On 30/01/2014 22:57, Daniel Berlin wrote:> On Thu, Jan 30, 2014 at 2:34 PM, Alp Toker <alp at nuanti.com> wrote: >> On 30/01/2014 22:06, Daniel Berlin wrote: >>> Actually, the policy actually says the right thing, you removed a >>> sentence, which says: >>> "Please contact the oversight group for more details." >> >> To be clear, I didn't remove this sentence -- it's some way down in the >> paragraph and pertains to receiving details from the oversight group, >> whereas the section about providing notification pertains to "us", the LLVM >> community. >> >> This is grammatically clear and unambiguous so if it's not the intention, >> that needs to be reworded. > Don't take this the wrong way, but most people would still understand > that this probably means "don't talk about patents except to the > oversight group".Hi Dan, I'm not sure if you realize, but that paragraph reads as an open invitation to notify and discuss patents on the development and commits lists.> There are roughly no open source projects where the > rule is "talk about patents all you like on random development mailing > lists"There are plenty of projects around the world the where that is absolutely the rule. Your remarks suggest that you've worked on a limited range of projects without the perspective it takes to accurately word a developer policy like this for an international audience. "Most people" doesn't cut it here and we need to set out our expectations explicitly before we start turning away new contributors and telling them to "hit the road" for something they said. I can't stand by while new contributors receive abuse for some violation of a rule that's not even written in the LLVM developer policy. We should assume that contributors come from a background that's varied, inclusive and different to the norms in our immediate circles, and aim to provide them with accurate and helpful guidance in the developer policy and reflects our expectations. Flaming people when they try to engage our community because they have a different legal system or interpretation. What next, turn people away because they have a funny name? Reject patches because their skin color is different to yours? I'm disappointed that you've tried to defend what is clearly repeated and inexcusable behaviour by Chandler towards people who are graciously trying to help out. In so far as there is a community, we must stand up distance ourselves from behaviour like that. (This is becoming OT for cfe-dev, moving the thread to llvm-dev. Let's refocus into a more productive mode and and roll a patch already?) Alp. -- http://www.nuanti.com the browser experts
Sorry about causing all this trouble. I misunderstood what was allowable conversation on this list, and honestly didn't even think that I was seriously discussing patents, I thought my statement was simply a statement of fact, but I realize that these are things on which professionals should be consulted. As to patent issues, I'll cross that bridge with the oversight group if I ever get something viable to be included in llvm. I would like to redirect my question simply to be: how do people feel about whether it should be implemented as Visual C++ SEH or barest-bones SEH. Being new, I'm also not sure whether that would be a CFE question or an LLVM question, if it is better suited to llvm-dev, please let me know. Thanks, JB On Thu, Jan 30, 2014 at 5:08 PM, Alp Toker <alp at nuanti.com> wrote:> > On 30/01/2014 22:57, Daniel Berlin wrote: > >> On Thu, Jan 30, 2014 at 2:34 PM, Alp Toker <alp at nuanti.com> wrote: >> >>> On 30/01/2014 22:06, Daniel Berlin wrote: >>> >>>> Actually, the policy actually says the right thing, you removed a >>>> sentence, which says: >>>> "Please contact the oversight group for more details." >>>> >>> >>> To be clear, I didn't remove this sentence -- it's some way down in the >>> paragraph and pertains to receiving details from the oversight group, >>> whereas the section about providing notification pertains to "us", the >>> LLVM >>> community. >>> >>> This is grammatically clear and unambiguous so if it's not the intention, >>> that needs to be reworded. >>> >> Don't take this the wrong way, but most people would still understand >> that this probably means "don't talk about patents except to the >> oversight group". >> > > Hi Dan, > > I'm not sure if you realize, but that paragraph reads as an open > invitation to notify and discuss patents on the development and commits > lists. > > > There are roughly no open source projects where the >> rule is "talk about patents all you like on random development mailing >> lists" >> > > There are plenty of projects around the world the where that is absolutely > the rule. > > Your remarks suggest that you've worked on a limited range of projects > without the perspective it takes to accurately word a developer policy like > this for an international audience. > > "Most people" doesn't cut it here and we need to set out our expectations > explicitly before we start turning away new contributors and telling them > to "hit the road" for something they said. > > I can't stand by while new contributors receive abuse for some violation > of a rule that's not even written in the LLVM developer policy. > > We should assume that contributors come from a background that's varied, > inclusive and different to the norms in our immediate circles, and aim to > provide them with accurate and helpful guidance in the developer policy and > reflects our expectations. Flaming people when they try to engage our > community because they have a different legal system or interpretation. > > What next, turn people away because they have a funny name? Reject patches > because their skin color is different to yours? I'm disappointed that > you've tried to defend what is clearly repeated and inexcusable behaviour > by Chandler towards people who are graciously trying to help out. In so far > as there is a community, we must stand up distance ourselves from behaviour > like that. > > (This is becoming OT for cfe-dev, moving the thread to llvm-dev. Let's > refocus into a more productive mode and and roll a patch already?) > > > Alp. > > > -- > http://www.nuanti.com > the browser experts > > _______________________________________________ > cfe-dev mailing list > cfe-dev at cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140130/f77b8892/attachment.html>
On Thu, Jan 30, 2014 at 4:08 PM, Alp Toker <alp at nuanti.com> wrote:> > On 30/01/2014 22:57, Daniel Berlin wrote: >> >> On Thu, Jan 30, 2014 at 2:34 PM, Alp Toker <alp at nuanti.com> wrote: >>> >>> On 30/01/2014 22:06, Daniel Berlin wrote: >>>> >>>> Actually, the policy actually says the right thing, you removed a >>>> sentence, which says: >>>> "Please contact the oversight group for more details." >>> >>> >>> To be clear, I didn't remove this sentence -- it's some way down in the >>> paragraph and pertains to receiving details from the oversight group, >>> whereas the section about providing notification pertains to "us", the >>> LLVM >>> community. >>> >>> This is grammatically clear and unambiguous so if it's not the intention, >>> that needs to be reworded. >> >> Don't take this the wrong way, but most people would still understand >> that this probably means "don't talk about patents except to the >> oversight group". > > > Hi Dan, > > I'm not sure if you realize, but that paragraph reads as an open invitation > to notify and discuss patents on the development and commits lists. > > >> There are roughly no open source projects where the >> rule is "talk about patents all you like on random development mailing >> lists" > > > There are plenty of projects around the world the where that is absolutely > the rule. > > Your remarks suggest that you've worked on a limited range of projects > without the perspective it takes to accurately word a developer policy like > this for an international audience.Actually, I've worked on hundreds of open source projects, and plenty of them very international, and worded a fairly large number of developer policies. You can't write one that works for everyone no matter what you do. But rather than say what you did, which comes off as a personal attack, a more productive thing to say it would have been "I would find the current policy better if it was worded like <x>"> > "Most people" doesn't cut it here and we need to set out our expectations > explicitly before we start turning away new contributors and telling them to > "hit the road" for something they said.I didn't see anyone do anything of the sort here.> > I can't stand by while new contributors receive abuse for some violation of > a rule that's not even written in the LLVM developer policy.Now we've gone from someone kindly asking someone else not to discuss something to abuse?> > We should assume that contributors come from a background that's varied, > inclusive and different to the norms in our immediate circles, and aim to > provide them with accurate and helpful guidance in the developer policy and > reflects our expectations. Flaming people when they try to engage our > community because they have a different legal system or interpretation.I haven't see any flaming here either ...> > What next, turn people away because they have a funny name? Reject patches > because their skin color is different to yours? I'm disappointed that you've > tried to defend what is clearly repeated and inexcusable behaviour by > Chandler towards people who are graciously trying to help out. In so far as > there is a community, we must stand up distance ourselves from behaviour > like that.This seems like a pretty strong digression and very odd comparison. I'm not sure what personal issues you have with chandler, but he kindly asked someone not to discuss stuff. The entire part above, both about chandler, and me, seem like highly unproductive personal attacks.> > (This is becoming OT for cfe-dev, moving the thread to llvm-dev. Let's > refocus into a more productive mode and and roll a patch already?)> > > Alp. > > > -- > http://www.nuanti.com > the browser experts >
On 31/01/2014 00:47, Jb Feldman wrote:> Sorry about causing all this trouble. I misunderstood what was > allowable conversation on this list, and honestly didn't even think > that I was seriously discussing patents, I thought my statement was > simply a statement of fact, but I realize that these are things on > which professionals should be consulted. As to patent issues, I'll > cross that bridge with the oversight group if I ever get something > viable to be included in llvm.Hi Jb, Thanks for holding on. This has happened before so I'd like to point out that Chandler Carruth, although a committer, doesn't speak for the project. We should have done a better job of politely moving the discussion towards a constructive conclusion providing some pointers towards the right forum. I'm going to follow this up once the dust settles so we have a better response in place.> > I would like to redirect my question simply to be: how do people feel > about whether it should be implemented as Visual C++ SEH or > barest-bones SEH. Being new, I'm also not sure whether that would be a > CFE question or an LLVM question, if it is better suited to llvm-dev, > please let me know.I think others will agree with me that we look forward to working on this feature with you and earlier patch contributors when the time comes. The feature would presumably cross into both projects so would be appropriate as a cross-posting to both list. Until that time comes I'd say it's expedient to focus on other areas. Does that work for you? Cheers, Alp.> > Thanks, > JB > > On Thu, Jan 30, 2014 at 5:08 PM, Alp Toker <alp at nuanti.com > <mailto:alp at nuanti.com>> wrote: > > > On 30/01/2014 22:57, Daniel Berlin wrote: > > On Thu, Jan 30, 2014 at 2:34 PM, Alp Toker <alp at nuanti.com > <mailto:alp at nuanti.com>> wrote: > > On 30/01/2014 22:06, Daniel Berlin wrote: > > Actually, the policy actually says the right thing, > you removed a > sentence, which says: > "Please contact the oversight group for more details." > > > To be clear, I didn't remove this sentence -- it's some > way down in the > paragraph and pertains to receiving details from the > oversight group, > whereas the section about providing notification pertains > to "us", the LLVM > community. > > This is grammatically clear and unambiguous so if it's not > the intention, > that needs to be reworded. > > Don't take this the wrong way, but most people would still > understand > that this probably means "don't talk about patents except to the > oversight group". > > > Hi Dan, > > I'm not sure if you realize, but that paragraph reads as an open > invitation to notify and discuss patents on the development and > commits lists. > > > There are roughly no open source projects where the > rule is "talk about patents all you like on random development > mailing > lists" > > > There are plenty of projects around the world the where that is > absolutely the rule. > > Your remarks suggest that you've worked on a limited range of > projects without the perspective it takes to accurately word a > developer policy like this for an international audience. > > "Most people" doesn't cut it here and we need to set out our > expectations explicitly before we start turning away new > contributors and telling them to "hit the road" for something they > said. > > I can't stand by while new contributors receive abuse for some > violation of a rule that's not even written in the LLVM developer > policy. > > We should assume that contributors come from a background that's > varied, inclusive and different to the norms in our immediate > circles, and aim to provide them with accurate and helpful > guidance in the developer policy and reflects our expectations. > Flaming people when they try to engage our community because they > have a different legal system or interpretation. > > What next, turn people away because they have a funny name? Reject > patches because their skin color is different to yours? I'm > disappointed that you've tried to defend what is clearly repeated > and inexcusable behaviour by Chandler towards people who are > graciously trying to help out. In so far as there is a community, > we must stand up distance ourselves from behaviour like that. > > (This is becoming OT for cfe-dev, moving the thread to llvm-dev. > Let's refocus into a more productive mode and and roll a patch > already?) > > > Alp. > > > -- > http://www.nuanti.com > the browser experts > > _______________________________________________ > cfe-dev mailing list > cfe-dev at cs.uiuc.edu <mailto:cfe-dev at cs.uiuc.edu> > http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev > >-- http://www.nuanti.com the browser experts
Alp, On Jan 30, 2014, at 4:08 PM, Alp Toker <alp at nuanti.com> wrote:> What next, turn people away because they have a funny name? Reject patches because their skin color is different to yours? I'm disappointed that you've tried to defend what is clearly repeated and inexcusable behaviour by Chandler towards people who are graciously trying to help out. In so far as there is a community, we must stand up distance ourselves from behaviour like that.I don’t have an opinion in this discussion, but this form of personal attack should be acceptable on the LLVM mailing lists. Please settle your disagreements with each other civilly. —Owen -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140130/019e8e84/attachment.html>
On Jan 30, 2014, at 5:12 PM, Owen Anderson <resistor at mac.com> wrote:> Alp, > > On Jan 30, 2014, at 4:08 PM, Alp Toker <alp at nuanti.com> wrote: > >> What next, turn people away because they have a funny name? Reject patches because their skin color is different to yours? I'm disappointed that you've tried to defend what is clearly repeated and inexcusable behaviour by Chandler towards people who are graciously trying to help out. In so far as there is a community, we must stand up distance ourselves from behaviour like that. > > I don’t have an opinion in this discussion, but this form of personal attack should be acceptable on the LLVM mailing lists. Please settle your disagreements with each other civilly.Erm, should *not* be acceptable. —Owen -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140130/475faa55/attachment.html>
I would like to say, in Alp's defense, that I definitely felt like it was more than just a polite request not to discuss something on the list. Being told to "talk to my lawyer" was not particularly constructive. I appreciate that Chandler was trying to keep the discussion from going any farther, but I also appreciate that Alp was trying to stick up for me and turn it into positive changes for the list and policy documentation. I would also like to mention that until today, I had not read the page that Alp linked to: http://llvm.org/docs/DeveloperPolicy.html. I would suggest putting a link to that on the first page of http://clang.llvm.org/get_involved.html so that when people sign up for the list, they know what is expected of them, which is something I've still had a hard time figuring out. Thanks, JB On Thu, Jan 30, 2014 at 6:12 PM, Owen Anderson <resistor at mac.com> wrote:> Alp, > > On Jan 30, 2014, at 4:08 PM, Alp Toker <alp at nuanti.com> wrote: > > What next, turn people away because they have a funny name? Reject patches > because their skin color is different to yours? I'm disappointed that > you've tried to defend what is clearly repeated and inexcusable behaviour > by Chandler towards people who are graciously trying to help out. In so far > as there is a community, we must stand up distance ourselves from behaviour > like that. > > > I don't have an opinion in this discussion, but this form of personal > attack should be acceptable on the LLVM mailing lists. Please settle your > disagreements with each other civilly. > > --Owen > > _______________________________________________ > cfe-dev mailing list > cfe-dev at cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev > >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140130/061e37b4/attachment.html>
<responding to llvmdev too> On Jan 30, 2014, at 4:08 PM, Alp Toker <alp at nuanti.com> wrote:>> There are roughly no open source projects where the >> rule is "talk about patents all you like on random development mailing >> lists" > > Your remarks suggest that you've worked on a limited range of projects without the perspective it takes to accurately word a developer policy like this for an international audience.I agree, we absolutely need to be more clear about this on the developer policy. The problem with discussion on the mailing list is twofold: 1) for almost everyone (myself included), patent discussion is *speculation*, because they are not qualified patent attorneys. Patent law doesn't make sense, and applying logic to it only makes things worse. 2) when patents get discussed, invariably people get irritated about the state of the patent system and the discussion spirals off in an off-topic and destructive direction. I'd rather that the community stay productive and focused on technical matters, while people who understand this discuss it off-list. It also happens that this specific topic is one that is well understood by people who care a lot about such things.> I can't stand by while new contributors receive abuse for some violation of a rule that's not even written in the LLVM developer policy.Yes, the policy needs to be more clear.> What next, turn people away because they have a funny name? Reject patches because their skin color is different to yours? I'm disappointed that you've tried to defend what is clearly repeated and inexcusable behaviour by Chandler towards people who are graciously trying to help out. In so far as there is a community, we must stand up distance ourselves from behaviour like that.We're all friends here. Harassment will not be tolerated, and I'm sorry if you took the responses about this issue that way. You're absolutely right that the policy should be made crystal clear, and that future instances of this can be handled simply by pointing to the policy. I don't speak for Chandler, but I'm pretty sure that he didn't mean to drive anyone away - he was just being a little over-eager in enforcing the poorly documented policy we have. -Chris