I add a few lines regarding to the coding style currently I discover so far, but other convention may require more discussion if needed. Please correct me and help me add new one for that section as appropriate.
Thanks. 2011/1/24 Edward J. Yoon <[email protected]>: > Yup, pls feel free to open new wiki page on http://wiki.apache.org/hama > > Let's clarify our coding standards. > > 2011/1/24 Chia-Hung Lin <[email protected]>: >> Is there any chance that this information can be described in HAMA >> developer page? I went through the website and checked Apache >> Contributor Documentation, but seems there is no mention about such >> convention (maybe I miss it somewhere else). So I make use of coding >> style referred to at >> http://www.kernel.org/doc/Documentation/CodingStyle for local >> variables. >> >> I will prevent this next time. >> >> Thanks for the advice. >> >> 2011/1/21 Edward J. Yoon <[email protected]>: >>> Great! >>> >>> And here's my one comment for you about instance/reference variables >>> naming. I think, they should say at least something about the object. >>> For example, >>> >>>>> GroomServerStatus gss; >>> >>> The "groomServerStatus" is much better and more readable than "gss" in >>> my opinion. :) >>> >>> 2011/1/21 Chia-Hung Lin <[email protected]>: >>>> I've updated HAMA 346 with a patch that should fix the null value >>>> returned by whichJob(). >>>> >>>> 2011/1/21 Edward J. Yoon <[email protected]>: >>>>> Thanks! >>>>> >>>>> 나의 iPhone에서 보냄 >>>>> >>>>> 2011. 1. 20. 오후 11:06 Chia-Hung Lin <[email protected]> 작성: >>>>> >>>>>> I check out source from svn with the revision r1058109 and this issue >>>>>> also happens on my testing machines. >>>>>> >>>>>> It looks like the completion of task status is not consistent between >>>>>> GroomServers and BSPMaster. For insatnce, GroomServers report Tasks >>>>>> but JobInProgress consider tasks already been completed. But the >>>>>> detail may require more examine I will fix this as soon as possible. >>>>>> >>>>>> 2011/1/19 Edward J. Yoon <[email protected]>: >>>>>>> Today, I noticed that there was a bug in the patch of HAMA-346. >>>>>>> >>>>>>> After first job execution completed, the BSPMaster.whichJob() returns >>>>>>> null. So the BSP system always throws error message "Fail to >>>>>>> communicate with BSPMaster for reporting.". >>>>>>> >>>>>>> Simply, I'll add this code : "if (jip != null) { }". >>>>>>> >>>>>>> What do you think? >>>>>>> >>>>>>> -- >>>>>>> Best Regards, Edward J. Yoon >>>>>>> http://blog.udanax.org >>>>>>> http://twitter.com/eddieyoon >>>>>>> >>>>> >>>> >>> >>> >>> >>> -- >>> Best Regards, Edward J. Yoon >>> http://blog.udanax.org >>> http://twitter.com/eddieyoon >>> >> > > > > -- > Best Regards, Edward J. Yoon > http://blog.udanax.org > http://twitter.com/eddieyoon >
