Your comment about a locked screen showing the contents of the desktop is probably related to a screensaver taking a screen capture and using it in the screen saver. This can be fixed by running xscreensaver-demo (I think) program to get the configuration for the screen savers. You may have to install the xscreensaver package to get this program. When it comes up, it will warn you that xscreensaver isn't running -- that's OK, just ignore it. You can still set the options for the various screen savers.
This is necessary because about a year ago (or so) Gnome decided to replace the standard xscreensaver with their own screen saver system. In good Gnome fashion, there are almost no actual configuration options. However, it still actually uses the xscreensaver programs, which have a lot of configuration options, including where to get image sources (that's on one of the tabs). Tell it not to use a screen capture. Personally, I have it use images from my camera directory, so I'll see pictures I've taken in the various screen savers.
Considering that xscreensaver has had this options dialog for a long time, I don't know why Gnome couldn't either use it, or make one like it. But I guess that would expose useful options to users.
You're right that Ubuntu has been fairly bug filled. My personal annoyance with Ubuntu 8.10 comes with Gnome 2.24: they released a version of Gnome with a half re-implemented session management. They ripped out the mostly working implementation and replaced it with a version that cannot save or restore session state! Now, I'm all for replacing poor implementations, which as I've read the previous session management code was. However, it's really bad to take out mostly working code and replace it with a "** (gnome-session-properties:4618): DEBUG: Session saving is not implemented yet!" printout.
But frankly, I yet to see any well written large scale C++ project.
I supposed this depends on your definition of "well written". KDE/Qt is almost entirely C++. I believe that most higher level MS products like Office and Visual Studio are C++. You and I may not like MS, but some of their products do run well, though that doesn't necessarily speak to the quality of the code.
I believe Mozilla/Firefox has large amounts of C++ code, ss does OpenOffice.
I can say for sure that I run a lot more C++ programs on a daily basis than Java ones.
There was an article in my local paper not to long ago about a rule change at the county level that would enable the local police departments to collect more revenue from parking tickets they wrote. I can't find a reference to it online, but basically, before the change, it cost about about as much in an officer's time to write the ticket as the department received from the ticket. The change meant they would essentially profit from each ticket. The article quoted a police officer who said how great the change was. Before, they would only write tickets for really egregious things, like parking in a handicapped spot, but now they can write tickets for all sorts of things.
It's too bad they don't teach people what a conflict of interest is anymore.
There are some distros that do that (the AC mentioned Linspire) and there have been window managers that have done that in the past. Over a decade ago, I saw some public computer terminals that were running fvwm95 (or something similar) on RedHat being used for Internet access. Even being familiar with Linux, the only reason I even noticed that they weren't just Windows machines was that the Windows logo on the start menu was backwards.
But the real reason that "Linux" by and large doesn't just mimic the Windows interface is because the Windows interface isn't really that great. Now in reality, much of the interface is the same -- windows, menus, icons, double-clicking, selection, scrolling, etc are all consistent between the two interfaces. What you're left with are things like the window decorations and appearance, and there are themes for both Gnome and KDE that mimic the Windows look and feel, at least to a large extent.
Frankly, I don't think anyone with Windows experience would have any trouble at all sitting down and using a Linux desktop. They may not know how to use some "advanced" features like virtual desktops or widgets, but I bet they would learn quickly. Then they'd probably start to feel limited by the MS interface.
The biggest fear in switching to Linux is the actual switch and getting things to work, as well as software and hardware compatibility. Sadly, most hardware manufacturers still only test with Windows, and there's usually at least one piece of hardware on a system that just doesn't work on Linux. I'm surprised at how much on my Ubuntu laptop works, though it's not perfect -- I can't use the microphone to record sound. And while things like Wine are good, it doesn't always work, and even when it does, it's still kind of a hack.
So, I think that the actual interface is a relatively minor reason why people don't use Linux more.
I love Gentoo, but that's not been my experience at all. I dread running 'emerge world', since I have no idea what will break while it's working. Granted, 95% (or more) of the emerge goes really smoothly. But it's that other part that is trouble.
For example, this past weekend, I went through the process. It seems that the files that were in the com_err package got moved to the e2fsprogs package (or something like that) so that the new version of e2fsprogs blocked because com_err was installed. OK, so I have to manually unmerge com_err first. No problem. Except for the fact that after I unmerge com_err, wget doesn't work, so emerging e2fsprogs fails while trying to download the new version of the package. Fortunately, I can re-merge com_err without problems, so wget works again, user 'emerge -f' to fetch the files for the e2fsprogs packages, re-unmerge com_err, and then emerge e2fsprogs. Usually every time I upgrade there's something like that that causes pain. Even if it's only the "minor pain" of having to find all the packages that linked with the old version of some library that now need to rebuilt to link with the new version.
Also, while not strictly Gentoo's fault, I also ran out of disk space while emerging OpenOffice 3 (and why is OO3 unmasked while Firefox 3 is not) because I only had 3 or 4 GB free on / and OpenOffice needs more like 6 or so to compile. But maybe I wouldn't have been so low on free space if I didn't have 16 GB of old packages in/usr/portage/distfiles that never seem to get cleaned up. Though I supposed keeping those around can be a good thing -- if the com_err package wasn't already there, I would have been in real trouble when I tried to re-emerge it earlier.
This isn't to dissuade anyone from using Gentoo, because it is a nice system. As long as you know what you're doing and are willing to work around problems that are thrown at you.
The company I work for recently (last 6 months or so) switched to Norton from McAfee. We also use ClearCase for or software revision control. It turns out that there is some bad interaction between the two, and running applications out of a dynamic ClearCase view can sometimes lock up or blue screen the machine. This doesn't happen very often, maybe 0.5% chance or less. But we have hundreds of unit tests that we run, which are built in the ClearCase view and would be run there. So, over the course of testing, you have a pretty good chance of causing the error.
Unrelated, but this pales in comparison to the big problem I have building on Windows now. I got a new build machine and some thing (some driver?) is causing some sort of process ID starvation. We use GNU make in Cygwin to build the software, and make creates lots of processes to do its work. I don't know what the correct behavior should be, but on some of our machines, newly created processes get ever increasing IDs (but not sequential) as time goes on, until some ceiling is reached. After that, it becomes harder and harder to make new processes. Presumably, this is because the Windows kernel asks whatever it is to make a new process ID, and that either fails, or returns an ID that something else rejects. Maybe it thinks it's still in use or something. But the result is that a new process fails to start, and the build stops. I don't think this problem is related to Norton. I found a single Internet posting on a Cygwin list that blamed an ATI driver, which this machine has. But the machine is newer than the switch to Norton, so it could be a bad interaction between Norton and the ATI driver (and maybe ClearCase too).
Fortunately, I can ignore all these fun things and just work on Solaris and Linux. Even the ancient version of Gnome (2.0.2) on my Solaris machine is preferable to dealing with the problems on Windows:P
Gentoo's init system is like this (and has been for quite a while). It doesn't do stuff in parallel by default, but I think there is an option to enable that.
I think I remember hearing something about Ubuntu and/or Debian also trying to create something like that, as well.
This is outrageous! I don't think I can vote for the Senator running for president that voted for that bill that goes completely the wrong way on copyright reform, so I guess I'll have to vote for
The Senate vote was unanimous
Damn.
I wonder if any of the third party candidates opposed this bill...
Where I work, we have a number of Solaris(*) and Linux machines. Almost as a rule, you can count on the Linux machines being at least twice as fast as the Solaris machines. At some jobs, Linux is 10 times as fast as Solaris. A lot of this probably comes from the fact that the processors run at twice the speed (x86 vs. sparc), but it seems like the I/O goes faster as well on Linux.
But remember, anecdote != evidence, so YMMV.
(*) Mostly these are Solaris 9 machines. We have a few Solaris 10 machines, but they aren't that much faster in general.
The pull model really isn't a good idea, because that is what spammers are already trying to get people to do. They want you to open the email and click the link. A pull model just makes that automatic. Not to mention all the marketing people (pseudo-spammers) that would just love to know which of their recipients actually look at their emails, and how long they look at them, etc. I already get mailings (alumni stuff, etc) that are just links to a web page where I can read the actual letter.
And of course, "just use gmail" isn't really a solution. It only works until someone figures out how to get through gmail's filters, or Google really sells out and starts allowing select "partners" to advertise to members directly. Though there is some irony in the idea that you can avoid email advertising by using a system that has ads in the email viewer. I'm not saying anything bad about Google or gmail, just pointing out the irony:)
If the email from the stranger is properly signed, then you look at it. If it is not spam, then you mark the sender as someone you trust. If it is spam, then you mark the sender as a spammer, and further messages from that identity are rejected. Spammers cannot afford to make a new key pair for every spam and ensure that the public key gets to the PKI that you use to validate messages.
Of course, there are challenges -- users misconfiguring their keys, the PKIs not being available, the PKIs themselves being overloaded with spammers trying to push keys there anyway, etc. Frankly, I think these are all challenges that can be overcome, and digital signing of email would go a long way towards helping with spam.
Oh, and someone else mentioned SPF. The problem with that is that it only signs and trusts the server. I don't care (or rather, I shouldn't care) about the server sending the email. I care about the user sending the email. If everyone digitally signed their emails, we could theoretically go back to open relays without any additional problems, since we would be verifying the actual person sending the email, not whatever machine they happened to be using.
For C, C++, perl, and make, the man (and info) pages are usually the quickest and easiest for me. The C man pages are pretty standard. The C++ pages I get from the Sun compiler. They're decent, but not great. They at least say what's available on each class. The perl pages are the same as what's online, but are often a little easier to get to through man (at least for me, especially if I don't have access to the Internet). The 'make' info page pretty much has everything I could want to know about that insanely convoluted (but interesting) language.
Why are those on by default? Because they are nice features! I like the fact that Firefox remembers the dozens (possibly hundreds) of stupid logins that I have to various sites. I like the fact that cookies allow the site to remember who I am or my preferences so I don't have to log in each time I go there. I occasionally use the history to look for a site I visited earlier and can't remember the address to.
In general, I use Firefox in my account, and no one (other than root) can get to any of that information. And that information can be very useful to me.
Now, on a public computer, making the privacy mode default does make a lot of sense. But there's no reason why it should be turned on for everyone all the time.
The self-signed certificate LOOKS *EXACTLY* the same as a MAN IN THE MIDDLE ATTACK.
Only if there is a MITM attack the first time you connect to the site. And every other time you connect to the site. I admit, there is a potential danger. But the same danger exists for SSH. The first time you connect to your ssh server, you download the server's key. Maybe you verify it, but now it's in the ssh client's cache of keys. The next time you connect to the server, the client compares the key to the cached version. If they're different, it lets you know something weird is going on. Otherwise, it connects without issue because the only way there could be a MITM attack the second (and third, and so on) time you connect is if there was always been a MITM attack going on.
As others have pointed out, the security of connections goes like:
CA signed cert > self signed cert > no cert
However, FF3 makes the self signed cert a bigger PITA to deal with than no cert. There are lots of ways Firefox could make it clear that the self signed cert is not completely safe without subjecting users to confusing warning screens.
I don't know about that. We have a bunch of Solaris clients here at work, and some people were complaining that their browsers were crashing occasionally. I had them install adblock to block flash ads, and that solved their problems.
I wouldn't be surprised if most of the "flash is unstable" comments come from people having problems with poorly written flash advertisements. Certainly, the flash player shouldn't cause as much trouble as it can, but installing adblock to filter out the bad flash definitely helps.
I agree that the problem is unsolvable, but not for the reason you stated. The reason that software will always have usability problems is because of the author's point #4 (yes, I did RTFA):
4. Usability is hard to measure.
More specifically, it's hard to know what is "usable" and what is not. And something that is perfectly usable for one person may be very difficult for another. Obviously, there are degrees to this -- some software can be really hard to use for everybody and some software can be fairly easy to use. Often this depends on what the software does -- is it a game for children (probably easy) or is it a video editor (probably hard). In addition, some programs are hard to learn or require domain knowledge to use correctly. Is vi's usability poor because it's hard to learn? What about Blender? Is the GIMP really unusable, or is it just different than expected?
The author also points out in #8 that developers scratch their own itch, and once the program is usable for them and people like them, they may not feel the need to improve it much more. There has to be a leap from "here's the code to a neat program I wrote that solves my problem -- see if you can use it" to "here is a product that most people can use". That leap takes a fair amount of effort, and I don't blame anyone for not taking it.
In the end, though, usability in OSS will always be "poor" because no matter how good it is, there will always be some program that isn't as good or is just not meant for the masses to use.
But quite frankly, I don't know how anyone can complain about the usability of Free Software when there is so much unusable stuff coming out of places like Microsoft, Adobe and Apple (yes, even Apple -- I still remember the horrors of having to use the QuickTime player back when I ran Windows). Linux is just more usable for me. Maybe that's personal, I don't know. But in general I don't find FOSS usability "poor".
Yes, it's C++, so include changes can cause big compiles.
It's hard to say exactly how many people use the baseline. There are probably 2 dozen active developers, but probably more than 50 individuals who have worked on it. As far as releases, there is the core system which is released about once a quarter, with interim releases done about once a month. Then there are more program specific addons that use the core that release on their programs schedule. Some of the programs do internal developer releases once a week on their own program branch. Typically, those program addons upgrade to the next available quarterly release when it is appropriate for them. However, most of the changes that affect the core come from the programs using it, so it's kind of like an open source system in the company (sadly, it is not public open source).
It has taken a while, and a lot of script work (mostly on my part when I'm not doing real work) to get everything running mostly smoothly. Though we're starting to run into problems with ClearCase slowing down because of too many branches on some files. I'm not quite sure how we're going to solve this problem in the long term. Maybe corporate will come down and demand we switch to something else (with no funding, just like when they switched us to ClearCase) and make that problem moot:)
I imagine that those small teams of people are probably working on a piece of software that takes a few minutes to compile completely. Try working on a piece of software that takes hours to compile. In that situation, when one of your teammates makes a change to a core file that is included all over the place and the whole team has to stop working while everybody recompiles, you'll learn why branches are good.
Where I work, we probably take branching a little to far. Changing the code involves a minimum of 2 branches, but most changes actually touch 4 branches. Each developer has a work branch where (s)he can do whatever (s)he wants. When the work is complete, those changes are merged to a "changeset" branch (we have a different internal name that wouldn't make sense here) where it is reviewed by other developers. After it is reviewed, it is merged to an integration branch with other changesets for initial testing. This is where conflicts are resolved (yep - through the creation of "merge" branches). When integration is complete, the changeset (or merged changeset) branches are merged to a delivery branch. That is compared to the integration branch to ensure that it is correct (and no one did anything sneaky on the integration branch) before the delivery branch is merged back to "main" (or the "trunk" or whatever, we use ClearCase, so we call it "main").
Is our branching extreme? Maybe. But it does allow developers to work without causing pain to others and without having to worry about others causing them pain. This is useful because we have many developers working in the same code for different projects. It is also useful to isolate the changes into a single branch for easy reviewing (all of our code must be peer reviewed by at least one other developer). We have a lot of scripts to simplify the branch creation and merging, and we usually manage to do a pretty good job. However, it is annoying to see a typo in a comment (or something equally small) and realize that it's not worth the effort required to fix it.
Same for me on Solaris at work. I even downloaded the source hoping to be able to compile a Cairo-free GTK version of Firefox 3, but the configure script only works with 'cairo-gtk2'. I don't even have GTK2 on this Solaris machine, so I'll just have to live with Firefox 2 here for now. It's unfortunate that the Cairo work changes caused many of the supported toolkits to be removed.
Forget the anonymous email. Like another poster said, that could potentially be traced back. And there's the possibility that the school could take it as a threat instead of assistance.
Better would be to just go to a local technology reporter. Preferably, a real reporter, who thinks of himself as a journalist. The kind of person who would rather go to jail than reveal his anonymous source. Explain to that person why you want to be anonymous (fear reprisals). Any responsible journalist would understand that and protect your identity.
Even better, by going to the press, you're a protected whistleblower. It's sad that you have to get a third party involved to fix a simple issue, but that is probably the safest course.
Possible dangers are that the reporter either won't be a professional or will not understand the issues involved and will take the school's side (that you're a dirty hacker trying to cause problems). The best way to avoid those problems are to research the particular reporter to see if he has done anything like that in the past and if the organization he works for has any policies regarding whistle blowers / anonymous sources.
(Note: slashcode changed "%s" into "%25s" in the URL-- replace that with a just "%s")
Then put "g" in the keyword box. You can search Google by typing "g whatever i want to google" in the URL bar. The one I really like is my "w" one for Wikipedia:
UML is not a tool. UML is a notation used for communicating software design ideas between developers. If a portion of UML is not widely known, to the point that only one or two people on a software team readily understand it, then it is essentially not worth it and it might as well not exist. Remember that UML should be used to help developers talk to other developers about the software they are developing. UML for the sake of UML (i.e., not using it for effective communication) is pretty worthless.
Of course, UML used for communication like design and documentation, especially at a high level, is a good thing. Just don't go UML crazy and think that every little detail of the system has to be documented in UML. You'll probably end up spending more time doing UML than you will making the actual system.
Amen. CSS is a much better (and easier) way to go. I couldn't imagine trying to manage all the enters and exist with Javascript manually (though I'm not a Javascript expert -- maybe there's an easy way?).
The only problem I had was having two rows in a table that logically went together. I wanted the mouse over either of the rows to highlight both. I could get it to highlight both rows if the mouse was over the top one (I think it was something like "tr:hover + tr"), but I couldn't get it to highlight the top row if the mouse was over the bottom. There just wasn't any way to select a tr whose next tr was "hover".
I guess maybe Javascript would have worked, but what I was working on wasn't that important, so I lived with the highlight not working perfectly.
"We know the picture... ends in an.exe, which is not shown" to the user, Royal says.
Even a well educated userbase that knows that running random executables could get confused if the file name is shown as "blah.jpg". The decision at Microsoft to hide extensions by default was nice in theory, but it has ended up confusing more users because sometimes the extension is on the file name (mostly on web sites) and sometimes it is not. I don't know why MS keeps hiding the extensions by default. It should be clear to them by now that it causes more harm than good.
Slashdot Mirror
Your comment about a locked screen showing the contents of the desktop is probably related to a screensaver taking a screen capture and using it in the screen saver. This can be fixed by running xscreensaver-demo (I think) program to get the configuration for the screen savers. You may have to install the xscreensaver package to get this program. When it comes up, it will warn you that xscreensaver isn't running -- that's OK, just ignore it. You can still set the options for the various screen savers.
This is necessary because about a year ago (or so) Gnome decided to replace the standard xscreensaver with their own screen saver system. In good Gnome fashion, there are almost no actual configuration options. However, it still actually uses the xscreensaver programs, which have a lot of configuration options, including where to get image sources (that's on one of the tabs). Tell it not to use a screen capture. Personally, I have it use images from my camera directory, so I'll see pictures I've taken in the various screen savers.
Considering that xscreensaver has had this options dialog for a long time, I don't know why Gnome couldn't either use it, or make one like it. But I guess that would expose useful options to users.
You're right that Ubuntu has been fairly bug filled. My personal annoyance with Ubuntu 8.10 comes with Gnome 2.24: they released a version of Gnome with a half re-implemented session management. They ripped out the mostly working implementation and replaced it with a version that cannot save or restore session state! Now, I'm all for replacing poor implementations, which as I've read the previous session management code was. However, it's really bad to take out mostly working code and replace it with a "** (gnome-session-properties:4618): DEBUG: Session saving is not implemented yet!" printout.
But frankly, I yet to see any well written large scale C++ project.
I supposed this depends on your definition of "well written". KDE/Qt is almost entirely C++. I believe that most higher level MS products like Office and Visual Studio are C++. You and I may not like MS, but some of their products do run well, though that doesn't necessarily speak to the quality of the code.
I believe Mozilla/Firefox has large amounts of C++ code, ss does OpenOffice.
I can say for sure that I run a lot more C++ programs on a daily basis than Java ones.
I agree.
There was an article in my local paper not to long ago about a rule change at the county level that would enable the local police departments to collect more revenue from parking tickets they wrote. I can't find a reference to it online, but basically, before the change, it cost about about as much in an officer's time to write the ticket as the department received from the ticket. The change meant they would essentially profit from each ticket. The article quoted a police officer who said how great the change was. Before, they would only write tickets for really egregious things, like parking in a handicapped spot, but now they can write tickets for all sorts of things.
It's too bad they don't teach people what a conflict of interest is anymore.
There are some distros that do that (the AC mentioned Linspire) and there have been window managers that have done that in the past. Over a decade ago, I saw some public computer terminals that were running fvwm95 (or something similar) on RedHat being used for Internet access. Even being familiar with Linux, the only reason I even noticed that they weren't just Windows machines was that the Windows logo on the start menu was backwards.
But the real reason that "Linux" by and large doesn't just mimic the Windows interface is because the Windows interface isn't really that great. Now in reality, much of the interface is the same -- windows, menus, icons, double-clicking, selection, scrolling, etc are all consistent between the two interfaces. What you're left with are things like the window decorations and appearance, and there are themes for both Gnome and KDE that mimic the Windows look and feel, at least to a large extent.
Frankly, I don't think anyone with Windows experience would have any trouble at all sitting down and using a Linux desktop. They may not know how to use some "advanced" features like virtual desktops or widgets, but I bet they would learn quickly. Then they'd probably start to feel limited by the MS interface.
The biggest fear in switching to Linux is the actual switch and getting things to work, as well as software and hardware compatibility. Sadly, most hardware manufacturers still only test with Windows, and there's usually at least one piece of hardware on a system that just doesn't work on Linux. I'm surprised at how much on my Ubuntu laptop works, though it's not perfect -- I can't use the microphone to record sound. And while things like Wine are good, it doesn't always work, and even when it does, it's still kind of a hack.
So, I think that the actual interface is a relatively minor reason why people don't use Linux more.
I love Gentoo, but that's not been my experience at all. I dread running 'emerge world', since I have no idea what will break while it's working. Granted, 95% (or more) of the emerge goes really smoothly. But it's that other part that is trouble.
For example, this past weekend, I went through the process. It seems that the files that were in the com_err package got moved to the e2fsprogs package (or something like that) so that the new version of e2fsprogs blocked because com_err was installed. OK, so I have to manually unmerge com_err first. No problem. Except for the fact that after I unmerge com_err, wget doesn't work, so emerging e2fsprogs fails while trying to download the new version of the package. Fortunately, I can re-merge com_err without problems, so wget works again, user 'emerge -f' to fetch the files for the e2fsprogs packages, re-unmerge com_err, and then emerge e2fsprogs. Usually every time I upgrade there's something like that that causes pain. Even if it's only the "minor pain" of having to find all the packages that linked with the old version of some library that now need to rebuilt to link with the new version.
Also, while not strictly Gentoo's fault, I also ran out of disk space while emerging OpenOffice 3 (and why is OO3 unmasked while Firefox 3 is not) because I only had 3 or 4 GB free on / and OpenOffice needs more like 6 or so to compile. But maybe I wouldn't have been so low on free space if I didn't have 16 GB of old packages in /usr/portage/distfiles that never seem to get cleaned up. Though I supposed keeping those around can be a good thing -- if the com_err package wasn't already there, I would have been in real trouble when I tried to re-emerge it earlier.
This isn't to dissuade anyone from using Gentoo, because it is a nice system. As long as you know what you're doing and are willing to work around problems that are thrown at you.
The company I work for recently (last 6 months or so) switched to Norton from McAfee. We also use ClearCase for or software revision control. It turns out that there is some bad interaction between the two, and running applications out of a dynamic ClearCase view can sometimes lock up or blue screen the machine. This doesn't happen very often, maybe 0.5% chance or less. But we have hundreds of unit tests that we run, which are built in the ClearCase view and would be run there. So, over the course of testing, you have a pretty good chance of causing the error.
Unrelated, but this pales in comparison to the big problem I have building on Windows now. I got a new build machine and some thing (some driver?) is causing some sort of process ID starvation. We use GNU make in Cygwin to build the software, and make creates lots of processes to do its work. I don't know what the correct behavior should be, but on some of our machines, newly created processes get ever increasing IDs (but not sequential) as time goes on, until some ceiling is reached. After that, it becomes harder and harder to make new processes. Presumably, this is because the Windows kernel asks whatever it is to make a new process ID, and that either fails, or returns an ID that something else rejects. Maybe it thinks it's still in use or something. But the result is that a new process fails to start, and the build stops. I don't think this problem is related to Norton. I found a single Internet posting on a Cygwin list that blamed an ATI driver, which this machine has. But the machine is newer than the switch to Norton, so it could be a bad interaction between Norton and the ATI driver (and maybe ClearCase too).
Fortunately, I can ignore all these fun things and just work on Solaris and Linux. Even the ancient version of Gnome (2.0.2) on my Solaris machine is preferable to dealing with the problems on Windows :P
I'm not sure exactly how it works, but my 64 bit version of Firefox in Ubuntu uses some sort of 64 bit wrapper to run the 32 bit flash plugin.
The npwrapper.libflashplayer.so library is what is loaded by Firefox.
However the magic works, it means I can run the Flash plugin from my 64 bit browser.
Gentoo's init system is like this (and has been for quite a while). It doesn't do stuff in parallel by default, but I think there is an option to enable that.
I think I remember hearing something about Ubuntu and/or Debian also trying to create something like that, as well.
This is outrageous! I don't think I can vote for the Senator running for president that voted for that bill that goes completely the wrong way on copyright reform, so I guess I'll have to vote for
The Senate vote was unanimous
Damn.
I wonder if any of the third party candidates opposed this bill...
Where I work, we have a number of Solaris(*) and Linux machines. Almost as a rule, you can count on the Linux machines being at least twice as fast as the Solaris machines. At some jobs, Linux is 10 times as fast as Solaris. A lot of this probably comes from the fact that the processors run at twice the speed (x86 vs. sparc), but it seems like the I/O goes faster as well on Linux.
But remember, anecdote != evidence, so YMMV.
(*) Mostly these are Solaris 9 machines. We have a few Solaris 10 machines, but they aren't that much faster in general.
The pull model really isn't a good idea, because that is what spammers are already trying to get people to do. They want you to open the email and click the link. A pull model just makes that automatic. Not to mention all the marketing people (pseudo-spammers) that would just love to know which of their recipients actually look at their emails, and how long they look at them, etc. I already get mailings (alumni stuff, etc) that are just links to a web page where I can read the actual letter.
And of course, "just use gmail" isn't really a solution. It only works until someone figures out how to get through gmail's filters, or Google really sells out and starts allowing select "partners" to advertise to members directly. Though there is some irony in the idea that you can avoid email advertising by using a system that has ads in the email viewer. I'm not saying anything bad about Google or gmail, just pointing out the irony :)
If the email from the stranger is properly signed, then you look at it. If it is not spam, then you mark the sender as someone you trust. If it is spam, then you mark the sender as a spammer, and further messages from that identity are rejected. Spammers cannot afford to make a new key pair for every spam and ensure that the public key gets to the PKI that you use to validate messages.
Of course, there are challenges -- users misconfiguring their keys, the PKIs not being available, the PKIs themselves being overloaded with spammers trying to push keys there anyway, etc. Frankly, I think these are all challenges that can be overcome, and digital signing of email would go a long way towards helping with spam.
Oh, and someone else mentioned SPF. The problem with that is that it only signs and trusts the server. I don't care (or rather, I shouldn't care) about the server sending the email. I care about the user sending the email. If everyone digitally signed their emails, we could theoretically go back to open relays without any additional problems, since we would be verifying the actual person sending the email, not whatever machine they happened to be using.
I thought I'd post this because I hadn't seen anyone else mention this site. The best HTML / CSS reference I've found is at blooberry.com.
HTML: http://www.blooberry.com/indexdot/html/index.html
CSS: http://www.blooberry.com/indexdot/css/propindex/all.htm
For C, C++, perl, and make, the man (and info) pages are usually the quickest and easiest for me. The C man pages are pretty standard. The C++ pages I get from the Sun compiler. They're decent, but not great. They at least say what's available on each class. The perl pages are the same as what's online, but are often a little easier to get to through man (at least for me, especially if I don't have access to the Internet). The 'make' info page pretty much has everything I could want to know about that insanely convoluted (but interesting) language.
Why are those on by default? Because they are nice features! I like the fact that Firefox remembers the dozens (possibly hundreds) of stupid logins that I have to various sites. I like the fact that cookies allow the site to remember who I am or my preferences so I don't have to log in each time I go there. I occasionally use the history to look for a site I visited earlier and can't remember the address to.
In general, I use Firefox in my account, and no one (other than root) can get to any of that information. And that information can be very useful to me.
Now, on a public computer, making the privacy mode default does make a lot of sense. But there's no reason why it should be turned on for everyone all the time.
The self-signed certificate LOOKS *EXACTLY* the same as a MAN IN THE MIDDLE ATTACK.
Only if there is a MITM attack the first time you connect to the site. And every other time you connect to the site. I admit, there is a potential danger. But the same danger exists for SSH. The first time you connect to your ssh server, you download the server's key. Maybe you verify it, but now it's in the ssh client's cache of keys. The next time you connect to the server, the client compares the key to the cached version. If they're different, it lets you know something weird is going on. Otherwise, it connects without issue because the only way there could be a MITM attack the second (and third, and so on) time you connect is if there was always been a MITM attack going on.
As others have pointed out, the security of connections goes like:
CA signed cert > self signed cert > no cert
However, FF3 makes the self signed cert a bigger PITA to deal with than no cert. There are lots of ways Firefox could make it clear that the self signed cert is not completely safe without subjecting users to confusing warning screens.
I don't know about that. We have a bunch of Solaris clients here at work, and some people were complaining that their browsers were crashing occasionally. I had them install adblock to block flash ads, and that solved their problems.
I wouldn't be surprised if most of the "flash is unstable" comments come from people having problems with poorly written flash advertisements. Certainly, the flash player shouldn't cause as much trouble as it can, but installing adblock to filter out the bad flash definitely helps.
I agree that the problem is unsolvable, but not for the reason you stated. The reason that software will always have usability problems is because of the author's point #4 (yes, I did RTFA):
More specifically, it's hard to know what is "usable" and what is not. And something that is perfectly usable for one person may be very difficult for another. Obviously, there are degrees to this -- some software can be really hard to use for everybody and some software can be fairly easy to use. Often this depends on what the software does -- is it a game for children (probably easy) or is it a video editor (probably hard). In addition, some programs are hard to learn or require domain knowledge to use correctly. Is vi's usability poor because it's hard to learn? What about Blender? Is the GIMP really unusable, or is it just different than expected?
The author also points out in #8 that developers scratch their own itch, and once the program is usable for them and people like them, they may not feel the need to improve it much more. There has to be a leap from "here's the code to a neat program I wrote that solves my problem -- see if you can use it" to "here is a product that most people can use". That leap takes a fair amount of effort, and I don't blame anyone for not taking it.
In the end, though, usability in OSS will always be "poor" because no matter how good it is, there will always be some program that isn't as good or is just not meant for the masses to use.
But quite frankly, I don't know how anyone can complain about the usability of Free Software when there is so much unusable stuff coming out of places like Microsoft, Adobe and Apple (yes, even Apple -- I still remember the horrors of having to use the QuickTime player back when I ran Windows). Linux is just more usable for me. Maybe that's personal, I don't know. But in general I don't find FOSS usability "poor".
Yes, it's C++, so include changes can cause big compiles.
It's hard to say exactly how many people use the baseline. There are probably 2 dozen active developers, but probably more than 50 individuals who have worked on it. As far as releases, there is the core system which is released about once a quarter, with interim releases done about once a month. Then there are more program specific addons that use the core that release on their programs schedule. Some of the programs do internal developer releases once a week on their own program branch. Typically, those program addons upgrade to the next available quarterly release when it is appropriate for them. However, most of the changes that affect the core come from the programs using it, so it's kind of like an open source system in the company (sadly, it is not public open source).
It has taken a while, and a lot of script work (mostly on my part when I'm not doing real work) to get everything running mostly smoothly. Though we're starting to run into problems with ClearCase slowing down because of too many branches on some files. I'm not quite sure how we're going to solve this problem in the long term. Maybe corporate will come down and demand we switch to something else (with no funding, just like when they switched us to ClearCase) and make that problem moot :)
I imagine that those small teams of people are probably working on a piece of software that takes a few minutes to compile completely. Try working on a piece of software that takes hours to compile. In that situation, when one of your teammates makes a change to a core file that is included all over the place and the whole team has to stop working while everybody recompiles, you'll learn why branches are good.
Where I work, we probably take branching a little to far. Changing the code involves a minimum of 2 branches, but most changes actually touch 4 branches. Each developer has a work branch where (s)he can do whatever (s)he wants. When the work is complete, those changes are merged to a "changeset" branch (we have a different internal name that wouldn't make sense here) where it is reviewed by other developers. After it is reviewed, it is merged to an integration branch with other changesets for initial testing. This is where conflicts are resolved (yep - through the creation of "merge" branches). When integration is complete, the changeset (or merged changeset) branches are merged to a delivery branch. That is compared to the integration branch to ensure that it is correct (and no one did anything sneaky on the integration branch) before the delivery branch is merged back to "main" (or the "trunk" or whatever, we use ClearCase, so we call it "main").
Is our branching extreme? Maybe. But it does allow developers to work without causing pain to others and without having to worry about others causing them pain. This is useful because we have many developers working in the same code for different projects. It is also useful to isolate the changes into a single branch for easy reviewing (all of our code must be peer reviewed by at least one other developer). We have a lot of scripts to simplify the branch creation and merging, and we usually manage to do a pretty good job. However, it is annoying to see a typo in a comment (or something equally small) and realize that it's not worth the effort required to fix it.
Same for me on Solaris at work. I even downloaded the source hoping to be able to compile a Cairo-free GTK version of Firefox 3, but the configure script only works with 'cairo-gtk2'. I don't even have GTK2 on this Solaris machine, so I'll just have to live with Firefox 2 here for now. It's unfortunate that the Cairo work changes caused many of the supported toolkits to be removed.
Forget the anonymous email. Like another poster said, that could potentially be traced back. And there's the possibility that the school could take it as a threat instead of assistance.
Better would be to just go to a local technology reporter. Preferably, a real reporter, who thinks of himself as a journalist. The kind of person who would rather go to jail than reveal his anonymous source. Explain to that person why you want to be anonymous (fear reprisals). Any responsible journalist would understand that and protect your identity.
Even better, by going to the press, you're a protected whistleblower. It's sad that you have to get a third party involved to fix a simple issue, but that is probably the safest course.
Possible dangers are that the reporter either won't be a professional or will not understand the issues involved and will take the school's side (that you're a dirty hacker trying to cause problems). The best way to avoid those problems are to research the particular reporter to see if he has done anything like that in the past and if the organization he works for has any policies regarding whistle blowers / anonymous sources.
It's great that Opera works for you, but I've been doing this in Firefox for years. All you have to do is setup a bookmark with a URL like:
(Note: slashcode changed "%s" into "%25s" in the URL-- replace that with a just "%s")Then put "g" in the keyword box. You can search Google by typing "g whatever i want to google" in the URL bar. The one I really like is my "w" one for Wikipedia:
Works great for looking something up there.UML is not a tool. UML is a notation used for communicating software design ideas between developers. If a portion of UML is not widely known, to the point that only one or two people on a software team readily understand it, then it is essentially not worth it and it might as well not exist. Remember that UML should be used to help developers talk to other developers about the software they are developing. UML for the sake of UML (i.e., not using it for effective communication) is pretty worthless.
Of course, UML used for communication like design and documentation, especially at a high level, is a good thing. Just don't go UML crazy and think that every little detail of the system has to be documented in UML. You'll probably end up spending more time doing UML than you will making the actual system.
Amen. CSS is a much better (and easier) way to go. I couldn't imagine trying to manage all the enters and exist with Javascript manually (though I'm not a Javascript expert -- maybe there's an easy way?).
The only problem I had was having two rows in a table that logically went together. I wanted the mouse over either of the rows to highlight both. I could get it to highlight both rows if the mouse was over the top one (I think it was something like "tr:hover + tr"), but I couldn't get it to highlight the top row if the mouse was over the bottom. There just wasn't any way to select a tr whose next tr was "hover".
I guess maybe Javascript would have worked, but what I was working on wasn't that important, so I lived with the highlight not working perfectly.
Even a well educated userbase that knows that running random executables could get confused if the file name is shown as "blah.jpg". The decision at Microsoft to hide extensions by default was nice in theory, but it has ended up confusing more users because sometimes the extension is on the file name (mostly on web sites) and sometimes it is not. I don't know why MS keeps hiding the extensions by default. It should be clear to them by now that it causes more harm than good.