Hey just to add to my previous email: having the python interface is a top priority IMHO. This will enable us to support data scientists :)
On Sat, 15 Jan 2022 at 10:39 AM jorge Arnulfo Quiané Ruiz < [email protected]> wrote: > Hi all, > > Thanks Bertty for the nice summary. I fully agree with Zoi and Alex, SQL > interface and nice documentation are the top priorities yet the test > coverage is also an important aspect we must address. > > — Jorge > > On Thu, 13 Jan 2022 at 11:05 AM Alexander Alten <[email protected]> wrote: > >> Hi, >> >> +1 for Zoi’s comment. First we need to lower the bar for entering the >> wayang ecosystem. >> >> Cheers, >> —Alex >> >> >> — >> Alexander Alten >> PPMC Apache Wayang >> [email protected] >> >> @apachewayang >> >> > On 13. Jan 2022, at 11:00, Zoi Kaoudi <[email protected]> wrote: >> > >> > Hi all, >> > I would definetely prioritize finishing the Python and/or SQL API and >> the improvement of the documentation/installation as imo these two are the >> main obstacles for people to really start using Wayang. >> > Best >> > -- >> > Zoi >> > >> > Στις Πέμπτη, 13 Ιανουαρίου 2022, 12:15:31 π.μ. CET, ο χρήστης Bertty >> Contreras <[email protected]> έγραψε: >> > >> > Hi folks, >> > >> > I hope you are safe and had lovely "parties" during the last week of >> > December. >> > >> > 2021 was a pleasant year for Apache Wayang(Incubating); we first >> released. >> > We grow the numbers of "stars" in Github. >> > We get new people contributing. Obviously, we want the same for 2022; >> many >> > of you contacted me to chat about the future of Wayang; I collected your >> > ideas, and with all of them, we created the following roadmap. >> > >> > In the chat, most of you pointed out that Wayang needs refactoring to >> > support different elements in plug and play mode. The main reason for >> this >> > point is to smoothly incorporate the new optimizers, new "internal >> > workflows" that will help support Stream processing. During the coming >> > weeks, we will share with you the design of the modifications where we >> can >> > have a long discussion on where to start over this vital feature. >> > >> > It is not news that we have some issues with Travis in long waiting >> times >> > to run the compilation and validation of the code. However, during the >> 2021 >> > @CalvinKirs helps set up the first Github action to validate the >> licenses >> > on the files, creating a new opportunity that we want to explore that >> > incorporates the Github Actions in the daily workflow Wayang. >> Nevertheless, >> > we want also to contribute to the apache community with the creation of >> > automatization that could help other projects, for example, when you >> > generate a 'TODO' in the code, this automatically creates the issue in >> > Github and/or Jira to help with the visibility of the missing >> components. >> > >> > On the other hand, we need to start working on the weak points of >> Wayang. >> > The current missing elements that Wayang has are the documentation and >> the >> > coverage of the unitary tests. For the documentation, the first plan is >> to >> > bring the architecture and other descriptions that we have published in >> > different papers and create from there the documentation. Still, in this >> > opportunity, we need to extend it as much as possible and make >> > straightforward explanations, with good examples and precious insights >> for >> > the new contributors and the adopters of the project. >> > >> > As the second part of the documentation, we will need to build several >> > scripts that can help the user of Wayang make their installation and >> > execution; the user just needs to care about the platform's usage. >> > >> > The unitary test has a lack in terms of coverage. One way of solving >> this >> > problem could be to create a significant number of 'TODOs' explaining >> the >> > test and the characteristics they need to increase the coverage. The new >> > tests also help develop new functionalities to ensure that we are not >> > broken code in future modifications. >> > >> > To solve the previous point, we will publish almost daily many issues >> > related to it during the coming weeks, and we will request to hold some >> > small-time windows of your time and work on it. So if each of us >> dedicates >> > a short time to it, it will help narrow the gap that Wayang has. >> > >> > Suppose you arrive at this point asking where are the new plans and new >> > features for Wayang. Of course, many of you are currently working in the >> > Python API, SQL API and EMMA API; they are so important because they >> help >> > in the inclusion of new communities and make the environment of Wayang >> more >> > robust. However, one of the features that can enrich the user >> experiences >> > will be running Wayang in a server mode, not just like a program of one >> > run. Nevertheless, if you have an idea for a feature, please feel free >> to >> > write an issue with the features, then we can discuss it in the thread. >> > >> > One last idea is to have Github issues and Jira as a mirror. This idea >> > comes from the assumption that many developers are looking at the code; >> > they are focusing on the Github repository; having the code and issues >> on >> > the same platform may help with visibility and motivate someone. Also, >> > having everything in the same place will be nice for some of you, as you >> > commented to me. >> > >> > I hope this email was not boring for you, but there are many ideas and >> > plans for Apache Wayang(Incubating). It wasn't easy to make it shorter, >> but >> > if you have any comments or ideas, just reply to this email to discuss >> this >> > topic that is relevant for the future of Wayang. >> > >> > Best regards, >> > Bertty >> >>
