Re: OT - I ran....

On 2007-09-14, Rowland McDonnell <real-address-in-sig@xxxxxxxxxxxxxxx> wrote:
Whiskers <catwheezel@xxxxxxxxxxxxx> wrote:

[...]

But most Linux users (like me) are quite happy to use
non-free and non-open-source stuff if it does what we want or need.

Sensible. But there are problems; I recall reading of a bloke who used
to maintain a Linux driver for a particular firm's Webcams.
Unfortunately, the `Linux community' decided that because it included
proprietary software in object code only form (which the Linux driver
maintainer had persuaded the commercial firm to let him use for this
free driver, with reasonable restrictions such as not letting the source
code out), it couldn't be included in Linux distros. Since that meant
you'd have to re-compile your kernel to get it fitted (so I read), it
was too much bother for most people and so lost most of its usefulness.

Some distros are maintained by people who insist on 'purity' (as defined
by them); you won't find 'proprietary' or 'commercial' or 'closed-source'
stuff in their installation files - No Sir!! But other distros are more
flexible and happily use 'what works', or give the installer the choice.
(Mandriva Free is a 'pure' distro which can easily have non-FOSS stuff
added if you want it; the 'boxed sets' and other variations of Mandriva
have such things as "MP3" support and proprietary hardware support built
in from the start).

Normally, a kernel can be compiled with 'modules' for particular tasks
built in, or you can have them as seperate entities which are run as and
when required. Compiling a kernel isn't all that big a deal, particularly
if you're developing a distro anyway - many distros have 'customised
kernels'.

The maintainer decided that since he was getting so much opposition from
the rabid fundamentalists, he was going to drop the whole lot.

Perhaps he wanted to remain part of a 'purist team' for other things? For
example, Debian is a 'purist' distro with a large number of highly
accomplished developers involved; splitting from such a community over
something relatively unimportant would be a big step for someone also
involved (or wanting to be involved) in lots of other projects in that
community. 'If you want to be in our team, you have to play by our
rules', which seems reasonable.

The purist approach would be to develop a wholly independent driver for the
hardware, or get the manufacturer to release the source code or at least
the detailed specifications of the kit, or to simply not buy or use
anything made by that firm.

Reminds me a bit of `Cindy's Newsmailler' (or whatever it was called).
It was a quite nice free (as in beer *and* speech) email and Usenet
application for Macs, but written using Hypercard
<http://en.wikipedia.org/wiki/Hypercard>.

I looked at it and didn't use it because it `rang home' so that Cindy
could keep track of who was using her software. But that was not the
problem for many people. No, what happened was that she got piles of
nasty abuse because it wasn't a `proper' application, but written in
Hypercard. Her husband explained on the Web page that she'd had so much
nastiness on the subject that she'd decided to stop development.

I wish these idiots would *THINK* a bit before doing what they do.

Yes.

My
preferred web browser is Opera, which is 'free as in beer' although not
necessarily 'free as in speech', and certainly isn't open source.

I'm not enough of a programmer for the open source aspect to be
particularly significant for me, although I do like the idea that anyone
who knows how and cares to bother can find and fix defects in a program or
come up with useful variations or additions to it.

I like that too. But it strikes me that the way in which typical modern
FOSS projects are handled does not usually result in high quality at the
moment.

I don't know what a 'typical modern FOSS project' is. Mozilla (Firefox,
Seamonkey, Thunderbird, ...) is vastly different from Jedsoft (one
person).

There does need to be a central controller to keep the thing on track
and up to quality - not necessarily a single person, but on the other
hand you can't have too many people trying to do things differently or
you get `committee type' problems. I think the best method is to copy
the military `general staff' idea, where you get agreement on `how to do
it' amongst a group of people, and they all do their bit but singing
from the same song ***, so to speak. Smallish groups of *good*
technical people seem to do this fairly naturally, so it seems.

I think most Linux-related, Gnu, etc, projects, are either one person or a
well-regulated community with a controlling committee or person to keep
the 'official version' on track. Of course, anyone who wants to take
development in a direction that the official version isn't going, is free
to start their own 'fork'; that happens fairly often.

[...]

But an awful lot of Windoze users do indeed get
very cross at the suggestion that they might not have made the best
choice.

Or defensive about not having made a choice at all but not wanting to
admit such ignorance.

I think that's more likely.

(That can produce some of the most rabid nonsense
on the 'net)/

Indeed - like Harley riders. Apparently, according to this magazine
article, H-Ds get there almost as quickly as sports bikes because sports
bike riding involves `an accordion of mad acceleration, heavy braking
for corners, slowing up to analyze the traffic, planning your overtake
and wazzing off into the distance. Harley riding lets you see the
traffic, plan your overtake as you approach, glide past and carry on
cruising at a steady 75mph, maybe just losing 10mph or so for the
turns.'

That's more about riding (or driving) style than 'hardware', surely? I've
seen some pretty daft riding done on a Harley.

I mean, Linux users are perfectly prepared to admit that the ease of use
of MacOS X has a lot going for it; they'll counter with the fact that
they're not tied into particular hardware and have full control over
absolutely everything and access to the source code of the OS - which is
a huge advantage over Macs, but the advantages are very different:
either you really need the one, or you really need the other. I supose
there are fewer points to argue about between Linux and the MacOS.

I consider Mac OS X to be an excellent example of what a good Unix-type
operating system can be. Given enough resources, there's no reason why
something just as good couldn't be produced by someone else, and now that
some computer makers are starting to offer PCs with pre-installed Linux as
an alternative to Windows we might see that developing into a close
hardware/software integrated design that is every bit as 'appliance-like'
as Apple Mac machines can be for those who want them to be.

I get the idea that there are quite a few people on the free Unix side
of things who have exactly the same idea.

I think it's coming. I think that to do it properly, you'd need
something that looked a bit like a research group or small start-up
company to come up with something reallly polished, mind - and for that
to work, all the components needed to make it really polished would have
to be in existence already. But I think it'll come.

I'm rather looking forward to it - at the moment, there's only one
decent, widely used, fully user friendly OS out there, and that is not
healthy.

The free Unixes are clearly moving in the more user-friendly direction,
mind. I just wish they'd get there *NOW*! ;-)

It's a lot harder to get a new start-up going in a relatively mature
market, which personal computing now is, than it was when Commodore,
Apple, et al, were starting out. The 'One Laptop per Child' project is
one approach; Dell now offer pre-installed 'Ubuntu' as an alternative to
Windows; HP have had a quiet line in Linux support for 'Large Enterprise'
customers for a fair while (and I bought this HP laptop a few years ago
'reconditioned' by HP with no OS installed - and an HP-branded Mandrake
(now Mandriva) installation CD tucked into the packaging); Linspire hasn't
folded yet (although I think that particular project is on the wrong track
by trying to be too much like Windows in ways that harm security and
encourage 'bad practice').

Intel and nVidia are pretty good at supporting Gnu/Linux.

Even without that, it is certainly possible to have a Linux-based system
that is just as integrated as any Mac

Umm... *nearly*. You'd be amazed, you really would.

I'll admit that it takes work; but Apple did it, so anyone can if they put
the work in! Few users bother, but an OEM could (and really should).
But the current target to beat is Windows rather than OS X, in practical
terms, and that's a much lower hurdle. Even OEM Windows systems are not
all that well integrated with the hardware, in my experience - and adding
or removing hardware or software on a Windows system can be a nightmare.

- once it has been installed and
configured on hardware that is chosen for its operability with Linux.
That's the part that is often missing with current Linux systems, of
course - we have to make do with computers built for operability with
Windows or Mac and do all the installation and configuration for
ourselves, or hope that we can find a distro that will do it for us and
which we can live with.

Apple's always used standard stuff to make Macs (just not the so-called
`industry standards' (i.e., `one firm's products that happen to dominate
the marketplace) prevalent on the dark side) so if it's built to work
with Macs, it's generally quite easy to get it to work with other
things. Now Macs have gone all Unixy, interoperability is dead
straightforward a lot of the time. And the switch to x86 as increased
the ease of interoperability still more.

So, for example, there are plenty of bits of software for Unix -
including Linux - which have been ported to MacOS X 'cos it's so dead
easy. You can get a Linux distro for a Mac and install it dead easily
if you like.

Yes, you could have KDE or Ratpoison instead of, or alongside, your OS X
interface, if you really wanted to (although it would be a rather eccentric
thing to do <G>).

It's not just a one-way thing - and from where I'm sat, the `straight'
Unix world does seem to be better provided for generally than the Mac
world, honest. There is very little that's designed to work with Macs -
not properly, anyway. For example, my recently bought printer (HP1320N
for `Network' i.e,. Ethernet) come with software that, on installation,
broke the OS. I contacted HP and they instructed me to not install the
software as the supplied instructions tell you to because they know it
breaks the OS. This information was not supplied either with the
printer or on HP's Website - it's like they didn't give a damn about Mac
users. The manual is wrong is several important respects - instructions
for clearing one fault result in an automatic paper jam, for example.
And lots of things simply don't work right due to obvious sloppy bugs.

That's a black mark against HP, and they should fix it quickly. Incorrect
documentation and drivers are worse than not having any at all. (I wonder
how much of the problem is really Apple not giving HP enough accurate
information?)

[...]

--
-- ^^^^^^^^^^
-- Whiskers
-- ~~~~~~~~~~
.

Quantcast