Yes there is and just the spin that in attempts to explain away
vulnerabilities.
Anyone who can't grasp the simple difference between Beta and Stable
releases is beyond help.
I'll agree with you on the state or Windows API. I will also add that
most of the GTK stuff I've seen looks like crap. WxWindows would look
much better.
That's another way to go. Or there's always python, java, or bunches
of other languages that'll get you a portable GUI. Which one should
become the standard that we force all open-source developers to use,
and how do we make them to toe the line?
Tucows is not the only game in town.
Then where is this game? I'm not talking about obscue freeware
applications now, but the kind of stuff that people commonly use. Say,
the kind of applications talked about daily in this newsgroup. How
many of those can you get source code for?
As I said, there are not very many but they are not non-existant.
That's why I said, "virtually".
These things are not "virtually unknown" by Linux, etc. They are very
rare in comparison to windows, sure.
You have a real problem with the word "virtually", but that aside...
how many times have you heard of Linux or *BSD systems being infected
by email worms?
this a bit. For several years now the best and easiest transport
method for viruses, worms, etc has been email. How many Linux, etc
servers do you know of have users pulling up email clients and reading
it on the server?
Quite a few, actually. Distributions like SuSE and Mandrake make it
trivially easy to set up your own webserver. Many Linux users I know
have set up their own little friends-and-family webservers on the same
machine they use as the household desktop. I've done a little of that
myself - when I was putting together the family picture archive I
created a small webserver on my desktop machine so various relatives
could scan and upload their own photos, and add comments to each
other's submissions.
The combination of Windows providing these bells
and whistle to wow the consumer and having the dominate (something
like 90%) number of desktop users makes them the big target. And
don't misunderstand. A lot of it has to do with a poor programming
discipline in regards to security on the end of MS.
Market share has nothing to do with it - Linux has over 5% of the
desktop market and around 30% of the server market, yet it doesn't
have even 0.01% of the worms, trojans, and viruses that Windows
suffers.
Easier to use and having more features does not inherently dictate
anything whatsoever about the security of the system. In fact in many
cases usability of the system is inversely proportional to the
security of that system. For instance a system that requires no
password for it's use is more usable than one that doesn't. Which one
is more secure?
Ask someone how usable that non-passworded system is when malware or
one user's mistake takes out the everybody else's files along with the
OS, and puts their computer out of commission for a week. Usability is
quite often a question of just a little up-front convenience versus
big hassles somewhere down the road. Microsoft users have been
conditioned only to look at the up-front conveniences without
considering the longer-range implications.
Take the argument up with the security experts. I'm just repeating
their words. I'm sure that you know much more than people who have
been working as professional security engineers for over 20 years.
It doesn't take a security expert to see the almost daily reports of
windows worms, many of which have cost businesses billions of dollars
and caused panic on the Internet from their traffic alone. The Linux,
*BSD, and Unix sites are devoid of such panics except to the extent
that we worry about our connections being overloaded by traffic from
all the Windows worms.
Vulnerabilities that allow hackers to break in are another matter, but
if Linux is so much more secure than Windows in the
virus-worm-and-trojan department, that's a pretty good indication that
its networking will also more secure.
The web server doesn't need a GUI. However the rest exposes the
typical Linux mentality. No GUI's for word processors, paint
programs, media programs, web browsers? You REALLY don't want Linux
to succeed on the desktop do you?
Since the subject was security and not GUI's, I figured you'd be smart
enough to figure out that I was saying "Since when does a web server
need a GUI? Since when does a web server need a calculator program?
Since when does a web server need a word processor?" etc.
As for CLI-vs-GUI, I use both and don't much care for any OS that
doesn't give me both. Even when I'm working at the command line I'd
rather do it in an xterm because X-Windows has cleaner fonts than the
blocky ones provided by the video BIOS.
Case in point. WINE. WINE is primarily a mechanism to provide users
with the ability to run Windows programs of which the equivalent does
not exist in Linux either at all or in usability. The primary point
providing usability.
The only usability failure here is Microsoft's and the proprietary
software industry's delibrate use of closed, proprietary, and
non-portable data formats to lock customers into their software and
the Windows platform. WINE lets those customers switch from Windows to
Linux without giving up access to *THEIR* data.
WINE is more-often used by Windows refugees to run a few old favorites
that they just refuse to give up. It honestly doesn't run enough
applications well enough to be a full Windows replacement. WINE was
improving rapidly for awhile, but I think everyone lost interest as
the Linux applications base came up to speed.
Guess what - that opens Linux up to some extent
to some of the same vulnerabilities of Windows. It has been reported
that someone running Outlook Express under WINE became a victim of a
typical email virus that OE was susceptible to.
No, Linux isn't opened up - Windows applications are, which is what
you would expect on any windows-environment simulator that can run
Windows software (including Windows malware). If you go out of your
way there is a means to give Windows applications access to your Linux
files, but by default WINE is a sort of jail. Windows viruses can only
damage the Windows applications in the simulated Windows environment.
It is for just this reason that people who like to play with Windows
viruses often do so using WINE.
With WINE, you can also use user- and file-permissions to protect
Windows 9x files, even though Windows 9x doesn't have user
permissions. I chown and chmod my Windows excutables and DLL's so
they're writeable only by root, and run WINE only as a user, so no
Windows virus can alter the executable. Wow, Windows 98 with file
permissions. Neat, huh?
Another advantage of WINE - though it no longer matters now that WinXP
is out - is that you don't have to worry about unstable Windows
applications bringing down the entire OS. At worse, they crash WINE.
But a lot of people have also noticed that problematic Windows
applications that work under WINE are often also more stable than
under Windows.
If they didn't need it they wouldn't use it. I see it used all the
time.
Of course, bells and whistles are designed for people to play with.
That doesn't mean they really need them, though.
Again this mentality will not go far in the battle for the desktop.
Don't get me wrong. I see beauty and security in simplicity. However
the fact of the matter is, a significant factor why people are buying
this stuff is because of the bells and whistles.
"People are buying this stuff" mostly because Microsoft killed off
most of its commercial competitors through non-competitive dirty
tricks and left consumers with no say in the matter. Windows was not
shaped by consumers exercising free choice in a free market brimming
with choices. Microsoft's design decisions were dictated primarily by
their desire to sabotage competitors' products and by the marketing
necessity of providing some justification for an expensive upgrade
treadmill, not by the need to win sales by making a better product.
Never in their history has Microsoft faced a competitor on equal
footing, for the simple reason that they've found a way to kill off
every possible competitor before they gained enough market share to
compete on product quality.
You may not like it.
To some extent I don't like it, however to compete you have to come to
grips with it. I see Linux gradually implementing a lot of "useless"
bells and whistles on it's desktop as well. Why? They want a chunk
of the desktop market. Otherwise they wouldn't do it.
There are bells and whistles that open up serious vulnerabilities,
and there are bells and whistles that carry little risk. The latter
are fine. I don't believe in depriving users of the former if they
just insist on shooting both feet off - but there's a difference
between giving people what they demand, and trying to create demand
for something they don't really need and that carries serious security
consequences.
Basically on the consumer level, desktop security has been extremely
poor since the beginning with Windows. Until the release of XP home.
I'm not saying XP home is the epitome of a secure system. It is not.
It is, however, a vast improvement over the previous consumer desktop
offering. That being said, MS is still way behind and playing
catchup.
I don't consider XP an improvement at all. Worms still run rampant,
we still hear about a new vulnerability almost every day. But I never
got that far - WPA alone was enough to send me to open source. Then
there's DRM, and all the other attempts to take ownership of our own
machines, software, and data away from us. If the hackers aren't
trying to get you, it's Microsoft or the RIAA. I don't consider that
an improvement.
Not all large corporations have the same needs.
How many corporations need executable content in their emails? Can you
point to anyone who uses this feature enough to justify the daily
deluge of Windows worms that corporations and end-users the world over
have to deal with every day or the billions of dollars that email
worms cost businesses every year?
These are mostly aestic "features" (except for the task bar). We
weren't talking about those so I'm not sure what your point is.
My point is that Linux doesn't do as much copying from Windows as it's
so often accused of by Linux detractors. I could have used more
technical examples but they're more complicated to explain.
About the only things I can think of where Microsoft got there first
were WPA , DRM, and embedding executable code in document formats. I
for one can live without all of those. But as usual under OSS, anybody
who wants them under Linux is welcome to start a project.
I've seen it and used it. Overall it looks very good. But again
that's just aesthetics. However the standardization/organization of
virtually every desktop that I used in Linux quite frankly is poor.
version blah has the cancel button on the left, version blah+1 has the
cancel button on the right. Version blah has a terminal launch button
on the task bar, version blah+1 does not. Version blah has a somewhat
useful package manager, version blah+1 has no package manager.
Applications don't have the same look and feel and even basic
operations are different. Consistency in the various GUI's isn't there.
I don't disagree on the package mangers. For everything else you're
making mountains out of molehills. So the OK button looks a little
different, or is on the other side of the dialog box. Big deal. The
only place where I have a usability problem with consistency is
cut-and-paste, which was fine until some developers, in an attempt to
appease Windows refugees, started replacing the usual X-Windows
mouse-oriented technique with MS-Windows bindings. Now I never know
which method to use. However if I don't like that, I don't have to use
the applications that don't conform to the way I think things should
be done.
Standards, by their very nature, must resist change. They require that
each new idea has to be sold to the decision-makers and then herded
through the standardization process and officially adopted before it's
allowed to be implemented in applications. Standards can also keep an
OS from evolving - imagine, for example, if Linux's official standards
were all set by the geeks who dominated the Linux movement 5 years go.
Would we have user-friendly installations now? Would we even have a
KDE or a GNOME?
In open source, the people at the top deliberately keep standards to
an absolute minimum so that new ideas can be freely offered and users
can decide for themselves which ones will become "standard". The
minimal standards that are enforced only ensure portability between
Linux systems. Everything else is allowed to free-run in a belief that
the best ideas will bubble to the top over time.
In this respect OSS is not really that different from proprietary
software, except that the direction of proprietary development is
determined by the desire to maximize profits. In a truly free market
this would be laudible and lead to great things, but the reality is
that "getting ahead" nowadays is more a matter of destroying
competition by methods having nothing to do with free-market
competiton, and then forcing locked-in users onto upgrade treadmills
by producing shoddy products.
The mere existence of the OSS movement is proof of the failure of the
free market to meet consumer needs. All those people, working on all
those projects, represent pent-up demand. Microsoft refused to satisfy
that demand and they also refused to allow other companies to do so.
So, the demand bypassed the marketplace.
And apparently a lot of Linux developers don't understand how
to put together a decent RPM. I get this RPM and install it. Now
where is it in the menuing system or "start menu". Nowhere to be
found. I have to open a command line up and search for the GUI
executable. Apparently they don't have the common sense to realize
that if the application has a GUI front end that someone might
actually want to use it and therefore there should be a link to it
easily available from the desktop. I've installed many RPM's and
scant few of them did this. I could go on with more but I think I've
made the point. There is much more to a very usable desktop and GUI
design than just looks and a few basic useful mechanism
I won't disagree there. I've never cared for most of the Linux
installers (except for Debian's apt-get) and originally went to the
BSD's precisely because of their ports systems. If you're not familiar
with the BSD ports systems, here's how you install Mozilla under
FreeBSD:
cd /usr/ports/www/mozilla
make install
All dependencies are fetched and installed automatically, with MD5's
being computed for all fetched files and compared to an on-system
database before the files are even unpacked. Beats the heck out of
RPM! There are also automated tools to update your applications to the
very latest versions.
BSD's are for more-technical users so as you might imagine its fans
prefer to admin their systems from the command line, but it would be
child's play to write a GUI installer for this system. In fact there
is one - pib - but it keeps falling behind due to lack of interest.
Gentoo is emulating the BSD ports system with their 'portage', and a
lot of other Linux distributions are thinking of adopting the change
for themselves. Something, I might add, that would not have happened
if RPM had been adopted and forced on all distributions as a
"standard" back when it had everyone's eye. This is why in Linux,
standards are allowed to bubble up from the users through a
competitive process, rather than being dictated to all by a group
vote.
I often wonder if Linux users are just lucky in having people at the
top who understand this, or if linux became so popular because the
people at the top *do* "get it"?
You have a point, but from my perspective stuff like Gnome would not
exist if it weren't for the pressure to try to come up to speed in
usability to Windows and the Mac. In the last five or six years Linux
usability from the GUI perspective has dramatically improved. But to
claim that it has arrived to be equal is not true. There is plenty of
work still to be done.
The capabilities already exist, many (though not enough to satisfy the
Windows refugees) applications have adopted them, but there are a lot
of applications floating around that do things other ways. Many people
have been doing things differently for years and don't want that to
change, and Linux isn't about forcing them into line. That's why the
KDE group, for example, decided to start from scratch and develop
their own applications conforming to their own standard. Despite that,
KDE applications also work under other window managers, and vice
versa. It's like being able to run Apple software under Windows and
vice versa. No doubt if that happened you'd complain that it's bad
that all the applications no longer looked and felt exactly the same
now?
Windows users get all starry-eyed over odd-shaped Windows applications
software with futuristic widgets, yet you seem to think they can't
handle the relatively minor differences between X-Windows applications
written using different GUI toolkits.
Part of the problem is with the nature of Open Source. It has it's
good points and drawbacks. One of the things that inhibits the
ability to achieve the usefulness that Windows, Mac has is the
resistance to embrace GUI standardization wherever possible.
The resistance exists because the userbase is still made up mostly of
technical users. But unlike Microsoft and Apple, we don't force
everyone to do things one way. So, Windows-like environments and
applications are not banned from development by some standard written
by and for geeks. Similarly, if Linux should become outrageously
popular and the ex-Windows crowd takes it over, we geeks won't be
forced to suffer Windows-type standards.
That's something Windows users really have trouble with - multiple
standards. The Windows mindset is that you MUST have a GUI standard
and it must be THE standard, and nobody is allowed to deviate. Well,
what if I want to run a GUI on an underpowered machine? What if I
value speed over artistry, or vice versa? What if I want to run a
CLI-only web server, but the OS powers-that-be have decided to do away
with the GUI?
Part of Linux's appeal is that it does allow so many different
standards to co-exist. The price you pay for many choices, is that you
don't have one big fascist standard forcing everyone to toe one
person's or group's (or monopoly's) One and Only Standard, but many
smaller standards freely adopted by various groups according on their
needs or preferences. Sometimes you just have to accept that free
speech also means you have to tolerate the other guy's rantings.
In fact, it would be impossible to enforce such a standard because
open-source is *OPEN*. If people don't like your standard, they'll
tell you to take a hike and fork the code into a project that works
exactly the way they want it to. Wherever there's a need and enough
interest to support the effort, the OSS concept allows software to
evolve to fill the need.
With
Windows/Mac, the user knows where all the controls are likely to be
and that if they install a program links to it will appear in the
Start menu, etc, etc. The other related thing is that of the
independent spirit. Standards are resisted because it "stifles my
work/creativity/whatever". That is true to some extent, but to spurn
standardization is to inhibit usability.
Again, you're taking a short-sighted view of usability. One single
fits-all standard makes life easy up-front, but it also means that a
lot of people don't get exactly what they need and it doesn't even
allow software to be introduced and rapidly evolve to fit that need.
Case in point. There was an article posted on /. last week where
someone suggested that the Linux desktop should be standardized to
increase usability.
He deserved to be roasted. Since no one window manager suits all
needs, some users should, according to this idiot, be prevented from
having a GUI that works the way they prefer even if they and 10,000
other volunteer programmers with similar needs wanted to get together
and write it on their own free time.
