Article View: sci.image.processing
Article #1284Re: Image Processing Software for Unix
From: Walter Roberson
Date: Sun, 25 Jun 1995 00:00
Date: Sun, 25 Jun 1995 00:00
74 lines
4517 bytes
4517 bytes
In article <3shfp1$4a5@crcnis3.unl.edu>, Steven Kuznicki <kuznicki@cse.unl.edu> wrote: |rubinsnk@is2.nyu.edu (Kalman Rubinson) writes: | |>Gilbert S Pajela (gsp@blue.seas.upenn.edu) wrote: |>: I am looking for a commercial image processing package, similar to NIH |>: Image for the Macintosh, that runs on a Unix platform. | |Better yet, Khoros is public domain and has full support ( a CD commercial |version for V2.1 is supposed to be made available). It has tons of |features and developer support to create your own routines. Check out |their newsgroup at comp.soft-sys.khoros The entire package can be ftp'd. Khoros is NOT public domain! I quote from the license: RIGHT TO REDISTRIBUTE Permission to distribute this Software using electronic media and at no cost is granted, provided that the Software containts no modifications, provided that the distribution is not assigned any price or value in any way, provided that the distribution is not combined with any product or service in any way, and provided that the distributed Software is identical to the Software that is currently being distributed at no cost via computer network by KRI. DERIVATIVE WORKS Permission to prepare derivative works or to modify the Khoros Software for internal use only is granted to Licensee. However, all rights to such modifications and derivative works, including all copyrights, are the property of KRI. New software developed by Licensee as a result of using the Khoros Software belongs to the author and is not bound by this license aggreement. However, any Khoros Software that is Loaded with the author's software is bound by this license agreement. [Back to commentary] If Khoros *were* public domain, then KRI ==COULD NOT== assert ANY control over its use or derivations from it. "Public domain" means just that: a work is given to the Public to use as they see fit. If even *one* Right is reserved by the author of the work, the piece is not Public Domain!! Note the bit in the license about copyrights on derivative works. Suppose you trivially modify a Khoros program to (say) double some internal limit. The new program is legally a derivative work. Because KRI reserves the copyright on derivative work, and explicitly bans distribution of any modified version of the software, you would have to seek the permission of KRI before you could hand your trivially modified program to your co-worker in the next office who had asked you to make the change in the first place! (Yes, this is "internal use", but "internal use" does not necessarily require copying of the derived work; and since it is -possible- for the co-worker to be given a copy of the original and an exact set of instructions on how to modify the software to get the derived version, IMHO the license must be read narrowly as not allowing even internal copying of a derived version.) I would also note that *after* gzip'ing all the postscript files, khoros takes approximately 350 Mb of disk space. It also takes a lot of memory, and has a nasty habit of creating files, sockets, and shared memory arenas that it never cleans up after itself. We also detected an infinite loop in 'craftsman' whereby it will continually retry to create a directory if the first attempt fails... this will eventually fill up the audit logs if the reason for the failure was that the user does not have permission to create the directory 'craftsman' thinks it should be creating :-( Overall, given these license restrictions and the high resource use of khoros, I personally would recommend looking at other packages first, and using khoros only if the others will not do what you want. Walter Roberson roberson@ibd.nrc.ca
Technical Details
Message-ID:
<3skjss$62@canopus.cc.umanitoba.ca>Path:
arc1-ca.pugleaf.net!unknown!not-for-mailReferences:
<3s75n7$7iq@netnews.upenn.edu> <3s7cer$1ra@cmcl2.NYU.EDU> <3shfp1$4a5@crcnis3.unl.edu>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 10aee9,e3869bc6e1a18ff5 X-Google-Attributes: gid10aee9,public X-Google-Thread: 103f76,e3869bc6e1a18ff5 X-Google-Attributes: gid103f76,public X-Google-Thread: fa6b6,e3869bc6e1a18ff5 X-Google-Attributes: gidfa6b6,public From: roberson@hamer.ibd.nrc.ca (Walter Roberson) Subject: Re: Image Processing Software for Unix Date: 1995/06/25 Message-ID: <3skjss$62@canopus.cc.umanitoba.ca> X-Deja-AN: 105054734 references: <3s75n7$7iq@netnews.upenn.edu> <3s7cer$1ra@cmcl2.NYU.EDU> <3shfp1$4a5@crcnis3.unl.edu> organization: National Research Council of Canada newsgroups: sci.image.processing,alt.image.medical,bionet.software