I'm ditching my ipod mini (2g) - and will buy an SE W800i phone. I have a k750i - which is a superb phone. For one thing it uses standards (png for transparency on themes, jpg for images). Connects out of the box via usb in FreeBSD. 2Mpixel camera. Nice menus. I'm selling it too along with my ipod, so I can buy a w800. No more struggle with the nightmarish interface of itunes. Just drag and drop my music dirs into an mp3 dir on the phone. Comes with an 512Mb card, which can be upgraded to a 2Gb one. That's good enough for music and photos. And sound quality on this baby is infinitely better than on the ipod, thanks to these earbuds and the excellent audio player. Has a built in fm radio with rdf support as well. And believe it or not, it is perfectly usable - in fact, given it's functions (organizer, phone, camera, walkman) the interface is an UI marvel. Costs as much as an ipod btw.
Despite to the raves I read here on slashdot, the ipod was a HUGE disappointment for me - I guess I'm not the target audience. I'm more concerned about sound quality and features than the fancy click-wheel. Give me something that I can figure out easily (the W800 works while the phone is switched off, providing 30h long playback. The ipod mini's battery life sucked big time as well), is small, has at least 2Gb space, and doesn't need a separate program just to copy files to it. W800 provides me with that - and much much more (actually, the camera is pretty good as well). Yeah, I'm absolutely anti-ipod. So my advice is: don't buy an ipod. Buy something much much better for the same money. If you don't need a new phone, buy a player that supports ogg and flac (not just crappy mp3s - without gapless playback support! and AACs). The ipod is overrated.
That benchmark is from 2003... since then, it appears that ext2/3 (and xfs) has improved a lot (the new benchmark mentions that). Anyway, thanks for the link - I'm still not convinced about the capabilities of reiser4 though, neither as a stable/reliable fs, nor as something that outperforms the competition. I don't see how much more CPU can help it... afterall, it did use more CPU to perform worse than the others, and 500Mhz is not much, but it should be enough, especially for a test where nothing is done but copying files.
But the difference is HUGE - utilizing more CPU power, reiser4 underperforms every filesystem discussed. Do you say the situation would be reversed if running on a better CPU? How much better we are talking here? I mean if the difference would have been small, I could expect some improvement if moving to a more modern hardware, but there is a really big gap between ext2/ext3 (+XFS) performance and Reiser4 - and 500Mhz is not that slow, especially doing nothing else but copying files! Let's assume that reiser4 is a safe bet from every single point of view except this (it isn't - for production servers no sane person would chose something as experimental as reiser4). Then we have to make sure that there is always spare CPU cycles available (let's say 30% is always IDLE) for the filesystem? How can you guarantee that on a busy webserver or database? No, I don't think reiser4 looks good at all, no matter how much CPU power you throw at it - it can make results somewhat better, but for now, it seems reiser4 is not a safe bet at all.
Hmmm.... I don't unerstand your outburst: "OK, so ext3 is not the fastest filesystem on earth" - what? I looked at the benchmarks, and it appears that ext2/3 wins every single test that matters to me: find dirs, files, untar, tar, copy tarballs(s), the kernel source tree (ext2/3 now outperforms reiser3). I understand your points, and everything you wrote is true, but there is no need to defend ext3, for it performs admirably (did you RTFA?).
THANKS!!!! I'll tell you later why:) (Not because you're right, sorry, but because the link you provided gave me an excellent idea - inspiration - to write about something totally unrelated to the topic at hand:)
There are at least 12 distinct operating systems in their list - Solaris, Cisco, SCO Unixware, OpenBSD, FreeBSD, NetBSD, HP-UX, AIX, HP Tru64, MacOS X, Linux variants like SuSE, Debian, Gentoo, RedHat (I counted Linux as one, even though most of the vulns. are found in their specific configuration/management tools). Add an arbitrary number of applications: KDE and GNOME, that in itself has more apps that are counted for Windows, every free SQL database server, mail server, (LotusDomino for Christ's sake!), imap client, ftp client, ftp server, etc...
Now we have a comparison of a single operating system (Windows) + apps running on it with at least 12 distinct operating systems + 10x the number of apps that was counted for windows. The result is rather surprising: there are JUST 4x more bugs in 12 operating systems + 10x more apps than in windows + windows apps alone! This result is much more unfavorable for Microsoft than to any Unix/Linux OS!
Of course, the fallacy of the comparison is that it suggests that Linux or Unix is an Operating System. For someone who does not look at the details, it might seem that installing a specific Linux or Unix operating system is more risky - hey, there are more bugs found in Linux/Unix, that's what the article says! In fact, the opposite is true, if you look at the details.
Not that the comparison is useful in any way - why are Safari bugs counted at all? Safari runs on OS X only, so you can't just dump safari bugs into linux/unix bugs category (how retarded is that?). Why are bugs found in SuSE YAST counted as Linux bugs? They have nothing to do with linux or unix - they are specific to one operating system: SuSE linux (the same applies for all the bugs counted in Debian, RedHat, Gentoo, etc.) Not to mention the duplications: Eric Raymonds "Fetchmail POP3 Client Buffer Overflow" is counted 5 times for linux and BSDs. There are duplications for windows as well though. In other words, this list or comparison is pretty much unusable.
395 megs of memory is too much. K web browser never uses that much. In fact, konqi's memory usage if far below firefox's. The least amount of ram you need for kde is 192. With 256, it should work smoothly (you can even have some konq. instances preloaded). Using purely KApps makes the experience smoother than with WinXP. Now if you start firefox (which is a memory hog) or openoffice, and $insert_app_here, and you find yourself running out of ram, don't blame kde!
This needs no special tuning whatsoever. Plain vanilla KDE will work fine without any tweaking on a puter with 256Megs. My main machine has 512, and even after extensive use, my swap partition isn't even touched. That with lots of apps loaded by default: skype, amarok, kmail, 4 preloaded instances of konqi, etc. My system begins swapping only if I start up firefox or ooo-build. (Or perhaps krita with an 50meg PNG:)
KDE's memory management is very efficient. In fact, considering what it does, I would say that I'd expect higher memory usage. Of course, we can throw numbers around here with little or no way to back up our claims, I realize that, but if you check the specs of people running kde (on forums) you'll see that configs like a 700Mhz duron with 256Mb RAM (I mentioned this in another post) is enough. I don't know where your K browser using 384Mb RAM comes from (well, except if you pull it out of your ass). Actually I made some screenies of kde 3.4.3 here. One of the screenshots displays memory usage. If you check the clock, you'll see that it shows the state of memory after opening a lot of apps, including scribus, with images loaded, etc (and you'll see what I have running in my systray). So I don't understand people who report excessive memory usage of KDE - it is either FUD, or they should switch distroes:)
You have absolutely no idea of what you're talking about... Slow down KDE even more? What version did you use last time? 2.2? With every release since 3.0 KDE is getting better and better perfomance on old hardware. I'm happily using it on a 700Mhz duron with 256Mb SDRAM (not my main machine though). Please don't spread FUD about KDE if you haven't used it for years...
On the other hand, if KDE is slow for you (on hw with speck >= to my duron conf.), than you screwed up your config (or your distro screwed up kde). KDE permorms admirably well these days...
If you're using Mambo currently, do you need to switch to Joomla? The answer is no, as far as my research for this article shows.
Hmmm... You don't have to do much research to see that the future lies with Joomla. Basically the entire development team - the same team that made mambo great - left and they are working on Joomla now. How safe is to stay with a product that has "we are looking for developers" on their website for months? Especially since joomla! offers a clear migration path... Basically the first release is latest mambo with trademarks stripped out, so the sooner one switches the better...
If we compare the "roadmap" of the two projects, joomla has a clearer vision of the future, so yeah, I don't think mambo is a safe bet from what I've seen.
Yeah, but how would the code exploiting the buffer overflow propagate? I'm not familiar with how these exploits work, so this might be a stupid question: how would the code exploiting the vulnerability run? Will it be launched autamitaclly by opening an email?
Yeah, but how could code run - without asking for a password first? I'm not saying it can't, because I don't know, so I'm just curious... in the above scenario (/home mounted with noexec). Would it run from/tmp (but firefox/konqi/whatever can have - and I think it does have - its cache elsewhere)?
Well, I don't know how it did that... It might have been when I downloaded the prog, although the prog itself wasn't infected. Temporary files stored if firefox cache might be another possibility.
Free-av is... well free - as in no activation code, no time limit, etc. You have to update manually, but that's the only drawback. Otherwise it is very light on resources, and it never ever failed me (it is very effective). Works on very old machines (I use it in a small computer lab I administer, on some celeron 300s with 64Mb ram) without even feeling that it's there. It does one thing, and it does it well: it finds viruses and some other malware. So yeah, it's nice I think, compared to the alternatives (if Symantec's products were my only choice, I'd rather have viruses lol:) For firewall, I use WinXP's SP2 default firewall (big relief compared to the bloat some other firewalls have become) which is again - perfect for a desktop machine. This combo works well for me (free-av + default fw) - but then, I rarely use windows (usually 50% time spent in win is updating the system, the rest is testing out software for the lab I administer).
Never ever visit astalavista from windows, not even in Firefox - even using firefox, free-av catched ~10 viruses that tried to execute while only visiting the site, and searching for my lost cd key (well, lost CD to be precise, taht came with my TV card, with the only app that worked for me).
Well, for one thing (yeah, this is a boring argument, but still holds true) most linux distroes actively discourage accounts with elevated privileges for day to day use.
Take for instance (K)Ubuntu, that actively discourages using the root account. System wide infection is impossible, the problem is localized for the user's home directory. Better still, Average Joe, who is most likely to get infected, would not need execute privileges in his/her home directory at all, so/home can be mounted noexec for instance, without affecting day to day use of his/her computer. In this case, no remote execution vulnerability will have the chance of infecting even the home directory. I don't see how a remote exec vulnerability in a core functionality of the OS like libpng (which is a core functionality for desktop users) could work in this case... The wmf vuln. relies on downloading files and running them on the puter - how could that work in the kubuntu scenario? Running sg. from/tmp perhaps?
So I see more difficulty for such exploits in *nix based systems, but I can be easily convinced to the contrary...
Linus on slashdot: "Screw slashdot." Details here. This is getting better and better - and there is some very sound reasoning (not for the screw./ part) in the linked post above. If anyone is still interested.
However, he wants people to use KDE, based solely on personal preference, which is nothing more than zealotry
Have you read the entire discussion? Linus gave specific examples of what he considers usability "failures" in GNOME. Is that zealotry? Why don't you reflect on the points he raised instead of calling him a zealot. Who comes across as a zealot here? - read the post he answers to... Or read very carefully this post (focus on the content, not the author) - is it reasonable enought? Actually, what bothers Linus is the "usability zealotry" - whenever GNOME devs have to defend one decision or another (missing input field from file dialog, print dialog issues, spatial by default) we always here usability this and usability that, without any substantial evidence or empirical research to back it up... just some outlandish theory. Now that is what is so annoying, that I'm not surprised in the least that Linus seems a bit vexed:) He is not known for his diplomatic skills, he never was, but as soon as his straightforward opinion does not coincide with some GNOME users, he is labeled as zealot?
I was wondering how long it would take for this discussion to come up on Slashdot. It's noteworthy really only because Linus comes across as a 13 year old arsehole
Diplomacy is not Linus' forte - and it never was, and we know that. Just so it happens that his opinion does not coincide with folks who invested a lot of emotion in supporting GNOME... and now, suddenly, he "comes across as a 13 year old". Sure.
On the other hand, Havoc later admitted that some missing functionality was NOT a usability decision. But whenever it comes down to defending some very questionable choices, I always hear these bogus "usability" arguments: it is less confusing this way, less bloated, $insert_bogus_usability_argument_here. And that is precisely Linus' gripe. Kurt Pfeife (I hear now: but he is obviously biased because he is a KDE guy - but shut up and listen to the argument) summed it up well:
"Frederic told that the options from the PPD file are intentionally mot
listed in the printing dialog, the usability team of GNOME was against
listing these options. They clutter the dialog and can be more confusing
than useful to the user.
This is not a wise decision."
Because it means there is no usability at all for many features: you
can not *use* them, they are not there at all, they are forbidden to
the user by some higher usability being.
Usubility for a given feature starts to become debate-able only where
the feature is already present, where something can be done *at all*.
Before, there is the land of un-usability.
What type of printers do you think users in many enterprise environ-
ments are used to? It is the 60 pages per minute model, that can do
duplex and stapling and punching and cover-sheets-on-cardbox-paper
and watermark and foo and bar and morestuff..
.
I was one of the guys who pushed for adding all features (which the
underlying CUPS provided) into the user interface of KDEPrint.
And I know for a fact that KDE's power in printing matters (given to
it by CUPS) was the one feature that determined a pro-KDE decision
in Linux desktop rollouts in Europe.
I also know that the printing dialog of KDE can be improved a *lot*
from where it is now. However, this is much more difficult than
just removing most features and declare their un-usability to be
the new religion of usability.
We'll work on that for KDE 4, but without removing any feature (we
will rather be adding some more, because the wonderful CUPS 1.2
gives them to us for free).
Cheers,
Kurt
Computers are about empowering users. Now there is a difficulty here - you want to enable your users to do more, and do it easily. And that is what makes UI design an art: how can you provide more without confusing the user? GNOME's answer is simple (again, this is Linus' point): take functionality away, or don't provide it at all. A very easy way to cut the gordian knot, but it is not the right way imho. The right way is to do proper research (not on a very limited set of computer illiterate users, who don't know that the right mouse button is there for a reason - results won't be representative at all) - and organize your functionality. That is what the last two paragraphs are about, that is Linus' point, whether you like it or not, whether you agree with it or not. Calling him names or ridiculing him is kinda ironic: you accuse him as coming across as a 13 year old, yet this is what you do in your own post. You don't reflect on the points raised by Linus. You probably didn't even read the entire thread (otherwise you'll see that Havoc's post comes off as a kind of confession - yeah it was not a usability question at all, and that is where Linus' criticism is mainly targeted at). And apparently you are +5 insightful. Way to go folks!
You are an expert in operating system kernels. Please keep to what you do best. Users will vote with their own desktop. There is no need for you to teach people what GUI and desktop to use.
Yeah, well, he didn't make an announcment or a press release you know... He voiced his opinion on a mailing list - and I think Linus is pretty good at that:)
Incidentally, I had exactly the same experience. I migrate users to free software, and we offer two choices: FreeBSD backend, Kubuntu desktop. Why? The same reasons he cites. In the past two years, we heard a lot of "usability" noise from GNOME devs, and imho they are all bogus. Why? Because people throw around words like "usability" too easily, leading to circular or unsubstantial arguments, while real usability studies are not conducted at all. I haven't read a serious usability study for a long time. (maybe this will change with openusability and all). And no, I don't consider a study conducted with people who are absolute computer illiterate (not knowing that the right mouse button is good for something) representative. They are a very specific subset of users, they are NOT the majority, and making design decisions based on experiments conducted on this very small subset of the userbase is WRONG. That is Linus' point. Is he politically correct? Of course not (" This "users are idiots, and are confused by functionality" mentality of
Gnome is a disease. If you think your users are idiots, only idiots will
use it.")
My girlfriend is absolutely computer illiterate: she thinks (well, thought) that Office is the OS that runs on his laptop. Being lazy and all she often sits down to my computer (instead of opening her laptop) to browse the net. Sometimes she doesn't even notice that instead of firefox, she is using konqueror. There is a small set of functionality that users expect at specific areas of your screen: first buttons should be back and forward, they expect an input field for URLs at the top, maybe a google search bar... and that's it. If they are there, they are not really "confused" because there are additional buttons (kget, print, even cervisia) to the right side. They don't even notice it. It is the same with the file dialog: were users really bothered by the input field? I very much doubt that - and just like Linus, I was not aware of ctrl + L until someone told me here on./. And in the past years, I hear one bogus "usability" claim from these so called "usability experts" after another (spatial nautilus anyone?) No evidence, no empirical study, just "we say so as usability experts" with some outlandish theory to back it up... so yeah, I think he is right on spot (and yeah, yeah, we know, diplomacy is not his forte).
Win* usage 2003 oct (overall: nt, 98, 2k, xp, etc) - 93,2 %
Win* usage 2005 oct - 90% - down more than 3% in two years... not much, I admit, but not bad either, considering the rate of development (and the state of desktop linux even two years ago...).
Where did that 3% go? Apparently 1-1 to Linux and Mac, and 1 to something other (in my case, I'm in other with FreeBSD).
Good point, afterall, most of the packages aimed specifically for kids come from kde-edutainment (geography modul was added just recently, and is present in 3.5).
The article also reminded me of this blogI read sometimes ago. Its author (a kde dev) works for a company who provides linux solutions for - among other things - schools. The idea of an Office Suite aimed specifically at kids is very interesting, and you can see nice mockups of such a possibilities. Seeing the pictures commented Forget the children. Make this for my 58 year-old mother - and there is a part two where the developer further elaborates on the idea. Actually they are looking for volunteers to implement it:
And finally some words about the implementation. If you want this to become a reality, please volunteer your time and skill. The current engines of KWord and Karbon are great starting points, and both of these programs will probably be little more than new shells (skins if you like) on top of them.
If you volunteer to help with Kids Office, I can almost guarantee you your 15 minutes of fame. I have been approached by magazine editors who wondered if I was implementing it already and when it would be finished. So don't hesitate or be shy. You can mail me or Danny, and you could go into #koffice on irc where we hang out sometimes.
Perhaps the strong focus on education (the edutainment package has a very active developer community behind it) was one of the reasons why Mark Shuttleworth "promoted" KDE (or rather, Kubuntu) to a tier 1 status when the Novell thing happend a few weeks ago.
I used pagemaker years ago (on MacOS 7.5.x) - although only casually - at a company I worked for (for a brief period of time) and I didn't know it even exists today... A month ago I began designing a new website, where we would issue a quarterly newsletter, and a yearly book (in pdf format). I haven't thought of using any specific software for the newsletters at that time, though I had scribus installed (because it was there in ports, and I heard about it on some forum or something). So 3 weeks ago I tried it, and what surprised me was that it was very easy to use, and it is just perfect for the job. I don't use publishing (and never did) software "professionally" - and I assumed that these software are fairly complex, and generally, not for the casual desktop user.
Scribus changed that perception, and that's why I think it deserves a little marketing: it is usable with little effort for Joe Average desktop user, if he or she wants to quickly compile a document or presentation that looks cool. This is not to say it is not complex, because it is, but you don't have to learn everything to produce better looking documents than you could in oo.o, and faster! A larger userbase might lead to more feedback on missing features (or bugs) and faster development perhaps. Also, it is a useful application not only for people in publishing business, but average desktop users as well...
I was surprised at PageMaker as well... But actually we do have a good alternative to PageMaker if the need for it is real: Scribus!
Scribus exists for some time now, and still, it was only a few months ago that I heard about it, and only 3 weeks ago that I installed it. And I was absolutely amazed at its quality (and standard support, even the latest pdf specs, including embedded script support!). What OSDL needs is a central place where all the great apps likely to be needed in an enterprise/office environment should be listed. There aren't that many actually, but how many of the respondends who missed PageMaker knew about scribus? Seeing the quality of the app, I was surprised at the lack of marketing/hype this application receives. Even their website... well, it is not bad, but it isn't good either.
Right now marketing focuses on distributions or tools, and rarely on the application stack (except FF and OO.o). GNOME does a good job at marketing as well. But KDE? They formed a new marketing group only recently, and as they started to look around, they found plenty of examples for KDE use in business, like this one (Dutch Record Shop Chain Migrates 1000 PCs to KDE). Or take a look at this initiave: part 1, part 2.. Or another example here. Quote:
second, i have the inkling that we have a lot of small and medium sized business deployments out there. personally i count anything under 500 seats to be in that umbrella. at the table (which i picked at random) i ate lunch at in munich during trolltech dev days there i found myself sandwiched between two such examples. while eating the rather amazingly good food, i discovered that on my right was a fellow who works for a company that makes linux based satelite t.v. transmission software (sky t.v. is amongst their clientelle) and they use qt for their in-house engineering tools. on my left were three men from a vienese company that writes kde software for a group of five private hospitals. these hospitals all run kde on the desktop and everything from patient records to x-rays is handled on them.
Cases like these will convince businesses to adopt linux solutions, and as the article says, not necessarily because lower costs, but because of the quality of the software out there. But there needs to be a central place that enumerates and provides a short description of the application stack (I think 10-15 desktop apps, no more, that are essential for business) as well as provides examples for the various scenarious where free software can be put to use. KDE in hospitals controlling everything including x-rays, a music record chain with desktop locked down via kiosk to include the 4 necessary apps, satellite tv transmission software - this is staggering if you think of it, and it should be shown to CIOs and PHBs (make a nice newsletter/booklet in scribus for instance:)))
I switched from oo when 1.4.2 came out (with some bugfixes to odt support). Works fine, and I like the integration (into the desktop, and the other parts of koffice - I use kword mostly). I rarely use tables in my documents, and that's where most of the issues lie with kword - or at least it did in the past.
Someone mentioned konqi as a viable alternative now for ff... indeed, konqi has been an excellent browser in the 3.4.x series already: easy on resources, and fast as hell. But to become a real competitor, it needs to be ported to windows. With QT4, that would probably happen. What we really need is lightweight, feature rich and modern Office competitor, and koffice could become one. It is already being ported to qt4 and windows (I saw a demo of kexi on windows). Karbon and Krita received a lot of attention lately, and it shows. Krita is beginning to replace GIMP for me here: it has a very nice and usable UI, lots of features (excellent layer control). Karbon has come a long way as well - it is not yet a viable replacement for Inkscape, but it's getting there. Now only if the rest of the office suite received the same amount of attention... I know, I know: developers, developers, developers:))
Another nice app is Scribus! It is the best free desktop publishing application, and it is relatively easy to use. It is also needs more marketing I think - there is simply no free software alternative for scribus, yet I haven't heard about it until a month ago. After installing it I was absolutely surprised at its quality!
Slashdot Mirror
Despite to the raves I read here on slashdot, the ipod was a HUGE disappointment for me - I guess I'm not the target audience. I'm more concerned about sound quality and features than the fancy click-wheel. Give me something that I can figure out easily (the W800 works while the phone is switched off, providing 30h long playback. The ipod mini's battery life sucked big time as well), is small, has at least 2Gb space, and doesn't need a separate program just to copy files to it. W800 provides me with that - and much much more (actually, the camera is pretty good as well). Yeah, I'm absolutely anti-ipod. So my advice is: don't buy an ipod. Buy something much much better for the same money. If you don't need a new phone, buy a player that supports ogg and flac (not just crappy mp3s - without gapless playback support! and AACs). The ipod is overrated.
That benchmark is from 2003... since then, it appears that ext2/3 (and xfs) has improved a lot (the new benchmark mentions that). Anyway, thanks for the link - I'm still not convinced about the capabilities of reiser4 though, neither as a stable/reliable fs, nor as something that outperforms the competition. I don't see how much more CPU can help it ... afterall, it did use more CPU to perform worse than the others, and 500Mhz is not much, but it should be enough, especially for a test where nothing is done but copying files.
But the difference is HUGE - utilizing more CPU power, reiser4 underperforms every filesystem discussed. Do you say the situation would be reversed if running on a better CPU? How much better we are talking here? I mean if the difference would have been small, I could expect some improvement if moving to a more modern hardware, but there is a really big gap between ext2/ext3 (+XFS) performance and Reiser4 - and 500Mhz is not that slow, especially doing nothing else but copying files! Let's assume that reiser4 is a safe bet from every single point of view except this (it isn't - for production servers no sane person would chose something as experimental as reiser4). Then we have to make sure that there is always spare CPU cycles available (let's say 30% is always IDLE) for the filesystem? How can you guarantee that on a busy webserver or database? No, I don't think reiser4 looks good at all, no matter how much CPU power you throw at it - it can make results somewhat better, but for now, it seems reiser4 is not a safe bet at all.
Hmmm.... I don't unerstand your outburst: "OK, so ext3 is not the fastest filesystem on earth" - what? I looked at the benchmarks, and it appears that ext2/3 wins every single test that matters to me: find dirs, files, untar, tar, copy tarballs(s), the kernel source tree (ext2/3 now outperforms reiser3). I understand your points, and everything you wrote is true, but there is no need to defend ext3, for it performs admirably (did you RTFA?).
THANKS!!!! I'll tell you later why :) (Not because you're right, sorry, but because the link you provided gave me an excellent idea - inspiration - to write about something totally unrelated to the topic at hand :)
In other words:
There are at least 12 distinct operating systems in their list - Solaris, Cisco, SCO Unixware, OpenBSD, FreeBSD, NetBSD, HP-UX, AIX, HP Tru64, MacOS X, Linux variants like SuSE, Debian, Gentoo, RedHat (I counted Linux as one, even though most of the vulns. are found in their specific configuration/management tools). Add an arbitrary number of applications: KDE and GNOME, that in itself has more apps that are counted for Windows, every free SQL database server, mail server, (LotusDomino for Christ's sake!), imap client, ftp client, ftp server, etc...
Now we have a comparison of a single operating system (Windows) + apps running on it with at least 12 distinct operating systems + 10x the number of apps that was counted for windows. The result is rather surprising: there are JUST 4x more bugs in 12 operating systems + 10x more apps than in windows + windows apps alone! This result is much more unfavorable for Microsoft than to any Unix/Linux OS!
Of course, the fallacy of the comparison is that it suggests that Linux or Unix is an Operating System. For someone who does not look at the details, it might seem that installing a specific Linux or Unix operating system is more risky - hey, there are more bugs found in Linux/Unix, that's what the article says! In fact, the opposite is true, if you look at the details.
Not that the comparison is useful in any way - why are Safari bugs counted at all? Safari runs on OS X only, so you can't just dump safari bugs into linux/unix bugs category (how retarded is that?). Why are bugs found in SuSE YAST counted as Linux bugs? They have nothing to do with linux or unix - they are specific to one operating system: SuSE linux (the same applies for all the bugs counted in Debian, RedHat, Gentoo, etc.) Not to mention the duplications: Eric Raymonds "Fetchmail POP3 Client Buffer Overflow" is counted 5 times for linux and BSDs. There are duplications for windows as well though. In other words, this list or comparison is pretty much unusable.
This needs no special tuning whatsoever. Plain vanilla KDE will work fine without any tweaking on a puter with 256Megs. My main machine has 512, and even after extensive use, my swap partition isn't even touched. That with lots of apps loaded by default: skype, amarok, kmail, 4 preloaded instances of konqi, etc. My system begins swapping only if I start up firefox or ooo-build. (Or perhaps krita with an 50meg PNG :)
KDE's memory management is very efficient. In fact, considering what it does, I would say that I'd expect higher memory usage. Of course, we can throw numbers around here with little or no way to back up our claims, I realize that, but if you check the specs of people running kde (on forums) you'll see that configs like a 700Mhz duron with 256Mb RAM (I mentioned this in another post) is enough. I don't know where your K browser using 384Mb RAM comes from (well, except if you pull it out of your ass). Actually I made some screenies of kde 3.4.3 here. One of the screenshots displays memory usage. If you check the clock, you'll see that it shows the state of memory after opening a lot of apps, including scribus, with images loaded, etc (and you'll see what I have running in my systray). So I don't understand people who report excessive memory usage of KDE - it is either FUD, or they should switch distroes :)
On the other hand, if KDE is slow for you (on hw with speck >= to my duron conf.), than you screwed up your config (or your distro screwed up kde). KDE permorms admirably well these days...
Oh, btw, wasn't the Jem Report sponsored by Mambo?
Hmmm... You don't have to do much research to see that the future lies with Joomla. Basically the entire development team - the same team that made mambo great - left and they are working on Joomla now. How safe is to stay with a product that has "we are looking for developers" on their website for months? Especially since joomla! offers a clear migration path... Basically the first release is latest mambo with trademarks stripped out, so the sooner one switches the better...
If we compare the "roadmap" of the two projects, joomla has a clearer vision of the future, so yeah, I don't think mambo is a safe bet from what I've seen.
Yeah, but how would the code exploiting the buffer overflow propagate? I'm not familiar with how these exploits work, so this might be a stupid question: how would the code exploiting the vulnerability run? Will it be launched autamitaclly by opening an email?
Yeah, but how could code run - without asking for a password first? I'm not saying it can't, because I don't know, so I'm just curious... in the above scenario (/home mounted with noexec). Would it run from /tmp (but firefox/konqi/whatever can have - and I think it does have - its cache elsewhere)?
Well, I don't know how it did that... It might have been when I downloaded the prog, although the prog itself wasn't infected. Temporary files stored if firefox cache might be another possibility. Free-av is ... well free - as in no activation code, no time limit, etc. You have to update manually, but that's the only drawback. Otherwise it is very light on resources, and it never ever failed me (it is very effective). Works on very old machines (I use it in a small computer lab I administer, on some celeron 300s with 64Mb ram) without even feeling that it's there. It does one thing, and it does it well: it finds viruses and some other malware. So yeah, it's nice I think, compared to the alternatives (if Symantec's products were my only choice, I'd rather have viruses lol :) For firewall, I use WinXP's SP2 default firewall (big relief compared to the bloat some other firewalls have become) which is again - perfect for a desktop machine. This combo works well for me (free-av + default fw) - but then, I rarely use windows (usually 50% time spent in win is updating the system, the rest is testing out software for the lab I administer).
Never ever visit astalavista from windows, not even in Firefox - even using firefox, free-av catched ~10 viruses that tried to execute while only visiting the site, and searching for my lost cd key (well, lost CD to be precise, taht came with my TV card, with the only app that worked for me).
Well, for one thing (yeah, this is a boring argument, but still holds true) most linux distroes actively discourage accounts with elevated privileges for day to day use. Take for instance (K)Ubuntu, that actively discourages using the root account. System wide infection is impossible, the problem is localized for the user's home directory. Better still, Average Joe, who is most likely to get infected, would not need execute privileges in his/her home directory at all, so /home can be mounted noexec for instance, without affecting day to day use of his/her computer. In this case, no remote execution vulnerability will have the chance of infecting even the home directory. I don't see how a remote exec vulnerability in a core functionality of the OS like libpng (which is a core functionality for desktop users) could work in this case... The wmf vuln. relies on downloading files and running them on the puter - how could that work in the kubuntu scenario? Running sg. from /tmp perhaps?
So I see more difficulty for such exploits in *nix based systems, but I can be easily convinced to the contrary...
Linus on slashdot: "Screw slashdot." Details here. This is getting better and better - and there is some very sound reasoning (not for the screw ./ part) in the linked post above. If anyone is still interested.
Have you read the entire discussion? Linus gave specific examples of what he considers usability "failures" in GNOME. Is that zealotry? Why don't you reflect on the points he raised instead of calling him a zealot. Who comes across as a zealot here? - read the post he answers to... Or read very carefully this post (focus on the content, not the author) - is it reasonable enought? Actually, what bothers Linus is the "usability zealotry" - whenever GNOME devs have to defend one decision or another (missing input field from file dialog, print dialog issues, spatial by default) we always here usability this and usability that, without any substantial evidence or empirical research to back it up ... just some outlandish theory. Now that is what is so annoying, that I'm not surprised in the least that Linus seems a bit vexed :) He is not known for his diplomatic skills, he never was, but as soon as his straightforward opinion does not coincide with some GNOME users, he is labeled as zealot?
Diplomacy is not Linus' forte - and it never was, and we know that. Just so it happens that his opinion does not coincide with folks who invested a lot of emotion in supporting GNOME... and now, suddenly, he "comes across as a 13 year old". Sure.
On the other hand, Havoc later admitted that some missing functionality was NOT a usability decision. But whenever it comes down to defending some very questionable choices, I always hear these bogus "usability" arguments: it is less confusing this way, less bloated, $insert_bogus_usability_argument_here. And that is precisely Linus' gripe. Kurt Pfeife (I hear now: but he is obviously biased because he is a KDE guy - but shut up and listen to the argument) summed it up well:
Computers are about empowering users. Now there is a difficulty here - you want to enable your users to do more, and do it easily. And that is what makes UI design an art: how can you provide more without confusing the user? GNOME's answer is simple (again, this is Linus' point): take functionality away, or don't provide it at all. A very easy way to cut the gordian knot, but it is not the right way imho. The right way is to do proper research (not on a very limited set of computer illiterate users, who don't know that the right mouse button is there for a reason - results won't be representative at all) - and organize your functionality. That is what the last two paragraphs are about, that is Linus' point, whether you like it or not, whether you agree with it or not. Calling him names or ridiculing him is kinda ironic: you accuse him as coming across as a 13 year old, yet this is what you do in your own post. You don't reflect on the points raised by Linus. You probably didn't even read the entire thread (otherwise you'll see that Havoc's post comes off as a kind of confession - yeah it was not a usability question at all, and that is where Linus' criticism is mainly targeted at). And apparently you are +5 insightful. Way to go folks!Yeah, well, he didn't make an announcment or a press release you know... He voiced his opinion on a mailing list - and I think Linus is pretty good at that :)
Incidentally, I had exactly the same experience. I migrate users to free software, and we offer two choices: FreeBSD backend, Kubuntu desktop. Why? The same reasons he cites. In the past two years, we heard a lot of "usability" noise from GNOME devs, and imho they are all bogus. Why? Because people throw around words like "usability" too easily, leading to circular or unsubstantial arguments, while real usability studies are not conducted at all. I haven't read a serious usability study for a long time. (maybe this will change with openusability and all). And no, I don't consider a study conducted with people who are absolute computer illiterate (not knowing that the right mouse button is good for something) representative. They are a very specific subset of users, they are NOT the majority, and making design decisions based on experiments conducted on this very small subset of the userbase is WRONG. That is Linus' point. Is he politically correct? Of course not (" This "users are idiots, and are confused by functionality" mentality of Gnome is a disease. If you think your users are idiots, only idiots will use it.")
My girlfriend is absolutely computer illiterate: she thinks (well, thought) that Office is the OS that runs on his laptop. Being lazy and all she often sits down to my computer (instead of opening her laptop) to browse the net. Sometimes she doesn't even notice that instead of firefox, she is using konqueror. There is a small set of functionality that users expect at specific areas of your screen: first buttons should be back and forward, they expect an input field for URLs at the top, maybe a google search bar... and that's it. If they are there, they are not really "confused" because there are additional buttons (kget, print, even cervisia) to the right side. They don't even notice it. It is the same with the file dialog: were users really bothered by the input field? I very much doubt that - and just like Linus, I was not aware of ctrl + L until someone told me here on ./. And in the past years, I hear one bogus "usability" claim from these so called "usability experts" after another (spatial nautilus anyone?) No evidence, no empirical study, just "we say so as usability experts" with some outlandish theory to back it up... so yeah, I think he is right on spot (and yeah, yeah, we know, diplomacy is not his forte).
Win* usage 2003 oct (overall: nt, 98, 2k, xp, etc) - 93,2 %
Win* usage 2005 oct - 90% - down more than 3% in two years... not much, I admit, but not bad either, considering the rate of development (and the state of desktop linux even two years ago...).
Where did that 3% go? Apparently 1-1 to Linux and Mac, and 1 to something other (in my case, I'm in other with FreeBSD).
The article also reminded me of this blogI read sometimes ago. Its author (a kde dev) works for a company who provides linux solutions for - among other things - schools. The idea of an Office Suite aimed specifically at kids is very interesting, and you can see nice mockups of such a possibilities. Seeing the pictures commented Forget the children. Make this for my 58 year-old mother - and there is a part two where the developer further elaborates on the idea. Actually they are looking for volunteers to implement it:
Perhaps the strong focus on education (the edutainment package has a very active developer community behind it) was one of the reasons why Mark Shuttleworth "promoted" KDE (or rather, Kubuntu) to a tier 1 status when the Novell thing happend a few weeks ago.Scribus changed that perception, and that's why I think it deserves a little marketing: it is usable with little effort for Joe Average desktop user, if he or she wants to quickly compile a document or presentation that looks cool. This is not to say it is not complex, because it is, but you don't have to learn everything to produce better looking documents than you could in oo.o, and faster! A larger userbase might lead to more feedback on missing features (or bugs) and faster development perhaps. Also, it is a useful application not only for people in publishing business, but average desktop users as well...
Scribus exists for some time now, and still, it was only a few months ago that I heard about it, and only 3 weeks ago that I installed it. And I was absolutely amazed at its quality (and standard support, even the latest pdf specs, including embedded script support!). What OSDL needs is a central place where all the great apps likely to be needed in an enterprise/office environment should be listed. There aren't that many actually, but how many of the respondends who missed PageMaker knew about scribus? Seeing the quality of the app, I was surprised at the lack of marketing/hype this application receives. Even their website ... well, it is not bad, but it isn't good either.
Right now marketing focuses on distributions or tools, and rarely on the application stack (except FF and OO.o). GNOME does a good job at marketing as well. But KDE? They formed a new marketing group only recently, and as they started to look around, they found plenty of examples for KDE use in business, like this one (Dutch Record Shop Chain Migrates 1000 PCs to KDE). Or take a look at this initiave: part 1, part 2.. Or another example here. Quote:
Cases like these will convince businesses to adopt linux solutions, and as the article says, not necessarily because lower costs, but because of the quality of the software out there. But there needs to be a central place that enumerates and provides a short description of the application stack (I think 10-15 desktop apps, no more, that are essential for business) as well as provides examples for the various scenarious where free software can be put to use. KDE in hospitals controlling everything including x-rays, a music record chain with desktop locked down via kiosk to include the 4 necessary apps, satellite tv transmission software - this is staggering if you think of it, and it should be shown to CIOs and PHBs (make a nice newsletter/booklet in scribus for instance
Someone mentioned konqi as a viable alternative now for ff ... indeed, konqi has been an excellent browser in the 3.4.x series already: easy on resources, and fast as hell. But to become a real competitor, it needs to be ported to windows. With QT4, that would probably happen. What we really need is lightweight, feature rich and modern Office competitor, and koffice could become one. It is already being ported to qt4 and windows (I saw a demo of kexi on windows). Karbon and Krita received a lot of attention lately, and it shows. Krita is beginning to replace GIMP for me here: it has a very nice and usable UI, lots of features (excellent layer control). Karbon has come a long way as well - it is not yet a viable replacement for Inkscape, but it's getting there. Now only if the rest of the office suite received the same amount of attention ... I know, I know: developers, developers, developers :))
Another nice app is Scribus! It is the best free desktop publishing application, and it is relatively easy to use. It is also needs more marketing I think - there is simply no free software alternative for scribus, yet I haven't heard about it until a month ago. After installing it I was absolutely surprised at its quality!
Mod parent funny pls.