On Mon, Apr 7, 2014 at 11:08 AM, Pat Ferrel <[email protected]> wrote:
> Right! Even more black box is needed. Mahout is not for scientists, it is > for app devs, some of which are trying desperately to learn the math. Not > people coming from R and yearning to do scalable app dev. > > I guess I've already said that the Mahout reboot should be first on Spark > and leave the other engines to work out their own integrations. > > On Apr 7, 2014, at 10:45 AM, Sebastian Schelter <[email protected]> wrote: > > A few questions here. > > @Dmitriy I very much share your vision. I just think that our targeted > userbase is more than people wanting to implement their own algorithms > using high level constructs in a modern language like Scala. > > I think there is still a huge demand for a blackbox of algorithms that > allows to easily build a model without having to know too much about the > underlying math. Our recommenders are a good example. Provide data in a > simple CSV format, throw ItemSimilarityJob on that and use the item > similarites for recommendations. > > Do you suggest we should leave the blackbox stuff to MLBase/Oryx and > solely focus on providing high level ML constructs? > No. I suggest nothing of the kind. I would still expect healthy stream of blackbox contributions to be important. The ML environment provided by generalist contributors (1) helps to attract blackbox contributions from specialist contributors by making it easier; (2) facilitates user devs to assemble and customize blackbox building blocks into application with ease (as we know, 90% of application effort is feature prep and vectorization effort, along with custom business rules and metrics -- and we just want to hand out the end user means to do it quickly and painlessly. Blackbox building bricks will never cover multitude of those entirely); (3) provides reliable differentiation of Mahout from other similar attempts. I am just saying report fails to mention it and leaves an impression we have no difference in vision from mllib, let alone MLI. This is simply not true, at least in my case. > @Sean How much can you agree on the vision I suggested? It meets your > demand of having a plan to solve the problems with the MR codebase (by > getting rid of it in the near future) and provides a direction for Spark as > the new underlying execution system, with optional support for Stratosphere > and H20, if those communities manage to convince us that it is worth to > integrate. > > --sebastian > > > > > 2014-04-07 19:29 GMT+02:00 Pat Ferrel <[email protected]>: > The document does not mention the state of the existing Spark work in the > snapshot codebase. Shouldn't this be noted? > > On Apr 7, 2014, at 5:06 AM, Sebastian Schelter <[email protected]> wrote: > > I think we should mention the redesign/rework of the website and the > completion of the move from the old wiki to Apache CMS. > > --sebastian > > On 04/07/2014 02:04 PM, Grant Ingersoll wrote: > > Here is my proposed report. For the most part, I think the only right > thing to do vis-a-vis the Board is to report that we are in the midst of a > healthy (yes, I believe it is, for the most part healthy and normal) > discussion on where to go next. > > > > PMC Members: this is checked into SVN at > https://svn.apache.org/repos/asf/mahout/pmc/board-reports/2014/board-report-apr.txt. > It is due on Wednesday. If you object to this approach of reporting, > please let me know ASAP and suggest alternatives. > > > > === Apache Mahout Status Report: April 2014 === > > > > ----- > > > > Apache Mahout has implementations of a wide range of machine learning and > > data mining algorithms: clustering, classification, collaborative > filtering > > and frequent pattern mining > > > > Project Status > > -------------- > > > > The project continues to have a large and active user base. While > > the developer base has continued to grow, there is a very active > > and healthy debate going on about where Mahout goes next. Please > > see the Issues section below for more details. > > > > Community > > --------- > > > > * Andrew Musselman was voted in as new committer. > > * No changes to the PMC in the reporting period. > > > > * The main issue concerning the community right now is the addition > > of new contributions from 0xData and the integration of Mahout with > Spark. > > > > Community Objectives > > -------------------- > > > > Our goal is to build scalable machine learning libraries. See the Issues > > section below for the debate in the community about our objectives. > > > > > > Releases > > -------- > > > > In addition to an ongoing debate on Mahout's future, the community is > actively > > working on integrating Mahout with Scala/Spark, updating > > documentation, and bringing in new code and committers to update the > core project. > > > > > > Issues > > ------ > > The Mahout community is at a crossroads in terms of where > > to go next. While the project has a broad number of users and interested > > parties, most committers are trying to maintain the code base on a purely > > part time basis, when the amount of work to sustain these users > > clearly points to it needing to > > be full time. Furthermore, much of our original code base is written > > for Hadoop MapReduce 1.0, which many in the community have come to > realize > > is not well-suited for solving the kinds of problems that Mahout has set > > out to solve. There have been several lengthy discussions and prototypes > > going on to work out next directions along the lines of the Spark and > > 0xData contributions (there are numerous threads on the [email protected] > > mailing list.) > > > > The PMC does not think this requires Board intervention at this time > > as the debate is, as far as we can tell, healthy. We do, however, > > expect that this debate will take some time to resolve and may mean we > > won't be shipping a 1.0 release any time soon. We will keep the Board > > apprised of our next steps as we work through the process. > > > > > > > > > > On Apr 7, 2014, at 4:53 AM, Grant Ingersoll <[email protected]> wrote: > > > >> To Sean's point, if Mahout were "my company", I would do the following, > albeit pragmatic and not so pleasant thing, assuming, of course, I had the > $$$ to do so: > >> > >> 1. Clean up existing code with a laser focus on a few key areas > (Sebastian's list makes sense) using a part of the team and call it 1.0 and > ship it, as it has a number of users and they deserve to not have the rug > pulled out from under them. > >> > >> 2. Spin out a subset of the team to explore and prototype 2.0 based on > two very positive and re-energizing looking ideas: > >> a. Scala DSL (and maybe Spark) > >> b. 0xData > >> > >> All of the work for #2 would be done in a clean repo and would > only bring in legacy code where it was truly beneficial (back compat. can > come later, if at all). > >> It would then benchmark those two approaches as well as look at > where they overlap and are mutually beneficial and then go forward with the > winner. > >> > >> 3. Once #2 is viable, put most effort into it and maintain 1.0 with as > minimal support as possible, encouraging, neh -- actively helping -- 1.0 > customers upgrade as quickly as possible. > >> > >> The tricky part then becomes how do you make sure to still make your > sales #'s while also convincing them that your roadmap is what they are > really buying. > >> > >> If I didn't have the $$$ to do both of these (i.e. we need a massive > turn around and we have one last shot), I would be all in on #2. > >> > >> ----------------------------------- > >> > >> That being said, Mahout is not "my company". Heck, Mahout is not even > a "company", so we don't need to be bound by company conventions and > thought processes, even if that fits with all of our individual day jobs. > And, thankfully, we don't have any sales numbers to make. > >> > >> We are chartered with one and only one mission: produce open source, > scalable machine learning libraries under the Apache license and community > driven principles. We are not required by the Board or anyone else to > support version X for Y years or to use Hadoop or Scala or Java. We are > also not required to implement any specific algorithms or deliver them on > specific time frames. We are also not required to provide users upgrade > paths or the like. Naturally, we _want_ to do these things for the sake of > the community, but let's be clear: it is not a requirement from the ASF. > We are, however, required, to have a sustaining community. > >> > >> ------------------------------------ > >> > >> I personally think we should start clean on #2, throwing off the > shackles of the past and emerge 6-9 months later with Mahout 2.0 (and yes, > call it that, not 0.1 as Sebastian suggests, for marketing reasons) built > on a completely new and fresh repository, likely bringing in only the > Math/collections underpinnings and maybe the build system. This new > repository would have only a handful of core algorithms that we know are > well implemented, sustainable and best in class. > >> > >> I think we should look at the lead up to 0.9 as an experiment that > proved out a lot of interesting ideas, including the fact that Mahout > proved there is vast interest in open source large scale machine learning > and that it is the benchmark for comparison. Not many other ML projects > can say that, even if they have better technical implementations or are > less fragmented. Once you realize something has outlived it's usefulness > in software, however, there is no point in lingering. > >> > >> That being said, at least for the foreseeable future, I am not in a > position to contribute much code. So, from my perspective, the ASF > Meritocratic approach takes over: those who do the work make the > decisions. If you want something in, then put up the patch and ask for > feedback. If no one provides feedback, assume lazy consensus and move > forward. Nothing convinces people better than actual, real, executing > code. For my part, I am happy to continue to work the bureaucratic side of > things to make sure reports get filed, credentials get created, etc. and > the occasional patch. I hope one day I will have time to contribute again. > >> > >> I will follow up w/ a separate email on what I am going to put in the > Board Report. > >> > >> On Apr 7, 2014, at 1:52 AM, Sean Owen <[email protected]> wrote: > >> > >>> No, it's about the opposite. I'm referring to the default, current > >>> state of play here. > >>> > >>> The issues for a vendor are demand and supportability. Do people want > >>> to pay for support of X? Can you honestly say you have expertise to > >>> support and influence X over at least a major release cycle (12-18 > >>> months)? The latter needs a reasonably reliable roadmap and > >>> continuity. > >>> > >>> I'm suggesting that in the current state, demand is low and going > >>> down. The current code base seems de facto deprecated/unsupported > >>> already, and possibly to be removed or dramatically changed into > >>> something as-yet unclear. Nobody here seems to have taken a hard > >>> decision regarding a next major release, but, the trajectory of that > >>> decision seems clear if the current state remains the same. > >>> > >>> From my perspective, "middle-ground" new directions like adding a bit > >>> of H2O, a bit of Spark, leaving bits of M/R code around, etc. are only > >>> worse. I can see why there may be a little renewed demand for the new > >>> bits, but then, why not go all in on one of them? > >>> > >>> Because a substantially all-new direction is a different story. If a > >>> "Mahout2O" or "Spahout" ("Mark"?) emerges as a plan, I could imagine a > >>> lot of renewed demand. And a clearer underlying roadmap sounds > >>> possible. It would remain to be seen, but there's nothing stopping > >>> those ideas from becoming part of a distro too. > >>> > >>> > >>> On Mon, Apr 7, 2014 at 6:22 AM, Ted Dunning <[email protected]> > wrote: > >>>> Please be explicit here. It sounds like you are saying that if > Mahout goes > >>>> in the proposed new direction that Cloudera will drop Mahout. > >>>> > >>>> Is that what you mean to say? > >> > >> > > > > -------------------------------------------- > > Grant Ingersoll | @gsingers > > http://www.lucidworks.com > > > > > > > > > > > > > > >
