pklausler

9 ms·

151.
by pklausler 1y ago
Judging Fortran by looking at Fortran 77 is preposterously uninformative.
152.
by pklausler 1y ago
10 never implies the start of a character; those begin with 0 or 11.
153.
by pklausler 1y ago
See also the code in LLVM Flang for binary/decimal conversions. Minimal-digit decimals don’t come up in Fortran as often as other languages, but I use them when allowed by the output format. Interestingly, the last digit of a minimal
154.
by pklausler 1y ago
As an old-timer, I worry that it may have been forgotten that all the major systems vendors used to have extensive shelves of high-quality technical manuals for their software, written by their in-house teams of professional technical write
155.
by pklausler 1y ago
The author discusses finding “unique” values, but I think they meant “distinct”.
156.
by pklausler 1y ago
Been to the Ark Encounter in Kentucky yet?
157.
by pklausler 1y ago
Maybe the parents shouldn't let their offspring go feral on the exhibits?
158.
by pklausler 1y ago
Fortran doesn't disallow aliasing per se of dummy arguments to functions or subroutines, but we do restrict what can be done with aliases in many cases. The details matter.
159.
by pklausler 1y ago
Every programming language has a practical definition: it is the intersection of the sets of features that are accepted by the various relevant production compilers and interpreted identically enough to be portable to all of them. Formal la
160.
by pklausler 1y ago
Ugly people like to blame the mirrors.
161.
by pklausler 1y ago
Interval arithmetic makes good intuitive sense when the endpoints of the intervals can be represented exactly. Figuring out how to do that, however, is not obvious.
162.
by pklausler 1y ago
Obviously; but the point was, if the language were to limit the circumstances in which circular structures can arise, one could exploit that fact.
163.
by pklausler 1y ago
You'd love Haskell, which uses curly braces for many constructs, but also has rules by which they are implied by indentation -- so in practice you only ever see them on records.
164.
by pklausler 1y ago
It could be made to work in a language like Haskell, where cyclic structures can arise only in limited circumstances (recursive `let` groups) and can be given a distinct runtime representation.
165.
by pklausler 1y ago
How do you know that climate scientists are as unreliable as some economists?
166.
by pklausler 1y ago
How do you know that?
167.
by pklausler 1y ago
I miss going to some conferences, but I dislike being "on duty" as a representative of my employer, and the travel / hotel / away-from-home hassles just aren't worth it any more for going on my own dime to attend pr
168.
by pklausler 1y ago
BYTE was best before Jerry Pournelle showed up.
169.
by pklausler 1y ago
Please see https://docs.nvidia.com/cuda/parallel-thread-execution/index... .
170.
by pklausler 1y ago
So it's a "SIMD lane" that can itself perform actual SIMD instructions? I think you want a metaphor that doesn't also depend on its literal meaning.
171.
by pklausler 1y ago
strace might help, if you have it.
172.
by pklausler 1y ago
Even the first release of FORTRAN had statement functions.
173.
by pklausler 1y ago
In practice, though, I can't recall any HPC codes that want to use IEEE-754 infinities as valid data.
174.
by pklausler 1y ago
It's possible for the sign bit of a NaN to be changed by a "non-arithmetic" operation that doesn't trap on the NaN, so don't put anything precious in there.
175.
by pklausler 1y ago
I am allowed to disagree with the article.
176.
by pklausler 1y ago
52 bits of payload, and at least one bit must be set.
177.
by pklausler 1y ago
Division by zero is an error and it should be treated as such. "Infinity" is an error indication from overflow and division by zero, nothing more.
178.
by pklausler 1y ago
People need to understand rounding better, especially the topic of when rounding can happen and when it can't for the basic operations. Updating with a concrete example: The Fortran standard defines the real MOD and MODULO intrinsic
179.
by pklausler 1y ago
I fall squarely into the 80-column limit camp, at least for C++, but I admit an important exception: string literals used for user-visible messages, like error messages in a compiler. It's important to be able to grep for them reliabl
180.
by pklausler 1y ago
Fixed-form Fortran source is 72 columns. Anything after column 72 is ignored without error as a comment.

More ›