mkidd

I looked at several alternatives before settling on GrapheneOS. Many made claims that sounded good but only GrapheneOS had deep technical documentation. I'm not an Android developer so some of it goes over my head but I've written detailed software documentation. I know you don't write what is in their FAQ or revision history if you don't know what you are doing.

I've been using GrapheneOS for about a year and it has worked well. Only have two minor issues:

1) I'd like the ability to have timestamp data added to the EXIF image info. There is an option to have the GPS data added, so timestamp really ought to be an option too, though I certainly understand why some users would choose neither. I'll probably end up writing a Perl driver for exiftool to add the timestamp based on the image filename which is YYYYMMDD_HH_MMSS_MSEC.

2) There aren't many fonts and Roboto supplies 1037 glyphs. The card suit symbols are missing which matters to a bridge player. E-books that display correctly almost everywhere don't on GrapheneOS unless the e-book explicitly includes a font with glyphs for the suits.

I've used Calibre for years and I like it a lot. It quite good for creating or editing e-books (oh the ghastly things you see if you look inside an e-book...)

The Calibre UX is fine. The real UX divide isn't desktops / laptops vs. smartphones / tablets, but rather content creators vs. content consumers. Modern UX is oriented to content consumers but even there manages to waste too much space and has become too flat and minimal. But once your application is at least partly for content creators, a dense interface is ideal and comforting.

I've haven't experienced any data loss with Calibre except when the e-book editor crashes while checking a book for errors. This has always been due to memory exhaustion (2 GB limit) when checking books that were unnecessarily complex (see “ghastly things” above).

SumatraPDF opens EPUBs almost instantaneously, but at the cost of not rendering every aspect correctly. As I recall the documentation is clear about this limitation.

I like the PocketBook app a lot and use it to read books on smartphones. But it has some basic rendering quirks; for example it can ignore margin-top and margin-bottom on paragraphs, seemingly because it does not apply CSS precedence rules correctly. It took me five e-mail exchanges, including sending a very simple hand-rolled EPUB illustrating the problem, to convince the developers that there is a problem.

I don't have the latest statistics, but from 2009–2015 the average ACBL member age increased 0.44 years / year and was never less than 0.33 years / year. So the reality is nearly halfway to the joke.

An interesting pair is Saloons of the Old West (Richard Erdoes) and Jacob Hamblin Peacemaker (Pearson H. Corbett), not for contrasting views but rather contrasting visions of the U.S. West. Saloons details the all purpose role of the saloon, as social center, post office, traveling preacher's pulpit, etc, in support of a rapacious gold rush mentality to extract each resource as quickly as possible and move on. Hamblin, by contrast is an early Mormon pioneer repeatedly sent by Brigham Young to establish new towns in West as the Mormons made a bid for their own Zion. The early Mormons really were different from the rest of the folks heading West, very much intending permanent settlements and a farm based economy. The book title comes from Hamblin's command of Native American languages and ability to regularly make peace with the Native Americans.

Peacemaker has references to the laying on of hands and other religious hoo-haw, but just ignore that and read it for the interesting historical document that it is.

Go is a fabulous game (and AlphaGo is a wondrous software—I read the first technical paper immediately). But it's another thing I'd have to do online. Seattle had a go center in the 1990s that still seems to be around. San Diego had a few players who met up in Hillcrest or Carlsbad, both an annoying drive away. Who knows if there are any go players in Salt Lake? My one go playing friend lives a time zone away. It's far easier to find f2f bridge than f2f go.

In the early days of Amazon, a call center employee pointed me to a deal on go set shipped directly from Japan. Beautiful slate and shell stones. I haven't played a dozen games on it. I almost feel unworthy of owning it as a perhaps 12 kyu player.

I'm the author. What a weird surprise to see it at 4 am on Hacker News.

If you play bridge on BBO you might be interested in my BBO Helper browser extension. It's popular with strong players and streamers (yes, there are bridge streamers!), but there is functionality to benefit players of all levels.

https://lajollabridge.com/Software/BBO-Helper/

The last evening bridge game at the Soledad Club was held April 7, 2016 about a year after I wrote the referenced article, when it hit the five table break even point.

The pandemic was hard on bridge. About half of the ACBL members made the transition to online play, primarily BBO. This looks like a sea change. Players are returning to f2f play but the ACBL lists only 1,855 bridge clubs today; there were 2,721 in 2019, a decline of 32%. Players are also returning to the tournaments but table counts are only 50–80% of pre-pandemic. Matt Baylow has been carefully tracking the situation:

https://medcitybridge.com/R_other%20pages/work/F2F%20VACB%20... https://medcitybridge.com/R_other%20pages/work/F2F%20Updates...

I've run many of the smaller (sectional) tournaments and thus know the many fixed expenses. Even a 20% decline in table count will affect the viability, choice of venue, or card fees. The bigger tournaments (regionals and nationals) depend on room night guarantees made to the host hotel in exchange for using their convention space; I'm not sure how that is shaking out.

This is a huge pet peeve for me. Shouldn't this be a “solved” style problem in 2023! I'd gladly fire every tenth web designer responsible for this problem regardless of their other merits.

Small fonts are one thing. There are browser hot keys to change the font size. And I'm a huge fan of the "Zoom Text Only" option in Firefox where the rest of page doesn't go flying off the edge of my viewing area. But low contrast text is a hugely annoying. What are my options? Open the developer tools and tweak the CSS every time? Try to find a good browser extension that solves the problem? Write my own? And some of these solutions are inconvenient or impossible on mobile.

I understand using low contrast text (and small fonts) for tedious but necessary legal disclaimers. But if you use it elsewhere, I'll move on—either you have nothing to say and don't really want people to read the text, have a poor designer, or are generally incompetent.

I second frutiger's recommendation of Elliptic Tales: Curves, Counting, and Number Theory. There should be more science and math books at this level but it's rare—either you're assumed to be starting from scratch or to be at least a graduate student in the field. I also recommend Jim Baggott's Perfect Symmetry: The Accidental Discovery of Buckminsterfullerene as another example at this rarely seen intermediate level. (Arguably all of Scientific American was at this level 50 years ago, but that era is long gone.)

This article is paywalled but here are the articles from Bridge Winners that it references:

https://bridgewinners.com/article/view/confession-of-a-self-... https://bridgewinners.com/article/view/confession-of-a-self-... https://bridgewinners.com/article/view/a-chance-to-clear-the... https://bridgewinners.com/article/view/curtis-cheek-suspende...

The problem is real and the national organization (ACBL), run by seventy and eighty year-olds, has not found the courage to address it.

There is no way to prevent the exchange of unauthorized information (i.e. not conveyed by a bid or a card played) online, but the full record of the play makes it easy to detect cheating with a high degree of statistical confidence, indeed overwhelming after hundreds of hands. This is entirely practical. It's not even "big data". A 16-core desktop should suffice to process a day's worth of data from Bridge Base Online (BBO) in under 15 minutes, and update statistics in a database. Nicolas Hammond, cited in the article, has done the analysis. I'm confident it works, because I can see step by step how I would replicate his work, well enough to verify.

But nothing is ever easy... There is bad blood between Nicolas Hammond and the ACBL. To the ACBL's credit, they moved quickly to develop a computer based scoring program at the dawn of the PC era. And 30+ years later, they are still using the same ACBLscore program written in Pascal by a single smart self-taught (now retired) programmer, without so far as I can ascertain the benefit of revision control.

Nine years ago, the ACBL tried to replace ACBLscore. Mr. Hammond's software company was hired to do the work. It didn't go well. The ACBL has never given an honest accounting of the failure to its membership. Mr. Hammond hides behind the NDA he signed with the ACBL. From experience, I'm sure there was shared responsibility on a project chock full of weird legacy issues. But it's impossible to apportion the blame.

From a programming view point, Mr. Hammond's cheating detection pipeline is simple and clean, and free of legacy issues, a whole different world from ACBLscore+. But the ACBL does not understand this, will not bury the hatchet, fears lawsuits, and has insiders and sponsors who don't want cheating investigated too carefully.

So a great intellectual game is dying. The ACBL brings more nails for the coffin. They deserve the bad publicity.