Thank you very much to describe in detail. I totally understand your question now. There is no way to solve your question for now but it's really interesting issue. so Could you make a Jira issue for this with your experience? Thanks!
2017-07-05 10:16 GMT+09:00 小野圭二 <onoke...@gmail.com>: > Thx for your reply, Hyung. > > Thank you for introducing some docs, but i knew them already.And my point > was each notebook execute but hide/invisible the code. > The zeppelin auth system manage every notebooks at once. That feature > effect in the case of all user share same zeppelin system, like cloud > computing. > > Against it, how about in the case of just sharing notebook? > For example i set strict roles into my notebook, and share it with > someone who has their zeppelin. After they download and import my strict > auth notebook into their zeppelin, they can arrange my roles setting by > themselves. In this case, they can see my hided code on their zeppelin. > > I do not misunderstand about the zeppelin auth system, do i ? > > To prevent this issue, may need another authenticate/management feature on > Zeppelin, this is my point. :^) > > What for it, imagine such a case. > In the case of selling my big data analyze application through zeppelin > eco-system, i write and test my code on my zeppelin. Then export it and > upload to sharing site. > May someone see my sharing notebook on the site, then they download and > import it into their zeppelin, and execute it, > > Ok that is good, fantastic, viva Zeppelin, if i am a volunteer programmer. > > But i would like to get some revenue from my code if someone go down my > analyzing, hopefully. > In this case, i just want to provide the result of analyzing, not my > algorithm. > Even my client share my notebook on his zeppelin, my code is protected by > independent authentication, encryption and so on. > > This is my point. :-) > Sorry for getting longer. > > -Ono > > > 2017-07-04 23:35 GMT+09:00 Hyung Sung Shim <hss...@nflabs.com>: > >> Hello onokeiji. >> >> You can refer http://zeppelin.apache.org/docs/0.7.2/quickstart/explo >> rezeppelinui.html#note-toolbar if you want to hide code in the notebook. >> and http://zeppelin.apache.org/docs/0.7.2/security/notebook_ >> authorization.html#authorization-setting is relate to lock code your >> note. >> >> Hope this helps. >> >> >> 2017-07-04 22:05 GMT+09:00 小野圭二 <onoke...@gmail.com>: >> >>> Hey, how do you think about my idea of BLACKBOX-NOTEBOOK? >>> >>> -Ono. >>> >>> >>> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> >>> ウイルス >>> フリー｡ www.avast.com >>> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> >>> <#m_-8962903669315512691_m_3961825290808015369_m_-7798815867934407536_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> >>> >>> 2017-06-30 10:38 GMT+09:00 小野圭二 <onoke...@gmail.com>: >>> >>>> Hi, >>>> >>>> I wonder if this mail title would be realize. >>>> I mean, when I do not want to be changed my notebook programing code, >>>> but just give a way to change the query of programing by dynamic forms. >>>> >>>> 1. are there way to hide all code in the notebook? >>>> 2. are there rock system to prevent changing the code in notebook? >>>> >>>> If No.1 would be available, it makes easy to provide commercial codes >>>> and/or secret codes to widely. >>>> In fact, from the view of code writers, especially in the case of >>>> commercial products, they do not want to provide any source code to their >>>> customers definitely. Another hand, it is costly to force R or Python >>>> programer to write Java or Scala program to change their naked codes to >>>> binary code. >>>> >>>> I have a image Script programers write and test their >>>> R/Python/Spark-Shell/SQL.... codes on notebooks and be it to >>>> BLACKBOX-NOTEBOOK, then upload it to github/svn.... where endusers who want >>>> to use those notebooks could access by themselves. The endusers down load >>>> those BLACKBOX-NOTEBOOK and import into their Zeppelin and it works fine. >>>> Sound like happy the both. >>>> >>>> No.2 seems like easier than No.1, because the issue is depend on >>>> authorization. But it is incomplete, because the codes reveal to the >>>> endusers. >>>> >>>> I saw zepl.com as well. >>>> Seems like Zeppelin and ZELP are aimed to help collaboration between >>>> analysts. That is good. And if adding the above NO.1 feature into it, they >>>> will be the best, i think. >>>> >>>> Do not You think so? >>>> >>>> -Ono >>>> >>> >>> >> >
