From strenholme.usenet at gmail.com Fri Apr 15 14:52:14 2011 From: strenholme.usenet at gmail.com (Sam Trenholme) Date: Fri, 15 Apr 2011 13:52:14 -0500 Subject: New MaraDNS and Deadwood snapshots Message-ID: In today's snapshot of MaraDNS, I have fixed an error reported on this list by Bryan Duff [1]. In particular, instead of replying with "not implemented" when given an IXFR or AXFR query over UDP, MaraDNS now returns "truncated". In today's snapshot of Deadwood, I have refined the heuristics of how to handle empty packets by only treating them as "this host does not exist" if the user asked for an AAAA (IPv6 IP) record. This resolves a problem Deadwood has resolving the archive.org names when one of the three upstream DNS servers replied to requests with a REFUSED DNS reply. They can be downloaded here: http://www.maradns.org/download/2.0/snap/ http://www.maradns.org/deadwood/snap/ The next day I will work on MaraDNS is two weeks from today: Friday, April 29 2011. - Sam [1] http://woodlane.webconquest.com/pipermail/list/2011-March/000807.html From wayne.kroncke at tiscali.co.uk Sat Apr 16 01:59:26 2011 From: wayne.kroncke at tiscali.co.uk (wayne at tiscali) Date: Sat, 16 Apr 2011 06:59:26 +0100 Subject: New MaraDNS and Deadwood snapshots In-Reply-To: References: Message-ID: <4DA9303E.5050607@tiscali.co.uk> thanks for continuing on this project, sam. hope we get an update to deadwood windows build at some point. Best Regards, Wayne Kroncke On 15 Apr 2011 19:52, Sam Trenholme wrote: > ... > > In today's snapshot of Deadwood, I have refined the heuristics of how > to handle empty packets by only treating them as "this host does not > exist" if the user asked for an AAAA (IPv6 IP) record. This resolves a > problem Deadwood has resolving the archive.org names when one of the > three upstream DNS servers replied to requests with a REFUSED DNS > reply. > ... > The next day I will work on MaraDNS is two weeks from today: Friday, > April 29 2011. > > - Sam > > [1] http://woodlane.webconquest.com/pipermail/list/2011-March/000807.html From strenholme.usenet at gmail.com Sat Apr 16 09:20:25 2011 From: strenholme.usenet at gmail.com (Sam Trenholme) Date: Sat, 16 Apr 2011 08:20:25 -0500 Subject: New MaraDNS and Deadwood snapshots In-Reply-To: <4DA9303E.5050607@tiscali.co.uk> References: <4DA9303E.5050607@tiscali.co.uk> Message-ID: > hope we get an update to deadwood windows build at some point. http://maradns.org/blog Which shows this blog entry: http://samiam.org/blog/20110218c.html - Sam From sebastiano at datafaber.net Thu Apr 28 18:16:49 2011 From: sebastiano at datafaber.net (Sebastiano Pilla) Date: Fri, 29 Apr 2011 00:16:49 +0200 Subject: [MaraDNS list] Deadwood cache file is not created Message-ID: <4DB9E751.2000409@datafaber.net> I'm having an issue with Deadwood on my system, I would like to use the persistent cache feature but the cache file is not created. When starting Deadwood, I see the following messages in the log: Apr 29 00:01:08 echo /opt/maradns2.0.02/sbin/deadwood: Deadwood version 3.0.02 Apr 29 00:01:08 echo /opt/maradns2.0.02/sbin/deadwood: Deadwood: A DNS UDP non-recursive cache (IPv4-only) Apr 29 00:01:08 echo /opt/maradns2.0.02/sbin/deadwood: We bound to 1 addresses Apr 29 00:01:08 echo /opt/maradns2.0.02/sbin/deadwood: Could not open hash at cache/deadwood_cache for reading I can understand this message if this was the very first run of Deadwood on the box, but I get it on all subsequent runs when it should have created a cache file. The /etc/dwood3rc file is: bind_address="127.0.0.1" cache_file="cache/deadwood_cache" chroot_dir="/var/run/deadwood" dns_port=53 filter_rfc1918=1 maximum_cache_elements=1024 maradns_uid=99 maradns_gid=99 recursive_acl="127.0.0.1/8" reject_mx=0 root_servers={} root_servers["."]="198.41.0.4," root_servers["."]+="192.228.79.201," root_servers["."]+="192.33.4.12," root_servers["."]+="128.8.10.90," root_servers["."]+="192.203.230.10," root_servers["."]+="192.5.5.241," root_servers["."]+="192.112.36.4," root_servers["."]+="128.63.2.53," root_servers["."]+="192.36.148.17," root_servers["."]+="192.58.128.30," root_servers["."]+="193.0.14.129," root_servers["."]+="199.7.83.42," root_servers["."]+="202.12.27.33" tcp_listen=1 timeout_seconds=30 timeout_seconds_tcp=30 upstream_servers={} upstream_servers["."]="8.8.8.8, 8.8.4.4" verbose_level=3 The directories exist, ownership and permissions look correct to me: [root at echo run]# ll /var/run | grep deadwood drwxr-xr-x 3 root root 4.0K Apr 29 00:00 deadwood [root at echo run]# ll /var/run/deadwood/ total 4.0K drwxr-xr-x 2 nobody nobody 4.0K Apr 29 00:00 cache [root at echo run]# ll /var/run/deadwood/cache/ total 0 I would expect to find a file /var/run/deadwood/cache/deadwood_cache, but Deadwood does not create this file. This is a CentOS 5.6 x86 box, though I have the same issue on CentOS 5.6 x64. The user and group "nobody" exist and they have uid and gid as 99. I'm perhaps missing something in the configuration, or the permissions on the directories aren't exactly what they should be... In any case I'm not seeing what should I do to fix the problem. Best regards Sebastiano Pilla From strenholme.usenet at gmail.com Fri Apr 29 10:14:56 2011 From: strenholme.usenet at gmail.com (Sam Trenholme) Date: Fri, 29 Apr 2011 09:14:56 -0500 Subject: [MaraDNS list] Deadwood cache file is not created In-Reply-To: <4DB9E751.2000409@datafaber.net> References: <4DB9E751.2000409@datafaber.net> Message-ID: It's your lucky day--today is the one day every two weeks when I babysit MaraDNS (answer email, fix bugs, etc.) > I would like to use the > persistent cache feature but the cache file is not created. I have verified that I can create a cache file with the configuration you shared on the list in Scientific Linux 5 [1] (as an aside, have you had any issues using Deadwood as a fully recursive client instead of piggybacking off of Google's DNS servers?) So, some things to check: * Is Deadwood being killed with a TERM signal and not a KILL signal (kill {pid of Deadwood} instead of kill -9 {pid of Deadwood})? * Does your hard disk have enough free space to write the cache file (remember, root is given more hard disk space than users, so root can still write files when other users can not)? * Does your system have the appropriate files in /etc/init.d and /etc/rc.d to ensure Deadwood is properly stopped when the system is rebooted? - Sam [1] http://samiam.org/blog/20110324.html From sebastiano at datafaber.net Fri Apr 29 10:37:02 2011 From: sebastiano at datafaber.net (Sebastiano Pilla) Date: Fri, 29 Apr 2011 16:37:02 +0200 Subject: [MaraDNS list] Deadwood cache file is not created In-Reply-To: References: <4DB9E751.2000409@datafaber.net> Message-ID: <4DBACD0E.3060404@datafaber.net> Sam Trenholme wrote: > It's your lucky day--today is the one day every two weeks when I > babysit MaraDNS (answer email, fix bugs, etc.) > >> I would like to use the >> persistent cache feature but the cache file is not created. > > I have verified that I can create a cache file with the configuration > you shared on the list in Scientific Linux 5 [1] (as an aside, have > you had any issues using Deadwood as a fully recursive client instead > of piggybacking off of Google's DNS servers?) I haven't tried it, but now that you mention it I will set up Deadwood as a fully recursive resolver. I'll give it a week to see how things go and I'll report back to the list. > > So, some things to check: > > * Is Deadwood being killed with a TERM signal and not a KILL signal > (kill {pid of Deadwood} instead of kill -9 {pid of Deadwood})? That was it! I had a 'kill -9' in my shutdown script, as soon as I replaced it with a simple 'kill' and I stopped the daemon I saw the cache file created in the configured directory. Thank you very much for your help. Sebastiano Pilla From Bradley at NorthTech.US Fri Apr 29 18:45:21 2011 From: Bradley at NorthTech.US (Bradley D. Thornton) Date: Fri, 29 Apr 2011 15:45:21 -0700 Subject: [MaraDNS list] AlphaDNS... Message-ID: <4DBB3F81.8030800@NorthTech.US> -----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 I was fumbling through things, and stumbled over what was claimed to be a 'fork' of MaraDNS: http://alfadns.org/wiki/A_quick-start_guide_for_experienced_UNIX_and_Linux_admins I don't see anywhere that Jefsey has forked anything at all, yet it's not too bad as far as a third party HowTo goes. I've been using your product Sam, since it first came out (almost, anyway), and have been pleased with the performance and simplicity for many years. As an amusing note, what Jefsey doesn't realize, is that when he contracted me to build and operate his short-lived rootsystem years ago, I used MaraDNS for the hidden master he slaved off of, fed from my PacificRoot zonefile. Now it has come full circle and he's espousing the benefits of using MaraDNS like it's a fresh discovery. Well Sam, it's good to hear that you're back in the states, and I certainly hope you pick up development (and not just maintenance) of MaraDNS when the need arises. For now everything is just dandy, but to think that you won't EVER resume active development as future technological demands require it would sadden me. I'm glad there's an loldns, Bernstein painted himself into a corner as far as I'm concerned and letting it go PD is what he should have done much sooner, but even so, I personally prefer MaraDNS for my own zones, and use named more for slaving zones served by MaraDNS than the other way around most of the time. Kindest regards, - -- Bradley D. Thornton Manager Network Services NorthTech Computer TEL: +1.760.666.2703 (US) TEL: +44.203.318.2755 (UK) http://NorthTech.US -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Find this cert at x-hkp://pool.sks-keyservers.net iQEcBAEBAwAGBQJNuz+BAAoJEE1wgkIhr9j3mMsH/RG8W5HV1NMAg8bGL92ye0Lq aPI0gJ7XFW8jTbUS1jkPJFLuIFrvj72B350SkiEyqG6uTYc2Wx3YR0IRZiHdg/BT qLKISQg9R11Pj3WjxsapS8fEuo/Nm4Iyjz5GhL7PSGmMq+3spN3X7E5XRrgsaGuh odO6bqlFU0EhxxiDLUTknHvpbPJWfBdnBy22NAMutMOiuANig6SnCZTLHwjc8mS0 4DXRNuOTXhruep5fIjScatIjcM0xBpjaJIqAtRfo8NNwMUIYXm06zIfpfU2ARgd0 NUupL0u+l5KINlxvOYgdR7id7sQMwaKaSaqLx9cO9osYcvxWG8xsRPzbxiKghHs= =Fci7 -----END PGP SIGNATURE----- From rick at linuxmafia.com Fri Apr 29 19:06:35 2011 From: rick at linuxmafia.com (Rick Moen) Date: Fri, 29 Apr 2011 16:06:35 -0700 Subject: [MaraDNS list] AlphaDNS... In-Reply-To: <4DBB3F81.8030800@NorthTech.US> References: <4DBB3F81.8030800@NorthTech.US> Message-ID: <20110429230635.GZ16041@linuxmafia.com> Quoting Bradley D. Thornton (Bradley at NorthTech.US): > I was fumbling through things, and stumbled over what was claimed to > be a 'fork' of MaraDNS: > http://alfadns.org/wiki/A_quick-start_guide_for_experienced_UNIX_and_Linux_admins > I don't see anywhere that Jefsey has forked anything at all, yet it's > not too bad as far as a third party HowTo goes. It seems that Jefsey aims to serve a specialised niche, ML-DNS, which expands to either multilayered DNS, and involves DNS being more closely integrated into 'extended naming functions' at higher layers in the software stack. At a guess, he's still in the design phase. > I'm glad there's an loldns, Bernstein painted himself into a corner as > far as I'm concerned and letting it go PD is what he should have done > much sooner, but even so, I personally prefer MaraDNS for my own zones, > and use named more for slaving zones served by MaraDNS than the other > way around most of the time. I will not dwell here on the legal pitfalls of PD declarations, as that's far beyond the scope of this mailing list, but I've covered that here: http://linuxmafia.com/faq/Licensing_and_Law/public-domain.html For those people who consider even an unproblematic one-line licence like 'Copyright (C) 2011 Owner Name. Do whatever you want with this work.' too lengthy, loldns might not appeal, as Joshua Small uses for it a quirky permissive licence of his devising. However, Mark Johnson's fork 'zinq-djbdns' and Gerrit Pape's fork Debian djbdns/dbndns are both professed to be public domain. -- Rick Moen The Doctor: "I feel disoriented." rick at linuxmafia.com Sarah Jane Smith: "This is the disorientation centre." McQ! (4x80) The Doctor: "That makes sense." From Bradley at NorthTech.US Sat Apr 30 07:59:21 2011 From: Bradley at NorthTech.US (Bradley D. Thornton) Date: Sat, 30 Apr 2011 04:59:21 -0700 Subject: [MaraDNS list] AlphaDNS... In-Reply-To: <20110429230635.GZ16041@linuxmafia.com> References: <4DBB3F81.8030800@NorthTech.US> <20110429230635.GZ16041@linuxmafia.com> Message-ID: <4DBBF999.6000804@NorthTech.US> -----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 Wow, That was a good read Rick, and at the risk of being flamed for going off-topic on Sam's list, I've got some points you might be able to address for me :) 1.) This was pretty much the nail in the coffin that turned my stomach (and thinking) around on the GPL: http://www.topology.org/linux/gpl.html I had been vasilating over dumping my support for the FSF's [L]GPL(s) for some time already, and although I benefit from it, and believe in free software, my concerns with it have centered around the psychological warfare and bullying that RMS is legendary for. Specifically, as it was related by Ulrich Drepper of glibc notability. I've created a 'few' things that were more non-trivial than other things that I dutifully glued the GPL to, and although these larger projects are technically still at this time under the GPL, it's a rather moot point since no one has ever seen them but me. The client software for these applications simply accesses the API, and those clients were GPL'd, yet my backends were proprietary as far as I was concerned, yet if I wanted to release them, I considered that the GPL would have been the mechanism I used for licensing. Now, however, the GPL merely turns my stomach. Don't get me wrong, and I've already stated above that I have reaped many benefits from the GPL, and I seek out and use GPL'd software too, but to release any of my works under this license nowadays is simply incomprehensible. And it's not Stallman's bullying or at times vitriolic disposition either that sways me away from it, it is simply the notion that the GPL v7 could effectively close source any and all GPL'd software that kept that little line, "or any later verion of the GPL" (paraphrased, of course. It's not the point that I could simply remove that line. The point is that the authority for the license is now ceded to Stallman. Maybe that's a good thing, because you inadvertently imply that regardless, there may be cupability - why else would our justice system prohibit the divestiture of ownership? I'm not saying that RMS would radically change a future GPL to anything so atrocious as, "This work was once free but no more sucker" - but he (or someone else leading the FSF) could, and licensing a work under the auspices of perpetuity to an organization I have no say so over is rather odd, to me anyway, in this day and age. I don't even really believe that that will happen - but again, it could. 2.) You've really opened up my eyes wrt the notion of 'public domain' and toward the end of that page (I didn't follow any of the links but read your entire page) Bernstein shows once again that he should stick with code and not engage in shadetree lawyering - he puts a lot of words in peoples mouths (Maybe he would have made a good atty). 3.) Here's what I'm leaning towards now, as I note that the original (4 clause?) BSD licenses are not considered free by the FSF: http://www.topology.org/LICENCE.ak or http://webknotes.sourceforge.net/cgi-bin/wkn/browse.cgi?theme=happyfree&LICENSE Going back to topology.org's rant, I also found this notion a bit disturbing regarding the BSD license: A disadvantage of this BSD license is that the modifier of this code is: 1. permitted to modify the code and still claim it is the same software [they say "with or without modification"!]; 2. not required to provide a copy of the unaltered original when distributing the altered version; 3. permitted to distribute binary without source, even if the binary is from modified source. This means that the author could be blamed for other people's modifications. You don't implicitly raise the issue of culpability on the page you supplied the link for, but like I said before, I felt a loose implication of such on your part. I'm really interested, especially since you don't even mention it, what your 'take' and/or interpretation of the Artistic license is. Basically, I'm a selfish human, and although I want to offer my works (most of them) to others to use (for the most part) as they see fit, There's a couple of cannons I want them to adhere to. A.) I want my name on it. no matter where the software ends up. I think it was an earlier BSD license that was cited somewhere as ridiculous because there were like 40 copyright holders as it was passed down through the ages and improved upon. That in and of itself doesn't bother me, but if there were five hundred copyright holders in the list... well, that's a little silly. B.) I want it to be clear that you can change the code, do with what thou wilt, as long as my original, unmodified code is included - like the GPL in that respect. C.) I want it clear that I don't warrant it for any particular merchantability or what-ev, and if it's useful, kewl, if not, that's kewl to, and if it blows up on your customers ERP system it's on you, not me, and if you otherwise break it, you get to keep both pieces. D.) I want it clear that you can't do what the warning against BSD above suggests, meaning that you can't say this is my program when you changed the code (my code has to be included with any redistribution of the modified code). and E.) I don't even care if mACROsFOT integrates it into something they're going to charge zillions of dollars to license - but they have to include my original source (free source) along with their borgware. um... Aside from the vane bit about "I want my name on it forever!", I think that most developers in the FOSS world seek something quite similar - we just don't interpret the licenses the same way. Of course, I'm not going to call my license a "Modified Artistic License". I'll call it "The Bradley License", or "The Watson License" after my dad, or, "The Alveda License", after my mom. I guess the selfishness on my part is just that I want the nod somewhere down the line when someone says, "Who was the dude who originally came up with this kewl app that I'm using?". That's all, just a feather in my cap, a lack of culpability, a prohibition of plagiarism, and an assurance that (legally, at least), you'll always be able to get to the source I originally wrote and that you can have it for free (not withstanding charging someone for the favor of putting it on an 8" floppy for them, of course). I take this legal stuff in small chunks, but think about it a lot. I'm a geek, and as such this copyright mumbo jumbo has a tendency to make my eyes glaze over. If you don't dignify this post with a response, know this: That at the very least you have cleared up some major mis-conceptions on my part about "Public Domain", and the fact that one cannot divest a work of the notion of ownership (since even if they send off the message in a bottle, an offspring of theirs can claim rightful ownership of it later). So for all practical purposes - There's no such thing as public domain AFAIC from here on out. Thanks again Rick, that was a real eye opener :) Copyright 2011, Bradley D. Thornton - Do with this email as thou wilt and in the spirit of Bill and Ted, "Be Excellent to each other". On 04/29/2011 04:06 PM, Rick Moen wrote: > Quoting Bradley D. Thornton (Bradley at NorthTech.US): > >> I was fumbling through things, and stumbled over what was claimed to >> be a 'fork' of MaraDNS: >> http://alfadns.org/wiki/A_quick-start_guide_for_experienced_UNIX_and_Linux_admins >> I don't see anywhere that Jefsey has forked anything at all, yet it's >> not too bad as far as a third party HowTo goes. > > It seems that Jefsey aims to serve a specialised niche, ML-DNS, which > expands to either multilayered DNS, and involves DNS being more closely > integrated into 'extended naming functions' at higher layers in the > software stack. At a guess, he's still in the design phase. > >> I'm glad there's an loldns, Bernstein painted himself into a corner as >> far as I'm concerned and letting it go PD is what he should have done >> much sooner, but even so, I personally prefer MaraDNS for my own zones, >> and use named more for slaving zones served by MaraDNS than the other >> way around most of the time. > > I will not dwell here on the legal pitfalls of PD declarations, as > that's far beyond the scope of this mailing list, but I've covered that > here: http://linuxmafia.com/faq/Licensing_and_Law/public-domain.html > > For those people who consider even an unproblematic one-line licence like > 'Copyright (C) 2011 Owner Name. Do whatever you want with this work.' > too lengthy, loldns might not appeal, as Joshua Small uses for it a > quirky permissive licence of his devising. However, Mark Johnson's fork > 'zinq-djbdns' and Gerrit Pape's fork Debian djbdns/dbndns are both > professed to be public domain. > - -- Bradley D. Thornton Manager Network Services NorthTech Computer TEL: +1.760.666.2703 (US) TEL: +44.203.318.2755 (UK) http://NorthTech.US -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Find this cert at x-hkp://pool.sks-keyservers.net iQEcBAEBAwAGBQJNu/mZAAoJEE1wgkIhr9j3UvAIAJ2BvZZlx63Nkm2W3HKW5UPT VkLuoChvxLNFYuN3jKF0FyvQ8+3Ixuv1vbKutfhHuUrcYwjA7BRPh74JFEaIc1Kr ocvuUGfqJHc9JTUPbBVUTsouqA8QsAiJKYRh0tyxL/TTTSMY19Qa0mJ8udoMtX7c SdbLLIP1O+ceBG95sDQAVacckfFlGPCmUbVNxOwauDYXwXCgBSc4XolWir9GqLz2 7NvnDhwT3dAtf7e0F4ndttnkfI3rywBANwmeKeYakXSK4DWLsLEeSZFZ9t1mEgWN diQ1RXvmeS2/my4KcxTg0tUDHDQsxg0d7h8JiNwHMqgkd3fc3AKECrQo0hQS2XM= =+1kI -----END PGP SIGNATURE----- From strenholme.usenet at gmail.com Sat Apr 30 11:37:29 2011 From: strenholme.usenet at gmail.com (Sam Trenholme) Date: Sat, 30 Apr 2011 10:37:29 -0500 Subject: [MaraDNS list] AlphaDNS... In-Reply-To: <4DBBF999.6000804@NorthTech.US> References: <4DBB3F81.8030800@NorthTech.US> <20110429230635.GZ16041@linuxmafia.com> <4DBBF999.6000804@NorthTech.US> Message-ID: Some thoughts on this thread: * Developing MaraDNS again. I will never say never because I don't know how I will feel when I wake up tomorrow or the next day. Based on what I feel right now, this will only happen if either 1) Someone pays me to develop MaraDNS 2) I hit gold and retire young. The reason for this is because there just isn't any money in it for me: no direct money because it's, well, Free software; little "resume-ware" money because plain C without C++'s classes isn't a very marketable skill any more. [1] That said, I have some good news for the MaraDNS community: I have changed employers [2] and no longer have a conflict of interest receiving money for MaraDNS. I will probably have donations set up again, but on a strictly "tip jar" basis. * GPL vs. BSD vs. Public domain: I originally made MaraDNS public domain; I changed this to a simple 2-clause BSD license because of concerns people had with the legality of making it public domain. The appeal of the GPL is because a company can release some GPL code and have their business model be "If you don't want this to be GPL, pay us". * On Microsoft: When I was young and idealistic, I had a real "Micro$oft sucks" attitude; I only used Linux on the desktop. Today, I regret thinking that way. I have seen people proclaim the year of the Linux desktop for years and it has never happened and never will happen. [3] Not to get completely off topic, but configuring things in Linux is just too much of a pain in the butt for me. For example, in Ubuntu, the touchpad on this S10-3 has this annoying "click whenever I stop moving my finger on the touchpad" behavior. Fixing it in Linux would require Google searches and editing text files. This isn't an issue in Windows since it comes with a driver for the touchpad; if it didn't, fixing it is a simple matter of downloading the driver then "Next", "I agree", "Next", "Install", "OK". If I'm going to do Google searches and learning how to edit text files, it better be something that can pay me, such as learning Javascript and jQuery. Back to MaraDNS: Deadwood is still fairly new and I am definitely interested in knowing about domains that don't resolve with Deadwood but do resolve with other DNS servers. - Sam [1] Unless one is writing Linux kernel modules, which MaraDNS isn't. [2] Since you asked: I needed to be able to spend more time in Mexico with my wife until she has permission to come with me to the US. [3] Linux's chances of becoming a desktop operating system died in 2002: http://set.tj/+kd6a From rick at linuxmafia.com Sat Apr 30 12:30:22 2011 From: rick at linuxmafia.com (Rick Moen) Date: Sat, 30 Apr 2011 09:30:22 -0700 Subject: [MaraDNS list] AlphaDNS... In-Reply-To: <4DBBF999.6000804@NorthTech.US> References: <4DBB3F81.8030800@NorthTech.US> <20110429230635.GZ16041@linuxmafia.com> <4DBBF999.6000804@NorthTech.US> Message-ID: <20110430163021.GI16041@linuxmafia.com> Quoting Bradley D. Thornton (Bradley at NorthTech.US): > That was a good read, Rick. Glad to help. > at the risk of being flamed for going off-topic on Sam's list, I've > got some points you might be able to address for me :) We can't continue much on the topic, as this has nothing to do with MaraDNS. Also, I steer entirely away from licence advocacy, anyway: Each coder is entitled to decide whatever terms he/she will use for works, and logically that should be whatever terms best suit the coder's needs. My comments tend to be attempts to explain law and licence mechanics, in cases where they are unclear. > 1.) This was pretty much the nail in the coffin that turned my stomach > (and thinking) around on the GPL: http://www.topology.org/linux/gpl.html Licence advocacy. ;-> > [...] it is simply the notion that the GPL v7 could effectively close > source any and all GPL'd software that kept that little line, "or any > later verion of the GPL" (paraphrased, of course. That is actually not the case. Think about the licence mechanics: Any attempt to introduce a restrictive v7 licence would immediately lead to persistent (and dominant) forks under the earlier, less restrictive version. > 3.) Here's what I'm leaning towards now, as I note that the original (4 > clause?) BSD licenses are not considered free by the FSF: If you don't mind another mild correction, FSF never asserted any BSD licence variant was non-free. They asserted that the 'advbertising clause' in the original U.C. Regents licence text produced long-term practical effects they described as 'obnoxious'. For example, all commercial advertising literature for BSDi's BSD OS (if memory serves) had to include a huge list in very small type of all contributor names. [snip two modifications of Artistic License 1.0] Hmm. Larry's (Larry Wall's) Artistic License 1.0 had some very serious legal problems on account of vague wording, leading (among other things) to adverse U.S. Court of Appeals decision Jacobsen v. Katzer. Artistic License 2.0 fixed those problems. [things most important to you:] > A.) I want my name on it. no matter where the software ends up. Retaining copyright notices (in the _source code_ of any redistribution or derivative works) is a hard requirement of copyright law itself. You don't even need any particular licence for that. On the other hand, if you also want to require that your name be displayed at runtime, or in product advertising for any commercial products based on the work, that's a different matter. The latter was what the 4-clause BSD licence's advertising clause required. The former? Hmm, well some Web 2.0 firms have been using a class of licences derisively described as 'badgeware' licensing, but wary observers consider them a bit sleazy and problematic in various ways. We shouldn't get into that. [snip B, C, D, and E] The major licence best suited for your requirements is probably Artistic License 2.0. http://www.perlfoundation.org/artistic_license_2_0 > Of course, I'm not going to call my license a "Modified Artistic > License". I'll call it "The Bradley License", or "The Watson License" > after my dad, or, "The Alveda License", after my mom. Your decision, but, in general, adopting a one-off licence is usually a bad idea for adoption of your software by other coders and possibly even by users: Between two codebases, one with a familiar and acceptable licence and one with an unfamiliar licence, the former has inherent appeal. If you want to learn more about licensing, I can recommend this one by OSI's founding chief counsel, Larry Rosen: http://www.rosenlaw.com/oslbook.htm From jefsey at jefsey.com Sat Apr 30 22:29:14 2011 From: jefsey at jefsey.com (JFC Morfin) Date: Sun, 01 May 2011 04:29:14 +0200 Subject: [MaraDNS list] AlfaDNS... In-Reply-To: <20110429230635.GZ16041@linuxmafia.com> References: <4DBB3F81.8030800@NorthTech.US> <20110429230635.GZ16041@linuxmafia.com> Message-ID: <7.0.1.0.2.20110501033631.0623a8a0@jefsey.com> Hi! Bradley, nice to get you on line at a right place. Time flows. Yes it took all that time to force my vision of diversity support into the IETF (swallowed but very far from being digested). At 01:06 30/04/2011, Rick Moen wrote: >It seems that Jefsey aims to serve a specialised niche, ML-DNS, which >expands to either multilayered DNS, and involves DNS being more closely >integrated into 'extended naming functions' at higher layers in the >software stack. At a guess, he's still in the design phase. Yes. There are many issues involved as ML-DNS is to be building a top of IDNA2008 (RFC 5890-5895) we consensually approved with it in mind. I had to clarify with IAB if this was IETF scope or not, and since it is not (RFC 6050 is the IETF current border) how to liaise (through the http://iucg.org/wiki iucg at ietf.org low traffic mailing list) and how to organize myself for the years to come. alfaDNS is also a way to explore an ML-DNS prototype, including external development like IDNA2008 libraries that are on the verge to be released. Also, I investigate the proper architecture and language for the port. I definitely agree with Sam about Linux. My background for network development years ago was Tymnet and then QNX which unfortunately is commercial. My area is the IUI (Intelligent Use Interface between the user's plug and the digital ecosystem systems), so I have the experience that I may not necessarily be tied to Windows or linux. Last but not least I was tied by personal issues for the last two years what limited my working capacity. Slow but determined. I am glad Sam is able to spend sponsored time again. jfc From strenholme.usenet at gmail.com Sat Apr 30 22:39:35 2011 From: strenholme.usenet at gmail.com (Sam Trenholme) Date: Sat, 30 Apr 2011 21:39:35 -0500 Subject: [MaraDNS list] AlfaDNS... In-Reply-To: <7.0.1.0.2.20110501033631.0623a8a0@jefsey.com> References: <4DBB3F81.8030800@NorthTech.US> <20110429230635.GZ16041@linuxmafia.com> <7.0.1.0.2.20110501033631.0623a8a0@jefsey.com> Message-ID: > I am glad Sam is able to spend sponsored time again. To clarify: Right now, I do accept donations (To donate: Send money via PayPal to this email address). But, it is strictly a "Tip jar" donation--a token of appreciation for my hard work with MaraDNS. I am currently not offering "money for code" nor "money for support" with MaraDNS; should this change, I will let people on the list know. - Sam