Rick's b.log - 2023/05/21
It is the 24th of November 2024 You are 3.142.255.23, pleased to meet you!

Funny I should...

...mention sleazy slimy Tories. Suella Braverman, who was already shown the door for sending an official email from her personal email account (technically a security risk thus a breach of the rules) turns out to have tried to weasel her way out of the repussions of being caught speeding - yet another not-terribly-surprising case of "one rule for us, another for them".
It is perhaps worth remembering that in her previous resignation, she said "I have made a mistake; I accept responsibility; I resign". So, will she do the decent this time and sod off once and for all?

AndFTP terribly slow

I started the AndFTP app on my Android laptop, and selected the directories and files that I wanted to take a copy of. I then asked it to do exactly that.

That was yesterday evening. This afternoon, the thing was still chundering through the pictures. It had made it as far as summer 2016. Which means if I left it, I'd be looking at leaving it running all night again.

The reason this is happening is that, for some reason, it pauses for about thirty seconds in between each file. As the files are all fairly small, this means a lot of time thumb-twiddling.

I just fired up the desktop PC and set WinSCP to grab a copy. Running at around 250KB/sec, it reckons (so long as the signal doesn't drop out) it'll be done in a little over half an hour...and indeed it did in a little over half an hour what AndFTP couldn't manage in around 15 hours.
Bonus - I have a copy on µSD as well as the portable.

I tried to register with the Lysesoft forum to ask if this was some sort of anti-leach thing that could perhaps be disabled, but a "security question" was to ask what's the most recent version of Android. It did not accept "12" nor did it accept "Snow Cone". If it wants 12.this.that.whatever, get stuffed, I'm not looking up the exact latest build just to sign into a forum... (assuming of course it isn't woefully out of date)

Finding rice

My next stop was Satsuki. They don't stock this particular rice, but there was an alternative that looked good (Akita Komachi), however the other things I looked at were out of stock.

My final port of call was mon panier d'Asie, who not only had this exact rice, but it was a 2kg bag. The problem is, the delivery information on the site said that Colissimo delivery (up to 4kg) was a flat rate of €5,60. When I added the rice, it automatically set the postage to €13,20. I went through part of the order process, and it seems fixed at that price which, frankly, is ridiculous - it's almost as much as the product I am wanting to purchase.

There was no contact form on the site, it gave a phone number and a postal address. I did eventually discover an email address, so I wrote an email to ask about why the postage is so different to what it should be, and...

The mail system <contact@monpanierasie.com>: host
mgra.mail.ovh.net[...ip address...] said:
550 sorry, user over quota [mx1]
(in reply to RCPT TO command)

The expression the Frenchies would use is "tu n'es pas sérieux". While this literally translates as "you're not serious", it carries a connotation of a lack of competence. Which about sums up my feelings.

I'll pop back over to Satsuki in a couple of weeks.

You only use 10% of your brain

It is actually roughly true. You do only use about 10% of your brain. But there is a massive caveat that changes everything.

You only use about 10% of your brain at any given time.

Pretty much your entire brain is functional. It's not like there's a massive lump of grey matter stuck inside our heads that for some reason we're normally unable to utilise. There are parts for processing speech, for processing music, for seeing, for deciding if you really want that pair of shoes or power drill, and even for daydreaming. Many people have done many things while inside a scanner to see which parts of the brain light up, which helps to try to get some sort of idea of how such a small thing actually works, can manage to acquire language, move our bodies in interesting ways, and even come up with new ideas from abstract thoughts. We're a long way from understanding the human brain, but every study gets us a little closer.

Sadly, if one was to find a way to use more of their brain at any on time, the results would not be an innate understanding of quantum fluctuations or the ability to manipulate matter or time.
I don't remember the exact figures, but the brain is a very "costly" (as in resources) organ. In normal use, it's something like 7% of body mass while consuming between a fifth and a quarter (that's 20-25%) of the body's energy, blood, etc.
So if you were to manage to use that much more of your brain, it wouldn't give you superhuman properties, it would likely just give you a seizure, with potential for irreparable damage. It is supposed to run with that amount of load. Finding a way to push it harder would be like overclocking a processor. It may work, but only up to the point of catastrophic failure. Luckily for most of us, we have a built in rate limiter so the grey matter can function for the lifespan of a person.

That's not to say it doesn't go wrong - alzheimers and dementure are two well known failure modes. I wonder if anybody has done a wide-scale study on the lifestyles of people who contract such maladies. Is it, perhaps, the brain wearing out after a lifetime of stress?

I have written this in bits and pieces, while doing some cleaning and tidying up - including washing the kitchen window for the first time in thirty years (don't ask).
I also did a full inventory of my larder and made sure the spreadsheet matched reality (it mostly did).
Now I think I'll do a macaroni cheese with a decaf tea and wind down. It's a long work week coming. Pfff...

Depressing factoid - longest day (Summer Solstice / Litha) in just a month. Already...!

Your comments:

Anon, 21st May 2023, 22:57

Just had a look at the HTML source for this page. I think the most polite thing I can say is "legacy code"!    It's important to emphasise that's an observation, not a criticism. I used to (and we're talking within the last few years) have to write HTML 4.01 code with loads of 'hacks' and 'workarounds' for crappy browsers (mainly IE) that didn't follow the standards (IE) and broke correctly written pages (IE), then when you fixed the page to work on those broken browsers (IE) it then broke on every other browser that actually followed the standards.    (Do you get the impression that as someone who does web sites part time, I don't like Internet Explorer very much?)    Fortunately IE has now been put out to pasture, the default browser on Windows is Edge (which uses the Webkit engine, same as Chrome and Safari). Webkit is standards compliant. Gecko is standards compliant.    Have you looked into tweaking the blog script to output HTML5 rather than 4.01? It has some very nice features, for example it defines elements for "header", "footer", "main", "nav" and "aside", rather than using DIVs. Just altogether cleaner design.    I re-factored some of the sites I was working on into HTML 5. It got rid of virtually all the "boiler plate" HTML code that did the page layout. The empty "template" for the page was reduced from about 3 or 4KB down to about half a K, with all presentational elements moved into the CSS. Google loves it as it doesn't have to wade through all the boiler-plate HTML to get to the actual content.    Also benefits from much faster page loads (especially on mobile devices) and much easier to make a "mobile friendly" version of the site using pure CSS.    Just a suggestion, when you've got a couple of hours to spare to modify the scripts etc.

Rick, 22nd May 2023, 05:48

It was originally designed to be compatible with Fresco and NetSurf, with a nod towards IE.  Now, it's aimed at a mix of NetSurf, Firefox, and Chrome.  But it is important that it looks "right" on RISC OS (unlike, say, the ROOL GitLab that simply doesn't work on NetSurf!).

David Pilling, 22nd May 2023, 15:58

Hmm Google says "Have the UK ever won Eurovision?" generation of Googlers can't believe.    Sophie etc is rumoured to be the BBC's target. Not sure bringing folk back is a good idea, memories of Cliff Richard.    Pity that 'repussion' is not a word.    The W3C checker found 31 errors and 2 warnings. Nice to fix what it finds but life goes on just fine without fixing anything. Just interesting to see what it finds.    cannot generate system identifier for general entity "keitai"    element "DEFTAG" undefined

Rick, 22nd May 2023, 18:05

Hi,    What line is giving the system identifier error? That happens when there's "&name=" but the parser is expecting "&name=", which it ought not be doing in URLs, surely?    The fake tags "deftitle" and "deftag" are used by the blog system, and it was a pain in the arse to parse the content to omit them, so I just send the whole thing. Not strictly compliant, but browsers are supposed to skip what they don't understand.    In the part that I wrote, I'll imagine it would whinge about br with no / after, ditto img. Whether or not I add the trailing slash depends upon whether or not I remember, which is usually not as it wasn't needed until HTML decided to be more like XML.    As you say, life goes on. I aim for it looking (mostly) correct on the NetSurf, Firefox, and Chrome. I dropped IE ages ago, and while I used to support Safari I don't any more as my iThingy is only iOS 7 which doesn't support SNI. 🤷 Whatevs.

David Pilling, 23rd May 2023, 02:30

"What line is giving the system identifier error? "    <link rel="alternate" media="only screen and (max-width: 640px)" href="https://heyrick.eu/blog/index.php?diary=20230521&keitai=1" >    There's a load of stuff about the problem, to quote one bit:    "Entity references start with an ampersand (&) and end with a semicolon (;). If you want to use a literal ampersand in your document you must encode it as "&" (even inside URLs!). "    Oddly enough when I tried the checker on my web site, it complained about the /'s after br etc. That is a convention which came and then went.

Anon, 23rd May 2023, 10:03

David is correct, the & sign has to be encoded as an entity even in a URL string. As for the trailing slash on an empty element (such as img, br, etc etc), this was recommended in HTML4, mandatory in XHTML and XML, and depricated in HTML 5. As Rick would say, go figure.    Safari and Chrome are both Webkit based, so what works on one will (theoretically) work on the other.    Generally I test stuff on Firefox (Gecko) and Edge (Webkit). I don't use Edge regularly, but it has the same rendering engine as Chrome and Safari so is useful for testing. Although one nice feature in Edge - it can generate a QR code on-the-fly for the current page.    As for RISC OS - if I recall, Netsurf supports CSS. Fresco? Is that still going? Used to be a decent browser back in the 90s until Acorn Browse turned up. (A real shame Browse never had CSS or Javascript support though, CSS in particular.)    If you're that worried, just use user-agent detection to spit out table-based markup for the layout if the script detects "Fresco" in the user-agent string. Otherwise send the HTML5 and CSS version. (Trust me, it's several orders of magnitude quicker to load and render when the HTML contains only content, with all the presentational elements in the CSS file.)    As for ArcWeb... is that even still a thing?

Anon, 23rd May 2023, 10:08

(Can't edit comments to add something further, never mind...)    I also noticed you have '<base target="_top">' in the code so you can 'break out' of a frameset. I recall coming across a Javascript hack to do this, however nowadays you just put:    document.write('<base target="_top">')    in the block of Javascript, just before the line "if (self != top)". Because it's written into the document using Javascript it won't fail validation.    I'd still consider switching to HTML5; it's been around for nearly a decade and a half now.