Stas Sergeev <s...@list.ru> writes: > 22.07.2013 10:29, Eric W. Biederman пишет: >> Stas Sergeev <s...@list.ru> writes: >> >>> 21.07.2013 20:09, solarflow99 пишет: >>>> I like gpl 3 too, if gpl3 existed long ago dosemu probably would have >>>> used it. But it would be a shame to drag things down over licensing. >>> I don't think anything can be dragged down this time. >>> Things have changed. Firstly, the discussion now happens >>> in a public ML, instead of an IRC as it was 10 years ago. >>> Secondly, there is now the Git, which allows us to _look_ >>> into the proposed things, changing them quickly etc, while >>> in the past we had to argue without much of a demonstration >>> of our ideas, and the misunderstandings ensued. >>> And thirdly, there is no more permanent disagreement >>> between me and Bart on the code simplicity vs compatibility. :) >>> Simplicity won, compatibility (with ancient linux environments) died. >>> My goal is to simply resolve the misunderstandings that >>> are all there in the COPYING.DOSEMU. This is a very bad >>> place for misunderstandings IMHO. :) >> Hans did get upset that distributions were retarring his tarball and >> making it impossible for people to verify that what he shipped was >> actually what the distributions had, and took an interesting take on the >> GPLv2. > Hi Eric, this is not the main point of that discussion. > Even though my proposal contains also what you are > talking about: > http://sourceforge.net/p/dosemu/code/ci/49870d19fd987a507a03130c4ddf1faccf27ac3b/ > the main crux is this one: > http://sourceforge.net/p/dosemu/code/ci/a1dff2a15698efc4f400812993ec819ea7091439/ > which is an entirely different problem from disallowing the > distributors to do things.
Clause 5 is definitely needed as a clarification. The essence of clause 5 is that if your have a weird lawyer that interprets the GPL this way you are still fine, because we the authors explicitly say it is not a problem. Use of a program is not always allowed. You have to make a copy of a program to use it, and furthermore it can be considered a perforamnce. Both things that depending of your exact copyright law are regulated. FSF does not regulate how the GPL can be used they just give advice so FSF guidelines arguments are bollox. I don't know how old this text is but unless it is recent it should stand. >> Arguably Bart as the current maintainer is free to relax that part of >> the interpretation. > This would be nice too. > We seem to have forgotten about this part of my proposal > during the discussion. > >>> Considering the dire >>> consequences of these misunderstandings, like the "GPLv2 >>> only" instead of "GPLv2 or later", it IMHO really needs to be >>> resolved once and for all. >> The fact that the GPLv3 has more restrictions and is not compatible >> with the GPLv3 I find a royal pain. > You meant GPLv2 in at least one instance above. :) Yes I meant the fact the GPLv3 has more restrictions and is not compatible with the GPLv2 I find a royal pain. > I think it is incompatible intentionally, ie FSF always suggested > to use "GPLv2 or later" to make everything compatible. People > are free to stick to the "only" version of some license, opting > themselves to stay incompatible with every further version. > So IMHO this is not an FSF's fault, but the deliberate choise This is exactly the FSF's fault. The GPLv3 could simply have been a clarification in wording. The FSF produced a GPLv3 that is less permissive and fewer people are willing to adopt it, which in turn reduced the value of the GPL by causing license fragmentation. > of whoever changed "or later" to "only". It is good that FSF > gives such an option to people (it have to), but it is bad that > they use it. No. The FSF has not proved as untrustworthy as some have feared but they have proved they make strange weird choices that are inappropriate for some application domains. The GPLv3 seems to suffer from second system effect pretty badly. I can just about understand the GPLv2 the GPLv3 is lost in insane legalease and weird restrictions. I very much admire the purity of stance of the FSF but it is dangerous to follow the path they chart blindly. >>> Aside from resolving the misunderstanding, of couse an >>> extra steps can be done, like asking people about relicensing >>> etc, but for now just removing the misunderstandings would >>> be already a big achievement. >> For a project as old as dosemu if we can't get out a release that >> makes networking easy I can't see a change in licencing being realistic >> or relevant. > You have probably forgot how long ago the last release > was made. :) There are many improvement even without > the networking. For instance, x86_64 fixes (cpuemu) from > Bart are very important as everyone uses x86_64 these days. > As for networking, we'll get something here for sure. > The question is only will Bart find the time for the release > or not. :) > Still it would be interesing to take a look into VDE or > OpenVSwitch projects before making the decision on what > we need with networking... > Also, maybe I am underestimating the importance of such > a things, but I wonder if it is really a crucial functionality to > have. It is indeed a very nice one, have to be added (and we > work on it in git), but nevertheless TAP already allows you > to do all this and even more. The point was simply that dosemu does not have timely releases, and that in general there is very little energy put into the project at this point. The code is 20+ years old at this point, that is a lot of history and a lot of developers to talk to. If very important things are being forgotten like why there was an explicit runtime linking exception I think you could get into hot water pretty fast. On the flip side making a release that works on x86_64 is probably crucial to keeping things going. Eric ------------------------------------------------------------------------------ See everything from the browser to the database with AppDynamics Get end-to-end visibility with application monitoring from AppDynamics Isolate bottlenecks and diagnose root cause in seconds. Start your free trial of AppDynamics Pro today! http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk _______________________________________________ Dosemu-devel mailing list Dosemu-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dosemu-devel
