I could be totally wrong here, but I was under the impression that digital cameras don't even have a shutter. My thinking is something along the lines of a traditional film camera uses a shutter to maintain complete darkness on the film to prevetn the chemical reactions which take place upon exposure to light and the the shutter opening is able to regulate the photo-chemical reaction by limiting exposure. However a digital camera has a sensor which doesn't rely on absolute dark to prevent exposure. It simply captures all the time and the "shutter" button triggers a data capture routine in the camera's electronics which then process the captured images as a composite of all image data captured within the "shutter speed" window. Perhaps there is still some ancillary shuttering mechanism in a DSLR to prevent light from entering through the view finder and washing out the image that passes through the penta prism in the body; but I'm just guessing here.
There is no risk of "screwing up" your camera, the hack loads the "firmware" into volatile ram in such a way that simply deleting the file from your mem card will revert your camera to the original state.
Which is another ridiculousness in an already stupid argument, since the default plugins include several which have no obvious functionality nor any obvious way to configure or learn of the functionality. For instance the "DBus Example" plugin which aside from apparently providing some example of dbus integration doesn't mention what exactly it is supposed to be using dbus for...
They use a standard connector, it is just not in the standard location for a 3.5" form factor. If you were to remove the drive from it's heatsink/3.5" adaptor you'd likely find the connector offset is identical to a standard 2.5" sata drive.
You must be high if you think that the odds of a failed disk aren't significant. I average about 2 disks a month in an environmentally controlled datacenter using enterprise class equipment. I don't lose data because I "OMG DO THE MATH" and realize that data security is important when you have data worth keeping. OTOH if you are so instant that the data you have on your disk/s is so worthless that a loss wouldn't affect you then I'm curious why it'd be important to squeeze any performance out of your storage subsystem since effectively you'd only be speeding up access to your worthless data.
Alas, this all does have an upside. My job will remain secure as long as people like you don't see the big picture and remain DEU instead of applying your 31337 skillz to real systems.
Say that 1 expensive drive has a reliability rating of 2u (arbitrary units, where higher is better) and you are advocating using two cheaper drives with a reliability of 1u each, then striping them gives a combined reliability of 1u/2 or.5u... maybe you get performance approaching the 1 expensive disk but at a cost of 75% reliability. Realistically the target market for these drives is deploying them in arrays with parity and spares. I personally don't see any performance increase when going to 2 drives vs a single since, like everyone else doing a job like mine, I'm more concerned with reliability than balls out performance; to wit 2 drives to me never get striped, a single drive is only usable for something non-critical maybe page/swap, 2 drives is always going to be deployed as a mirror pair where I see a bit of an upswing in read perf, 3+ drives will always be an array with 1+ parity drives and 1+ spare drives.
Storage infrastructure is one of those areas of system work where the nut is much harder to crack than most squirrels would realize;)
It's even eaiser than that on 8.04 HH ubuntu, since the restricted drivers are enabled by default. The only way you'd even know they are being used is because a small icon will pop up in the "tray" and when you hover over it you will see the message about how ubuntu is using an "driver which is restricted due to licensing" or some such.
Well, I've only seen this behavior in a very limited set of circumstances. Namely when you install from usb media or add/remove a drive from an existing system. However if you take a moment to actually read the lengthy usage notes at the top of the menu.lst in ubuntu you will see the following:
## BEGIN AUTOMAGIC KERNELS LIST ## lines between the AUTOMAGIC KERNELS LIST markers will be modified ## by the debian update-grub script except for the default options below
## DO NOT UNCOMMENT THEM, Just edit them to your needs
and just below that
## default grub root device ## e.g. groot=(hd0,0) # groot=(hd0,0)
If you have had this problem many times before and have edited the menu.lst I'd have expected you to have noticed this. At any rate this is not only an infrequent problem for most configurations, but it is one which can manifest itself in windows as well. Perhaps you'll recall a game update for one of the big windows game titles earlier this year would overwrite c:\boot.ini and create a similar situation; however victims of this problem weren't doing anything as necessarily invasive as a kernel update AND they were left with a fix that requires them to edit a file which is not only uncommented, but also hidden in the root directory of their drive. The defacto solution is the same as the one you've been using, i.e. go find the rescue/install disk.
I have a hard time considering arguments against linux which have nearly identical arguments that go the other way. Why is it that everyone expects that linux will seamlessly and effortlessly manage a multi boot system when they don't hold windows to anywhere near the same standard. As someone else in this thread mentioned, you can basically go through the ubuntu installer hitting enter on every screen and successfully resize your windows partition, setup ubuntu with the default options AND end up with a perfectly good multiboot setup.
What is a "mainstream wifi card"? Perhaps an intel abg chipset as is found in anything with intel's centrino branding? Or an Atheros chipset as is found in products from d-link, airlink, iogear, linksys, and just about every other network equipment manuf. out there? Or Maybe a Realtek or Atmel as you'd find in basically anything not possessing either of the 2 aforementioned chipsets. All those are supported out of the box in Ubuntu linux, those that aren't supported directly are in the minority and have poor support in windows too (I know I'm not the only person whose been surprised to find that an oddball marvell or orinco wifi "driver" installer replaces the entire windows wireless stack including the wireless network management interface.)
See my other post in this thread about how "edge cases" are not what makes a mainstream OS. (hint: search for "10 button mouse")
Configuring a ten button mouse is a seriously edge case scenario. If you are using these sort of issue to differentiate what a "mainstream OS" is and isn't then you are shooting way over the target. By definition a mainstream OS is one that hits solidly in the middle of the user base's needs. To that end you'll find that out of the box Ubuntu support for 98% of pointing devices is not only there and quite capable, but actually exceeds what is offered out of the box on windows. For instance touchpad devices are automatically detected and are configurable for both horiz/vert scroll edge actions, dwell events, and tap clicks. All 3 of these features on windows require 3rd party software even though those features are clearly within the 98% of mainstream users wants/needs/expectations. Furthermore if one were to click on the add/remove program item in the application menu and type touchpad into the search box you would find that you can easily, with a single click, add reconfigurability for many other types of touchpad events; while on windows not only would you not have a simple "add/remove program" interface but you'd also not have any easy indicator as to where you'd go to find the appropriate software to configure these features. Don't believe me? type "touchpad software" into google and let your Grandma choose between the 2.3 million results that you get; or heck ask her what a "synaptics touchpad" is.
for example setting up wireless is a bit messy (not as slick as windows) and the power management sucks a bit. This has not been my experience at all. I've run the last 3 generations of ubuntu and HH support for both wireless networking and power managment is, in my experience, at least as simple as on windows and in the case of wireless I'd say it is "slicker" in that it has worked fine out of the box including support for all the major encryption schemes AND vpn AND didn't require any futzing with the crap 3rd party network manager stuff that I've experienced on windows xp. Atheros and intel 3945 ABG chipsets work flawlessly for me and suspend resume functionality didn't require any configuration to work exactly how I'd expect i.e. closing the lid suspends, plugging/unplugging AC dims the screen and changes the processor speed, etc
YMMV, but I'd think that my (non-technical windows using) mother would feel right at home on HH.
Additionally the inclusion of a firefox, an office package, cd burning software and about of course solitaire means the majority of users won't ever want for any additional 3rd party software. Although users will find that the package and update management is far and away fast and more complete than anything windows has to offer.
To be fair it had only a single hit yesterday, which was of course the GP post... Mostly I was so amused by the theory I immediately did a google search to determine if this was an original idea or if GP was just a super genius. Turns out, super genius...
I did read the article AND I disagree with your assessment. It is not obvious that the material of the thing is unimportant, it cannot be determined weather the gauge or cross-sectional shape of the material used is of import. It doesn't specify whether soldering vs. crimping connection will have an impact on the performance nor whether the length of the lead from the connector will affect the signal received. I know that all these things matter as I have dabbled briefly in CB radio "tweaking and peaking" maybe a dozen years ago and small tweaks to the setup of a given rig could yield huge differences in the outcome. I'm not saying that reproducing what they have done is impossible, I'm just saying that reproducing their results would require better documentation and/or a greater basic knowledge of antenna design/construction than is presented on their website.
What this is article is clearly missing is construction directions and a BoM to make replication of their design possible without a background in the field of antenna design/construction. While I'm sure that their design works well for them. I don't see it being very useful to the average garage tinkerer to have this GPL'd without any sort of assembly/construction guidance. For something as fiddly as an antenna things like materials, construction techniques, connector positioning and design can make something which should work great end up not living up to it's potential.
Good call! Also, the parent poster's car analogy is exactly like the processor comparison except that the parent poster also seems to be taken in by the over simplification of the comparison. There are many cars with big HP numbers which are absolute pigs when it comes to performance and likewise there are many cars without huge HP numbers which are seriously fast. A 60hp aircooled vw motor in a formula-Vee could run circles around a 300hp V8 in a 1972 Olds Toronado.
The problem with a ticket system is that there is no public oversight as to what the relative frequency of specific complaints is vs. the manner with how they are acted upon. By using a public forum for the airing of complaints as well as the responses to those complaints it allows for a much higher level of transparency. Sure the volume may be hard to manage and some of the complaints irrational or impractical, but the "wisdom of the masses" should become readily apparent when common complaints far outnumber the ridiculous ones and the TSA is able to address these complaints in batches which leaves less room for differing answers or inconsistent solutions.
I've seen plenty of complaints about UAC, but I'll be damned if I can figure out what the problem is. I find it to not be a hassle that priviledge escalations require notification and approval of the user and that, in the same way my ubuntu installs work, input from other applications to the UAC dialog is locked out to prevent apps from authorizing themselves. I actually can't think of a single example of a UAC dialog appearing that felt excessive nor a time when UAC has prevented me from using the computer in the ways that I've wanted to use it.
The only other gripe I have about Vista is that there is no option the windows updates dialog to have security patches install automatically but require interaction for other updates. IIRC correctly this was an option in XP. It probably wouldn't bug me except that the windows defender definitions get updated at least once every 48 hours and I would really prefer for it to just notify me that the update applied successfully vs. me having to manually click through 2 or 3 panels to approve an update which is really a no-brainer.
Notable changes in SP1Hot fixes and patches rolled up in SP1Release Notes document
Unlike most of the chatter I've read on/. I've been mostly satisifed with my Vista install so far. The only real problems I've experienced is the repackaging of some of the SDK tools such as graphedit which used to be available as standalone, but the 64 bit vista specific version is only available as part of a multi-gig sdk download... Also some vendors have been slow to ship good drivers although I suspect that MS requiring a 64 bit for the "vista compatible" label and not requiring a 32 bit version will in time result in a better driver base.
A multiprotocol client will always be a kludge, some are slightly more elegant than others, but a kludge none the less.
That said, changing protocols on their end doesn't really buy anything for anyone, unless they also support jabbers network interoperability. I.E. an aim user should be able to message directly to a google chat user without both users requiring an account on the others' service.
I've tested that use case as well. I can't profess to understand the exact method by which the two interact, but I experienced similar results with both zfs and ufs.
Non-Sparc hardware... If I pay for Sparc hardware from Sun I expect Sun's version of Sparc Solaris to have, at a minimum, feature parity with X86 solaris on a beige x86 hardware. A little over 3 grand buys me an T1000 spec'd for my task OTOH that same 3 grand buys a much beefier beige box which also runs solaris AND has more features...
Slashdot Mirror
I could be totally wrong here, but I was under the impression that digital cameras don't even have a shutter. My thinking is something along the lines of a traditional film camera uses a shutter to maintain complete darkness on the film to prevetn the chemical reactions which take place upon exposure to light and the the shutter opening is able to regulate the photo-chemical reaction by limiting exposure. However a digital camera has a sensor which doesn't rely on absolute dark to prevent exposure. It simply captures all the time and the "shutter" button triggers a data capture routine in the camera's electronics which then process the captured images as a composite of all image data captured within the "shutter speed" window. Perhaps there is still some ancillary shuttering mechanism in a DSLR to prevent light from entering through the view finder and washing out the image that passes through the penta prism in the body; but I'm just guessing here.
There is no risk of "screwing up" your camera, the hack loads the "firmware" into volatile ram in such a way that simply deleting the file from your mem card will revert your camera to the original state.
Which is another ridiculousness in an already stupid argument, since the default plugins include several which have no obvious functionality nor any obvious way to configure or learn of the functionality. For instance the "DBus Example" plugin which aside from apparently providing some example of dbus integration doesn't mention what exactly it is supposed to be using dbus for...
They use a standard connector, it is just not in the standard location for a 3.5" form factor. If you were to remove the drive from it's heatsink/3.5" adaptor you'd likely find the connector offset is identical to a standard 2.5" sata drive.
You must be high if you think that the odds of a failed disk aren't significant. I average about 2 disks a month in an environmentally controlled datacenter using enterprise class equipment. I don't lose data because I "OMG DO THE MATH" and realize that data security is important when you have data worth keeping. OTOH if you are so instant that the data you have on your disk/s is so worthless that a loss wouldn't affect you then I'm curious why it'd be important to squeeze any performance out of your storage subsystem since effectively you'd only be speeding up access to your worthless data.
Alas, this all does have an upside. My job will remain secure as long as people like you don't see the big picture and remain DEU instead of applying your 31337 skillz to real systems.
Say that 1 expensive drive has a reliability rating of 2u (arbitrary units, where higher is better) and you are advocating using two cheaper drives with a reliability of 1u each, then striping them gives a combined reliability of 1u/2 or .5u... maybe you get performance approaching the 1 expensive disk but at a cost of 75% reliability. Realistically the target market for these drives is deploying them in arrays with parity and spares. I personally don't see any performance increase when going to 2 drives vs a single since, like everyone else doing a job like mine, I'm more concerned with reliability than balls out performance; to wit 2 drives to me never get striped, a single drive is only usable for something non-critical maybe page/swap, 2 drives is always going to be deployed as a mirror pair where I see a bit of an upswing in read perf, 3+ drives will always be an array with 1+ parity drives and 1+ spare drives.
;)
Storage infrastructure is one of those areas of system work where the nut is much harder to crack than most squirrels would realize
It's even eaiser than that on 8.04 HH ubuntu, since the restricted drivers are enabled by default. The only way you'd even know they are being used is because a small icon will pop up in the "tray" and when you hover over it you will see the message about how ubuntu is using an "driver which is restricted due to licensing" or some such.
I have a hard time considering arguments against linux which have nearly identical arguments that go the other way. Why is it that everyone expects that linux will seamlessly and effortlessly manage a multi boot system when they don't hold windows to anywhere near the same standard. As someone else in this thread mentioned, you can basically go through the ubuntu installer hitting enter on every screen and successfully resize your windows partition, setup ubuntu with the default options AND end up with a perfectly good multiboot setup.
See my other post in this thread about how "edge cases" are not what makes a mainstream OS. (hint: search for "10 button mouse")
Configuring a ten button mouse is a seriously edge case scenario. If you are using these sort of issue to differentiate what a "mainstream OS" is and isn't then you are shooting way over the target. By definition a mainstream OS is one that hits solidly in the middle of the user base's needs. To that end you'll find that out of the box Ubuntu support for 98% of pointing devices is not only there and quite capable, but actually exceeds what is offered out of the box on windows. For instance touchpad devices are automatically detected and are configurable for both horiz/vert scroll edge actions, dwell events, and tap clicks. All 3 of these features on windows require 3rd party software even though those features are clearly within the 98% of mainstream users wants/needs/expectations. Furthermore if one were to click on the add/remove program item in the application menu and type touchpad into the search box you would find that you can easily, with a single click, add reconfigurability for many other types of touchpad events; while on windows not only would you not have a simple "add/remove program" interface but you'd also not have any easy indicator as to where you'd go to find the appropriate software to configure these features. Don't believe me? type "touchpad software" into google and let your Grandma choose between the 2.3 million results that you get; or heck ask her what a "synaptics touchpad" is.
Atheros and intel 3945 ABG chipsets work flawlessly for me and suspend resume functionality didn't require any configuration to work exactly how I'd expect i.e. closing the lid suspends, plugging/unplugging AC dims the screen and changes the processor speed, etc
YMMV, but I'd think that my (non-technical windows using) mother would feel right at home on HH.
Additionally the inclusion of a firefox, an office package, cd burning software and about of course solitaire means the majority of users won't ever want for any additional 3rd party software. Although users will find that the package and update management is far and away fast and more complete than anything windows has to offer.
To be fair it had only a single hit yesterday, which was of course the GP post... Mostly I was so amused by the theory I immediately did a google search to determine if this was an original idea or if GP was just a super genius. Turns out, super genius...
Flammeus Fortuitus
GOOGLEWHACK!
I did read the article AND I disagree with your assessment. It is not obvious that the material of the thing is unimportant, it cannot be determined weather the gauge or cross-sectional shape of the material used is of import. It doesn't specify whether soldering vs. crimping connection will have an impact on the performance nor whether the length of the lead from the connector will affect the signal received. I know that all these things matter as I have dabbled briefly in CB radio "tweaking and peaking" maybe a dozen years ago and small tweaks to the setup of a given rig could yield huge differences in the outcome. I'm not saying that reproducing what they have done is impossible, I'm just saying that reproducing their results would require better documentation and/or a greater basic knowledge of antenna design/construction than is presented on their website.
What this is article is clearly missing is construction directions and a BoM to make replication of their design possible without a background in the field of antenna design/construction. While I'm sure that their design works well for them. I don't see it being very useful to the average garage tinkerer to have this GPL'd without any sort of assembly/construction guidance. For something as fiddly as an antenna things like materials, construction techniques, connector positioning and design can make something which should work great end up not living up to it's potential.
Good call! Also, the parent poster's car analogy is exactly like the processor comparison except that the parent poster also seems to be taken in by the over simplification of the comparison. There are many cars with big HP numbers which are absolute pigs when it comes to performance and likewise there are many cars without huge HP numbers which are seriously fast. A 60hp aircooled vw motor in a formula-Vee could run circles around a 300hp V8 in a 1972 Olds Toronado.
The problem with a ticket system is that there is no public oversight as to what the relative frequency of specific complaints is vs. the manner with how they are acted upon. By using a public forum for the airing of complaints as well as the responses to those complaints it allows for a much higher level of transparency. Sure the volume may be hard to manage and some of the complaints irrational or impractical, but the "wisdom of the masses" should become readily apparent when common complaints far outnumber the ridiculous ones and the TSA is able to address these complaints in batches which leaves less room for differing answers or inconsistent solutions.
err, that should say "only gripe besides the ones from my earlier post"
I've seen plenty of complaints about UAC, but I'll be damned if I can figure out what the problem is. I find it to not be a hassle that priviledge escalations require notification and approval of the user and that, in the same way my ubuntu installs work, input from other applications to the UAC dialog is locked out to prevent apps from authorizing themselves. I actually can't think of a single example of a UAC dialog appearing that felt excessive nor a time when UAC has prevented me from using the computer in the ways that I've wanted to use it.
The only other gripe I have about Vista is that there is no option the windows updates dialog to have security patches install automatically but require interaction for other updates. IIRC correctly this was an option in XP. It probably wouldn't bug me except that the windows defender definitions get updated at least once every 48 hours and I would really prefer for it to just notify me that the update applied successfully vs. me having to manually click through 2 or 3 panels to approve an update which is really a no-brainer.
Notable changes in SP1 Hot fixes and patches rolled up in SP1 Release Notes document /. I've been mostly satisifed with my Vista install so far. The only real problems I've experienced is the repackaging of some of the SDK tools such as graphedit which used to be available as standalone, but the 64 bit vista specific version is only available as part of a multi-gig sdk download... Also some vendors have been slow to ship good drivers although I suspect that MS requiring a 64 bit for the "vista compatible" label and not requiring a 32 bit version will in time result in a better driver base.
Unlike most of the chatter I've read on
Gah my mod points expired! Someone mod this up... NN!=GA
Sounds like a broken dhcp server...
A multiprotocol client will always be a kludge, some are slightly more elegant than others, but a kludge none the less.
That said, changing protocols on their end doesn't really buy anything for anyone, unless they also support jabbers network interoperability. I.E. an aim user should be able to message directly to a google chat user without both users requiring an account on the others' service.
I've tested that use case as well. I can't profess to understand the exact method by which the two interact, but I experienced similar results with both zfs and ufs.
Non-Sparc hardware... If I pay for Sparc hardware from Sun I expect Sun's version of Sparc Solaris to have, at a minimum, feature parity with X86 solaris on a beige x86 hardware. A little over 3 grand buys me an T1000 spec'd for my task OTOH that same 3 grand buys a much beefier beige box which also runs solaris AND has more features...