Re: What is the legal basis for enforcing release policies at ASF?
On Thu, Aug 20, 2015 at 8:52 AM, Jim Jagielski j...@jagunet.com wrote: A snapshot is not a release. Licenses kick in at distribution/ release. Lets just imagine if Jim, VP Legal is actually correct in his interpretation, and that there are no AL 2.0 licenses applicable to our source code repositories, svn or git. Quoting http://apache.org/licenses/LICENSE-2.0 ... 2. Grant of Copyright License. Subject to the terms and conditions of this License, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable copyright license to reproduce, prepare Derivative Works of, publicly display, publicly perform, sublicense, and distribute the Work and such Derivative Works in Source or Object form. No, you may not modify the sources or derive those that reside within version control of the ASF, until and upon the time when the project has blessed that project as a release. Patches to others' contributions to source code control are not within the scope of this imaginary non-license application. 3. Grant of Patent License. Subject to the terms and conditions of this License, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable (except as stated in this section) patent license to make, have made, use, offer to sell, sell, import, and otherwise transfer the Work, where such license applies only to those patent claims licensable by such Contributor that are necessarily infringed by their Contribution(s) alone or by combination of their Contribution(s) with the Work to which such Contribution(s) was submitted. If You institute patent litigation against any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed. No, you may absolutely not test the code that has been committed to source control without a patent license, which you do not have, until that time when the ASF blesses the work and calls it a release. 4. Redistribution. You may reproduce and distribute copies of the Work or Derivative Works thereof in any medium, with or without modifications, and in Source or Object form None of that, it's all straight out, none of it applies to your work at the ASF until the release is blessed. That includes passing off a patched fork of a security fix to a reporter who claimed there was a defect in the earlier release. 5. Submission of Contributions. Unless You explicitly state otherwise, any Contribution intentionally submitted for inclusion in the Work by You to the Licensor shall be under the terms and conditions of this License Except when it isn't in Ross's and our VP Legal's own minds... 6. Trademarks. This License does not grant permission to use the trade names, trademarks, service marks, or product names of the Licensor, except as required for reasonable and customary use in describing the origin of the Work and reproducing the content of the NOTICE file. Which wasn't a right in the first place, so no change here under any interpretation... 7. Disclaimer of Warranty. Unless required by applicable law or agreed to in writing, Licensor provides the Work (and each Contributor provides its Contributions) on an AS IS BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied, including, without limitation, any warranties or conditions of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A PARTICULAR PURPOSE. You are solely responsible for determining the appropriateness of using or redistributing the Work and assume any risks associated with Your exercise of permissions under this License. Except that perhaps the ASF is liable, under our VP Legal's interpretation, for works which do reside in source control and were not, in fact, released to the general public? [Ad nauseam 8. and 9.] Let's just not go this direction, because it is plainly false. Jim, it would truly be helpful if you spoke up for or in contradiction to your earlier statements, here... Cheers, Bill
[DISCUSS] HAWQ Incubation Proposal
Hi! I would like to start a discussion on accepting HAWQ into ASF Incubator. The proposal is available at: https://wiki.apache.org/incubator/ApexProposal and is also attached to the end of this email. Please note, that this proposal is very complementary to the desire of HAWQ's sister project (MADlib) to join ASF Incubator: http://madlib.net/pipermail/user/2015-August/ http://madlib.net/pipermail/devel/2015-August/ I've volunteered to help MADlib community and we're currently working on a separate proposal to be submitted later next week. If you're interested in monitoring progress of that please see updates to: https://github.com/madlib/madlib/wiki/MADlib-ASF-Incubator-Proposal and later: https://wiki.apache.org/incubator/MADlibProposal Thanks in advance for your time and help. Thanks, Roman. == Abstract == HAWQ is an advanced enterprise SQL on Hadoop analytic engine built around a robust and high-performance massively-parallel processing (MPP) SQL framework evolved from Pivotal Greenplum DatabaseⓇ. HAWQ runs natively on Apache HadoopⓇ clusters by tightly integrating with HDFS and YARN. HAWQ supports multiple Hadoop file formats such as Apache Parquet, native HDFS, and Apache Avro. HAWQ is configured and managed as a Hadoop service in Apache Ambari. HAWQ is 100% ANSI SQL compliant (supporting ANSI SQL-92, SQL-99, and SQL-2003, plus OLAP extensions) and supports open database connectivity (ODBC) and Java database connectivity (JDBC), as well. Most business intelligence, data analysis and data visualization tools work with HAWQ out of the box without the need for specialized drivers. A unique aspect of HAWQ is its integration of statistical and machine learning capabilities that can be natively invoked from SQL or (in the context of PL/Python, PL/Java or PL/R) in massively parallel modes and applied to large data sets across a Hadoop cluster. These capabilities are provided through MADlib – an existing open source, parallel machine-learning library. Given the close ties between the two development communities, the MADlib community has expressed interest in joining HAWQ on its journey into the ASF Incubator and will be submitting a separate, concurrent proposal. HAWQ will provide more robust and higher performing options for Hadoop environments that demand best-in-class data analytics for business critical purposes. HAWQ is implemented in C and C++. == Proposal == The goal of this proposal is to bring the core of Pivotal Software, Inc.’s (Pivotal) Pivotal HAWQⓇ codebase into the Apache Software Foundation (ASF) in order to build a vibrant, diverse and self-governed open source community around the technology. Pivotal has agreed to transfer the brand name HAWQ to Apache Software Foundation and will stop using HAWQ to refer to this software if the project gets accepted into the ASF Incubator under the name of Apache HAWQ (incubating). Pivotal will continue to market and sell an analytic engine product that includes Apache HAWQ (incubating). While HAWQ is our primary choice for a name of the project, in anticipation of any potential issues with PODLINGNAMESEARCH we have come up with two alternative names: (1) Hornet; or (2) Grove. Pivotal is submitting this proposal to donate the HAWQ source code and associated artifacts (documentation, web site content, wiki, etc.) to the Apache Software Foundation Incubator under the Apache License, Version 2.0 and is asking Incubator PMC to establish an open source community. == Background == While the ecosystem of open source SQL-on-Hadoop solutions is fairly developed by now, HAWQ has several unique features that will set it apart from existing ASF and non-ASF projects. HAWQ made its debut in 2013 as a closed source product leveraging a decade's worth of product development effort invested in Greenplum DatabaseⓇ. Since then HAWQ has rapidly gained a solid customer base and became available on non-Pivotal distributions of Hadoop. In 2015 HAWQ still leverages the rock solid foundation of Greenplum Database, while at the same time embracing elasticity and resource management native to Hadoop applications. This allows HAWQ to provide superior SQL on Hadoop performance, scalability and coverage while also providing massively-parallel machine learning capabilities and support for native Hadoop file formats. In addition, HAWQ's advanced features include support for complex joins, rich and compliant SQL dialect and industry-differentiating data federation capabilities. Dynamic pipelining and pluggable query optimizer architecture enable HAWQ to perform queries on Hadoop with the speed and scalability required for enterprise data warehouse (EDW) workloads. HAWQ provides strong support for low-latency analytic SQL queries, coupled with massively parallel machine learning capabilities. This enables discovery-based analysis of large data sets and rapid, iterative development of data analytics applications that apply deep machine learning –
Re: [DISCUSS] HAWQ Incubation Proposal
Hi Roman, Great news! BTW, it might be a invalid URL for the proposal. Should be https://wiki.apache.org/incubator/HAWQProposal ? Thanks, Youngwoo On Fri, Aug 21, 2015 at 12:14 PM, Roman Shaposhnik r...@apache.org wrote: Hi! I would like to start a discussion on accepting HAWQ into ASF Incubator. The proposal is available at: https://wiki.apache.org/incubator/ApexProposal and is also attached to the end of this email. Please note, that this proposal is very complementary to the desire of HAWQ's sister project (MADlib) to join ASF Incubator: http://madlib.net/pipermail/user/2015-August/ http://madlib.net/pipermail/devel/2015-August/ I've volunteered to help MADlib community and we're currently working on a separate proposal to be submitted later next week. If you're interested in monitoring progress of that please see updates to: https://github.com/madlib/madlib/wiki/MADlib-ASF-Incubator-Proposal and later: https://wiki.apache.org/incubator/MADlibProposal Thanks in advance for your time and help. Thanks, Roman. == Abstract == HAWQ is an advanced enterprise SQL on Hadoop analytic engine built around a robust and high-performance massively-parallel processing (MPP) SQL framework evolved from Pivotal Greenplum DatabaseⓇ. HAWQ runs natively on Apache HadoopⓇ clusters by tightly integrating with HDFS and YARN. HAWQ supports multiple Hadoop file formats such as Apache Parquet, native HDFS, and Apache Avro. HAWQ is configured and managed as a Hadoop service in Apache Ambari. HAWQ is 100% ANSI SQL compliant (supporting ANSI SQL-92, SQL-99, and SQL-2003, plus OLAP extensions) and supports open database connectivity (ODBC) and Java database connectivity (JDBC), as well. Most business intelligence, data analysis and data visualization tools work with HAWQ out of the box without the need for specialized drivers. A unique aspect of HAWQ is its integration of statistical and machine learning capabilities that can be natively invoked from SQL or (in the context of PL/Python, PL/Java or PL/R) in massively parallel modes and applied to large data sets across a Hadoop cluster. These capabilities are provided through MADlib – an existing open source, parallel machine-learning library. Given the close ties between the two development communities, the MADlib community has expressed interest in joining HAWQ on its journey into the ASF Incubator and will be submitting a separate, concurrent proposal. HAWQ will provide more robust and higher performing options for Hadoop environments that demand best-in-class data analytics for business critical purposes. HAWQ is implemented in C and C++. == Proposal == The goal of this proposal is to bring the core of Pivotal Software, Inc.’s (Pivotal) Pivotal HAWQⓇ codebase into the Apache Software Foundation (ASF) in order to build a vibrant, diverse and self-governed open source community around the technology. Pivotal has agreed to transfer the brand name HAWQ to Apache Software Foundation and will stop using HAWQ to refer to this software if the project gets accepted into the ASF Incubator under the name of Apache HAWQ (incubating). Pivotal will continue to market and sell an analytic engine product that includes Apache HAWQ (incubating). While HAWQ is our primary choice for a name of the project, in anticipation of any potential issues with PODLINGNAMESEARCH we have come up with two alternative names: (1) Hornet; or (2) Grove. Pivotal is submitting this proposal to donate the HAWQ source code and associated artifacts (documentation, web site content, wiki, etc.) to the Apache Software Foundation Incubator under the Apache License, Version 2.0 and is asking Incubator PMC to establish an open source community. == Background == While the ecosystem of open source SQL-on-Hadoop solutions is fairly developed by now, HAWQ has several unique features that will set it apart from existing ASF and non-ASF projects. HAWQ made its debut in 2013 as a closed source product leveraging a decade's worth of product development effort invested in Greenplum DatabaseⓇ. Since then HAWQ has rapidly gained a solid customer base and became available on non-Pivotal distributions of Hadoop. In 2015 HAWQ still leverages the rock solid foundation of Greenplum Database, while at the same time embracing elasticity and resource management native to Hadoop applications. This allows HAWQ to provide superior SQL on Hadoop performance, scalability and coverage while also providing massively-parallel machine learning capabilities and support for native Hadoop file formats. In addition, HAWQ's advanced features include support for complex joins, rich and compliant SQL dialect and industry-differentiating data federation capabilities. Dynamic pipelining and pluggable query optimizer architecture enable HAWQ to perform queries on Hadoop with the speed and scalability required for enterprise data warehouse (EDW)
Re: [DISCUSS] HAWQ Incubation Proposal
On Thu, Aug 20, 2015 at 8:33 PM, 김영우 (Youngwoo Kim) warwit...@gmail.com wrote: Hi Roman, Great news! BTW, it might be a invalid URL for the proposal. Should be https://wiki.apache.org/incubator/HAWQProposal ? Two may copy-paste buffers strike again :-( Thanks for spotting it so quickly. Yes it is: https://wiki.apache.org/incubator/HAWQProposal Thanks, Roman. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
[DISCUSS] Horn Incubation Proposal
Hi all, We'd like to propose Horn (혼), a fully distributed system for large-scale deep learning as an Apache Incubator project and start the discussion. The complete proposal can be found at: https://wiki.apache.org/incubator/HornProposal Any advices and helps are welcome! Thanks, Edward. = Horn Proposal = == Abstract == (tentatively named Horn [hɔ:n], korean meaning of Horn is a Spirit) is a neuron-centric programming APIs and execution framework for large-scale deep learning, built on top of Apache Hama. == Proposal == It is a goal of the Horn to provide a neuron-centric programming APIs which allows user to easily define the characteristic of artificial neural network model and its structure, and its execution framework that leverages the heterogeneous resources on Hama and Hadoop YARN cluster. == Background == The initial ANN code was developed at Apache Hama project by a committer, Yexi Jiang (Facebook) in 2013. The motivation behind this work is to build a framework that provides more intuitive programming APIs like Google's MapReduce or Pregel and supports applications needing large model with huge memory consumptions in distributed way. == Rationale == While many of deep learning open source softwares such as Caffe, DeepDist, and NeuralGiraph are still data or model parallel only, we aim to support both data and model parallelism and also fault-tolerant system design. The basic idea of data and model parallelism is use of the remote parameter server to parallelize model creation and distribute training across machines, and the BSP framework of Apache Hama for performing asynchronous mini-batches. Within single BSP job, each task group works asynchronously using region barrier synchronization instead of global barrier synchronization, and trains large-scale neural network model using assigned data sets in BSP paradigm. Thus, we achieve data and model parallelism. This architecture is inspired by Google's !DistBelief (Jeff Dean et al, 2012). == Initial Goals == Some current goals include: * builds new community * provides more intuitive programming APIs * needs both data and model parallelism support * must run natively on both Hama and Hadoop2 * needs also GPUs and InfiniBand support (FPGAs if possible) == Current Status == === Meritocracy === The core developers understand what it means to have a process based on meritocracy. We will provide continuous efforts to build an environment that supports this, encouraging community members to contribute. === Community === A small community has formed within the Apache Hama project and some companies such as instant messenger service company and mobile manufacturing company. And many people are interested in the large-scale deep learning platform itself. By bringing Horn into Apache, we believe that the community will grow even bigger. === Core Developers === Edward J. Yoon, Thomas Jungblut, and Dongjin Lee == Known Risks == === Orphaned Products === Apache Hama is already a core open source component at Samsung Electronics, and Horn also will be used by Samsung Electronics, and so there is no direct risk for this project to be orphaned. === Inexperience with Open Source === Some are very new and the others have experience using and/or working on Apache open source projects. === Homogeneous Developers === The initial committers are from different organizations such as, Microsoft, Samsung Electronics, and Line Plus. === Reliance on Salaried Developers === Few will be worked as a full-time open source developer. Other developers will also start working on the project in their spare time. === Relationships with Other Apache Products === * Horn is based on Apache Hama * Apache Zookeeper is used for distributed locking service * Natively run on Apache Hadoop and Mesos * Horn can be somewhat overlapped with Singa podling (If possible, we'd also like to use Singa or Caffe to do the heavy lifting part). === An Excessive Fascination with the Apache Brand === Horn itself will hopefully have benefits from Apache, in terms of attracting a community and establishing a solid group of developers, but also the relation with Apache Hama, a general-purpose BSP computing engine. These are the main reasons for us to send this proposal. == Documentation == Initial plan about Horn can be found at http://blog.udanax.org/2015/06/googles-distbelief-clone-project-on.html == Initial Source == The initial source code has been release as part of Apache Hama project developed under Apache Software Foundation. The source code is currently hosted at https://svn.apache.org/repos/asf/hama/trunk/ml/src/main/java/org/apache/hama/ml/ann/ == Cryptography == Not applicable. == Required Resources == === Mailing Lists === * horn-private * horn-dev === Subversion Directory === * Git is the preferred source control system: git://git.apache.org/horn === Issue Tracking === * a JIRA issue tracker, HORN == Initial Committers and Affiliations
Re: What is the legal basis for enforcing release policies at ASF?
On Aug 20, 2015 08:52, Jim Jagielski j...@jagunet.com wrote: Coming in late. A snapshot is not a release. Licenses kick in at distribution/ release. I want to fix FUD before it infests the rafters and subfloor. I really have never read something so stupid or ill phrased... Every contributor committing code to any ASF project, or even contributing it to us in public forums (including our mailing lists, our bug trackers, etc) is committing that code under the AL or has designated explicitly what licence it came in under (commit message: forked from BSD-licensed code base at {URL}.) It is generally AL code all the time. I don't know where you invented a 'kick-in' concept, but unless the committers are violating their ICLA/CCLA, nothing could be further from the truth. There is also a trademark issue as well... only the ASF can declare something as a release. There we agree :)
Re: What is the legal basis for enforcing release policies at ASF?
On 8/20/15, 5:27 PM, William A Rowe Jr wr...@rowe-clan.net wrote: It is generally AL code all the time. I don't know where you invented a 'kick-in' concept, but unless the committers are violating their ICLA/CCLA, nothing could be further from the truth. Committers sometimes make mistakes. IIRC, Justin recently caught a mistake where some files accidentally got their non-AL headers replaced with AL headers. Large codebase contributions, especially initial podling code grants might be messy as well until scrubbed and approved for an official ASF release. I know from experience. -Alex
Re: What is the legal basis for enforcing release policies at ASF?
On Aug 20, 2015 8:19 PM, William A Rowe Jr wr...@rowe-clan.net wrote: On Aug 20, 2015 7:39 PM, Alex Harui aha...@adobe.com wrote: On 8/20/15, 5:27 PM, William A Rowe Jr wr...@rowe-clan.net wrote: It is generally AL code all the time. I don't know where you invented a 'kick-in' concept, but unless the committers are violating their ICLA/CCLA, nothing could be further from the truth. Committers sometimes make mistakes. IIRC, Justin recently caught a mistake where some files accidentally got their non-AL headers replaced with AL headers. Large codebase contributions, especially initial podling code grants might be messy as well until scrubbed and approved for an official ASF release. I know from experience. We don't disagree on this point. Sometimes, they are caught through the release process, or by peer review. Other times, we must retract the claim we offered. Nothing changes the fact that code is either offered under the AL 2.0 or another license, unless the author/licensor changes their license retroactively. Your comment also hones in on the logical fallacy our VP fell into... While it may be true that the ASF granted its own AL 2.0 license to the release package, the ASF is unable to change component licenses in incompatible ways. And the warranty the ASF offers on an inaccurate license claims is - nil - c.f. AL 2.0 However, if our repositories are under another license, that VP needs to make public this information, because I never got the memo, and I must notify friends and the many companies I advise and consult to that they all need to cease looking at the ASF's repositories, and let their respective legal departments each sort this all out, if those repositories are licensed with terms and conditions differing from the AL 2.0.
Re: What is the legal basis for enforcing release policies at ASF?
This thread started as a discussion of Linux distros and trademarks. Perhaps I could try to return it there? If a distro takes a release of Apache X, compiles it with minimal changes that adapt it to the environment, and distributes it, I believe that it's a fine thing for them to call it simple Apache X, and acknowledge our marks. If a distro takes a release of Apache X, and make significant changes to it, and then distributes it, I believe that it's not OK with us for them to simply call it Apache X. I've seen some evidence that Gentoo Linux makes a regular habit of this, because their policies drive them to make some pretty scary changes in some cases. Others may not share my view. Further, if someone takes a snapshot (small 's') from source control and starts from that, with minimal changes, I think that this would also be trademark-acceptable, so long as they accurately describe what they did. The operative concept here, as Shane has taught it, is 'confusion in the marketplace.' If some third party behaves so as to cause confusion as to the identity of Apache X, there's a trademark issue. If not, not.
Re: What is the legal basis for enforcing release policies at ASF?
On Thu, Aug 20, 2015 at 9:37 PM, Ross Gardler ross.gard...@microsoft.com wrote: I do not agree with this interpretation when viewed from a legal angle (though I do agree from a trademark angle). I have a feeling that the root of my disagreement is the same as the root of Jim's earlier statement (though I may be mistaken). You've lost me already, but let's unwind this... There are two points of IP due diligence in an Apache project: At the point of contribution where the IP is validated by the committer and zero or more people who review the patch. The second phase of IP validation is at the point of release, where 3 our more PMC members validate that the foundation can legally release the code. No, 3 or more PMC members make a best-effort that the code meets our qualifications for release. Not being copyright and patent atty's, we presume they did not cast their votes based on a legal definition of due diligence. This means that taking a snapshot and building a release is *not* trademark-acceptable since the foundation, through the project PMC has not approved the release, therefore it is not an Apache release. That much we agree on... Only the ASF gets to say what is an ASF release and to do so requires a vote of the PMC. It has nothing to do with the number of changes made to what is in our repositories. It has everything to do with whether it's a release of the foundation. Accurate... So, in the strictest sense, distributions that make minor changes for their distribution should call it Bar powered by Apache Foo in order to differentiate it from an official release of the foundation. In the real world the question is, from a legal point of view, do we care? Here is where there is some room for interpretation, the httpd project can probably be built more than 10^9 different ways (I extrapolate this from a Chipotle drink cup that claimed the number of permutations of their quick-service faux-tex-mex menu) (lets ignore the fact that some people vote on releases without doing proper validation, that's why we require 3 +1 votes, the assumption is that at least one of them did the job properly) Define Proper, I haven't read that http://www.apache.org/dev/release/proper.html page yet. You still didn't comment on the license under which the repository is licensed, so this wasn't a terribly helpful post. From: William A Rowe Jrmailto:wr...@rowe-clan.net Sent: 8/20/2015 7:17 PM To: general@incubator.apache.orgmailto:general@incubator.apache.org Subject: Re: What is the legal basis for enforcing release policies at ASF? On Thu, Aug 20, 2015 at 9:03 PM, Benson Margulies bimargul...@gmail.com wrote: This thread started as a discussion of Linux distros and trademarks. Perhaps I could try to return it there? If a distro takes a release of Apache X, compiles it with minimal changes that adapt it to the environment, and distributes it, I believe that it's a fine thing for them to call it simple Apache X, and acknowledge our marks. If a distro takes a release of Apache X, and make significant changes to it, and then distributes it, I believe that it's not OK with us for them to simply call it Apache X. I've seen some evidence that Gentoo Linux makes a regular habit of this, because their policies drive them to make some pretty scary changes in some cases. Others may not share my view. Further, if someone takes a snapshot (small 's') from source control and starts from that, with minimal changes, I think that this would also be trademark-acceptable, so long as they accurately describe what they did. The operative concept here, as Shane has taught it, is 'confusion in the marketplace.' If some third party behaves so as to cause confusion as to the identity of Apache X, there's a trademark issue. If not, not. You summed this up to the best of my understanding ... +1. If our legal VP agrees (and retracts earlier FUD) it appears we are entirely in agreement.
Re: What is the legal basis for enforcing release policies at ASF?
I think it is somewhat amusing, that this is actually discussed ~20years after Apache group is formed. A newcomer must be flabbergasted that this isn't clear cut by now... ;-) // Niclas On Fri, Aug 21, 2015 at 10:37 AM, Ross Gardler ross.gard...@microsoft.com wrote: I do not agree with this interpretation when viewed from a legal angle (though I do agree from a trademark angle). I have a feeling that the root of my disagreement is the same as the root of Jim's earlier statement (though I may be mistaken). There are two points of IP due diligence in an Apache project: At the point of contribution where the IP is validated by the committer and zero or more people who review the patch. The second phase of IP validation is at the point of release, where 3 our more PMC members validate that the foundation can legally release the code. This means that taking a snapshot and building a release is *not* trademark-acceptable since the foundation, through the project PMC has not approved the release, therefore it is not an Apache release. Only the ASF gets to say what is an ASF release and to do so requires a vote of the PMC. It has nothing to do with the number of changes made to what is in our repositories. It has everything to do with whether it's a release of the foundation. So, in the strictest sense, distributions that make minor changes for their distribution should call it Bar powered by Apache Foo in order to differentiate it from an official release of the foundation. In the real world the question is, from a legal point of view, do we care? (lets ignore the fact that some people vote on releases without doing proper validation, that's why we require 3 +1 votes, the assumption is that at least one of them did the job properly) Sent from my Windows Phone From: William A Rowe Jrmailto:wr...@rowe-clan.net Sent: 8/20/2015 7:17 PM To: general@incubator.apache.orgmailto:general@incubator.apache.org Subject: Re: What is the legal basis for enforcing release policies at ASF? On Thu, Aug 20, 2015 at 9:03 PM, Benson Margulies bimargul...@gmail.com wrote: This thread started as a discussion of Linux distros and trademarks. Perhaps I could try to return it there? If a distro takes a release of Apache X, compiles it with minimal changes that adapt it to the environment, and distributes it, I believe that it's a fine thing for them to call it simple Apache X, and acknowledge our marks. If a distro takes a release of Apache X, and make significant changes to it, and then distributes it, I believe that it's not OK with us for them to simply call it Apache X. I've seen some evidence that Gentoo Linux makes a regular habit of this, because their policies drive them to make some pretty scary changes in some cases. Others may not share my view. Further, if someone takes a snapshot (small 's') from source control and starts from that, with minimal changes, I think that this would also be trademark-acceptable, so long as they accurately describe what they did. The operative concept here, as Shane has taught it, is 'confusion in the marketplace.' If some third party behaves so as to cause confusion as to the identity of Apache X, there's a trademark issue. If not, not. You summed this up to the best of my understanding ... +1. If our legal VP agrees (and retracts earlier FUD) it appears we are entirely in agreement. -- Niclas Hedhman, Software Developer http://zest.apache.org - New Energy for Java
Re: What is the legal basis for enforcing release policies at ASF?
On Thu, Aug 20, 2015 at 9:03 PM, Benson Margulies bimargul...@gmail.com wrote: This thread started as a discussion of Linux distros and trademarks. Perhaps I could try to return it there? If a distro takes a release of Apache X, compiles it with minimal changes that adapt it to the environment, and distributes it, I believe that it's a fine thing for them to call it simple Apache X, and acknowledge our marks. If a distro takes a release of Apache X, and make significant changes to it, and then distributes it, I believe that it's not OK with us for them to simply call it Apache X. I've seen some evidence that Gentoo Linux makes a regular habit of this, because their policies drive them to make some pretty scary changes in some cases. Others may not share my view. Further, if someone takes a snapshot (small 's') from source control and starts from that, with minimal changes, I think that this would also be trademark-acceptable, so long as they accurately describe what they did. The operative concept here, as Shane has taught it, is 'confusion in the marketplace.' If some third party behaves so as to cause confusion as to the identity of Apache X, there's a trademark issue. If not, not. You summed this up to the best of my understanding ... +1. If our legal VP agrees (and retracts earlier FUD) it appears we are entirely in agreement.
Re: What is the legal basis for enforcing release policies at ASF?
On Thu, Aug 20, 2015 at 9:11 PM, Christopher ctubb...@apache.org wrote: It sounds to me like you're saying that the license under which code is offered (to anybody who encounters it) is independent of the license declaration attached to the project. No, the license is that which was granted by the author, and I think you missed my followup by a few minutes, so I will quote myself here... Your comment also hones in on the logical fallacy our VP fell into... While it may be true that the ASF granted its own AL 2.0 license to the release package, the ASF is unable to change component licenses in incompatible ways. And the warranty the ASF offers on an inaccurate license claims is - nil - c.f. AL 2.0 However, if our repositories are under another license, that VP needs to make public this information, because I never got the memo, and I must notify friends and the many companies I advise and consult to that they all need to cease looking at the ASF's repositories, and let their respective legal departments each sort this all out, if those repositories are licensed with terms and conditions differing from the AL 2.0. Obviously, I think our VP Legal isn't taking his job seriously of advising the community on the specific legal particularities of the software we create, which is why I'm going to stand pat until someone offers up a compelling argument over why anyone is not able to take any of the AL 2.0 code out of ASF repositories, released or not, and re-purpose it for whatever they desire. But don't name it by Apache {foo} unless {foo} PMC sanctioned the release of the code. It's entirely in trademark law, and our license and copyright law gives them everything they need to utilize the code, released or not.
Re: What is the legal basis for enforcing release policies at ASF?
It sounds to me like you're saying that the license under which code is offered (to anybody who encounters it) is independent of the license declaration attached to the project. This makes sense to me, presuming that we still agree that the license declaration (header or license file) is the best way to communicate the license under which the code is offered. It seems to follow, then, that were saying that there are sometimes errors in the declaration, where it doesn't reflect what license the code is actually offered under (if any). Further, we're saying that this is hopefully less likely in a release, which has been vetted with greater scrutiny. Is that right? If so, then it seems to me that the question really becomes: is it sufficiently communicated by the very fact of being a snapshot (any state of the code other than in a release), that errors are possible in the license? I would think the answer is yes, personally. However, I'm not sure it really means much, because it's still reasonable for people to assume the license declaration is correct, until shown otherwise. It seems to me that the very fact that any license declaration is attached to the code at all, regardless of its state as a release or snapshot, shifts the burden of responsibility to actually demonstrate that the license does not apply. This is the reverse of the case when no obvious license declaration is made. The burden in that case is to show that the license does apply. Isn't that why we explicitly put headers on each file, in addition to the LICENSE file? To explicitly shift this burden to us in order to encourage free use of our software by others? On Thu, Aug 20, 2015, 21:19 William A Rowe Jr wr...@rowe-clan.net wrote: On Aug 20, 2015 7:39 PM, Alex Harui aha...@adobe.com wrote: On 8/20/15, 5:27 PM, William A Rowe Jr wr...@rowe-clan.net wrote: It is generally AL code all the time. I don't know where you invented a 'kick-in' concept, but unless the committers are violating their ICLA/CCLA, nothing could be further from the truth. Committers sometimes make mistakes. IIRC, Justin recently caught a mistake where some files accidentally got their non-AL headers replaced with AL headers. Large codebase contributions, especially initial podling code grants might be messy as well until scrubbed and approved for an official ASF release. I know from experience. We don't disagree on this point. Sometimes, they are caught through the release process, or by peer review. Other times, we must retract the claim we offered. Nothing changes the fact that code is either offered under the AL 2.0 or another license, unless the author/licensor changes their license retroactively.
Re: What is the legal basis for enforcing release policies at ASF?
AFAIK a SNAPSHOT has not been voted on and is therefore not a formal ASF release. So for example this would cover CI builds that deploy jars to the ASF Maven SNAPSHOT repo. On 20 August 2015 at 23:33, Mike Kienenberger mkien...@gmail.com wrote: On Thu, Aug 20, 2015 at 6:23 PM, Gavin McDonald ga...@16degrees.com.au wrote: So what do we do about all the rc1|rc2|rcx ,alphas, betas and Milestone ‘releases’ that are on our official mirrors right now? (Because they would have been voted on as a ‘’release’’ for the projects to put them there in the first place) Release means different things in different contexts. An ASF release is a product that a PMC has vetted to meet ASF release standards (builds from source, APL2 licensed) and has made available to end-users in our download services. This use of release deals with legal and can-be-modified promises made to the end-user. Various ASF projects also use release to mean something different -- a community-approved product that has a certain API and typically has no known issues. This use of release generally deals with technical aspects of the project, such as stability and reliability. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: What is the legal basis for enforcing release policies at ASF?
On Aug 20, 2015 7:39 PM, Alex Harui aha...@adobe.com wrote: On 8/20/15, 5:27 PM, William A Rowe Jr wr...@rowe-clan.net wrote: It is generally AL code all the time. I don't know where you invented a 'kick-in' concept, but unless the committers are violating their ICLA/CCLA, nothing could be further from the truth. Committers sometimes make mistakes. IIRC, Justin recently caught a mistake where some files accidentally got their non-AL headers replaced with AL headers. Large codebase contributions, especially initial podling code grants might be messy as well until scrubbed and approved for an official ASF release. I know from experience. We don't disagree on this point. Sometimes, they are caught through the release process, or by peer review. Other times, we must retract the claim we offered. Nothing changes the fact that code is either offered under the AL 2.0 or another license, unless the author/licensor changes their license retroactively.
RE: What is the legal basis for enforcing release policies at ASF?
I do not agree with this interpretation when viewed from a legal angle (though I do agree from a trademark angle). I have a feeling that the root of my disagreement is the same as the root of Jim's earlier statement (though I may be mistaken). There are two points of IP due diligence in an Apache project: At the point of contribution where the IP is validated by the committer and zero or more people who review the patch. The second phase of IP validation is at the point of release, where 3 our more PMC members validate that the foundation can legally release the code. This means that taking a snapshot and building a release is *not* trademark-acceptable since the foundation, through the project PMC has not approved the release, therefore it is not an Apache release. Only the ASF gets to say what is an ASF release and to do so requires a vote of the PMC. It has nothing to do with the number of changes made to what is in our repositories. It has everything to do with whether it's a release of the foundation. So, in the strictest sense, distributions that make minor changes for their distribution should call it Bar powered by Apache Foo in order to differentiate it from an official release of the foundation. In the real world the question is, from a legal point of view, do we care? (lets ignore the fact that some people vote on releases without doing proper validation, that's why we require 3 +1 votes, the assumption is that at least one of them did the job properly) Sent from my Windows Phone From: William A Rowe Jrmailto:wr...@rowe-clan.net Sent: 8/20/2015 7:17 PM To: general@incubator.apache.orgmailto:general@incubator.apache.org Subject: Re: What is the legal basis for enforcing release policies at ASF? On Thu, Aug 20, 2015 at 9:03 PM, Benson Margulies bimargul...@gmail.com wrote: This thread started as a discussion of Linux distros and trademarks. Perhaps I could try to return it there? If a distro takes a release of Apache X, compiles it with minimal changes that adapt it to the environment, and distributes it, I believe that it's a fine thing for them to call it simple Apache X, and acknowledge our marks. If a distro takes a release of Apache X, and make significant changes to it, and then distributes it, I believe that it's not OK with us for them to simply call it Apache X. I've seen some evidence that Gentoo Linux makes a regular habit of this, because their policies drive them to make some pretty scary changes in some cases. Others may not share my view. Further, if someone takes a snapshot (small 's') from source control and starts from that, with minimal changes, I think that this would also be trademark-acceptable, so long as they accurately describe what they did. The operative concept here, as Shane has taught it, is 'confusion in the marketplace.' If some third party behaves so as to cause confusion as to the identity of Apache X, there's a trademark issue. If not, not. You summed this up to the best of my understanding ... +1. If our legal VP agrees (and retracts earlier FUD) it appears we are entirely in agreement.
Re: apache binary distributions
On Thu, Aug 20, 2015 at 1:06 AM, William A Rowe Jr wr...@rowe-clan.net wrote: There are some special things here we do have absolute control over. If a project wants to provide the 'official' build, why not start signing the .jar? Good idea, but to be practical to users, the certificate for the signing needs to be part of the certificate chain of the JVM (otherwise those would be needed to be installed on every host). I don't know how willing infra would be to support PKI at ASF for this, otherwise many projects will be limited due to cost (I could be wrong by now and that there are totally free CAs) Cheers -- Niclas Hedhman, Software Developer http://zest.apache.org - New Energy for Java
Re: What is the legal basis for enforcing release policies at ASF?
On Thu, Aug 20, 2015 at 9:52 AM, Jim Jagielski j...@jagunet.com wrote: Coming in late. A snapshot is not a release. Licenses kick in at distribution/ release. Are you sure? When you have a public source control repo, with a LICENSE file at the top, I would think that this counts as a legal 'publication' under the terms of the license. if not, just what is the legal status of source code snipped from our repositories? There is also a trademark issue as well... only the ASF can declare something as a release. On Aug 6, 2015, at 8:50 PM, Roman Shaposhnik ro...@shaposhnik.org wrote: Hi! while answering a question on release policies and ALv2 I've suddenly realized that I really don't know what is the legal basis for enforcing release policies we've got documented over here: http://www.apache.org/dev/release.html For example, what would be the legal basis for stopping a 3d party from releasing a snapshot of ASF's project source tree and claim it to be a release X.Y.Z of said project? Thanks, Roman. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: What is the legal basis for enforcing release policies at ASF?
Coming in late. A snapshot is not a release. Licenses kick in at distribution/ release. There is also a trademark issue as well... only the ASF can declare something as a release. On Aug 6, 2015, at 8:50 PM, Roman Shaposhnik ro...@shaposhnik.org wrote: Hi! while answering a question on release policies and ALv2 I've suddenly realized that I really don't know what is the legal basis for enforcing release policies we've got documented over here: http://www.apache.org/dev/release.html For example, what would be the legal basis for stopping a 3d party from releasing a snapshot of ASF's project source tree and claim it to be a release X.Y.Z of said project? Thanks, Roman. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] HAWQ Incubation Proposal
Drill is implemented entirely in Java. This isn't core to the proposal, but it would be better corrected. On Thu, Aug 20, 2015 at 8:33 PM, 김영우 (Youngwoo Kim) warwit...@gmail.com wrote: Hi Roman, Great news! BTW, it might be a invalid URL for the proposal. Should be https://wiki.apache.org/incubator/HAWQProposal ? Thanks, Youngwoo On Fri, Aug 21, 2015 at 12:14 PM, Roman Shaposhnik r...@apache.org wrote: Hi! I would like to start a discussion on accepting HAWQ into ASF Incubator. The proposal is available at: https://wiki.apache.org/incubator/ApexProposal and is also attached to the end of this email. Please note, that this proposal is very complementary to the desire of HAWQ's sister project (MADlib) to join ASF Incubator: http://madlib.net/pipermail/user/2015-August/ http://madlib.net/pipermail/devel/2015-August/ I've volunteered to help MADlib community and we're currently working on a separate proposal to be submitted later next week. If you're interested in monitoring progress of that please see updates to: https://github.com/madlib/madlib/wiki/MADlib-ASF-Incubator-Proposal and later: https://wiki.apache.org/incubator/MADlibProposal Thanks in advance for your time and help. Thanks, Roman. == Abstract == HAWQ is an advanced enterprise SQL on Hadoop analytic engine built around a robust and high-performance massively-parallel processing (MPP) SQL framework evolved from Pivotal Greenplum DatabaseⓇ. HAWQ runs natively on Apache HadoopⓇ clusters by tightly integrating with HDFS and YARN. HAWQ supports multiple Hadoop file formats such as Apache Parquet, native HDFS, and Apache Avro. HAWQ is configured and managed as a Hadoop service in Apache Ambari. HAWQ is 100% ANSI SQL compliant (supporting ANSI SQL-92, SQL-99, and SQL-2003, plus OLAP extensions) and supports open database connectivity (ODBC) and Java database connectivity (JDBC), as well. Most business intelligence, data analysis and data visualization tools work with HAWQ out of the box without the need for specialized drivers. A unique aspect of HAWQ is its integration of statistical and machine learning capabilities that can be natively invoked from SQL or (in the context of PL/Python, PL/Java or PL/R) in massively parallel modes and applied to large data sets across a Hadoop cluster. These capabilities are provided through MADlib – an existing open source, parallel machine-learning library. Given the close ties between the two development communities, the MADlib community has expressed interest in joining HAWQ on its journey into the ASF Incubator and will be submitting a separate, concurrent proposal. HAWQ will provide more robust and higher performing options for Hadoop environments that demand best-in-class data analytics for business critical purposes. HAWQ is implemented in C and C++. == Proposal == The goal of this proposal is to bring the core of Pivotal Software, Inc.’s (Pivotal) Pivotal HAWQⓇ codebase into the Apache Software Foundation (ASF) in order to build a vibrant, diverse and self-governed open source community around the technology. Pivotal has agreed to transfer the brand name HAWQ to Apache Software Foundation and will stop using HAWQ to refer to this software if the project gets accepted into the ASF Incubator under the name of Apache HAWQ (incubating). Pivotal will continue to market and sell an analytic engine product that includes Apache HAWQ (incubating). While HAWQ is our primary choice for a name of the project, in anticipation of any potential issues with PODLINGNAMESEARCH we have come up with two alternative names: (1) Hornet; or (2) Grove. Pivotal is submitting this proposal to donate the HAWQ source code and associated artifacts (documentation, web site content, wiki, etc.) to the Apache Software Foundation Incubator under the Apache License, Version 2.0 and is asking Incubator PMC to establish an open source community. == Background == While the ecosystem of open source SQL-on-Hadoop solutions is fairly developed by now, HAWQ has several unique features that will set it apart from existing ASF and non-ASF projects. HAWQ made its debut in 2013 as a closed source product leveraging a decade's worth of product development effort invested in Greenplum DatabaseⓇ. Since then HAWQ has rapidly gained a solid customer base and became available on non-Pivotal distributions of Hadoop. In 2015 HAWQ still leverages the rock solid foundation of Greenplum Database, while at the same time embracing elasticity and resource management native to Hadoop applications. This allows HAWQ to provide superior SQL on Hadoop performance, scalability and coverage while also providing massively-parallel machine learning capabilities and support for native Hadoop file formats. In addition, HAWQ's advanced features include support for
Re: [DISCUSS] Horn Incubation Proposal
Hi, I am an initial committer of Apache(incubating) SINGA (http://singa.incubator.apache.org/) Both SINGA and the proposal follow the general parameter-server architecture: workers for computing gradients; servers for parameter updating. SINGA has implemented the model and data parallelism discussed in the Horn' proposal: multiple worker groups for asynchronous training---data parallelism; and multiple workers in one group for synchronous training---model parallelism. One feature of SINGA's architecture is that it can be extended to organize the servers in a hierarchical topology, which may help to reduce the communication bottleneck of servers organized in a flat topology. For the programming model, currently Horn proposes to support feed-forward models, e.g., MLP, auto-encoder, while SINGA supports all three categories of the known models, feed-forward models (eg MLP, CNN), energy models (eg RBM, DBM), and recurrent models (eg. RNN). SINGA provides good support for users to code, e.g., implement new parameter updating protocols or layers, and is being integrated with HDFS as well. We will submit the first release and full documentation to the mentors this weekend, and if ok, we will announce the first full release soon. The GPU version is scheduled for October release. Technical papers: http://www.comp.nus.edu.sg/~ooibc/singa-mm15.pdf http://www.comp.nus.edu.sg/~ooibc/singaopen-mm15.pdf and project website (which has more details than the Apache web site): http://www.comp.nus.edu.sg/~dbsystem/singa/ There are plenty of rooms for collaborations indeed... regards beng chin www.comp.nus.edu.sg/~ooibc On 2015-08-21 08:27, Edward J. Yoon wrote: Hi all, We'd like to propose Horn (혼), a fully distributed system for large-scale deep learning as an Apache Incubator project and start the discussion. The complete proposal can be found at: https://wiki.apache.org/incubator/HornProposal Any advices and helps are welcome! Thanks, Edward. = Horn Proposal = == Abstract == (tentatively named Horn [hɔ:n], korean meaning of Horn is a Spirit) is a neuron-centric programming APIs and execution framework for large-scale deep learning, built on top of Apache Hama. == Proposal == It is a goal of the Horn to provide a neuron-centric programming APIs which allows user to easily define the characteristic of artificial neural network model and its structure, and its execution framework that leverages the heterogeneous resources on Hama and Hadoop YARN cluster. == Background == The initial ANN code was developed at Apache Hama project by a committer, Yexi Jiang (Facebook) in 2013. The motivation behind this work is to build a framework that provides more intuitive programming APIs like Google's MapReduce or Pregel and supports applications needing large model with huge memory consumptions in distributed way. == Rationale == While many of deep learning open source softwares such as Caffe, DeepDist, and NeuralGiraph are still data or model parallel only, we aim to support both data and model parallelism and also fault-tolerant system design. The basic idea of data and model parallelism is use of the remote parameter server to parallelize model creation and distribute training across machines, and the BSP framework of Apache Hama for performing asynchronous mini-batches. Within single BSP job, each task group works asynchronously using region barrier synchronization instead of global barrier synchronization, and trains large-scale neural network model using assigned data sets in BSP paradigm. Thus, we achieve data and model parallelism. This architecture is inspired by Google's !DistBelief (Jeff Dean et al, 2012). == Initial Goals == Some current goals include: * builds new community * provides more intuitive programming APIs * needs both data and model parallelism support * must run natively on both Hama and Hadoop2 * needs also GPUs and InfiniBand support (FPGAs if possible) == Current Status == === Meritocracy === The core developers understand what it means to have a process based on meritocracy. We will provide continuous efforts to build an environment that supports this, encouraging community members to contribute. === Community === A small community has formed within the Apache Hama project and some companies such as instant messenger service company and mobile manufacturing company. And many people are interested in the large-scale deep learning platform itself. By bringing Horn into Apache, we believe that the community will grow even bigger. === Core Developers === Edward J. Yoon, Thomas Jungblut, and Dongjin Lee == Known Risks == === Orphaned Products === Apache Hama is already a core open source component at Samsung Electronics, and Horn also will be used by Samsung Electronics, and so there is no direct risk for this project to be orphaned. === Inexperience with Open Source === Some are very new and the others have experience using and/or working on
Re: [DISCUSS] Horn Incubation Proposal
multiple worker groups for asynchronous training---data parallelism; and multiple workers in one group for synchronous training---model parallelism. So, it's basically execution of the multiple asynchronous BSP (Bulk Synchronous Parallel) jobs. This can be simply handled within only single BSP job using region barriers as mentioned in proposal. Moreover, since Apache Hama is a general-purpose BSP framework on top of HDFS, it provides the data partition, locality optimization, job/task scheduling, messaging and fault tolerance in scalable way by nature. For the programming model, currently Horn proposes to support feed-forward There are plenty of rooms for collaborations indeed... Yeah, but still it can be more improved. Maybe we can discuss the simplified programming APIs and many others e.g., support GPUs together in the future. On Fri, Aug 21, 2015 at 1:13 PM, ooibc oo...@comp.nus.edu.sg wrote: Hi, I am an initial committer of Apache(incubating) SINGA (http://singa.incubator.apache.org/) Both SINGA and the proposal follow the general parameter-server architecture: workers for computing gradients; servers for parameter updating. SINGA has implemented the model and data parallelism discussed in the Horn' proposal: multiple worker groups for asynchronous training---data parallelism; and multiple workers in one group for synchronous training---model parallelism. One feature of SINGA's architecture is that it can be extended to organize the servers in a hierarchical topology, which may help to reduce the communication bottleneck of servers organized in a flat topology. For the programming model, currently Horn proposes to support feed-forward models, e.g., MLP, auto-encoder, while SINGA supports all three categories of the known models, feed-forward models (eg MLP, CNN), energy models (eg RBM, DBM), and recurrent models (eg. RNN). SINGA provides good support for users to code, e.g., implement new parameter updating protocols or layers, and is being integrated with HDFS as well. We will submit the first release and full documentation to the mentors this weekend, and if ok, we will announce the first full release soon. The GPU version is scheduled for October release. Technical papers: http://www.comp.nus.edu.sg/~ooibc/singa-mm15.pdf http://www.comp.nus.edu.sg/~ooibc/singaopen-mm15.pdf and project website (which has more details than the Apache web site): http://www.comp.nus.edu.sg/~dbsystem/singa/ There are plenty of rooms for collaborations indeed... regards beng chin www.comp.nus.edu.sg/~ooibc On 2015-08-21 08:27, Edward J. Yoon wrote: Hi all, We'd like to propose Horn (혼), a fully distributed system for large-scale deep learning as an Apache Incubator project and start the discussion. The complete proposal can be found at: https://wiki.apache.org/incubator/HornProposal Any advices and helps are welcome! Thanks, Edward. = Horn Proposal = == Abstract == (tentatively named Horn [hɔ:n], korean meaning of Horn is a Spirit) is a neuron-centric programming APIs and execution framework for large-scale deep learning, built on top of Apache Hama. == Proposal == It is a goal of the Horn to provide a neuron-centric programming APIs which allows user to easily define the characteristic of artificial neural network model and its structure, and its execution framework that leverages the heterogeneous resources on Hama and Hadoop YARN cluster. == Background == The initial ANN code was developed at Apache Hama project by a committer, Yexi Jiang (Facebook) in 2013. The motivation behind this work is to build a framework that provides more intuitive programming APIs like Google's MapReduce or Pregel and supports applications needing large model with huge memory consumptions in distributed way. == Rationale == While many of deep learning open source softwares such as Caffe, DeepDist, and NeuralGiraph are still data or model parallel only, we aim to support both data and model parallelism and also fault-tolerant system design. The basic idea of data and model parallelism is use of the remote parameter server to parallelize model creation and distribute training across machines, and the BSP framework of Apache Hama for performing asynchronous mini-batches. Within single BSP job, each task group works asynchronously using region barrier synchronization instead of global barrier synchronization, and trains large-scale neural network model using assigned data sets in BSP paradigm. Thus, we achieve data and model parallelism. This architecture is inspired by Google's !DistBelief (Jeff Dean et al, 2012). == Initial Goals == Some current goals include: * builds new community * provides more intuitive programming APIs * needs both data and model parallelism support * must run natively on both Hama and Hadoop2 * needs also GPUs and InfiniBand support (FPGAs if possible) == Current Status == ===
Re: apache binary distributions
On Thu, Aug 20, 2015 at 8:09 AM, Niclas Hedhman nic...@hedhman.org wrote: On Thu, Aug 20, 2015 at 1:06 AM, William A Rowe Jr wr...@rowe-clan.net wrote: There are some special things here we do have absolute control over. If a project wants to provide the 'official' build, why not start signing the .jar? Good idea, but to be practical to users, the certificate for the signing needs to be part of the certificate chain of the JVM (otherwise those would be needed to be installed on every host). I don't know how willing infra would be to support PKI at ASF for this, otherwise many projects will be limited due to cost (I could be wrong by now and that there are totally free CAs) That infrastructure now exists through code signing service by Symantec. One PMC member (or more) gets their own unique log in, pushes the artifact (.jar, in this example) to the service and is returned a signed artifact reflecting the ASF providence. The interesting thing is the actual cert is unique to the object, so if it is discovered that it was compromised, the signature can be revoked (good luck having sig revocations active at boot time, but otherwise this is quite useful.) And because there is a history, we know who precisely requested each object signing.
Re: What is the legal basis for enforcing release policies at ASF?
On Thu, Aug 20, 2015 at 7:23 AM, Benson Margulies bimargul...@gmail.com wrote: On Thu, Aug 20, 2015 at 9:52 AM, Jim Jagielski j...@jagunet.com wrote: Coming in late. A snapshot is not a release. Licenses kick in at distribution/ release. Are you sure? When you have a public source control repo, with a LICENSE file at the top, I would think that this counts as a legal 'publication' under the terms of the license. if not, just what is the legal status of source code snipped from our repositories? I agree with Jim that a snapshot is not a release. I also agree with him that licenses kick in at distribution. As to whether they kick in at distribution/release, I think that's a weird bit of wording, and I would be surprised if we are not all in agreement here. There were long threads on this topic back in 2007-2009 on legal-discuss@apache. http://markmail.org/message/jangmpbssvvd73az http://s.apache.org/6Wm http://markmail.org/message/xietapwmthvvknex http://s.apache.org/H6o Here's are a couple germane points from Roy: http://markmail.org/message/vbfjep4r2npkwufa http://s.apache.org/aXK Copyright law has no concept of software development. So, when a lawyer looks at http://svn.apache.org/repos/asf/httpd/httpd/trunk/ what the lawyer (or even layperson) sees is a website. http://markmail.org/message/44ezdre3se3ov5nu http://s.apache.org/MEC SVN is not a distribution point. Of course it is a distribution point. Distribution == copy to someone else. It isn't a release (an editorial decision by the ASF). Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: What is the legal basis for enforcing release policies at ASF?
On 20 Aug 2015, at 2:52 pm, Jim Jagielski j...@jagunet.com wrote: Coming in late. A snapshot is not a release. Licenses kick in at distribution/ release. Interesting. So what do we do about all the rc1|rc2|rcx ,alphas, betas and Milestone ‘releases’ that are on our official mirrors right now? (Because they would have been voted on as a ‘’release’’ for the projects to put them there in the first place) (Or are all those different to Snapshots somehow?) Gav… There is also a trademark issue as well... only the ASF can declare something as a release. On Aug 6, 2015, at 8:50 PM, Roman Shaposhnik ro...@shaposhnik.org wrote: Hi! while answering a question on release policies and ALv2 I've suddenly realized that I really don't know what is the legal basis for enforcing release policies we've got documented over here: http://www.apache.org/dev/release.html For example, what would be the legal basis for stopping a 3d party from releasing a snapshot of ASF's project source tree and claim it to be a release X.Y.Z of said project? Thanks, Roman. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org Gav... (( ( ( )\ ) )\ ) )\ ) ( )) )\(()/((()/( (()/( )\ ) ( ) ( /( ( (( /( ( ( ( _)( /(_))/(_)) /(_)) ( (()/( )( ( /( ( )\()))())\ ( )\()) ))\ )())\ )\ _ )\ (_)) (_))_| (_)) )\ ) /(_))(()\ )(_)) )\ (_))/(()\ /((_) )\ (_))/ /((_)(()\ /((_) (_)_\(_)/ __|| |_ |_ _| _(_/((_) _| ((_)((_)_ ((_)| |_ ((_)(_))( ((_)| |_ (_))( ((_)(_)) / _ \ \__ \| __| | | | ' \))| _|| '_|/ _` |(_-| _|| '_|| || |/ _| | _|| || || '_|/ -_) /_/ \_\ |___/|_||___||_||_| |_| |_| \__,_|/__/ \__||_| \_,_|\__| \__| \_,_||_| \___|
Re: What is the legal basis for enforcing release policies at ASF?
On Thu, Aug 20, 2015 at 6:23 PM, Gavin McDonald ga...@16degrees.com.au wrote: So what do we do about all the rc1|rc2|rcx ,alphas, betas and Milestone ‘releases’ that are on our official mirrors right now? (Because they would have been voted on as a ‘’release’’ for the projects to put them there in the first place) Release means different things in different contexts. An ASF release is a product that a PMC has vetted to meet ASF release standards (builds from source, APL2 licensed) and has made available to end-users in our download services. This use of release deals with legal and can-be-modified promises made to the end-user. Various ASF projects also use release to mean something different -- a community-approved product that has a certain API and typically has no known issues. This use of release generally deals with technical aspects of the project, such as stability and reliability. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org