Why do you still use Windows XP?

"J. P. Gilliver (John)" wrote in
:

There's this notion that W9X is the Neanderthal of the human species,
but actual Neanderthals are now known not to have been separate, but to
have been

Hmm. I wouldn't go as far as "known" ... results from, AIUI, one site
_suggest_ that to be the case, and even if it is ...

a part of the genetic legacy we all have now. So the question is: Do we
build

... we don't _all_ have it, only (roughly) the northern of us.


Well I hope I have some. A bit of extra genome can't be bad. Although if I
bleed green anytime soon I'd revise that view. (X-Files joke).

In message ,
"(PeteCresswell)" writes:
Per Char Jackson:
...I'm surprised to hear that a minor shuffling of folder
structures caused you more than a few seconds of downtime. It
shouldn't have taken longer than that to see what the new folder
structure was and get back on track.....

Maybe this is more a comment on my lack of gray matter... but, as
another computer user with decades of experience, I post rants on
a regular basis on the subject of Microsoft's gratuitously moving
the furniture around.

Yeah, coping with a given function may only be a few lost
minutes... but those minutes add up over many functions and
hopping back-and-forth between OS' doesn't help any either.

Indeed. A point not mentioned (I suspect they do see it, but would
rather not admit that they do) by new evangelists.

If it were only the few hours total that a retail user like me
lost, I guess it would be no big deal in the grand scheme of
things. But for a company with 20,000 employees - all of whom
are going to lose those few hours and maybe more if they wind up
going to some sort of class on the new sys..... I would say
that's a *very* big deal man hour-wise.

It is. FWIW, my employer (a large international company, over many
sites) is mostly still on XP (for the general desktops I mean, not the
dedicated hardware - that includes NT 3.5, 95, 3.x, DOS, BBC Master,
....); I don't think I've seen any Vista. The switch to 7 may be rolled
out some time this year. It'll mostly be managed by our IT supplier.

I would think that, for any new OS, Microsoft would have a
committee to approve UI changes. If, for instance, somebody
wants to rename the "Add-Remove Programs" control to something
else and bury it in a new hierarchy they would have to justify
that change to the committee in functional (not marketing...)
terms.

I would be most surprised of they don't actually have such a committee;
we end users may find their decisions incomprehensible, but I suspect
they are there. Otherwise we'd get a lot _more_ changes-for-changes'
sake.

But that doesn't seem tb the way it is... and the folks at MS
have probably forgotten more than I'll ever know... so go figure.

Hmm, I suspect they know more about the later stuff, but you (and I)
know more about some of the older (including underlying) stuff, just
because we're older - just from assuming some staff there eventually
retiring.
--
J. P. Gilliver. UMRA: 1960/1985 MB++G.5AL-IS-P--Ch++(p)Ar@T0H+Sh0!:`)DNAf

"Dogs come when they're called.
Cats have answering machines and may get back to you." - Phil Musiak

"J. P. Gilliver (John)" wrote in
:

It's generally agreed that Kerio bloated after 2.1.5, which is still
used by lots of us. (Eventually became - taken over by or renamed, I
can't remember - Sunbelt; a good product, but bigger, and also not I
think free. Nice newsletters though.) [IIRR KPF 2.1.5 doesn't work under
7.]


LnS isn't free either. Takeup would be huge if it was, I think. I was running
it on demo for a long time, then realised after updating it and seeing it
remove a small quirk I wanted gone, that I'd already found it indispensible
and it was high time I paid the man, so I did. No regrets about that at all,
except how long it was before I realised how long I'd left it. I think it
runs on MW7, but judging from posts here and some stuff I read elsewhere I
doubt I'll ever need to know. There'e no question for me now, if I need
another OS than W98, it will definitely be a Unix-based form, Linux, or more
likely OpenBSD if I can ever get adapted to it and some GUI shell I like
enough to stay with. Maybe a minimal WXP is viable, too, but later than that,
no way. Better to go with some new Windows written for an ARM chipped
machine, for me. I nearly got into programming for Windows Mobile on an ARM-
chipped PDA, and would have stayed if my sight hadn't meant the small screen
became unusable for me. Tablet machines may change that helpfully, but I'm
not eager to try yet.

KPF... that PF, personal firewall vs packet filter. I can imagine trademark
hawks fighting over that one some time.

In message ,
Lostgallifreyan writes:
[]
If anyone seeing this is interested in the API (or any of the higher level
systems), there's a small guide that works for me, and might help anyone who
otherwise finds it hard going. It's worth grabbing just for the plain
speaking it has to say about GDI leaks.

forgers-win32-tutorial.pdf (Less than 400KB, with helpful images.)
Google will find it I think, not sure where I got mine from.
'Forger' might just be his name, not his occupation.

Thanks for the pointer. From http://www.winprog.org/tutorial/ it's a
270k .zip, though that site also says "The translation and PDF versions
are unfortunately difficult to update, and are based on older versions
of the tutorial. Most of the content should be the same, but they are
missing recent updates and bug fixes." That page seems to be an online
HTML version (which presumably is being kept more up-to-date).
[]
With that tutorial, and an API reference, and a C coding reference, not a lot

and presumably a compiler (-:

else is needed other than a computer and Google to see if other people found
a better way to do something specific. There seems to be more than one way,
every time.
--
J. P. Gilliver. UMRA: 1960/1985 MB++G.5AL-IS-P--Ch++(p)Ar@T0H+Sh0!:`)DNAf

"Dogs come when they're called.
Cats have answering machines and may get back to you." - Phil Musiak

"J. P. Gilliver (John)" wrote in
:

About the overlapping but different interests, that relates to that
firewall bit I posted yesterday, trying to satisfy a wide user base can
make them unreliable, as conflicting interests arise. So modularity is
the only way

Ah. Thought you were talking about the OS in general, sorry.


Both.. I was thinking the firewall-specific context strengthened the general
context.

out, or alternative choices. Micosoft's problem is that for decades they
actively conspired against BOTH, which is a hell of a foot-shoot. It
likely explains why Apple are so dominant now too. And the increased
takeup of

Sorry, not sure I'm disentangling your sentence (I know, I shouldn't
talk!): are you saying modularity is what's helping Apple (i. e.
"apps"), or something else?


I don't know if Apple's stuff IS modular. Never used any. But I do know
that M$, conspiring against both modularity and choice, has driven people to
seek alternatives, hence the extreme growth in uptake of Apple and Linux
systems.

Whichever KEEPS all those users will likely be the one which responds most to
them. And you can't please them all equally with a monolithic system, ergo
the system that most favours internal choice (modularity) wins. OpenBSD knew
this at the outset; it's one reason I'm attracted to it.

"J. P. Gilliver (John)" wrote in
:

Linux. I think an even bigger change is going to come when people who
were

[I don't think Linux is going to make much impact any time soon. It's
best chance was a few years ago when netbooks started to appear, and a
few with a (very strapped down, i. e. the user wouldn't know _what_ OS
it was) Linux OS came - but they disappeared. From the high/main street,
I mean.]

used to microcontrollers as toys start reaching the age of 20 or so. Not
long now.

If you mean they will have experience in compact coding, I hope you are
right, but I fear not - even in microcontrollers, hardware has kept pace
with the requirements of bloat )-:.


Wasn't Android based on Linux? Or was it the ex-Psion thing, Symbian?

About those coders, I don't know, can't really predict that. But I do know
that using C and the raw API, while momentarily fretting over not having
access to wxWidget's SpinCtrl, I have to make my owm(!), this is tuning out
faster and easier to figure out that compiling wxWidgets on Windows is. Even
if I did figure that out, I'd forever be fighting against vagaries of other
people's code layers. It really IS liberating not to have to do that. If some
of these new coders get a taste for that, it could actually get fashionable.


Even if all they do is statically compile instead of dynamically, bloat may
be reduced in cases where few programs need a single common shared library
(DLL). The original principle of the DLL was to have as few as possible,
shared, to avoid duplicating of static linkages, which might bloat a
collection of software. But people ended up with 'DLL hell' and bloat anyway.
Evading the core issues hasn't helped at all, so those coders who learn to
cut to the chase might be the ones who lead the rest out of the mess. Even if
they repeat it, it would be better. Bloat better than bloat-squared, for want
of a better way to put it.

Hardware keeping pace with bloat? Curious. Likely true, but I hope it's
normally a case of filling available capacity. Good habits come from best use
of limited capacity. That's basically why I think the ARM chips will help a
bit. Big and fast enough to be very powerful, small enough to be unforgiving
of the REALLY bad habits. And most of their coders might be people who moved
up from 8 bit and 16 bit controllers where coding discipline is very
important.

"J. P. Gilliver (John)" wrote in
:

Ah, now there I agree with you that some people write far too much in
the way of comments - but that shouldn't lead to bloated code: a
compiler ought to remove comments, and AFAIK all do. There's the
half-way house of debug code, the compile-time flag (and which the wrong
use of was alleged to be the reason Netscape 6 was such a dog that it
killed the company), but that's not quite the same as comments.


Comments won't, but the tendency to sprawl dissolutely all over a virtual
page might. I like my code to occupy as small a space as reasonably
possible, in every sense. That way the function drives the form, and elegance
is a real thing, not a conceit. Code written like that tends to be self-
commenting, to a large extent.

Human perception is weird. We can barely hold in mind groups of disparate
forms of varied complexity, yet I read that a Beduin or Tuareg shepherd (I'm
being careful here not to sound entirely like Johnny English) can count a
herd in one glance. Whatever that means, it clearly implies that the more
compact the form, the more instantly a brain can grasp its nature, and avoid
making errors of judgement about it.

"J. P. Gilliver (John)" wrote in
:

Imagine having to rebuild part of a house, using the parts that survive
dismantling of the original structure. it is always easier to unwire
socket boxes and put them aside, than have to drag them all still
attached to original wiring and somehow make that fit the new walls! But
in a lof of high

But only if you made a note of which coloured wires went where (-:!


True, but that's why there is a consistent system. Like all the best
protocols, wiring codes are actually very simple. There are exceptions
(usually in parallel computing links, or in deliberate obfuscations for
burglar alarm wiring), but not many.

have. No-one would expect to drag the carpet from room to room with
furniture still in place on it. For whatever reason, people don't want
to beleive that

But equally, they wouldn't always move everything to another room just
to do a little dusting. (I've forgotten where we're going with this
analogy!)


That is true. For small frequent maintenance, it just needs to be less than
totally rigid. I'm not sure how to analogise that with code. Maybe because
it's usually easier to just copy from a new clean source at need. (I'm a BIG
fan of Ghost.

"J. P. Gilliver (John)" wrote in
:

Ah, your familiarity has made you forget. Going from 3.1 to 95 _did_
mean accepting quite a few changes

Indeed! I saw that a month after first seeing W31 on that second hand 486 I
mentioned. The change was immense, and extremely gratifying. I think many
remember that feeling, and it might account for the disappoinments in recent
times. Even when a newer M$ system pleases someone, I doubt the visceral
sense of increased power, liberation and comfort is ever as great as that
between W31 and W95. The next jump in bit-depth, from 32 to 64, was quite the
damp squib in comparison, and many people are complaining of restrictions,
not prasing a new freedom.

"J. P. Gilliver (John)" wrote in
:

And most users just fire up their
favourite applications and care little for the underlying OS - and use
the default folders for everything.


This is why I hope virtual machines that run efficiently on various hardware,
with no extra major OS layer to run them in, will occur. People have tried
i386 running native on ARM, but I'm not sure how much progress there is. But
it could be a lot more liberating in the end, than having to run a behemoth
OS before you can start thinking about running a VM! The VM ought to run
right on the hardware itself, like protected mode does now.