Sylpheed has good support for GnuPG, and is my favourite MUA on Linux.
The drawback is: I would like very much like to use the same e-mail client on Linux and Windows, but sylpheed is
only theoretically cross-platform. On ftp.gnupg.org, there is a w32 build of sylpheed
0.4.60 which is buggy like hell, and I have no idea how it was compiled (otherwise I would rebuild a newer version).
It's much easier to write network applications in Java than C, and cross-platform compatibility is far better.
Now that Apple has turned to BSD, there is Unix and Windows left for the desktop.
And last time I checked, JAP (Java Anon Proxy) worked ok on Win98 (except that Win98 occasionally crashes..), but had huge problems on Linux (incomplete page downloads) - which might tell something about the "portability" of Java.
Yes, there are alternatives (there is also
BerliOS, as some posters have mentioned), but none
of them offers the one thing that I personally
find most useful at SourceForge: the compile farm.
I do not host/offer my source code at SourceForge,
because my project uses strong encryption, and I prefer to host it in a country (Germany)
with less nebulous laws in that respect. However,
the compile farm is an invaluable resource for
testing compatibility with other platforms.
Actually, I can understand that the FreeBSD project may not
have the resources to offer a test box, but I fail
to understand why huge companies
like Sun can't be bothered to provide access to
some box to test open source code on their OS.
It is a pity that a third party (i.e. SourceForge)
has to step in to provide this service to the
open source community.
Stars form from collapsing gas clouds. During the collapse, the core will get very dense and hot.
However, if the mass is too low, the temperature
and density will never reach the point where
stable hydrogen fusion will ignite. The critical
mass is about 8 per cent of the Sun's mass, corresponding to some 84 Jupiter masses.
An object below this limit is substellar,
and may be either a Brown Dwarf or a planet. The
distinction between both is somewhat fuzzy, and
certainly to some degree arbitrary.
One could argue that Brown Dwarfs and planets
differ by their mode of formation - Brown Dwarfs
form from collapsing gas clouds, like normal stars, while planets form from accretion of material in a circumstellar disk. However,
establishing how a particular object has formed
is not an easy task, so this is not a very practical definition.
Because Brown Dwarfs cool down over time, they
eventually become cool enough for dust forming
in their atmospheres, and they may show atmospheric phenomena (e.g. dust clouds) similar
to those we
know from planets. This is one of the reasons why
surface imaging, either by direct imaging or with
Doppler imaging, would be very interesting.
I use junkbuster in the default setting, which
send a browser id string like "Mozilla/3.01Gold (Macintosh; I; 68K)". No problem with online banking, neither with any site I am really interested in.
People who learn the keyboard shortcuts for their apps do generally have far better performance.
Often stated, but wrong. Sorry, but the stopwatch doesn't lie. And it doesn't even take into account the lost time when somebody "rm's" the wrong directory or file or group of files.
As any study, this one also is rather uninteresting unless it is exactly specified (a) what
has been studied, (b) how the test persons have
been selected, (c) how the test was performed exactly, and (d) what data were gathered and how they were interpreted.
The cited website does not provide any information that would qualify it as a publication that is up to even the lowest standards in science. This does not imply that the study really is worthless or unscientific, but at any rate it is not possible to judge about
it (or its findings) from the website.
Currently, atmospheric pressure on the surface of Mars is about 6 millibar, which on Earth corresponds to a height of
35 km above sea level (4 times higher than Mt. Everest).
At such low pressure, some of the water in the soft tissues will vaporize and cause swelling of the human body (note that the blood will not vaporize, because it is always under sufficient pressure in the blood vessels). This can be prevented by
"a properly fitted elastic garment", but such garments are only know to work at pressures of
more than 20 millibars.
Which means that a lot of carbon dioxide
would be needed until you could walk around without a space suit.
I have Netscape 4.78 with junkbuster and default
settings, i.e. junkbuster sends "Mozilla/3.01Gold (Macintosh; I; 68K)" as browser name.
Result: when I try to access the home page of KPMG, I get redirected
to a error page, which, in a twist of circular logic, tells me that this object has been removed
and my be found here, where 'here' is exactly the same error page (!!!). Anyone knows whether KPMG has
already applied for a patent on self-referring error pages ?
(Note: if I bypass junkbuster, the
error page seems to work somewhat better...)
I hope their consulting skills are better than their webmasters.
Back in school, once my whole class was visiting Berlin (long before East and West Germany
got united again). We also did a bus trip through
East Berlin, and were feeling happy to live in
a free country when we noticed the many surveillance cameras there. Little did we know then...
It's still only a press release, but at least it
gives the reference to the original publication, which is a huge pro.
I'm always annoyed by sensational and completely
uninformative press stories that never give a reference to the original source
where you could get better information. Motto: dumb down
to the max, don't give readers a chance to get informed.
All in all, both the NASA press release and the
PR from STSci are incredibly dumbed down and uninformative (where the hell is the link to
the original article ?!?).
For those interested in the real stuff, the
preprint of the publication is available from the
main authors website (look for research papers) at CalTech.
Re:I gave the same answers but I beg to disagree
on
Freedom or Power?
·
· Score: 1
If a seller and user could not choose what kind of contract to carry out business, that's the lost of outmost freedom.
If seller and user are on equal level, yes. If the
seller in in a position of power, allowing him to
dictate the contract because the user has no viable alternative, no.
You forget that very often, as a consumer vs some
company, you have
zero choice concerning the 'kind of contract'. Accept the terms forced upon you, or live without that lovely toy you wanted to buy.
Actually perl-o-tine was one of the very few
apps I have ever seen that completely locked up my
Linux box - even login via ssh didn't work anymore, had to do a power cycle. After that, I rolled my own Perl script using ImageMagik. Things
may have improved since then, didn't try the latest release of gimp.
This first guy said outright that a lot of people have downloaded his application but few have submitted patches back to it.
Actually my own experience (from several open source projects) is that you cannot expect more than about one bug report per thousand
downloads, and much less useful patches. Which implies that only the top twenty (well, maybe the top fifty) out of the
many thousand projects listed on freshmeat get
enough patches/bug reports/overall attention to
make the developement really an open community
effort. All the other projects rely almost entirely on one person to push along, find bugs,
think about useful features, and get the project going. And they quietly die if that one person abandons them.
... and here is a tripwire
replacement which looks and feels like tripwire,
but never actually gets tripped. Do you believe the FBI isn't clever enough to come up with something like that ?
And the funny thing is that the worse article has
attracted already about six times as much posters... seems the newspapers are right in dumbing down stories to the least common denominator.
Apart from that, it is an interesting result, but only three sigma from the standard model, which
is not really too much if you want to announce
something groundbreaking. Sometimes even four sigma
results turn out to be just experimental
outliers.
I have a couple of friends with a Ph.D. or similar
level of education who are using Windows. Most
of them are/were not able to use it without the help of some more experienced friend.
On the other hand, a friend of mine recently
decided to install Linux (Debian) on her PC.
She lives 300 km away, so I had to support her
by phone, and we spent countless hours on the phone walking through the install procedure and the config files to get her system up and decently
running. Supporting a newbee on the phone is certainly the best way to discover how horrible
Linux can be.
Conclusion ? Grandma certainly cannot use Windows
without substantial help, but Linux is rather
unuseable as well. Personally, I learned to use Unix before I ever saw Windows, so I find Windows
pretty painful. I think neither Windows nor Linux
is user-friendly - it just depends on what you
are familiar with.
This is in fact one of the weakest spots of
current open source development: the dependency on
a very small number of commmercial supporters.
- a big centralized FTP/WEB hosting site... boosts VA's profile,
but is completely the contrary of what open source development really needs.
While sunsite/metalab/ibiblio/(whatever it will call itself next) does not provide the many services of sourceforge, it has much more mirrors,
and thus avoids the primary problem of SF: being
a single point of failure (remember themes.org ?).
SF is nice for development, but software should be stored elsewhere. Even metalab still has the problem of being US-based, and thus is subject to the obscure US crypto laws. In an ideal world, open source software would be stored on a distributed non-US system (and also protected by PGP signatures instead of reliance on SF not getting cracked). Unfortunately, for many developer kiddies, SF is just too convenient.
Wrong (well, maybe, at least). This quote
from Tognazzini's website is completely uninformative. In particular, it is far from clear
just which task was tested. Did the GUI users just need to click on an icon ? You can't
do 'tail myfile' with a single mouseclick. The
study by Tognazzini may be completely irrelevant
for most real-world tasks that are non-trivial
in even minor ways.
Furthermore, the whole quote is somewhat off-topic.
Tognazzini did not study effectiveness for real-world tasks, but rather user experience in terms of subjective time (thus his setup may well
have been completely inappropriate to test efficiency). He recommends strongly to
reduce subjective time, so the main conclusion from his study would be that irrespective whether
the CLI is faster or slower than a GUI, it is the better user interface because it provides a much more pleasant user experience.
Maybe encryption/privacy on the net goes down in the US, but at the same time it receives substantial funding by the German government.
This is not only true for GnuPG, which has funding
by the government (for the development of more user-friendly frontends, I think), but there is also a project
for the development of an open source anonymity service (JAP) as strong as (or even stronger than) the Freedom
anonymizer service, and there
is also the Sphinx project to build a PKI for the public authorities and maybe others.
One of the main drivers for the JAP project (and maybe others) seems to be that many consumers (at least in Germany)
apparently avoid E-commerce because of privacy concerns.
Mod me down if you dont' like this, but as a serious question: what real advantage do these
desktop environments have, after all ?
Almost everyone in our institute uses KDE, just because it is installed by default. However, I have never ever seen anyone using a feature not
provided by simple window managers as well (I have seen, however, people getting confused because
kvt is not a fully functional xterm replacement).
I have installed Ximian Gnome at home, mostly to test compliance for some app I have coded. I never find myself using any of its features. Actually,
the only reason I keep it on my disk is the nice
gdm login screen.
Also, I think you forgot a step... Shouldn't
you run make -n install before you actually install?
Good idea, wrong step. You should check the PGP
signature on the tarball. Or forget about the source code if no signature is provided. Or download from two locations and compare (this is the way I found a trojan in a tarball just recently.)
Although this is a fairly big event for
a calm, middle-age star like Sun, it is peanuts
compared to the events observed for younger and
more active stars of similar mass (the
brightest flare ever observed on a young solar-like star
released 10000 times more energy than any flare
on Sun).
Which implies that Earth has experienced much more impressive flare events when the Sun was young.
Also the qouted gigaton of mass loss is not really that much.
The Sun has 2x10^30 kg, and loses 5x10^9 kg per second (one from solar wind, four more from conversion of mass into the radiated energy). So one gigaton is just 200 seconds of normal mass loss.
Slashdot Mirror
The drawback is: I would like very much like to use the same e-mail client on Linux and Windows, but sylpheed is only theoretically cross-platform. On ftp.gnupg.org, there is a w32 build of sylpheed 0.4.60 which is buggy like hell, and I have no idea how it was compiled (otherwise I would rebuild a newer version).
I have a Casio fx-85. Never had to replace any batteries - it runs on solar cells, for many years already.
Now that Apple has turned to BSD, there is Unix and Windows left for the desktop. And last time I checked, JAP (Java Anon Proxy) worked ok on Win98 (except that Win98 occasionally crashes ..), but had huge problems on Linux (incomplete page downloads) - which might tell something about the "portability" of Java.
I do not host/offer my source code at SourceForge, because my project uses strong encryption, and I prefer to host it in a country (Germany) with less nebulous laws in that respect. However, the compile farm is an invaluable resource for testing compatibility with other platforms.
Actually, I can understand that the FreeBSD project may not have the resources to offer a test box, but I fail to understand why huge companies like Sun can't be bothered to provide access to some box to test open source code on their OS. It is a pity that a third party (i.e. SourceForge) has to step in to provide this service to the open source community.
Funny thing is, I recently discovered PHP and immediately liked it because it is so similar to my favourite 'hardcore' language - plain old C.
An object below this limit is substellar, and may be either a Brown Dwarf or a planet. The distinction between both is somewhat fuzzy, and certainly to some degree arbitrary. One could argue that Brown Dwarfs and planets differ by their mode of formation - Brown Dwarfs form from collapsing gas clouds, like normal stars, while planets form from accretion of material in a circumstellar disk. However, establishing how a particular object has formed is not an easy task, so this is not a very practical definition.
Because Brown Dwarfs cool down over time, they eventually become cool enough for dust forming in their atmospheres, and they may show atmospheric phenomena (e.g. dust clouds) similar to those we know from planets. This is one of the reasons why surface imaging, either by direct imaging or with Doppler imaging, would be very interesting.
I use junkbuster in the default setting, which send a browser id string like "Mozilla/3.01Gold (Macintosh; I; 68K)". No problem with online banking, neither with any site I am really interested in.
Often stated, but wrong. Sorry, but the stopwatch doesn't lie. And it doesn't even take into account the lost time when somebody "rm's" the wrong directory or file or group of files.
As any study, this one also is rather uninteresting unless it is exactly specified (a) what has been studied, (b) how the test persons have been selected, (c) how the test was performed exactly, and (d) what data were gathered and how they were interpreted.
The cited website does not provide any information that would qualify it as a publication that is up to even the lowest standards in science. This does not imply that the study really is worthless or unscientific, but at any rate it is not possible to judge about it (or its findings) from the website.
At such low pressure, some of the water in the soft tissues will vaporize and cause swelling of the human body (note that the blood will not vaporize, because it is always under sufficient pressure in the blood vessels). This can be prevented by "a properly fitted elastic garment", but such garments are only know to work at pressures of more than 20 millibars.
Which means that a lot of carbon dioxide would be needed until you could walk around without a space suit.
Result: when I try to access the home page of KPMG, I get redirected to a error page, which, in a twist of circular logic, tells me that this object has been removed and my be found here, where 'here' is exactly the same error page (!!!). Anyone knows whether KPMG has already applied for a patent on self-referring error pages ? ...)
(Note: if I bypass junkbuster, the error page seems to work somewhat better
I hope their consulting skills are better than their webmasters.
Back in school, once my whole class was visiting Berlin (long before East and West Germany got united again). We also did a bus trip through East Berlin, and were feeling happy to live in a free country when we noticed the many surveillance cameras there. Little did we know then ...
I'm always annoyed by sensational and completely uninformative press stories that never give a reference to the original source where you could get better information. Motto: dumb down to the max, don't give readers a chance to get informed.
For those interested in the real stuff, the preprint of the publication is available from the main authors website (look for research papers) at CalTech.
If seller and user are on equal level, yes. If the seller in in a position of power, allowing him to dictate the contract because the user has no viable alternative, no.
You forget that very often, as a consumer vs some company, you have zero choice concerning the 'kind of contract'. Accept the terms forced upon you, or live without that lovely toy you wanted to buy.
Actually perl-o-tine was one of the very few apps I have ever seen that completely locked up my Linux box - even login via ssh didn't work anymore, had to do a power cycle. After that, I rolled my own Perl script using ImageMagik. Things may have improved since then, didn't try the latest release of gimp.
Actually my own experience (from several open source projects) is that you cannot expect more than about one bug report per thousand downloads, and much less useful patches. Which implies that only the top twenty (well, maybe the top fifty) out of the many thousand projects listed on freshmeat get enough patches/bug reports/overall attention to make the developement really an open community effort. All the other projects rely almost entirely on one person to push along, find bugs, think about useful features, and get the project going. And they quietly die if that one person abandons them.
... and here is a tripwire replacement which looks and feels like tripwire, but never actually gets tripped. Do you believe the FBI isn't clever enough to come up with something like that ?
Apart from that, it is an interesting result, but only three sigma from the standard model, which is not really too much if you want to announce something groundbreaking. Sometimes even four sigma results turn out to be just experimental outliers.
On the other hand, a friend of mine recently decided to install Linux (Debian) on her PC. She lives 300 km away, so I had to support her by phone, and we spent countless hours on the phone walking through the install procedure and the config files to get her system up and decently running. Supporting a newbee on the phone is certainly the best way to discover how horrible Linux can be.
Conclusion ? Grandma certainly cannot use Windows without substantial help, but Linux is rather unuseable as well. Personally, I learned to use Unix before I ever saw Windows, so I find Windows pretty painful. I think neither Windows nor Linux is user-friendly - it just depends on what you are familiar with.
- a big centralized FTP/WEB hosting site ... boosts VA's profile,
but is completely the contrary of what open source development really needs.
While sunsite/metalab/ibiblio/(whatever it will call itself next) does not provide the many services of sourceforge, it has much more mirrors, and thus avoids the primary problem of SF: being a single point of failure (remember themes.org ?).
SF is nice for development, but software should be stored elsewhere. Even metalab still has the problem of being US-based, and thus is subject to the obscure US crypto laws. In an ideal world, open source software would be stored on a distributed non-US system (and also protected by PGP signatures instead of reliance on SF not getting cracked). Unfortunately, for many developer kiddies, SF is just too convenient.
Wrong (well, maybe, at least). This quote from Tognazzini's website is completely uninformative. In particular, it is far from clear just which task was tested. Did the GUI users just need to click on an icon ? You can't do 'tail myfile' with a single mouseclick. The study by Tognazzini may be completely irrelevant for most real-world tasks that are non-trivial in even minor ways.
Furthermore, the whole quote is somewhat off-topic. Tognazzini did not study effectiveness for real-world tasks, but rather user experience in terms of subjective time (thus his setup may well have been completely inappropriate to test efficiency). He recommends strongly to reduce subjective time, so the main conclusion from his study would be that irrespective whether the CLI is faster or slower than a GUI, it is the better user interface because it provides a much more pleasant user experience.
This is not only true for GnuPG, which has funding by the government (for the development of more user-friendly frontends, I think), but there is also a project for the development of an open source anonymity service (JAP) as strong as (or even stronger than) the Freedom anonymizer service, and there is also the Sphinx project to build a PKI for the public authorities and maybe others.
One of the main drivers for the JAP project (and maybe others) seems to be that many consumers (at least in Germany) apparently avoid E-commerce because of privacy concerns.
Almost everyone in our institute uses KDE, just because it is installed by default. However, I have never ever seen anyone using a feature not provided by simple window managers as well (I have seen, however, people getting confused because kvt is not a fully functional xterm replacement).
I have installed Ximian Gnome at home, mostly to test compliance for some app I have coded. I never find myself using any of its features. Actually, the only reason I keep it on my disk is the nice gdm login screen.
Good idea, wrong step. You should check the PGP signature on the tarball. Or forget about the source code if no signature is provided. Or download from two locations and compare (this is the way I found a trojan in a tarball just recently.)
Also the qouted gigaton of mass loss is not really that much. The Sun has 2x10^30 kg, and loses 5x10^9 kg per second (one from solar wind, four more from conversion of mass into the radiated energy). So one gigaton is just 200 seconds of normal mass loss.