative expression to positive expression.
He suggested to change the sentence from negative expression to positive expression. The former sentence written with a negative expression was also written by B.
Little to no benefit for these services. -> Single-cluster deployment is sufficient
If you don't deploy multi-cluster services and don't [want -> plan] to deploy them in the future, you can stop the survey here and submit the response.
That makes sense because it is not a problem of you want it or not. It sounds like asking whether you like it or not.
'Plan' sounds more neutral than 'want'.
We moved one question to the last. It was in the middle of the questions having good flows, disrupting them. Moving one question made the picture more natural and organic.
Somtimes I tend to make explanation verbose. There is no clear line between verbosity(-) and detailed explanation. It depends on the situation. But I am inclined to talk too much, being doubtful if people are on the same page with me. Being concise is as equally important as being descriptive when it comes to conveying the message clearly.
He really cares about the details. It was not possible to catch all the details at the first round of reviews even for him. You write it and stare it for some time. Then you will start to see what you want to revise. The clever and efficient way of working aligned with the former fact is to write the first draft as early as possible. The first draft should not be and cannot be a good one. Just make something.
'미국박사유학' 카테고리의 다른 글
What is good presentation? (0) | 2023.04.29 |
---|---|
measurement가 high level design에 영향을 주는가? (UIUC CS phd and B) (1) | 2023.04.17 |
(PhD Journal, Feb 24th 2023) Drummer Josh: '이 사람은 이 연주를 여기서 하기까지 어떻게 흘러왔을까' (0) | 2023.03.05 |
논문 내면서 (해)먹었던 음식들 (PhD Journal, Feb 6th 2023) (0) | 2023.02.07 |
(PhD Journal) Feb 6th 2023. The first paper submission as a PhD student. (0) | 2023.02.07 |