Re: Maple
- From: Craig Carey <research@xxxxxxxxx>
- Date: Sun, 26 Feb 2006 10:22:16 +1300
On Sat, 25 Feb 2006 04:43:57 GMT, "Richard Fateman"
<fateman@xxxxxxxxxxxxxxx> wrote:
"Craig Carey" <research@xxxxxxxxx> wrote in message
news:nhlvv11boiapeqg2i7qf82cdp7iqg00kdi@xxxxxxxxxx
This URL shows that Maple 10 only runs in Windows:
Craig: if you scroll down a little, you see Maple 10 runs on linux, unix,
macintosh, 32 bit, 64 bit, etc.
The rest of your comments seem to be mistaken as well.
RJF
http://www.freebsdfoundation.org/press/20041221-newsletter.shtml
| The FreeBSD Foundation Quarterly Newsletter / December 21, 2004
| Even after receiving notice of the termination of our license
| attempts to contact Sun to renegotiate the license have gone
| unanswered. For now, it is safe to assume that the Foundation
| will engage in another lengthy, and potentially costly, licensing
| negotiation before our binary distributions can continue.
http://www.freebsdfoundation.org/downloads/java.shtml
| [Modified: Tuesday, 15 November 2005 13:00:22] The FreeBSD Foundation
| has negotiated a license with Sun Microsystems to distribute FreeBSD
| binaries for the Java Runtime Environment (JRE) and Java Development
| Kit (JDK). These implementations have been made possible through the
| hard work of the BSD Java team, and contributors (particularly
| WorldGate Communications, Inc.) as well as through donations to the
| FreeBSD Foundation that supported hardware, developer time, testing
| resources, and license negotiation.
Some revenue figures of Worldgate, 2004-05: http://www.wgate.com/news/
The search engine of Sun, at http://java.sun.com/, has nothing to
show that Sun ever said that they didn't revoke the licence of the
"FreeBSD Foundation".
An uninteresting Sun document on licences
http://www.jcp.org/en/press/news/licensing_update?showPrint
....
| The SCSL license is broken into three levels: research use, essentially
| for evaluating and prototyping potential software applications;
| internal deployment, for limited distribution and testing; and
| commercial use, for selling products. Incompatible modifications can be
| made to the existing code without giving back to the Java community,
| but only compatible distributions may be distributed commercially.
|
| Some of the Java development community resisted SCSL?s code disclosure
| requirements and, many objected to SCSL?s ?clarity challenged? length
| and complexity. The JCP program spent the next five years working to
| simplify its requirements around licensing while protecting the
| community?s interest in platform compatibility. Today, there are much
| simpler licenses in use that are available to spec leads.
|
|
| Separate Licenses for Research and Distribution
|
| As one option, Sun has introduced two new companion licenses that are
| easier to understand and comply with: the front-end Java Research
| License (JRL) and the back-end Java Distribution License (JDL). Both
| are allowed for use in the JCP program.
|
| Used in conjunction with the TCK, these two newer licenses are the
| simplest way yet to license Java technology.
|
| The JRL is for researchers and universities [...]
|
| Next, when the code is shown to be stable, it?s time to license the TCK
| and test the product against the compatibility test suite for
| certification, which can include the right to use a Java compatibility
| logo with the product, applicable for some, but not all, JSRs. [...]
|
| Both the JRL and JDL are considered to be much easier to use than the
| TLDA or SCSL, and they are recommended by the JCP program for most
| projects today.
|
| Sun still offers the SCSL license for JDK 5.0 (JSR 176, Java 2 Standard
| Edition 5.0 ?Tiger?), but using the JDK under the JRL simplifies access
| to source code. [...]
....
Mr Laurent Bernardin is part of a company that says it can implement,
using Java:
http://www.nobilitas.com/en/implementation.shtml
In this paper:
A Java Framework for Massively Distributed Symbolic Computing
by Laurent Bernardin, submitted to Math. and Comp. in Simulation
in 1999,
completely censors out why Java would be selected (it is a thick
layer of unnecessary buggy code and the language can't reliably and
repeatedly get data out of a Windows TCP stack, especially if there
is a shortage of CPU and memory, and the test is comparative and
there run-time TCP thread bungling rate is required to be very
low.).
Here is Mr Bernadin Laurentin arguing the case:
(1) The case for use of Java in distributed symbolic algebra:
| 2 The Traditional Model
....
| Computer algebra systems like Maple are adequate for this model
| of distributed computation
....
A bit on the force named "urges". Inside the domain of knowledge.
| 4 The Technology
| We propose ...
| The urge to use Java is particularly important in the "divide and
| be conquered model" since the key for reaching millions of internet
| users at home and getting them to contribute idle cycles of their
| personal computers [...]
| Using Java, all they need is a decent browser (Netscape, Opera,
....
| 5 The Distribution Server
...
| For reasons discussed above, Java is a natural choice for
| implementing the client side of our framework.; it allows us to
| run on most modern web browsers without requiring a contributor
| to go through lengthy installation or downloading procedures.
| [...] While we have to pay a decreaese in efficiency [...]
| We also believe that the penalty for using Java over optimized
| C, which has already come down from an initial factor of around
| 100 to a factor of almost 2 using ... just-in-time compilers [3],
| will decrease even further in the near future.
Then the Netscape browser got less popular.
| a machine running a web browser can be coerced into donating
| CPU cycles without the machine's owner knowing. The only solution
| would be to disable the Java capabilities of his browser, but
| this is not always possible or desirable.
Mr Laurentin's paper is of early 1999.
The 233 MHz MMX Pentium CPU is said to have been released in June 1997
so PCs were very slow.
http://sfdm.ca.scad.edu/vsfx/faculty/jopasqua/Thesis/cputimeline.html
http://www.teachers.ash.org.au/dbrown/8162/timeline.htm
So the browser would unexpectedly slow down, and maybe it would
crash too. A solution was to disable Java.
I have the words "contribute" and "contributor", twice above.
The public contributes bug reports, using the their own
techonology of quality expectations, "decent" arguments etc.
A "penalty" for having a bug rteport thrown away, is the same
bug reappears in Maple version 11.
http://citeseer.ist.psu.edu/bernardin99symbolic.html
A Java Framework for Massively Distributed Symbolic Computing
Laurent Bernardin, Bruce Char, Erich Kaltofen
There is no suggestion that employees of Maplesoft, are not
'contributing' properly.
Craig Carey
.
- References:
- Prev by Date: Re: Maple
- Next by Date: Re: mupad laplace transform
- Previous by thread: Re: Maple
- Next by thread: Re: Metron: President's Letter
- Index(es):
Relevant Pages
|
