Three month of development for OS 5.x, which is a messy hybrid of old Dragonball code and new ARM stuff with mediocre development tools (e.g. POSE does not emulate OS 5.x machines, AFAIK) ? Are they nuts ? Or did they really think, that somebody is going to plop down money for the dev. kit, sign few NDAs to get the docs and implement it for $5k which probably does not even cover the price of the dev. kit.
Not to mention which Bluetooth card did they have in mind ? The ultra-proprietary Palm one ? As if there was anything else.
Unrealistic expectations doomed this project from the start, IMHO.
Just by chance, did you change your USER_AGENT string somehow ? If it is set to MSIE or something else than Mozilla/Gecko, Java will crash. This is a know bug of SUNs JDK:((
I really do wonder, why are all those people crazy about those 5.1, 6.0, AC-3 whatever systems.
It *is* possible to get 3D sound with just two speakers/headphones. Headphones are of course much preferable. Finally, humans have just two ears, not five or six. Trick is in the processing - the feeling of space is achieved not only by using intensity but also phase of the sound. The algorithms to do that are known, just Google for HRTF (head-related transfer function) - e.g. here.
If you have a good HRTF and a geometrical model of the space, you can recreate very accurate sound reproduction, with just two speakers/headphones.
EAX and DirectSound took a very rough approximation of HRTF and some rough approximation of the space (e.g. concert hall, church, etc.) and give you list of filters. The effects have nothing to do with reality and you will not get better spatial feeling using even twenty speakers. You do not take into account reflections, material on the walls, standing wave effects etc.
For people interested in accoustic, have a look here. I had a short course with prof. Rindel, who is one of the authors of the ODEON software (there is a free demo on the page) and the stuff is really impressive. It beats things like EAX or very expensive 5.1 setups hands down. If modelling of this sort was supported by hardware, that would be the real revolution in computer audio. BTW, this technology was used as a part of CAHRISMA EU project, which we participated in (for the virtual reality part), the stuff is pretty much usable in real time already ( CAHRISMA at DTU, CAHRISMA at our lab, Something on the VR aspects of the project
ABI is not the issue. The problem is module versioning, that's why you have to rebuild e.g. the nvidia module after each kernel replacement. Module versioning is a good thing, it prevents loading modules which will break your kernel because they do not use the same API.
Moreover, you are barking up the wrong tree here - what forces you to update your kernel ? Unless there is a big security hole (rare) or it doesn't work for you right, you do not have to run the bleeding edge. If it isn't broken, do not fix it.
OOo can access MySQL (or just any database) using ODBC or JDBC, creating the reports and mail merges is a breeze. With a bit of custom scripting, you can get even very advanced reports.
Or until your mail is not relayed through somebody else (think ISP or campus with centralized mail handling). Then you are lost.
This is a hack of similar sort as rejecting all mails where your name is not in To: or Cc: field. Worked, but for very short time and with great risk of trouble for legitimate mail.
You are wrong big time here. Blender can do plenty of things Maya can (we have both and also Max here). For learning the ropes, Blender is actually great, because you do not need shell out lots of money on HW and software (did you check the requirements to work comfortably with Maya or 3DS Max ?) and the user interface is a lot simpler (even though a bit weird). It is more than enough for most projects somebody interested about CG will attempt on his/her own.
However for doing professional CG, Blender will not help you, because nobody uses it (I mean studios, not enthusiasts), there is very little 3rd-party market for plugins, no export/import tools for industry standard files (which are Max and Maya ones), no support for commercial content production pipelines (CG production *does not* start and end with Maya or Max work! There is plenty of stuff involved around) etc.
Maya PLE is nice, if you want to learn Maya, but for nothing else. You cannot use the expertise with another software, because it works differently. The program is crippled, so you cannot really use it for anything except learning how to use the interface and how to use the modeling tools. As a personal training tool for designers aspiring for job at a studio using Maya it could be useful, but not for anything else.
Actually, if you are a student or a school, Alias has a special licensing program, where you can get a full Maya for free, if you are eligible. It is restricted by a special annually renewed license, for which you have to submit some projects you do with it/did with it + some paperwork, however it beats the crap out of the PLE edition. And it is available for Linux as well, unlike the PLE, which is only for Mac and Windows. However installing the thing on Linux is not fun - I tried to do it and it is really hair-pulling, head-banging experience - undocumented, buggy as hell, basically 1:1 converted IRIX installation to Linux. Forget about some rpms or such. Uses FlexLM from GlobeTrotter for license management, so it is a lot of fun.
Mandrake is now in the freeze and beta period for 10.1 release, which is due soon, so KDE 3.3 will be only in 10.2/11.0 => probably February/March 2005 for the Mandrakelinux Official.
It may be earlier than that in the 10.1 Community release, which receives updates more frequently, but I wouldn't hold my breath for that - Mandrake's KDE is heavily customized and it is a huge package to update.
Perhaps there will be some packages from contributors.
Cooker will most likely have 3.3 right after 10.1 release, when the freeze ends.
My T41p (T42 is out already) lasts 4-4.5 hours going full speed with the larger (9 cell) battery. With power management on (lower backlight, CPU throttling etc), it last cca 6 hours without problem. Should be more than enough for your needs. Another plus - the HW is completely Linux friendly, everything is supported and works.
SpamAssassin does not do any neural networks. It just matches rules against the mail and at the end totals the scores assigned by the rules. If the total is higher than some (arbitrary) threshold, mail gets tagged as spam. That's all. (ignoring Bayesian classifier in SA for now, however that is also treated just as a special case of a rule)
Neural networks do not use any rules - they work on feature vectors extracted from the input and send them through something like a state machine on steroids (I am sorry for AI people who cringe at this comment now..) where each transition has assigned weight. The weights are then used to decide, whether the mail is spam or not. These weights are in turn assigned by a learning process and adapted continuosly, whereas the SpamAssassin rules have static scores, which do not change.
Statistical filters (e.g. Bayesian schemes) work in a similar way - they compute a "spaminess" of each token (probability whether the token is more likely to appear in spam or clear mail) according to a database created by training. From the probabilities of the tokens, probability of the whole message being spam is computed - the formula for that is called Bayes formula, therefore the name of the method. If the probability is higher than a threshold, message is tagged.
How exactly is this news ? It seems that the author of the neural network idea didn't do his homework - e.g. DSPAM includes neural network as an experimental classifier already. And compared to the proposed C# solution, DSPAM is a widely used and mature product already.
Well, it depends. I owned a Visor, one Clie, Tungsten C and now I have a Zaurus C760. I have also cell phone which has some rudimentary PDA functions (calendar, notes, addressbook). It all comes down to what you want to use the device for.
If you need just to keep track of phone numbers and occassional appointment, then probably a PDA is an overkill for you. The same if you expect a PC-like functionality from it. It was just not intended for that use.
However, if you need an ultra-portable device, which is capable of decent networking (there are no cell phones with WiFi and data connections via e.g. GPRS are very slow and for exorbitant prices.), has usable screen for data entry and a bit of computing power to handle e.g. VPN, decent mail client and web browser, then probably a cell phone is not going to cut it. Also I want my cell phone small, not a huge brick it would have to be if the manufacturer wanted to accomodate large screen usable for PDA functions. Just look at the Treo smartphones, they are very big for a phone and the screens are still the good old crappy 160x160 Palm resolution (not sure about the recent 600 Treo). And Treos are probably the best PDA-phone combination (or smartphone if you want) that there is on the market today.
Compared to laptops, yeah, laptop is more powerful than any PDA, that's true. However, my Visor ran on one set of AAA batteries for two weeks with normal use and over 12 hours with intensive use in a day long meeting entering data via external keyboard. Even the most recent Centrino laptops have problems to last that long. Not to mention the portability - I am carrying my Zaurus all the time with me, my T41 Thinkpad has 2.5kg, which do not count as a brick yet, but are not something pleasant to constantly lug around neither.
Finally, usage patterns - with PDA, you can just whip it out, power it up, look up some phone number or whatever and turn it off again in seconds. That's about the time you manage to type something on your cell phone fighting with the clumsy interface or your laptop starts to be usable after waking up from suspend.
So, I do not think that because Sony exited the market with their horribly overpriced and often flaky PDAs, which refused to support e.g. Compact Flash because it would undercut the sales of their expensive Memorysticks (even in the NZ line, which had the slot for that - the slot can be used only for their proprietary and very expensive WiFi card), the whole market is going down the drain. There is a saturation in the market and little compeling reason to buy a new PDA if your old one still works fine. The amount of innovation brought by Palm, Sony and HP (iPaqs..) is abysmal in recent years, so no wonder that people do not buy. The largest peeves of the current PDAs - decent keyboard, decent display (Psion anybody?) and finally stable OS (yeah, both PalmOS and WinCE suck here) are still missing. Sharp is on good track with their C7xx line here, however the software leaves a lot to be desired and in the marketing department Sharp is shooting itself in the foot, IMHO. Fortunately, there is plenty of excelent free software for Zaurus available.
Calling PDAs fad which is dying out is at least bit premature, IMHO. If you have no use for it, do not buy one. For me it was a tremendous help, regardless of having a laptop and cell phone already. I like devices that take the "UNIX approach" - do just one thing and do it well. Laptop and cell phone do not do PDA role well, so why to push them there.
Well, I am bit tired of these "my mother" tests. There was also an aunt Tillie test, which was about the same. Both are actually completely irrelevant.
Why everybody encountering something that behaves differently than he/she expects starts complaining and claims that "it is the reason why Linux has an insignificant market share" ? Personally, I use Linux, Windows and sometimes Mac. Each is good for something else and they are (surprise!) tools, not a religion.
I am working with Linux most of the time but when I have to use Windows or Mac, I have problems because the things do not work as I am used to. Does it mean, that Windows or Mac will never have significant market share ? Actually, market share has nothing to do with usability but all with marketing, folks. You rarely get something sold or adopted just based on usability.
Back to the "mother test" - what is most important for adoption of Linux (or whichever OS or application) is not that, whether it behaves exactly the same as the thing you used before - if it did, why did you switch in the first place? Because it is "cool" ? I doubt it, that's only what the proverbial 13 year olds care about.
It all comes down to the motivation - is this new app or OS delivering something so new, that I can swallow the inconvenience of learning something different or putting up with something not working the way I was used to or am I just looking for an excuse why not switch ? If the answer is "yes, there is something that I need", then the app will get adoption regardless whether something silly as clipboard works the same as on Windows, Mac or whatever. If such problem deters you from using the application, you probably do not need its functionality enough and you did not need to switch to it in the first place.
An example for people complaining about Linux/Unix users being arrogant here. Disclaimer - the example holds in opposite direction as well, I didn't want to pick on some anti-americanism or some similar bull here.
In the U.S., most cars have A/C and automatic gearbox. Here in Europe, they mostly lack both. If an American comes to Europe and rents a car, discovering this fact, do you think that the clutch pedal will automagically disappear and the gearbox change to automatic just because he was used to have it that way at home ? No, it wont, the driver has to adapt and learn how to drive manually or rent another car. Does it mean, that such cars would not sell, because the manufacturers are arrogant and expecting the users to adapt ? Somehow doesn't compute neither. In the case of the car, it came down to the decision - "Do I need to drive so badly that I can put up with it or am I rather going to walk ?"
To conclude this my little rant, I agree, that the cut/copy-paste behavior on Linux is inconsistent sometimes and that there are applications which are broken and need to be fixed. However, this not Linux specific issue at all and hardly something preventing its adoption:-(
Yep, sure, however, if you flood the network with your packets, it will be probably noticed soon. Also in order to flood the network, you need a fast connection to the attacked host, reducing the problem to the immediate neighborhood of the attacked host. That's why this attack is not very practical to perform and the impact on most normal mortals is low (except for the BGP issue, which could take out your upstream router)
The problem with BGP, which is mentioned in the original article, is that it has a particularly bad failure mode - when a BGP session goes down, it is very expensive to restart it. To top it off, BGP uses very long transmission windows (because it transfers lot of data and that is more efficient with longer windows) and that makes it easier to squeeze-in the spoofed packet. This is quite messy affair, when it happens.
However, if somebody attacks your favorite web site in this way, I doubt that you are even going to notice it, since http sessions are stateless and comparably cheap to establish.
So, I think this is just a big scare for nothing, it is mainly BGP issue for now. It affects just people who know how to fix it (and are fixing it right now) and the machines involved are relatively few. Unlike some Windows RPC exploit which unleashed a massive world-wide worm in few minutes.
Actually, this is just a variation on an old attack with sequence number guessing. Some OSes have a very poor generator for these numbers (even though the vulnerability is known for ages) and it is possible to exploit it. Probably the most famous attack of this sort was Mitnick's break-in into Shimomura's machine ten years ago.
In practice, these attacks are quite tricky to perform, because they require good timing and quite a bit of skills, so they are quite rare. It is far easier to just exploit some common buffer overflow than this.
The impact could be not only DOS (by resetting the connection), but you can do also packet injection with potential for total system compromise. However, if the protocol used is encrypted (in Shimomura's case it wasn't, if I am not mistaken, Mitnick attacked rsh or rlogin), then the DOS is probably the worst thing that could happen to you. The encryption will reject the bogus packets, if properly implemented.
So, keep calm - this was here since at least 80-ties and there isn't much that you can do against it, except to use encryption. You can only hope, that your system vendor will decide to make their TCP stack less vulnerable (not likely to happen, since many systems still ship with the crappy sequence number generator!). Full fix is not possible anyway, unless you want to break the protocol.
The original poster was talking about spreadsheets, not pasting just text. I agree, the analogy was a bit exaggerated, but it makes a point - if you paste something that the application does not expect, you get poor results.
So, you don't use Gnome - not even any GTK apps? But you're qualified to say that a cut/paste problem doesn't exist on the Linux desktop?
I didn't say that - I said that I do not use Gnome as such, not that I do not use Gnome/GTK applications - e.g. Gimp or Totem, even OpenOffice (which is GTK-based, AFAIK).
The problems with cut&paste you are talking about are rudimentary bugs not a problem in general sense. Yes, it is annoying, but these bugs are being fixed and most applications do not suffer from them. This is a grossly overstated "problem".
Unix GUIs have worse problems, but none of them is cut&paste or X being slow. E.g. lack of good hardware support (ATI, please write better drivers!) or usability improvements (the infamous KDE menu...)
How come ? When I drag a file, I get a nice menu with 'Link here' entry, which creates symbolic link to the file. No KDE shortcut (.desktop file). I guess that your KDE is badly misconfigured.
Far slow on a modern hardware ? By what measure ? Compared to XP on my notebook (1.7GHz Centrino), it is lot faster, especially when browsing directories with lots of files. Not to mention things like MSIE mysteriously locking up and waiting for who-knows-what...
Pardon ? You are repeating the ancient FUD about X. When did you check the facts last time ?
ad 1) - dumb trolling on a stupid typo by the article poster
ad 2) cut&paste works fine, even with images and spreadsheets. Did you try OpenOffice or Koffice ? Probably not. If your Gnome has problems with it, that does not mean that *all* X-based UIs have problems with it. I guess that it works right even inside Gnome (although I do not use it myself), the standards for drag&drop are in place for very long time already. Interoperability between different applications could be better, but that holds for Windows and Mac as well. If you paste something from Excell into Photoshop, you are going to get less-than-stellar result too, because the application just does not expect that kind of data.
The bull about unix sockets is so ancient FUD, that I am not sure, whether it is even worth commenting on. Yes, even local clients use unix sockets, because you know what ? It is equaly fast or faster than anything else available (even shared memory). That something runs over socket does not inherently mean that it is inefficient. Not to mention the advantage, when you really need to run something over a network. Windows nor Mac are unable to do that without a costly 3rd-party add-ons (OK, XP has RDP now, but that is hardly the same thing).
Current X UIs are plenty fast, in many cases a lot faster than the Apple or Windows UIs, even though the latter run localy, direct on the hardware. How could that be true ? Could an application design be the issue ? No, let's just bash X instead, because it can do many things I do not use, so it has to be bloated and slow...
Actually, if you feel that the application is slow, in 99% it has nothing to do with X, it is a bug or sloppy coding in the application.
And the remark about rendering images off-screen in order to display them - are you sure, you know what are you talking about ? Any graphic engine has to unpack the image into a buffer somewhere. And most (if not all) UIs have nice libraries, which do this for you without having to bother with X pixmaps. Windows and Mac just hide this one step from you, but it happens anyway.
How the parent could have been moderated insightful is really beyond me:-( Mindless bashing of X seems to be really popular topic.
In fact, it does - what if MS files patents on the CIFS APIs ? Samba has no means how to license those, and even though, theoretically, the API is "open", it is not, in fact. And Microsoft is filling patents one after another these days. This is exactly what Register tries to point out and you have missed.
Regards,
Jan
Slashdot Mirror
Not to mention which Bluetooth card did they have in mind ? The ultra-proprietary Palm one ? As if there was anything else.
Unrealistic expectations doomed this project from the start, IMHO.
Just by chance, did you change your USER_AGENT string somehow ? If it is set to MSIE or something else than Mozilla/Gecko, Java will crash. This is a know bug of SUNs JDK :((
It *is* possible to get 3D sound with just two speakers/headphones. Headphones are of course much preferable. Finally, humans have just two ears, not five or six. Trick is in the processing - the feeling of space is achieved not only by using intensity but also phase of the sound. The algorithms to do that are known, just Google for HRTF (head-related transfer function) - e.g. here.
If you have a good HRTF and a geometrical model of the space, you can recreate very accurate sound reproduction, with just two speakers/headphones.
EAX and DirectSound took a very rough approximation of HRTF and some rough approximation of the space (e.g. concert hall, church, etc.) and give you list of filters. The effects have nothing to do with reality and you will not get better spatial feeling using even twenty speakers. You do not take into account reflections, material on the walls, standing wave effects etc.
For people interested in accoustic, have a look here. I had a short course with prof. Rindel, who is one of the authors of the ODEON software (there is a free demo on the page) and the stuff is really impressive. It beats things like EAX or very expensive 5.1 setups hands down. If modelling of this sort was supported by hardware, that would be the real revolution in computer audio. BTW, this technology was used as a part of CAHRISMA EU project, which we participated in (for the virtual reality part), the stuff is pretty much usable in real time already ( CAHRISMA at DTU, CAHRISMA at our lab, Something on the VR aspects of the project
Moreover, you are barking up the wrong tree here - what forces you to update your kernel ? Unless there is a big security hole (rare) or it doesn't work for you right, you do not have to run the bleeding edge. If it isn't broken, do not fix it.
OOo can access MySQL (or just any database) using ODBC or JDBC, creating the reports and mail merges is a breeze. With a bit of custom scripting, you can get even very advanced reports.
This is a hack of similar sort as rejecting all mails where your name is not in To: or Cc: field. Worked, but for very short time and with great risk of trouble for legitimate mail.
However for doing professional CG, Blender will not help you, because nobody uses it (I mean studios, not enthusiasts), there is very little 3rd-party market for plugins, no export/import tools for industry standard files (which are Max and Maya ones), no support for commercial content production pipelines (CG production *does not* start and end with Maya or Max work! There is plenty of stuff involved around) etc.
Maya PLE is nice, if you want to learn Maya, but for nothing else. You cannot use the expertise with another software, because it works differently. The program is crippled, so you cannot really use it for anything except learning how to use the interface and how to use the modeling tools. As a personal training tool for designers aspiring for job at a studio using Maya it could be useful, but not for anything else.
Actually, if you are a student or a school, Alias has a special licensing program, where you can get a full Maya for free, if you are eligible. It is restricted by a special annually renewed license, for which you have to submit some projects you do with it/did with it + some paperwork, however it beats the crap out of the PLE edition. And it is available for Linux as well, unlike the PLE, which is only for Mac and Windows. However installing the thing on Linux is not fun - I tried to do it and it is really hair-pulling, head-banging experience - undocumented, buggy as hell, basically 1:1 converted IRIX installation to Linux. Forget about some rpms or such. Uses FlexLM from GlobeTrotter for license management, so it is a lot of fun.
Mandrake is now in the freeze and beta period for 10.1 release, which is due soon, so KDE 3.3 will be only in 10.2/11.0 => probably February/March 2005 for the Mandrakelinux Official.
It may be earlier than that in the 10.1 Community release, which receives updates more frequently, but I wouldn't hold my breath for that - Mandrake's KDE is heavily customized and it is a huge package to update.
Perhaps there will be some packages from contributors.
Cooker will most likely have 3.3 right after 10.1 release, when the freeze ends.
Regards, Jan
My T41p (T42 is out already) lasts 4-4.5 hours going full speed with the larger (9 cell) battery. With power management on (lower backlight, CPU throttling etc), it last cca 6 hours without problem. Should be more than enough for your needs. Another plus - the HW is completely Linux friendly, everything is supported and works.
Well, I guess that fuxoft is right, it is most likely a shooting game :-)
BTW, folks, Frantisek was a great ZX Spectrum guru east of the Iron Curtain. That was quite a few years ago, but your games had quite a following :-)
SpamAssassin does not do any neural networks. It just matches rules against the mail and at the end totals the scores assigned by the rules. If the total is higher than some (arbitrary) threshold, mail gets tagged as spam. That's all. (ignoring Bayesian classifier in SA for now, however that is also treated just as a special case of a rule)
Neural networks do not use any rules - they work on feature vectors extracted from the input and send them through something like a state machine on steroids (I am sorry for AI people who cringe at this comment now ..) where each transition has assigned weight. The weights are then used to decide, whether the mail is spam or not. These weights are in turn assigned by a learning process and adapted continuosly, whereas the SpamAssassin rules have static scores, which do not change.
Statistical filters (e.g. Bayesian schemes) work in a similar way - they compute a "spaminess" of each token (probability whether the token is more likely to appear in spam or clear mail) according to a database created by training. From the probabilities of the tokens, probability of the whole message being spam is computed - the formula for that is called Bayes formula, therefore the name of the method. If the probability is higher than a threshold, message is tagged.
Jan
How exactly is this news ? It seems that the author of the neural network idea didn't do his homework - e.g. DSPAM includes neural network as an experimental classifier already. And compared to the proposed C# solution, DSPAM is a widely used and mature product already.
Regards, Jan
Well, it depends. I owned a Visor, one Clie, Tungsten C and now I have a Zaurus C760. I have also cell phone which has some rudimentary PDA functions (calendar, notes, addressbook). It all comes down to what you want to use the device for.
If you need just to keep track of phone numbers and occassional appointment, then probably a PDA is an overkill for you. The same if you expect a PC-like functionality from it. It was just not intended for that use.
However, if you need an ultra-portable device, which is capable of decent networking (there are no cell phones with WiFi and data connections via e.g. GPRS are very slow and for exorbitant prices.), has usable screen for data entry and a bit of computing power to handle e.g. VPN, decent mail client and web browser, then probably a cell phone is not going to cut it. Also I want my cell phone small, not a huge brick it would have to be if the manufacturer wanted to accomodate large screen usable for PDA functions. Just look at the Treo smartphones, they are very big for a phone and the screens are still the good old crappy 160x160 Palm resolution (not sure about the recent 600 Treo). And Treos are probably the best PDA-phone combination (or smartphone if you want) that there is on the market today.
Compared to laptops, yeah, laptop is more powerful than any PDA, that's true. However, my Visor ran on one set of AAA batteries for two weeks with normal use and over 12 hours with intensive use in a day long meeting entering data via external keyboard. Even the most recent Centrino laptops have problems to last that long. Not to mention the portability - I am carrying my Zaurus all the time with me, my T41 Thinkpad has 2.5kg, which do not count as a brick yet, but are not something pleasant to constantly lug around neither.
Finally, usage patterns - with PDA, you can just whip it out, power it up, look up some phone number or whatever and turn it off again in seconds. That's about the time you manage to type something on your cell phone fighting with the clumsy interface or your laptop starts to be usable after waking up from suspend.
So, I do not think that because Sony exited the market with their horribly overpriced and often flaky PDAs, which refused to support e.g. Compact Flash because it would undercut the sales of their expensive Memorysticks (even in the NZ line, which had the slot for that - the slot can be used only for their proprietary and very expensive WiFi card), the whole market is going down the drain. There is a saturation in the market and little compeling reason to buy a new PDA if your old one still works fine. The amount of innovation brought by Palm, Sony and HP (iPaqs ..) is abysmal in recent years, so no wonder that people do not buy. The largest peeves of the current PDAs - decent keyboard, decent display (Psion anybody?) and finally stable OS (yeah, both PalmOS and WinCE suck here) are still missing. Sharp is on good track with their C7xx line here, however the software leaves a lot to be desired and in the marketing department Sharp is shooting itself in the foot, IMHO. Fortunately, there is plenty of excelent free software for Zaurus available.
Calling PDAs fad which is dying out is at least bit premature, IMHO. If you have no use for it, do not buy one. For me it was a tremendous help, regardless of having a laptop and cell phone already. I like devices that take the "UNIX approach" - do just one thing and do it well. Laptop and cell phone do not do PDA role well, so why to push them there.
Regards, Jan
Single Button Launcher
and this: Teal Launch
and this: QuickSwitch
And many more similar Palm apps, which I used before (some of them are even free).
So, what is exactly new on this, that it warrants a patent ?
Well, I am bit tired of these "my mother" tests. There was also an aunt Tillie test, which was about the same. Both are actually completely irrelevant.
Why everybody encountering something that behaves differently than he/she expects starts complaining and claims that "it is the reason why Linux has an insignificant market share" ? Personally, I use Linux, Windows and sometimes Mac. Each is good for something else and they are (surprise!) tools, not a religion.
I am working with Linux most of the time but when I have to use Windows or Mac, I have problems because the things do not work as I am used to. Does it mean, that Windows or Mac will never have significant market share ? Actually, market share has nothing to do with usability but all with marketing, folks. You rarely get something sold or adopted just based on usability.
Back to the "mother test" - what is most important for adoption of Linux (or whichever OS or application) is not that, whether it behaves exactly the same as the thing you used before - if it did, why did you switch in the first place? Because it is "cool" ? I doubt it, that's only what the proverbial 13 year olds care about.
It all comes down to the motivation - is this new app or OS delivering something so new, that I can swallow the inconvenience of learning something different or putting up with something not working the way I was used to or am I just looking for an excuse why not switch ? If the answer is "yes, there is something that I need", then the app will get adoption regardless whether something silly as clipboard works the same as on Windows, Mac or whatever. If such problem deters you from using the application, you probably do not need its functionality enough and you did not need to switch to it in the first place.
An example for people complaining about Linux/Unix users being arrogant here. Disclaimer - the example holds in opposite direction as well, I didn't want to pick on some anti-americanism or some similar bull here.
In the U.S., most cars have A/C and automatic gearbox. Here in Europe, they mostly lack both. If an American comes to Europe and rents a car, discovering this fact, do you think that the clutch pedal will automagically disappear and the gearbox change to automatic just because he was used to have it that way at home ? No, it wont, the driver has to adapt and learn how to drive manually or rent another car. Does it mean, that such cars would not sell, because the manufacturers are arrogant and expecting the users to adapt ? Somehow doesn't compute neither. In the case of the car, it came down to the decision - "Do I need to drive so badly that I can put up with it or am I rather going to walk ?"
To conclude this my little rant, I agree, that the cut/copy-paste behavior on Linux is inconsistent sometimes and that there are applications which are broken and need to be fixed. However, this not Linux specific issue at all and hardly something preventing its adoption :-(
Yep, sure, however, if you flood the network with your packets, it will be probably noticed soon. Also in order to flood the network, you need a fast connection to the attacked host, reducing the problem to the immediate neighborhood of the attacked host. That's why this attack is not very practical to perform and the impact on most normal mortals is low (except for the BGP issue, which could take out your upstream router)
The problem with BGP, which is mentioned in the original article, is that it has a particularly bad failure mode - when a BGP session goes down, it is very expensive to restart it. To top it off, BGP uses very long transmission windows (because it transfers lot of data and that is more efficient with longer windows) and that makes it easier to squeeze-in the spoofed packet. This is quite messy affair, when it happens.
However, if somebody attacks your favorite web site in this way, I doubt that you are even going to notice it, since http sessions are stateless and comparably cheap to establish.
So, I think this is just a big scare for nothing, it is mainly BGP issue for now. It affects just people who know how to fix it (and are fixing it right now) and the machines involved are relatively few. Unlike some Windows RPC exploit which unleashed a massive world-wide worm in few minutes.
Actually, this is just a variation on an old attack with sequence number guessing. Some OSes have a very poor generator for these numbers (even though the vulnerability is known for ages) and it is possible to exploit it. Probably the most famous attack of this sort was Mitnick's break-in into Shimomura's machine ten years ago.
In practice, these attacks are quite tricky to perform, because they require good timing and quite a bit of skills, so they are quite rare. It is far easier to just exploit some common buffer overflow than this.
The impact could be not only DOS (by resetting the connection), but you can do also packet injection with potential for total system compromise. However, if the protocol used is encrypted (in Shimomura's case it wasn't, if I am not mistaken, Mitnick attacked rsh or rlogin), then the DOS is probably the worst thing that could happen to you. The encryption will reject the bogus packets, if properly implemented.
So, keep calm - this was here since at least 80-ties and there isn't much that you can do against it, except to use encryption. You can only hope, that your system vendor will decide to make their TCP stack less vulnerable (not likely to happen, since many systems still ship with the crappy sequence number generator!). Full fix is not possible anyway, unless you want to break the protocol.
Regards, JanLink to the old story
Here comes the mighty dupe ...
The original poster was talking about spreadsheets, not pasting just text. I agree, the analogy was a bit exaggerated, but it makes a point - if you paste something that the application does not expect, you get poor results.
I didn't say that - I said that I do not use Gnome as such, not that I do not use Gnome/GTK applications - e.g. Gimp or Totem, even OpenOffice (which is GTK-based, AFAIK).
The problems with cut&paste you are talking about are rudimentary bugs not a problem in general sense. Yes, it is annoying, but these bugs are being fixed and most applications do not suffer from them. This is a grossly overstated "problem".
Unix GUIs have worse problems, but none of them is cut&paste or X being slow. E.g. lack of good hardware support (ATI, please write better drivers!) or usability improvements (the infamous KDE menu ...)
How come ? When I drag a file, I get a nice menu with 'Link here' entry, which creates symbolic link to the file. No KDE shortcut (.desktop file). I guess that your KDE is badly misconfigured.
Far slow on a modern hardware ? By what measure ? Compared to XP on my notebook (1.7GHz Centrino), it is lot faster, especially when browsing directories with lots of files. Not to mention things like MSIE mysteriously locking up and waiting for who-knows-what ...
Pardon ? You are repeating the ancient FUD about X. When did you check the facts last time ?
ad 1) - dumb trolling on a stupid typo by the article poster
ad 2) cut&paste works fine, even with images and spreadsheets. Did you try OpenOffice or Koffice ? Probably not. If your Gnome has problems with it, that does not mean that *all* X-based UIs have problems with it. I guess that it works right even inside Gnome (although I do not use it myself), the standards for drag&drop are in place for very long time already. Interoperability between different applications could be better, but that holds for Windows and Mac as well. If you paste something from Excell into Photoshop, you are going to get less-than-stellar result too, because the application just does not expect that kind of data.
The bull about unix sockets is so ancient FUD, that I am not sure, whether it is even worth commenting on. Yes, even local clients use unix sockets, because you know what ? It is equaly fast or faster than anything else available (even shared memory). That something runs over socket does not inherently mean that it is inefficient. Not to mention the advantage, when you really need to run something over a network. Windows nor Mac are unable to do that without a costly 3rd-party add-ons (OK, XP has RDP now, but that is hardly the same thing).
Current X UIs are plenty fast, in many cases a lot faster than the Apple or Windows UIs, even though the latter run localy, direct on the hardware. How could that be true ? Could an application design be the issue ? No, let's just bash X instead, because it can do many things I do not use, so it has to be bloated and slow ...
Actually, if you feel that the application is slow, in 99% it has nothing to do with X, it is a bug or sloppy coding in the application.
And the remark about rendering images off-screen in order to display them - are you sure, you know what are you talking about ? Any graphic engine has to unpack the image into a buffer somewhere. And most (if not all) UIs have nice libraries, which do this for you without having to bother with X pixmaps. Windows and Mac just hide this one step from you, but it happens anyway.
How the parent could have been moderated insightful is really beyond meJan
In fact, it does - what if MS files patents on the CIFS APIs ? Samba has no means how to license those, and even though, theoretically, the API is "open", it is not, in fact. And Microsoft is filling patents one after another these days. This is exactly what Register tries to point out and you have missed. Regards, Jan