MATLAB, Cold Fusion, Conferences, XFCE

[tcpip]

MATLAB 2011a has been released as is typically with its parent company a pleasant miniconf was held at the Rendezvous Hotel in Melbourne, which has an ridiculous name, is ugly from the outside, but with some very nice deco features. The "great new product" for this release was the MATLAB coder, which allows conversion from MATLAB code (e.g., design and prototypes) to C code for actual implementation, a procedure which was often conducted manually. I asked whether it was capable of converting MATLAB code written for parallel computation to MPI-compatible C code; the speaker agreed that would be "awesome"... but no, it wasn't available. I further asked whether there was any plans to give the MATLAB coder export ability to other languages commonly used in engineering and the sciences, such as FORTRAN. This answer this time was "probably not".

Attended the "Singularity Salon" at the Melbourne University Graduate House on Thursday and participated in a very good discussion about some very interesting recent - and surprising - developments in cold fusion, a sufficiently important topic (OK, it's probably the most significant scientific discovery of all time - if it's true) that makes up the first 'blog post on lightbringers.net. Agenda for the Humanity Plus Conference was also discussed , where I get the Sunday morning slot (just as well I'm a morning person, right?) with a terse abstract. Also received confirmation today that my paper has been accepted for the New Zealand eResearch Symposium, focussing on teaching the value and practise of High Performance Computing, appearing at the same time my new passport came in the mail (this would possibly send a superstitious person into a flurry, I'm sure).

I have recently made a change to my default desktop environment in Linux on both my home and work machines. Ubuntu's latest release comes with Unity, which managed to break my GUI (I mean, NVIDIA, who uses that?). I am a little unhappy with Ubuntu in general for reasons that Jeff Waugh points out in detail, along with the refusual to implement some basic design issues. I thought about moving to a Fedora/GNOME 3 combination, but that didn't seem to be to my liking either. Finally, I have settled - and quite happily - with XFCE, a small, fast, mouse-capable WM. Oh, and yes it does have a cute rodent logo which had no effect at all on my choice.

Comments

[rialian.livejournal.com]

===XFCE is pretty much the only desktop I use now...and I am thinking of trying the new Mint based off Debian. I tried Lubuntu...nice, but not quite as usable for me at XFCE.

[tcpip.livejournal.com]

I would be tempted to use ratpoison/stumpwm as well being an extremely heavy CLI-orientated person. But a don't really want a serious learning curve for a desktop environment.

[tcpip.livejournal.com]

BTW, you're just the person who would have been keeping up with these recent developments in cold fusion. Any thoughts on this?

[rialian.livejournal.com]

===Actually, I am not as up on the higher level physics...more a permaculture and former social services geek, myself...

===I did host for a long time the work of Neil Boyd on rialian.com until relatively recently. We had to downsize the site, so we stopped hosting the information. Some other folks might be putting the stuff back up, if you had gone looking for it recently.

[rialian.livejournal.com]

===(I have had some folks mistake my writing for his, and vice versa)

[tcpip.livejournal.com]

Oh very much permaculture and social services, I see that... Just thought that this would be something that may have fallen into your inbox prior to now..

[ferrouswheel.livejournal.com]

I just upgraded to Natty, I wish I hadn't because I really don't want Ubuntu UI to be a clone of OSX.

I'll probably try XFCE or Xmonad when I get the time to set them up and learn how they work.

[tcpip.livejournal.com]

I did briefly consider xmonad, which has some nice features, but I decided in favour of something that wasn't entirely rodent free. Also the BSD license made me twitch. Ultimately, for desktop use, I prefer a small and fast rodent. XFCE fits the bill quite nicely.

[fluffyblanket.livejournal.com]

All that is far beyond my lamentably non-technical brain but I like the Art Deco and your spititualized , libertarian socialist ideas .

[tcpip.livejournal.com]

I thoroughly intend to keep my arts, sciences and politics in sync :)

The cold fusion thing really excites me... If it turns out to be true. The design looks fairly simple; the sort of thing that a local auto shop could put together.

Can you imagine energy at 0.1c per kilowatt/hour? Here's some comparisons...

If I was a praying person (or rather, if I wasn't constantly in prayer) I'd be on my hands and knees at a pulpit crying for this to be true..

[notthebuddha.livejournal.com]

Cool developments! Did they mention if the coder would be included with MATLAB's basic price?

[tcpip.livejournal.com]

Alas no... I comes with a hefty price tag (about $9K), which I suppose is fine if you're in a engineering company that's churning out a lot of embedded devices or similar..

It would be great if the same could be done for FORTRAN, R etc. and other languages that regularly find themselves interacting with MATLAB.

[notthebuddha.livejournal.com]

if I can ever get done with finals and my own multiprocessing project, I'll have to see if I can round up some other translators, and see if MATLAB-to-c-to-*-to-Fortran looks like something worth having.

[tcpip.livejournal.com]

Also... have a look at Octave; free and open-source implementation of the core functions of MATLAB.

[vaelynphi.livejournal.com]

Hrm... still skeptical of the 'cold fusion' device, but it sounds interesting.

[tcpip.livejournal.com]

Skepticism is healthy, especially with such dramatic claims and with apparently some guarded acceptance from people who are normally trustworthy.

This could very well be a troll of epic proportions... Or they may just have somehow managed to get erroneous data on several occasions. But both of those possibilities seem increasingly unlikely, leaving us with an interesting alternative.

[vaelynphi.livejournal.com]

Yeah, I'd love to read a paper by it, but frankly their explanation of 'actually we're not entirely sure how it works either' is oddly convincing.

It seems much more likely that someone would stumble upon such a thing than that they would design it--as two physicists interviewed about it said, the math is just too hard to do. It certainly seems, on the basis of any rational physical intuition, impossible--the Coulomb repulsion of the nucleus would be too great. However, plasmas behave quite strangely sometimes, and having seen so many papers on fairly accepted phenomena that seemed rather magical compared to 'physical intuition', well, I'm skeptical of my skepticism as well.

[tcpip.livejournal.com]

I can help a little bit with regard to papers:

http://www.nyteknik.se/incoming/article3144960.ece/BINARY/Download+the+report+by+Kullander+and+Ess%C3%A9n+%28pdf%29.

http://lenr-canr.org/acrobat/VillaMonthegamma.pdf

and of course, you might glean something from the patent application itself

http://www.wipo.int/pctdb/en/wo.jsp?WO=2009125444

[jujulilianan.livejournal.com]

hasn't matlab had matlab code to c as an expensive option for years?

[tcpip.livejournal.com]

Not that I know of. They were certainly promoting it as a new feature at the miniconf available as of 2011a.

[tcpip.livejournal.com]

Ahh, I'd quite forgotten about IceWM.. I used it for a while back in the days when I was a Mandrake fan (c2003) and thought it was pretty good...