Rather than trying to spend time figuring out what the open source restriction means, I will make it concrete. Is it okay to develop the algorithm using common python packages, and in particular SciKit Learn?
", "messages": [ { "id": "64493", "postDate": "02/18/2015 14:00:51", "content": "Rather than trying to spend time figuring out what the open source restriction means, I will make it concrete. Is it okay to develop the algorithm using common python packages, and in particular SciKit Learn?
", "rawMarkdown": "", "votes": null }, { "id": "64494", "postDate": "02/18/2015 14:03:42", "content": "have you seen the license of scikit-learn? Which rule says we cannot use it?
", "rawMarkdown": "", "votes": null }, { "id": "64495", "postDate": "02/18/2015 14:05:48", "content": "In this competition there is an open sources restriction. I never recall seeing this in a previous competition, and rather than researching I thought it would be more time cost effective to just ask.
\nScikit learn uses: Open source, commercially usable - BSD license. So, I assume it is okay to use in this competition.
", "rawMarkdown": "", "votes": null }, { "id": "64499", "postDate": "02/18/2015 15:19:20", "content": "I wouldn't call this a "restriction." A good majority of open source software falls into the OSI-approved categories.
\nThe intent is to avoid a situation where your solution is restrictive and/or costs money to use. The hosts intend for the results of this competition to be freely available for the benefit of anyone who wants to use it.
", "rawMarkdown": "", "votes": null }, { "id": "64503", "postDate": "02/18/2015 15:43:33", "content": "Thank you. I started this post foggy eyed in the morning before my 1st cup of coffee. I think I am clear on it now.
", "rawMarkdown": "", "votes": null }, { "id": "64705", "postDate": "02/21/2015 12:34:04", "content": "Some of the language used and conditions in the rules are confusing/strange.
\n\nSpecifically in reference to the point raised above, the rules don't seem to indicate that a submission depending on software licensed under a closed source license will be ineligible. The condition, copied below, only mentions open source licensed code that isn't approved.
\nOPEN-SOURCE CODE
\nA Submission will be ineligible to win a prize if it was developed using code containing or depending on software licensed under an open source license:
\n* other than an Open Source Initiative-approved license (see <http://opensource.org/>); or
* an open source license that prohibits commercial use.
Now I've actually looked at the rules, this also seems a little weird
\n\nBy accepting any Prize, each Participant receiving a Prize thereby agrees to use of his/her name, address, likeness and/or Prize information by Kaggle and Competition Sponsor for promotional purposes in any medium without additional compensation.
\n\nI can understand name and prize information, but address?
", "rawMarkdown": "", "votes": null }, { "id": "64708", "postDate": "02/21/2015 14:59:53", "content": "For legal reasons we can not approve or deny specific IP situations or uses, but the end requirement is there:
\nIf the Winner License Type for the Competition (see Winner License Type above) is Open Source License, then each Winner by accepting a Prize thereby:
\n* licenses their winning Submission and the source code used to generate the Submission under the MIT License (an open source software license commonly described at http://opensource.org/licenses/MIT), unless another open source license is chosen explicitly from the list of those approved by the Open Source Initiative at http://opensource.org/licenses/; and
* represents that he/she/it has the unrestricted right to grant that license.
I am not a lawyer, but the address is likely a requirement for compliance (e.g. to make sure you are not a resident of an enemy of the state, to comply with US sweepstakes law, etc.) We will not publish your address publicly.
", "rawMarkdown": "", "votes": null }, { "id": "64711", "postDate": "02/21/2015 16:07:24", "content": "The section you quote, at least to my reading, only covers the submission itself. If I release my solution under the MIT licence but it uses the Magic8Ball package only available as a closed-source binary it doesn't look like I violate either rule. Or, perhaps more accurately given IANAL, it is not obvious to me that such a situation would violate the rules of the competition.
\n\nWith regard to the address, the section I quote specifically deals with the use of personal information for promotional purposes. Totally reasonable that an address is required for legal and tax reasons, and these are covered in separate sections of the rules. My unease was specifically in reference to the use of these details for promotional purposes.
", "rawMarkdown": "", "votes": null }, { "id": "64716", "postDate": "02/21/2015 19:36:33", "content": "[quote=Jonathan Street;64711]
\nThe section you quote, at least to my reading, only covers the submission itself. If I release my solution under the MIT licence but it uses the Magic8Ball package only available as a closed-source binary it doesn't look like I violate either rule. Or, perhaps more accurately given IANAL, it is not obvious to me that such a situation would violate the rules of the competition.
\n[/quote]
\nIn the intro to the section:
\nAs a further condition of receipt of a Prize, each winning Participant thereby licenses their winning Submission and the source code used to generate the Submission according to the Winner License Type specified above
\n[quote=Jonathan Street;64711]
\n\n\nWith regard to the address, the section I quote specifically deals with the use of personal information for promotional purposes. Totally reasonable that an address is required for legal and tax reasons, and these are covered in separate sections of the rules. My unease was specifically in reference to the use of these details for promotional purposes.
\n[/quote]
\nGood point. It's outside my expertise to explain why that's there, but I can say we have never used and do not intend to use a full address for promotional purposes. Maybe it's just there to cover use of the city, e.g. "John Smith from Anytown, USA took first prize..."
", "rawMarkdown": "", "votes": null }, { "id": "64720", "postDate": "02/21/2015 22:26:46", "content": "I may well just be reading too much into the licensing. The language seems a little clunky (but then what T&Cs don't?) but as to whether there are genuine loopholes in there is beyond me.
\n\nFor the address, that seems reasonable. I have certainly never seen any communication from Kaggle I thought was questionable from a privacy standpoint.
", "rawMarkdown": "", "votes": null }, { "id": "64721", "postDate": "02/21/2015 22:37:04", "content": "[quote=William Cukierski;64716]
\n[quote=Jonathan Street;64711]
\nThe section you quote, at least to my reading, only covers the submission itself. If I release my solution under the MIT licence but it uses the Magic8Ball package only available as a closed-source binary it doesn't look like I violate either rule. Or, perhaps more accurately given IANAL, it is not obvious to me that such a situation would violate the rules of the competition.
\n[/quote]
\nIn the intro to the section:
\nAs a further condition of receipt of a Prize, each winning Participant thereby licenses their winning Submission and the source code used to generate the Submission according to the Winner License Type specified above
\n[/quote]
\nAs I understand it, nothing in the rules prohibits using closed-source tools/libraries.
\nSo if a competitor chooses to refactor some of their algorithms into a closed source library, that would be allowed. Unless you're saying that if Cleve Moler were to win, he would be required to open-source MATLAB...
", "rawMarkdown": "", "votes": null }, { "id": "65284", "postDate": "03/02/2015 19:52:46", "content": "I'm sorry, but I'm still not clear on this point. My (hesitant) understanding is that competitors are free to use proprietary software, as long as the solutions they generate are unencumbered by licensing issues.
\nIs this correct?
", "rawMarkdown": "", "votes": null }, { "id": "65416", "postDate": "03/04/2015 15:22:57", "content": "So can we use Matlab?
", "rawMarkdown": "", "votes": null }, { "id": "65471", "postDate": "03/05/2015 02:50:11", "content": "[quote=William Cukierski;64716]
\nIn the intro to the section:
\nAs a further condition of receipt of a Prize, each winning Participant thereby licenses their winning Submission and the source code used to generate the Submission according to the Winner License Type specified above
\n[/quote]
\nSo, it's acceptable to use proprietary software for analysis, as long as the model implementation software is free to give away, correct?
", "rawMarkdown": "", "votes": null } ], "comments": [ { "id": 64494, "author_name": "abhishek", "author_url": "", "post_date": "02/18/2015 14:03:42", "content": "have you seen the license of scikit-learn? Which rule says we cannot use it?
", "votes": null, "replies": [] }, { "id": 64495, "author_name": "joexjmmvhm", "author_url": "", "post_date": "02/18/2015 14:05:48", "content": "In this competition there is an open sources restriction. I never recall seeing this in a previous competition, and rather than researching I thought it would be more time cost effective to just ask.
\nScikit learn uses: Open source, commercially usable - BSD license. So, I assume it is okay to use in this competition.
", "votes": null, "replies": [] }, { "id": 64499, "author_name": "wcukierski", "author_url": "", "post_date": "02/18/2015 15:19:20", "content": "I wouldn't call this a "restriction." A good majority of open source software falls into the OSI-approved categories.
\nThe intent is to avoid a situation where your solution is restrictive and/or costs money to use. The hosts intend for the results of this competition to be freely available for the benefit of anyone who wants to use it.
", "votes": null, "replies": [] }, { "id": 64503, "author_name": "joexjmmvhm", "author_url": "", "post_date": "02/18/2015 15:43:33", "content": "Thank you. I started this post foggy eyed in the morning before my 1st cup of coffee. I think I am clear on it now.
", "votes": null, "replies": [] }, { "id": 64705, "author_name": "jstreet", "author_url": "", "post_date": "02/21/2015 12:34:04", "content": "Some of the language used and conditions in the rules are confusing/strange.
\n\nSpecifically in reference to the point raised above, the rules don't seem to indicate that a submission depending on software licensed under a closed source license will be ineligible. The condition, copied below, only mentions open source licensed code that isn't approved.
\nOPEN-SOURCE CODE
\nA Submission will be ineligible to win a prize if it was developed using code containing or depending on software licensed under an open source license:
\n* other than an Open Source Initiative-approved license (see <http://opensource.org/>); or
* an open source license that prohibits commercial use.
Now I've actually looked at the rules, this also seems a little weird
\n\nBy accepting any Prize, each Participant receiving a Prize thereby agrees to use of his/her name, address, likeness and/or Prize information by Kaggle and Competition Sponsor for promotional purposes in any medium without additional compensation.
\n\nI can understand name and prize information, but address?
", "votes": null, "replies": [] }, { "id": 64708, "author_name": "wcukierski", "author_url": "", "post_date": "02/21/2015 14:59:53", "content": "For legal reasons we can not approve or deny specific IP situations or uses, but the end requirement is there:
\nIf the Winner License Type for the Competition (see Winner License Type above) is Open Source License, then each Winner by accepting a Prize thereby:
\n* licenses their winning Submission and the source code used to generate the Submission under the MIT License (an open source software license commonly described at http://opensource.org/licenses/MIT), unless another open source license is chosen explicitly from the list of those approved by the Open Source Initiative at http://opensource.org/licenses/; and
* represents that he/she/it has the unrestricted right to grant that license.
I am not a lawyer, but the address is likely a requirement for compliance (e.g. to make sure you are not a resident of an enemy of the state, to comply with US sweepstakes law, etc.) We will not publish your address publicly.
", "votes": null, "replies": [] }, { "id": 64711, "author_name": "jstreet", "author_url": "", "post_date": "02/21/2015 16:07:24", "content": "The section you quote, at least to my reading, only covers the submission itself. If I release my solution under the MIT licence but it uses the Magic8Ball package only available as a closed-source binary it doesn't look like I violate either rule. Or, perhaps more accurately given IANAL, it is not obvious to me that such a situation would violate the rules of the competition.
\n\nWith regard to the address, the section I quote specifically deals with the use of personal information for promotional purposes. Totally reasonable that an address is required for legal and tax reasons, and these are covered in separate sections of the rules. My unease was specifically in reference to the use of these details for promotional purposes.
", "votes": null, "replies": [] }, { "id": 64716, "author_name": "wcukierski", "author_url": "", "post_date": "02/21/2015 19:36:33", "content": "[quote=Jonathan Street;64711]
\nThe section you quote, at least to my reading, only covers the submission itself. If I release my solution under the MIT licence but it uses the Magic8Ball package only available as a closed-source binary it doesn't look like I violate either rule. Or, perhaps more accurately given IANAL, it is not obvious to me that such a situation would violate the rules of the competition.
\n[/quote]
\nIn the intro to the section:
\nAs a further condition of receipt of a Prize, each winning Participant thereby licenses their winning Submission and the source code used to generate the Submission according to the Winner License Type specified above
\n[quote=Jonathan Street;64711]
\n\n\nWith regard to the address, the section I quote specifically deals with the use of personal information for promotional purposes. Totally reasonable that an address is required for legal and tax reasons, and these are covered in separate sections of the rules. My unease was specifically in reference to the use of these details for promotional purposes.
\n[/quote]
\nGood point. It's outside my expertise to explain why that's there, but I can say we have never used and do not intend to use a full address for promotional purposes. Maybe it's just there to cover use of the city, e.g. "John Smith from Anytown, USA took first prize..."
", "votes": null, "replies": [] }, { "id": 64720, "author_name": "jstreet", "author_url": "", "post_date": "02/21/2015 22:26:46", "content": "I may well just be reading too much into the licensing. The language seems a little clunky (but then what T&Cs don't?) but as to whether there are genuine loopholes in there is beyond me.
\n\nFor the address, that seems reasonable. I have certainly never seen any communication from Kaggle I thought was questionable from a privacy standpoint.
", "votes": null, "replies": [] }, { "id": 64721, "author_name": "emolson", "author_url": "", "post_date": "02/21/2015 22:37:04", "content": "[quote=William Cukierski;64716]
\n[quote=Jonathan Street;64711]
\nThe section you quote, at least to my reading, only covers the submission itself. If I release my solution under the MIT licence but it uses the Magic8Ball package only available as a closed-source binary it doesn't look like I violate either rule. Or, perhaps more accurately given IANAL, it is not obvious to me that such a situation would violate the rules of the competition.
\n[/quote]
\nIn the intro to the section:
\nAs a further condition of receipt of a Prize, each winning Participant thereby licenses their winning Submission and the source code used to generate the Submission according to the Winner License Type specified above
\n[/quote]
\nAs I understand it, nothing in the rules prohibits using closed-source tools/libraries.
\nSo if a competitor chooses to refactor some of their algorithms into a closed source library, that would be allowed. Unless you're saying that if Cleve Moler were to win, he would be required to open-source MATLAB...
", "votes": null, "replies": [] }, { "id": 65284, "author_name": "predictor", "author_url": "", "post_date": "03/02/2015 19:52:46", "content": "I'm sorry, but I'm still not clear on this point. My (hesitant) understanding is that competitors are free to use proprietary software, as long as the solutions they generate are unencumbered by licensing issues.
\nIs this correct?
", "votes": null, "replies": [] }, { "id": 65416, "author_name": "jiweiliu", "author_url": "", "post_date": "03/04/2015 15:22:57", "content": "So can we use Matlab?
", "votes": null, "replies": [] }, { "id": 65471, "author_name": "predictor", "author_url": "", "post_date": "03/05/2015 02:50:11", "content": "[quote=William Cukierski;64716]
\nIn the intro to the section:
\nAs a further condition of receipt of a Prize, each winning Participant thereby licenses their winning Submission and the source code used to generate the Submission according to the Winner License Type specified above
\n[/quote]
\nSo, it's acceptable to use proprietary software for analysis, as long as the model implementation software is free to give away, correct?
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "64493": "", "64494": "", "64495": "", "64499": "", "64503": "", "64705": "", "64708": "", "64711": "", "64716": "", "64720": "", "64721": "", "65284": "", "65416": "", "65471": "" }, "source": "meta" }