Hi, I read the discussion and all related issues/pull requests, and give some of my comments:
It is appreciated for cqzhang's contribution to iotdb, especially for the detailed bug feedback. This is the best contribution to the open source community. Being a TLP open source project of Apache Software Foundation, keeping diversity is the community's goal. Therefore, attracting more and more contributors contributing to IoTDB is what we always want. I have to say comments like "this issue may not be suitable for a newer to our system due to communication costs of these complicated cases" is ABSOLUTLY incorrect. It is very very harmful for the community. "newer" does not mean he/she knows little than "older", and does not mean his/her solution is worse than "older". However, "meritocracy" still takes effect, because hearing a person whom has make some successful experience is usual correct. The only way we need to obey is: accepting better quality of codes, which we still need to improve. For example, I find cqzhang gave a solution like "using 12 months to replace 1 year", which is accpeted in the final PR. This is a kind of contribution. In the closed PR (#11171), the most important shortcoming is lack of UT and IT. In the merged PR(#11323), it also has drawbacks, e.g., "DateTimeUtils.convertDurationStrToLong()" function semantic is still incorrect if the input parameter value is "y". Considering making it easy to let more developers join us, I advocate that: 1. For each public issue in Github/JIRA, claim you are working on it to let all others knowing that. If some other person think he/she want to take over, please also claims it asap, rather than submit PR directly. 2. Even though, the cases that two PRs solve the same issue may still occur. In this case, "better code wins" and meritocracy are still the best way. Best, ----------------------------------- Xiangdong Huang School of Software, Tsinghua University Weihao Li <18110526...@163.com> 于2023年10月31日周二 15:11写道： > > Hi cqzhang7, > As you can see, `not supporting group by nature year` and `not supporting > mixed units in group by` are two separate issues. > 1. For the first issue, we have discussed in > https://github.com/apache/iotdb/pull/11309, and the final PR was appended on > the discuss. > 2. For the second issue, your general idea is right, but there are still too > many other cases need to be considered, like us and ns time precision. After > discuss, we think this issue may not be suitable for a newer to our system > due to communication costs of these complicated cases. You can see the final > PR about this issue https://github.com/apache/iotdb/pull/11429. > Thanks for your issues and perfect fix ideas about them, maybe we can start > from some easy issues. Welcome to continue to contribute to IoTDB. > > > > At 2023-10-31 10:50:50, "张" <m18392456...@163.com> wrote: > > Dear all, > I am writing to share my experiences and concerns regarding my contributions > to IoTDB. I appreciate your time in reading this and hope that my feedback > will be taken constructively. > > > Here is my experience when submitting pr to apache IoTDB: > On 2023/09/13, I used iotdb at work and found a little problem with group by > year, so I submitted issues https://github.com/apache/iotdb/issues/11133 > On 2023/09/18, After this I submitted a useless PR > https://github.com/apache/iotdb/pull/11171 without reading the source code or > testing it myself, it was really the first time I submitted a PR for an open > source component, but fortunately @HTHou viewed the PR and still replied to > me, so I decided to read the source code and fix the issues. I realized that > iotdb currently does not support the "1mo1d" scenario based on group by > nature month. So I decided to try to finish this part as well. > > > > > > > > On 2023/10/11, After reading the source code, I found new bugs > https://github.com/apache/iotdb/issues/11286, I was motivated to fix them and > work on them over and over again, trying to do my best to make it better, and > then I submitted a new PR https://github.com/apache/iotdb/pull/11290, > expecting anyone to give better suggestions. However, I found that IoTDB > doesn't seem to welcome PRs from the outside, because I have been paying > attention to this matter in the past few days, and for internal PRs, It's > always handled in a timely manner, while for PRs from the outside, it seems > to be another kind of attitude. After I asked to the relevant people in the > community WeChat group, someone did review it on the same day, but after I > replied to them one by one, there was no more responses. > It's now October 31st, 2023, and another two weeks have passed. This > discrepancy in response time and engagement deeply concerns me. I fail to > understand why there is such a distinction between internal and external > contributions. I believe that this disparity could have a detrimental impact > on the promotion and adoption of IoTDB. People who face similar experiences > might have a hard time recommending IoTDB to his friends. > > > I kindly request that you address this issue and ensure that external > contributors receive fair and timely feedback. A more inclusive and > responsive approach will not only enhance the reputation of IoTDB but also > encourage more active participation from the open-source community. > > > Thank you for your attention to this matter. I look forward to your response > and hope for a positive resolution. > > > Best regards
