
Re: Abrilliant article that I will reference in future
That's Axel Nose you're thinking of.
20 publicly visible posts • joined 24 Mar 2016
Perhaps the vendors should help people decide for themselves, rather than deciding for them. I envision a classifier that would decorate displayed text to show its semantic categories. It would classify statements (descriptive, speculative, normative, imperative) as well as questions (inquiries, provocations, rhetorical ones). It could highlight weasel words (quantifiers, qualifiers, disclaimers) and indicate emotional charge.
The task seems more complex than syntax parsing but less daunting than sentiment analysis (also less attractive as a subversion target). It really should be doable with today's technology. The UX would have to be worked out, obviously, as well as a business model. Oh well, one can dream...
Sometimes the pressure to use easily digestible phrases results in such imprecision that it's just frustrating.
"Embedded carbon in concrete" sounds like someone figured out a way of making usable concrete with a substantial carbon fraction, effectively capturing and storing said carbon, preventing its release as carbon dioxide into the atmosphere (such technologies are actually being developed).
What's meant in the article instead is embedded carbon dioxide emissions, i.e. those released as the concrete was made. Which is something completely different. "Embedded carbon," really?
The difference between translations in dubbing vs. subtitles is natural, when you think about it. Dubbing optimizes for matching syllable counts (Japanese to English is a nightmare in this regard) while subtitles try to minimize the distraction by being quick and easy to read. The liberties translators take expose the dilemmas they face in pursuing these distinct goals.
I once contributed a set of translated subtitles to divxsubtitles.net for a rather chatty film (was it Being John Malkovich?) and it was fun to make them as brief as possible without leaving out essential details.
Back in the '90s I was once playing games on my dad's PC with my kid cousin. He knocked over a cup of something sugary, spilling it onto the desk. We cleaned it up immediately and thought nothing of it. About a week later, I picked up an unlabeled 3.5" floppy disk I found lying around and inserted it into the the PC's drive to see what it contained. The disk could not be read and, lo and behold, neither could any other disks after that...!
As it turned out, the little protective cover on the disk had got glued shut by the sugar and something in the drive broke when trying to move it. The drive had to be replaced. A nice little lesson in failure modes, I suppose.
There are ordinary desktop tools equivalent to the Google Authenticator. For Linux there's oathtool (https://www.nongnu.org/oath-toolkit/oathtool.1.html or your distro's package repository) and for Windows there's t2otp (https://www.token2.com/shop/page/t2otp-exe-command-line-totp-generator). Presumably there's something for the Apple crowd as well but I wouldn't know.
What about sending a balloon or an airship above the central structure and hooking it up to provide some lift? Yeah, I suppose the numbers don't check out. A quick search on the net puts the lifting capacity of hydrogen at 68 pounds per 1000 cu. ft. with helium even lower (can't be arsed to convert that to proper units). So offloading the cables in any meaningful way would require something much larger than the largest airships ever built.
Alternatively, repair crews could be lowered from an airship to work on the platform without the risk of falling down with it (well, a lower risk in any case). A helicopter could be used for that scenario as well, being somewhat less sensitive to windy conditions... but I wouldn't want to be on that repair crew in any case.
One other idea involving bags of air is to set up an airbag under the platform so that the damage to the mirror is at least somewhat reduced once the platform does fall. I'm afraid, though, that the mirror is far too sensitive even for that.
On the test-data-stochasticity spectrum this sits somewhere between coverage-guided fuzzing and old-school deterministic test inputs. Explicitly supplying a format specification spoils the elegance of it but if effectiveness improves, what's not to like? Oh, right - you have to have the format specification in the first place. Other than that, nice work indeed.
The Panasonic micro-hi-fi in my living room has a tiny segmented display that outputs a jerky scrolling HELLO when turned on and a jerky scrolling GOODBYE when shutting down. During CD playback it displays... minutes and seconds the current track has been playing, which is utterly useless. Either track number or time remaining would be more useful. I even suspect both of those could be displayed simultaneously (as long as the track is shorter than 10 minutes or its number is less than 100). Oh well...
Well, there is always something to improve. As an example, Thunderbird offers partial support for the Maildir storage format but it's languished for years in a semi-complete state. I even contacted the team to ask if I could make a donation specifically towards fixing the remaining issues. They had no way of assigning donations to Bugzilla tickets, unfortunately. I still sent some money their way, I think it's important for Thunderbird to keep going.
As for a visual refresh: as long as the budget is modest and well spent, why not? It did work as a PR hook, obviously, which is a good thing in itself.
@David Roberts
Proud owner of a Fairphone 1 here. Software support has been rather tragic for a simple reason. FP founders are mostly design/marketing folks and had no software chops when they started. They licensed a reference design from the manufacturing partner which saddled the FP1 with components supported only by binary blobs from open-source-hostile companies. Later they hired some coders to try upping the Android version from 4.2.2 to 4.4 but even that floundered in the end. FP2 has higher priority, I imagine.
Many FP1 owners seem to be disappointed but, being a first shot at the concept, I think the FP1 has been quite a success. It has certainly served me well.