On 15 Oct 2003 at 14:31, Phoebus R. Dokos (Φοίβος Ρ. Ντό wrote: (...) > Example 1. > > Case 1. Lets assume that IOSS is modified and drivers from other OSes are > capable to be ported. Lets also assume that these are linux drivers and by > default Free Software. Even if possible to integrate a Linux driver to > SMSQ/E we cannot do so as its license clashes with the SMSQ/E license (as > you mentioned yourself in Marcel's argument of distribution under two > separate licenses). That in itself closes the door to systems software > from other platforms. (Doesn't affect applications software of course)
In other words, you chose your example in such a way that we have two licences conflicting with each other. Why should the SMSQ/E one be the one to be changed? Why can't the other one be changed to suit SMSQ/E? To use a counter example, suppose the licence for the software to be incorporated had a provision that every user had to pay a million US $ every minute? Would we have to accept that, too? > Case 2. Lets take QLwIP that Peter wrote (and works as we can see from his > emails). If we want to give SMSQ/E TCP functionality by somehow > integrating QLwIP into SMSQ/E then we again clash with its license (QLwIP > is a port of A. Dunkels wIP that is Free Software). It is illegal to do so > therefore on the regular version of SMSQ/E. This is the same argument than the one before. You take something that has its own licence and ask that SMSQ/E be changed to accomodate it. > The only way that therefore you could integrate something like that with > the OS would be in the form of patches, something that was the cause of a > big brawl in this list when it came to that "Software Pirates in our > midst" thread if you recall. Just for the record, that is not what that row was about, IIRC. (...) > Example 2. (example itself snipped) Well, I'm sorry but I'm not sure I quite follow your example. Let me try to restate it to see whather I have understood it correctly. Please feel free to correct me if I have mistated it !!!!!! These people may sell a specific version of an OS. The OS is "normal" copyrighted software so it follows that: - These people have no access to the sources. - They may make no changes to the software. Yet they build a machine and an emulator on which this software may run. (I **presume** that the orginal author made the necessary changes). For some reason, the OS on both machines is incompatible. (again, I ** presume** that this is due to changes made by the original author, else I fail to see how there could be incompatibilities). The sources are then made open with the current SMSQ/E licence. Both of these people "patch" the software (apparently with the purpose of making it compatible to each other) whilst not having nor wishing (?) access to the source code which is now freely available. Question: is this legal? I'll take this example as stated without asking myself why these people, now having access to the source code, don't use it and, apparently, don't decide to talk to each other (or do they?). I'll also presume that they did not approach, for whatever reason, the original author, or whoever maintains the source code, to have their "patches" incorporated into the software. I'll further presume that the "patch" is some kind of software that adds to/replaces some parts of the original software, by changing the binaries as distributed. FInally, I'll also answer just from the point of view of the licence. Indeed, I don't want to get into the legalities of "Patching" software even though I understand that normally, you aren't allowed to change copyrighted software, and the argument may certainly be made that by "patching" it you are making such a change. There are, in my mind, three questions here: 1 - May you distribute the "patch"? 2 - May you distribute "patched" versions of the binaries? 3 - What about an agent? 1- For the first question, in short, then, the answer would be that the licence doesn't care. Teh licence, as it stands now, cares about - distribution and change of the source code - distribution of the binaries So, if you later make a patch which you supply separately, this isn't a problem with the licence. As mentioned above,the author may protest against patches being made, but that would be under general copyright law. 2- As to the second question: - What you are not allowed to do is distribute the binaries (with or without the patch) unless you are a reseller, so the point whether you may distribute a patched version is, in itself, moot unless you are a reseller (with one proviso, see below) - If you are a reseller, then you may distribute the official version. You may not distribute a patched version since that is not the official version. Proviso: If you had prior permission from the author to distribute an earlier (patched) version, and if that permission has not been withdrawn, you may still continue to distribute that. - You may distribute the source code under the usual provisions of the licence. 3- And finally, "agents". Basically, if you are entitled to appoint an agent, then that agent cannot have more rights that you do. Thus to the agent, the same as above applies. I hope this replied to your exact example. (of course, any resemblance to an actual situation would be a pure coincidence). > > You can see how confusing the status can be just by that. Not to me, if I understood your example correctly. > > Thirdly. here is why I think that internet distribution is necessary. > > Point 1. Linux (the DEFINITION of Free Software) would not be possible > without the Internet > Point 2. No OTHER Free Software, nor the GNU project would be possible > without the Internet Ah, so since Linux and other free software wouldn't have been posssible without the Internet, it follwos that SMSQ/E must be distributed on the Internet, too? > Point 3. Internet distribution of the sources can lure MC68K hobbyists to > the platform whereas Wolfgang Lenerz known and respected thoughout the QL > world but most probably an unknown entity to say an Atari ST user or a > simple coldfire board manufacturer/hobbyist will not (Unless of course he > changes his name to Linus Torvalds ;-) hehe just a joke here) I'll make an effort and call myself Wolfgangus Lenvalds, if that would help... > Point 4. Survival of a platform is based on adoption. If you throw a > really Free SMSQ/E on say Savannah or SourceForge you will get at least 10 > or 20 people to have a look-see. If you win half (sic!) of them you gained > a potential whole new tree of QL users (and developers). This is the same point as point 3 to my mind. It has been debated here before, and again, I just fail to see that happen. A (say) Linux aficionado with no prior QL connection would take up SMSQ/E? I dare say that you wouldn't win half of them - just none. (And, as an aside, I have had requests for the source code from people I have never heard of before, nor after...) > Point 5. A publicly available CVS server gives any project the "air" of > maintenance. A snail-mail only distribution does not (That of course > applies to a stagnant webpage with a couple of files to download but still > the exposure is greater). This "air" is attractive to many people. I could joke about a lot of hot "air" here. Seriously though, how serious can people be if they are just influenced by the air of something - what will they bring to us? Wolfgang
