mark and tim, just a few errors!!
these suggestions and corrections are outstanding. i am sorry in our rush to hit the short himss publishing deadline we did not circulate the document to the openhealth list. we'll create a revised draft and circulate it back to the list. with best regards, [wr] - - - - - - - - On Feb 23, 2007, at 2:26 AM, Tim Churches wrote: > Mark Preston wrote: >> Will Ross wrote: >>> The following link will send you to a four page introduction to Open >>> Source Software published by the HIMSS Open Source Work Group. >>> >>> http://www.himss.org/ASP/topics_FocusDynamic.asp?faid=190 >>> >>> Principal authors of this document, in alphabetical order, are: >>> >>> Neil Cowles >>> Guy Fisher >>> Nilish Gupte >>> Drew Ivan >>> Will Ross >>> >>> Corrections and suggested improvements are welcome and may be >>> sent to >>> my attention for the next edition. > >> Regarding the GPL your article states inter alia: >> "The fourth freedom is also a restriction: by changing the source >> code, >> the licensee agrees to release the changed code under the same free >> software license. In other words, the results must also remain as >> open >> source software, allowing the whole community to benefit from all >> improvements." >> >> This is incorrect. There is no compulsion to redistribute changes >> as far >> as I'm aware. If you decide to publicly redistribute changes then you >> must observe the fourth "freedom", but otherwise you can keep your >> changes to yourself. > > Yes, that is absolutely correct. If you don't distribute modified code > or derivative works outside "your organisation", then the GPL does not > require that the modified code or derivative works be shared or > otherwise made available to others. Thus it is perfectly OK to, say, > link or combine GPLed code with, say, proprietary, closed-source code, > provided the resulting derivative or modified work is only used inside > your organisation. > > Hmm, I'd better correct a few more errors (just a few...): > > P1, Para 1: source code is not always human-readable (eg any perl > code...just joking!), but teh wikipedia definition is slightly better > than yours: "Source code (commonly just source or code) is any > sequence > of statements and/or declarations written in some human-readable > computer programming language." Also, don't conflate "commercial > software" with "closed-source". Free, open source software can be > commercial, and closed source software can be free or non-commercial. > the use of the term "proprietary" is best avoided as well (even > though I > use it below), because "proprietary" means something is owned, and > open > source software (except public domain software) is absolutely owned by > its copyright holders. > > P1, Para 4: Lack of previous experience with OSS is identified as a > risk. Surely it is more a lack of technical expertise in the > particular > technology that is the risk, and that risk is independent of the > licensing arrangements of that technology? Also, capitalisation has > little to do with indemnification (against what? patent > challenges?) or > liability risks (risks of what? software failure or flaws?). Also, > open > source software is no more likely to infringe patents etc than closed > source software (see the $1.2 billion Microsoft has just been > ordered to > pay for MP3 patent infringements). In fact, it is less likely to > infringe patents because such infringements are spotted and usually > removed before damages can be claimed by the patent holders. > > p2 "All open source licenses stipulate that the source code is > available > for inspection and reuse." Better to say "Many open source licenses > stipulate that if the software is distributed or shared with others, > then the source code must also be shared or otherwise made > available for > inspection and reuse." I say "many" because popular licenses like the > MIT and BSD licenses don't stipulate anything about source code being > made available - see http://www.opensource.org/licenses/mit- > license.php. > It is really only the copyleft licenses that do that. > > That means that the next paragraph is also wrong - permissive open > source licenses don't require that source code be made available. > > However, you should note that the OSI open source definition doesn't > refer to "open source licenses", it refers to "open source software". > Thus, the MIT and BSD licenses can and are also used to distribute > closed-source software. Suggest that you don't talk about "open source > licenses" but instead "licenses commonly used for open source > software". > > P2, last para. Those four freedoms do not constitute copyleft. As the > FSF says at http://www.gnu.org/copyleft/ : "Copyleft is a general > method > for making a program or other work free, and requiring all modified > and > extended versions of the program to be free as well." - see the > rest of > that page and also the wikipedia entry on copyleft for correct > definitions. > > P3, para1 - error re GPL requiring distribution already noted above. > > P3, para2: The Mozilla and LGPL are both copyleft licenses, they > are not > permissive BSD/MIT-style licenses. The FSF terms them "weak" copyleft > licenses, while it reqards the GPL as a "strong" copyleft license. But > both the LGPL and the Mozilla license (and maybe some of the others > you > mention) are definitely not permissive licenses - they require > distribution of source code if modified versions are shared with > others > (and that is the core concept behind copyleft). The GPL is *not* the > only copyleft license (despite what the FSF may say - but then FSF > would > have conniptions about your use of the term "open source software" > - it > insists that "free software" is the correct term). > > p3, para3 is also wrong - see above. There is absolutely no problem > mixing GPL, Mozilla or any other copyleft code with proprietary, > closed > source code *within an organisation*. It is only if you want to > distribute the mixed code to others that you need to worry about what > the differences between open source licenses are. BSD-style licensed > code can be freely mixed with proprietary closed-source code and > distributed to others, and Mozilla- and LGPL-licensed code can be > freely > linked (but not mixed) with proprietary, closed-source code and > distributed to others, while GPL-licensed code distributed to > others can > only link to non-GPL code at runtime. > > p3 para4: see above - the GPL does not enforce "openness" of IP > added to > GPLed code, provided that the GPLed code with teh IP added to it not > distributed to others. Also, not all of the Linux kernel is under the > GPL, some is BSD-licensed, and most is licensed under a modified > version > of the GPL (with an extra clause) that makes it clear that > proprietary, > closed-source code is permitted to make runtime calls to the kernel > functions. > > p3 para5: LGPL is not permissive in the BSD sense, it just has weaker > copyleft provisions than the GPL. > > p3 last para: A bit misleading, because many of the most successful > open > source vendors do sell their software - either "commercially licensed" > versions as MySQL does, or extended partially-closed source > versions as > SugarCRM does, or some of the value-added PostgreSQL resellers do. > > p4 para1: I am not sure the assertions in that paragraph are true, and > are certainly not always true. And I don't think the paragraph helps > anyone in their understanding of FOSS. Suggest that you delete it. > > Refs: some very good essays at http://www.opensourcelaw.biz and also > lots of good and accurate articles on open source topics and > concepts on > wikipedia. In fact, you might do well to just re-use some of those > articles from wikipedia (with attributions, but you are permitted to > re-use them under the wikipedia Creative Commons license) rather than > use your own take on these concepts, which is just not quite > correct in > many respects as noted above. > > Sorry to be so negative but you did ask for corrections... others > should > feel free to correct my corrections! > > Tim C > > > > ------------------------ Yahoo! Groups Sponsor -------------------- > ~--> > Something is new at Yahoo! Groups. Check out the enhanced email > design. > http://us.click.yahoo.com/kOt0.A/gOaOAA/yQLSAA/W4wwlB/TM > -------------------------------------------------------------------- > ~-> > > > Yahoo! Groups - Join or create groups, clubs, forums & > communities. Links > > > > > [wr] - - - - - - - - will ross project manager mendocino health records exchange 216 west perkins street, suite 206 ukiah, california 95482 usa 707.462.6369 [office] 707.462.5015 [fax] www.mendocinohre.org - - - - - - - - "Getting people to adopt common standards is impeded by patents." Sir Tim Berners-Lee, BCS, 2006 - - - - - - - -
