Story and lessons from deciding not to submit to ASPLOS.

 

I am noting some keywords quickly before forgetting what I want to reflect.

 

It is not simply recording what happened but certainly it has some lessons, worth not forgetting.

 

Disclaimer: It is not exactly accurate and does not contain all context. I relied on my memory. And it is not for criticizing.

"I don't think we have finished our homework. It is not a complete paper and we know what's missing and we haven't tried enough to show them. ASPLOS+Sigcomm which will be one weak shot(ASPLOS) and one strong shot(Sigcomm) v.s. OSDI + NSDI two strong shots. Strategically, two strong shots seem better.

Also, reviewers might know who's paper it is. It is good to submit when it it complete considering reputation/bias.

We haven't shown why we need jumping strategy at the first place. We haven't answered questions like "why not focusing on updating the model(data) more more efficiently and quickly?", "Why do you want to solve the problem with jumping? Is it superior? then why? It is not clear. also show me the proof."

This is really good story. I haven't seen something like this before. We can make strong submission.

I don't think returning point has been reached. (meaning gain is bigger than loss when we delay the submission and improve more. loss: missing this submission opportunity. gain: chance to improve)

OSDI is high profile."

 

I think what he mentioned makes sense. And we actually decided not to submit.  However, he should have let us know about it much earlier. Clearly, he has a certain bar. I am not trying to say whether the bar is fair or unfair. As a important collaborator especially because he is my advisor, it is good know what he expects. It is usually something you as a student will cautiously think about. However, in my case my advisor does not really tell me what expectation he has and that makes me forgetful about it and possibly making me naive and the paper quality low which shouldn't happen. Since he will not tell me, I learned that I should be cautious about the bar by myself and sometimes try to read his mind? sounds kinda weird but the point is I shouldn't be naive.