Or Seamonkey for the Mac, you can configure it not to display images: /https://www.seamonkey-project.org/releases/

As someone with more than a million lines of production Fortran code, that is a non-trivial (and expensive!) undertaking. It's hard enough (i.e., near-impossible) getting EPA even to update their codes from 1977 Fortran to 1990 Fortran... "Physically painful" is a willfully-ignorang bigoted canard at least as regards the Fortran of the last 30 years. Besides, modern (post-1990, i.e., for the 30-year-old F90 Standard) Fortran is actually quite a nice language to program in -- much better than C or C++: for example, both strings and arrays are first-class citizens.

For the M1 based upcoming MACs that definitely should be true. You have to go really high-end (e.g. AMD's very latest 5950) to beat their M1 on, e.g., CFD codes.

Because of the overlap of its absorbtion spectrum with water vapor, the actual greenhouse effect of methane is only about 10% that of CO2. The "bigger greenhouse effect" happens only in the complete absence of water vapor, which doesn't happen on this Earth.

If you look at the science (physics, in this case), you will see that the absorbtion spectrum for methane is quite narrow, and its bands are "covered" by the bands for water vapor: i.e., in order to judge the greenhouse effectiveness of methane you must only apply it after the effects of water vapor have been applied -- a case in which the effective absorbtion is quite small.

See Bug 415876.

Use of "panning" improves my desktop productivity by about 20%, and I cannot forego that.

See, for example, Humidity as a non-pharmaceutical intervention for influenza A. Another Mayo Clinic study showed that schools maintaining 50% humidity had factor-of-three fewer influenza cases than the control group of schools.

Enshrouded-RNA viruses such as influenza, SARS, MERS, and CoVID-19 are known to be mildly sensitive to temperature, moderately sensitive to humidity, and quite sensitive to UV. This has been known by virologists in general since shortly after the 2003-vintage SARS outbreak, and has recently been confirmed for CoVID-19 by Homeland Security.

Health authorities should have been recommending humidifiers in public buildings for at least the last decade. That they have not done so has cost hundreds of thousands of lives from influenza deaths over the last decade.

Cut the R0 by a factor of 3 and CoVID-19 would not have been a pandemic. And this new swine-flu strain would not be a pandemic either.

Representation theory was the key insight for developing the SMOKE atmospheric-chemistry-emissions model. It was "somewhat" practical -- the prototype in 1993 replaced its database-style predecessor's 11-hour overnight Cray supercomputer run by 2 minutes 43 seconds on a desktop SPARC-2 workstation...

To be honest, it did also employ faster techniques for its set-up (not part of the 11 hours or 00:02:43): fast sorts, binary instead of linear searches, integer instead of character-string ID-lookups), but the non-setup part was almost entirely (sparse) matrix multiplies.

On the other hand, it is well established by the human-factors people that text is most readable if the effective line length -- from the first non-blank character in the line to the last non-blank character -- is at most 70 characters.

Those web pages that wrap paragraphs into 150-character lines drive me nuts. I refuse to read them as-is; they are in fact one of the major reason for "reader" extensions to web-browsers. Before these came into common availability, I wished I could punish the web-authors that did such things -- severely!

...over a manager's head...

to the prosecutor's office. The last time I heard, reckless endangerment was a felony.

Actually, the original study said " The best are 10x better than the average; the average is 10x better than the worst."

And in that visualization, imagine which variation gives the least coupling (shared data, shared assumptions, ...) between the various pieces!

See https://en.wikipedia.org/wiki/...

Reproducibility of (even published, peer-reviewed) social science results is extremely poor.

Yes.

And it gets away from what I think is the worst fault of Agile. As has been said long ago:

Too many cooks spoil the broth.

My experience is that Agile gives altogether too much room for ego--"I need to put in my two cents, no matter whether they are contradictory to having a unified vision."

Not my experience with environmental modelers. In particular, they have weaknesses with respect to (1) algorithms; and (2) how the implementation interacts with the hardware. Generally, it is coded in a "sloppy vector" style tailored for the limitations of 1980's vector supercomputers.

• Case in point # 1: in the mid-90s, recognizing that emissions modeling is a linear operator problem instead of a data processing problem not only made the code much more elegant, it also reduced the run-time from 11 supercomputer-hours to 163 Sparc-2 seconds.
• Case in point # 2: a little recoding to respect how memory access patterns interact with microprocessor based systems reduces the run-time of (my custom of) the WRF meteorology model by 40%
• Case in point # 3: restructuring the microphysics and the vertical-mixing codes in WRF and in CMAQ from "sloppy vector" style to "vertical-column style" reduces the scratch-variable memory usage for a typical case from about 500 KB (larger than L2 caches) to about 4KB (much smaller than L1, on almost anything), the number of lines of code by 30%, and the execution time by a factor of 8.