Posts by Havin_it 1227 publicly visible posts • joined 1 May 2008
Seems the real issue was the "hundreds of times" part, which was just sloppy coding. The webapp could have been patched before the switch was thrown, so it'd stop trying to pull content when it wasn't logged-in. Apparently this didn't occur to anyone (who still worked there).
Thanks for posting this background, I thought I was losing my marbles after reading the article.
We too got told in May '21 that Gridhost was going bye-bye and everything would be migrated to their CPanel platform. It was the predictable culmination of the preceding few years' litany of unheralded crap-outs (more niche and specific to what we were up to, none made headlines) and support tickets that invariably petered-out with words to the effect of: "We're not interested in solving this because this is a legacy platform anyway."
Established that their CPanel offering didn't do everything we needed anyway, so jumped ship as rapidly as possible. Quite amusing to hear that this too fell apart in their typical fashion. They may regret asking why we wanted to leave, I had quite a backlog of material built up by then.
NB. We were grandfathered-in from Hostroute, if anyone remembers them. Just imagine: knowledgeable support responses from those who'd actually built the platform! The past is truly another country.
Provided the code is documented in-line using the standard scheme most languages possess or have had imposed (and $DEITY help you anyway if you inherit a codebase that didn't adhere to this), any editor worth its salt will display the documentation in the autocomplete tooltip.
I'd say it is more the opposite way around in practice.
Corporates (the big-enough-to-be-influential ones) expect a high degree of lockdown capability *that they can control*. Letting MS or anyone external be gatekeeper of that lockdown, not so much.
It's cheapskate consumers who get properly reamed, because they aren't spending enough to have a vote, and most aren't bothered by the impositions the way power-users are anyway as it doesn't affect their lives (usually).
That's why the Corporate/Enterprise versions of Windows XP/2003 (can't speak for later iterations) had license keys that bypassed the fiddly activation process consumers were subjected to. That's why "business-grade" laptops have configurable SecBoot whereas cheap-shit consumer ones have it welded shut or so nobbled that it takes serious emotional investment to workaround in order to get Linux booting.
That would certainly be the most frictionless (if costlier) option, but the laptop travels with me and that "needing Windows" moment is by definition going to be unforeseen and most likely short-notice, so it's a bit easier than carrying that bagged drive (plus spludger, torx driver and whatnot) over hill and dale.
The sad fact is that laptops that are wedded to Windows often have no other way of performing low-level system operations. In some cases the BIOS may not even function correctly unless Windows at least remains in place. Likewise for some external hardware that needs Windows to initialise/update correctly but can otherwise function perfectly well under Linux. So your prescription is a bit glib.
(Now you may reasonably respond with "Then don't buy such cursed hardware!" and in spirit I agree, but for the laptop itself there is often a sizeable price/availability tradeoff in avoiding it, and for some bits of external hardware there may simply be no alternative at all. Again sad but true.)
I'm still of the "keep it just in case" orthodoxy but I'm quite glad I've never yet had to go there on this, my first Win10 machine (2017 model Thinkpad bought 2 years ago).
Just that brief visit to fiddle around and turn things off when setting up reminded me how much Windows nobbles the machine's performance. And that was while not mired in the interminable, and now un-declinable, updates. The next time I need to nip in there to do something, things will be made all the worse by it attempting to inflict 2+ years of updates. I'll want to get in and out quick enough it won't get the chance to apply any of them (with, as you say, the risk of new and hilarious destruction of my UEFI config/storage setup/who knows) but it'll be gruelling.
Suddenly I can see the appeal of a Linux-certified laptop ... shame they don't tend to certify for Gentoo :(
I've found that the vast majority of the dot-files that accumulate in my homedir aren't missed when I re-home (usually moving to new laptop). Most tend to be boilerplate default stuff that doesn't change much after install of whatever software. Also by convention they are almost never deleted when you uninstall/stop using the app in question, so genuinely useless cruft builds up (~/.kde4 dir etc).
That said, on every migration I dutifully scoop it all up and plonk it in its own folder in the new homedir, from which I dump the obvious cruft (thumbnail and shader caches etc), transpose the bits I actually know to be useful, and keep the rest for reference/in case I missed something. I've got three laptops' worth sitting around now - I'm a hoarder.
Another issue worth being aware of when directly re-homing a homedir as a whole partition is that just because you give your account the same name on the new system doesn't mean it'll have the same numeric UID, which is how file-ownership and permissions are dictated at filesystem level (not to mention ACLs if those are in use). Different Linuxes won't necessarily give even the first created user the same UID consistently, so you may need to do some chown'ing before you can get to work.
In addition, if like me you dealt with the /var issue mentioned above by moving Apache stuff into your homedir instead, the same issues will apply to any system-account file permissions used by those files (with even less likelihood of consistency between distros). Learned this one the hard way.
If Linksys appears on any poster relating to this it's likely a satirical one. They had no intention of any of this, but they went and copied some GPL code and inadvertently violated the license. When they got found out, they were (eventually) forced to open-source the firmware.
They learned their lesson and used VxWorks after that.
Oh, it's a thing. My local (in Leith) will do one on request.
For a while (~20yrs ago) there was a real silly-season where every choccy/confectionery item imaginable was given the treatment in the name of science.
Nowadays however ... I recently asked the aforementioned local if they'd do me a Crunchie (I remembered hearing these were especially good) but was declined. They told me they only do Mars bars because it is unique among mainstream choccy-bars in being certified nut-free, so on those rare occasions they actually do one it won't contaminate the fryer.
For one thing they have removed the "Back list", to wit: a long-press on the (browser or OS) back-button used to bring up a list of the previous pages for the current tab, allowing you to jump back multiple pages. Now it doesn't :(
Just my first severe pet-peeve after half an hour's cursory exploration. More to come, I'm sure.
I think you overlook the amount of resources MICROS~1 have to sink into making Windows compatible across every rando whitebox PC in all creation. (Never enough, but hey at least they try.)
That's their choice and the ubiquity it's earned them has made it worthwhile I guess, but I can't really blame Apple or anyone else for not wanting the hassle.
Would be nice to have included a cursory overview of how this tech actually works, no?
Sounds like something I could have used a week ago, when one of my cheapo buds leapt for freedom during, as luck would have it, the tiny section of my several-mile bike ride that involved unmaintained scrubland. Cue many perplexed passersby while I crawled slowly around with one ear to the ground hoping to pick out the faint strains of Prodigy's Warrior's Dance (the most percussive and trebly thing came to mind) from among the tussocks. Glad it narrowly missed the adjacent thistle-patch.
Yup. If there's one thing I'll always thank El Reg for it's having made me aware of the Squeezebox. Two original units left in service here (a pair of Radios operating as a stereo set). My original SB went dark a few years ago, but by then I had a Raspberry Pi Kodi box sat next to it, so just added the software client (Squeezeslave) to that and ran it into the amp. Miss the remote (and playing Tetris on the retro LCD display) but the phone app suffices; there is a Kodi module to provide UI, but I've never gotten it to play nice.
I suspect the software ecosystem may outlast the hardware! There was some concern when the online companion portal was shuttered that this would stop the server or devices from working, but this seems to have been averted. In any case there remains a strong dev and user community, so even if this should happen later a workaround can probably be found. With the server and client apps being open-source, there's always hope; even if/when bit-rot sets in to the point they can't be easily installed on modern devices/OSes, they can be kept as VM images or slapped on an RPi as an appliance.
For all of these reasons, I've never viewed the Sonos "experience" with anything beyond pity.
>If people kept calling you by the wrong pronouns
Here's the thing: You and Brandon 2 might have different ideas about what the "right" pronouns are for him/her/xim/etc.
What unsettles me about this issue is that people are staking a claim on a language feature whose historic function has been to convey information about their inherent characteristics, and wishing to make it instead convey information about their personal preferences.
I say preferences because, as I understand it, the pronouns that one wishes to use are entirely free-choice: the fact a person wants to go by "xe" does not map to anyone else's reasons for picking it.
As such, that person is obliging others to show deference to them in an entirely arbitrary manner; to convey recognition of characteristics that are not even defined except by them. To fail to do so, by dodging the issue with indeterminate pronouns or simply avoiding pronouns, is deemed abusive.
I recognise that we're in this mess due to gendered pronouns ever having been a thing in the first place, which in hindsight was a rubbish idea: how much simpler things would be if we did not expect pronouns to tell us anything about the subject, and I'd wholeheartedly endorse a campaign to eradicate them now if we could, but we are where we are.
I honestly feel that, apart from being senselessly coercive in respect of the above, this movement is self-harming in that it seeks to curtail others' ability to signal that they do not care about the subject's sex, gender or other identifications. Surely, in the world we strive for, 99% of the time such things will simply not be relevant?
It's not exactly new. They/them has long been a recognised pronoun for a person whose gender is unknown/unspecified in the speaker's frame of reference. (Ditto when number is not known, so the subject may be either singular or plural.) I assume that is what this form of usage derives from, and it's hardly a quantum leap. If a person wished to identify as nonbinary/gender-fluid, it makes a lot of sense to me, although I'm not prescribing it as such as I'm sure everyone who chooses it has their* own rationale.
*See? Perfectly grammatical and normative, and typed it without even thinking about it!
And you in turn are half-right.
It was indeed singular they/them she objected to (she was happy to use neopronouns if it came up). However this, according to a trans-community member who I believe was party to the original "discussion" and had a subsequent clear-the-air talk with Monica which she discussed in a post of her own, was for "personal" reasons that Monica did not wish to disclose.
In fact you have to run (build, configure and keep updated to maintain client compatibility) your own sync server (python), accounts server and content server (both nodeJS). With, last I looked, public docs that are an afterthought and you're doing very well indeed if you don't have to throw yourself on the mercy of the services-dev mailing list before long. Then there's the client configuration ... [twitch, dribble]
I gave it a red hot crack and it did work for a while but honestly, life's too short especially if you're not conversant in python, node (and mailing-lists). I thought I cared, but given my data is all encrypted client-side I really don't care enough to justify that much ongoing grief.
Agent Tick does have a valid point. What is not mentioned in TFA is that with 71 (and beyond), each update will create a new profile. The only way to retain access to saved passwords, autofill, etc. is by handing your data to Mozilla. There is no provision to do a local import from an older FF version.
Got a source for this? I can find no details about this anywhere.
The LO equivalent, under development deep in the TDF skunkworks, is called Stapley. When you let Stapley "help" you, you won't be able to undo the changes it makes, because it can't possibly have come out differently to what you wanted, right? Stapley knows best.
A couple of years after Stapley is unleashed, a motivated bedroom-coder will release a plugin called Staple-remover-y, which obliterates anything Stapley did that actually _did_ align with what you wanted, but still leaves your pages warped and slightly torn.
In the meantime, the user community will get used to a workaround which, had anyone the inclination to give it a name, would probably be called Reprint-the-whole-thing-and-just-use-a-paperclippy.
Nope to your nope (though nope to the quote too).
What you call "extensions" are plugins (e.g. Flash, Java, DRM modules etc). What you call "add-ons" are extensions, apps written in JavaScript targeting the WebExtensions API. These are a sub-type of add-on, the others being themes, search-engines and language-packs.
Humanity is wonderfully diverse, there's a matching pair (or perhaps more) of people for any particular 'kink'
These statements are in opposition to each other. The greater the diversity, the lower the chance of every kink occurring twice.
/forever alone in my crab salad paragliding kidney-massage fetish :(
The two are equally deserving because they demonstrate that the initial assumption was flawed.
If you want to put it that way. It was an educated projection of the advance of processing performance over three-and-a-half decades, which was always going to be a tall order since the kind of innovations that drive such advances are by definition not yet conceived of, and might even follow a complete paradigm-shift (quantum computing being an arguable example).
Even without someone discovering a mathematical shortcut, always a possibility,
Rivest assumed not, and that assumption has not yet been shown to be flawed.
with a slightly bigger hardware budget (like that of the NSA) the problem could be solved even faster.
Nope. The problem cannot be parallelised, so the performance of a single task is the bottleneck. So the only way an infinite sack of cash can help is by paying people lots of money to invent more efficient
hardware (which is no guarantee of results).
Rubbish. The article explains quite adequately the (very simple) reasoning: Edgium sends your data to Microsoft, not to Google. That data is, nowadays, the only compelling reason for being in the browser business (not counting those Mozilla loons who do it for fun).
And the investment required to put your own UI skin on someone else's engine is piss-all compared to maintaining your own engine. This move probably let them shed a few developer salaries.
I bailed from Pipex when Tiscali got it, that was never going to go well. To... Demon! Who proceeded to be borged by THUS -> C&W -> Vodafone. The Demon branding was upheld for a while and only finally binned a couple of years ago. Voda have generally stayed out of my way, so I've stayed put. In the way of these things, though, I expect I'll be subscribed to Disney Broadband before long :/
To my mind, however, Microsoft hit a high point for event audio with Windows 95. Subsequent releases of Windows have never reached such glorious levels of gratuitous audio; indeed, the startup and shutdown sounds these days are little more than clicks or bips.
How dare you overlook that Vista's "soundscapes" were imagineered by bleedin' Robert Fripp?
People gonna forget how rock'n'roll MICROS~1 used to be ;)
If you're on a multi-user machine, ensuring "your" stuff is insulated by the correct file permissions by default is valuable: I tried making a "personal" folder under C:\ (which last I looked is world-writeable by default - WTF?!) but ensuring the permissions were set to propagate properly was a screaming nightmare. Windows permissions editing ... shudder.
I stopped using [user]\Documents when ransomware became a thing, as this would likely be a default target. The weird symlinking shenanigans to which Documents and its ilk are subject make them problematic for doing backups as well.
As for AppData, I've never grokked the philosophical distinction between Local/LocalLow/Roaming, and it's pretty clear app devs don't either. I never know which one I'll find an app's settings in, let alone why. Some even use more than one of them(?)
At least the FOSS apps can be relied upon to ignore all of them and create a nice predictable [user]\.[appname] folder :)
As I recall, the main target of Linus's anger at the time was Andrew Tridgell, for coming out with the reverse-engineered BK client that provoked BK's owner to withdraw the free-beer client license for kernel devs. If git was named for any one person at that time it was probably Tridge.
However, the very swift appearance of git in the wake of BK's move did make me wonder whether Linus had already been working towards dumping BK for some time, having tacitly acknowledged the complaints from many around him about the risks of adopting it in the first place.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2023
