I can't see how those pads are that much better than a Targus Chill Pad. It's powered off your USB port using 1W of power.
Alternatively, you could use a Radio-Shack Rechargable battery pack (couldn't find URL, but catalog no. _was_ 23-047). It's about the size of 4 cassette tapes, ~12.6oz (~350g), output voltage selectable from 3-9v output and rechargable by plugging it into a wall output or from a 12v400ma source. If you are a doit-yourselfer, buy a 4 "cell" holder and wire it in series. Radio Shack sells battery adapter extension cord and heads singlely. You could choose capacity and weight by cell size (though note, I've often seen "D" rechargable cells with same ratings as the "C" indicating they've just stuck a "C" cell in a larger container.
Externally powered, this _should_ slightly increase laptop runtime (i.e. active external cooling => less internal fan use).
It holds the laptop on rubber feet about 7mm above 2 fans sucking air from center of underside and venting out the back.
noatime and no 8.3 file name generation is off in reg for NTFS so it already has a benefit there. Originally the disk was a replacement disk (7200RPM laptop drive vs 5400). I was looking at perf figures before I installed it as a system disk with the FAT32 and NTFS fs's in the sam loc during after reformatting the same partition.
I'm aware of the outer sectors reading slowly -- I see it painfully obviously where on a clean disk a "dd" (I have linux on the sys too, dual boot) will start at rate X and almost be down to.5X by the end of the disk.
I was running tests over USB2.0, Firewire and as an "extra" internal HD. Firewire and internal IDE HD w/NTFS were about.4 as fast as FAT32, USB2.0 was topping out at ~.6 of the speed on FAT32, but about.5 the speed when using NTFS and that's where NTFS was slowest.
Now, I get some of the bene's I need for username security by using a small NTFS partition (basically to install their unix extensions on to play w/), while getting perf bene's for files that don't need separate username management.
Most of my files don't need username management because it is a _laptop_. You still need a username/pw to log into the computer. The fs isn't exported (2-way 3rd party fw, default block).
For a _personal_ computer, I haven't found the overhead of NTFS (and problems easily reading/writing it from Linux). Like the original poster, I'd like rely on something other than windows code to have r/w access to my files. Theoretically, doesn't matter if you can use the virtualized Windows driver for r/w access -- you still need to have bought an authorized copy of Windows to run the driver on your machine. No?
Lots of comments on Apple vs. XP, but with Apple using a Unix core, I was slightly more interested on how it used run time tracing in background and tries to record caching signatures for apps and "fastlinks" for common libraries. Also boot and init.d tracing and moving all the startup programs into the same area of the disk (maybe near the beginning of a partition where read times are usually faster). Wouldn't have to pre-load a cache but having startup programs intelligently laid out would seem to help boot time. Using page-in behavior of programs traced in background to determine pre-fetch strategies might help program startup times. I dunno if Gnu libs/gcc use common lib[c,etc] addr's to speed runtime linkage. Anyone know??
The CPU wastes billions of CPU cycles -- doesn't mean "uses tons of CPU cycles". If I click on a program that isn't in cache, the CPU wastes tons of cycles unless you have it constantly running background programs. This usually isn't the case on a system designed for interactive user-use.
Anyone with a partial clue has been told to turn off as many as possible, or all background applications to speed up performance. So if there are no background applications, just how is the CPU *NOT* wasting time waiting on a READ? Writes are virtually free because of DMA while your program continues on, but READ's -- ya just gotta stop and wait.
As for the answer "just" being a faster HARD DISK -- will admit a 15K HD will feel nice for almost anyone, but the biggest delay these days -- internet latency. Bandwidth is continuing to be expanded, I just recently got mine upped by about 3x down/.75up at the cost of doubling the 10ms latency I had to the nearest router.
It isn't just internet delays -- but worse the latency in a connection to open and return info from a READ. Was just reading a slashdotted article on ARS.Technica...a standard template of the page was returned in under a second, but to display the article content took an additional 30-45 seconds.
The more widgets, the more time to fetch them. Certainly you can re-use a an HTTP connection sometimes, but then you're only getting 'serial' fetch performance.
It seems as more apps work with the network, the single app with worst latency is going to be web browswers.
While there is new HTML standards to allow for prefetch of data after the first read, it's seems it's rarely used. On some occasions, I've gotten noticably less frustrated spidering a site while fetching a cup of coffee, and then reviewing the site in my webcache that waiting 30-45 seconds between each page I want to read. Ug!
Well NTFS also has it's "sucking" points as in "sucking performance". More than one disk read/write test (Sandra, and one used by Magix) has shown my NTFS partition on the same physical hard disk to be 1/4th (that's 25%) of the speed of my FAT32 partition.
Both partitions were formatted using WinXP defaults, FAT32 at 16k segment size, and NTFS at a 1K segment size. NTFS has 86% free space, FAT32 has 38% free space. Both are defragmented and the results were duplicated on multiple hard disk drives, two at 5400RPM and one at 7200RPM. The best I could get out of a NTFS drive was "only" a 66% performance hit (that's 1/3rd the speed of FAT32), sized at same size and place on same disk, successive tests).
NTFS may be more reliable, but on Win2000 and WinXP, I submit that unplanned shutdowns are less frequent and only rarely do I have to run a chkdsk on bootup because a FAT32 partition wasn't unmounted cleanly.
It's not like in linux where the difference between ext2 and journaled filesystems 50-60% depending on the ops and the journaled FS's.
I was distinctly unimpressed. Of course your Mileage may vary.
Even if you have a gig of memory and never use more memory for programs than 300K, and even if you set the memory manager to give priority to running programs over file blocks, you will still see NT use swap.
If you don't give it swap, you won't be able to get various stats related to paging (as when it pages a program into memory) since the module that shows those stats also tries to show swap stats and basically does a divide by zero somewhere and crashes.
My linux server, on the other hand will use 'zero' swap (even if it is allocated, provided I never try to use more memory than I have physical). This was in the 2.4 era. Things may have changed in 2.6 to make it more like Win XP. Dunno.
Sorry, I've called Fry's to do a stockcheck. It doesn't take that long -- not even for them to do a physical stock check. It sounds more like the guy was just not wanting to wait on the phone for a non-commision sale.
There seems to be some questionable methodology for finding the source of SPAM unless it shifts radically in a few months. From an Information Week article written only a few months ago the top three were:
UNCTAD estimates that the majority of spam victims are in the USA but it also says that in March 2003 the USA was the source of 58.4% of spam, followed at a great distance by China (5.6%), the United Kingdom (5.2%), Brazil (4.9%) and Canada (4.1%).
The same source report (along with sources of digital attacks was mentioned in theInquirer.
I'm sure if I kept tracking links in google, I'd find other reports and
other percentages.
Maybe recent US lawsuits have, within a few months, forced most spammers off
shore or perhaps it's just part of the offshoring of US [SPAMMER] jobs.
;-/
In regards to my sig, a bit of dark thoughts: imagine if China became a strict Moslem nation. If stealing incurs the lost of a hand, I could see the penalty for sending spam to be the loss of one finger for each
separate mailing. Of course the worker would be fined and their boss. Eventually that might make a dent in spam. >:=}>
An update system that offers to overwrite/etc/passwd (and presumably every other security file) hardly seems like a safe or easy upgrade process.
I can't say my DoC (SuSE) has it better -- they don't ever seem able to upgrade my system in a sane or coherent manner. Last time around, it upgraded my squid 3.0 to squid2, tried, unsuccessfully to put my named in a basement mail (when it hadn't even been bad), but it was thrown in the basement w/o the root servers file and when the root servers all expired some large amount of time (~3-4 months) later, various TLD's started disappearing. It was bizzare watching large sections of internet just "go away" a few days before it completely consumed itself. Then I found the problem -- it hadn't copied in the root servers file from the previous upgrade (and/or didn't install a new copy). I tried grabbing some updates with their Yast Online solution, but it kept downloading copies of 8.2 binaries when I have 9.0 loaded. I never had 8.2 loaded -- I went straight from 8.1 to 9.0. Later, I found, buried in some paragraph of fine print somewhere that their updates only support updating from the immediately preceding version -- this was after it had removed all unknown packages fro the package database. At this point I had all the 8.1 packages installed, but no longer noted as "installed" in the database over which it automatically upgraded and installed about 10-15 packages out of the 100-150 it should have installed (I guess ~10-15 packages kept some same valid name). I'm always rather afraid to do an upgrade under SuSE as I know it will usually involve lots of pain.
On the flip side -- a fresh install of 9.0 for a never-used-linux user went real smooth -- they were able to navigate their way around after only one or two hiccups -- like buttons weren't where they used to be under Win, but I just told them they'd have to experiment a bit and find out how things were arranged differently. Once they experiemented some, they started finding what they needed surprisingly well.:-( &:-) -l
Old idea, new context and as others have mentioned, the tech isn't even equal to AM quality in many cases let alone FM stero.
But...like yeah, users can turn to the FM band, record a few hours direct to disk, then go back and edit out the talk/advertisements.
You'll get better quality recordings (if you have good FM receiver, and can pull in a strong quality FM station). If you live close to a college many run low-wattage stations geared just to the "right" age group...
Legal? Remember Ted Turner comparing digital video recorders that could skip advertising as video theft! Why do you think video corps are trying to make everything go digital and embed watermarks and require copy control tech in anything that can copy -- and can detect the watermarks even when played over an analogue medium.
Remember the DMCA? It'll be illegal to circumvent any of the copy protection they put in your recording devices. Maybe not a bad idea to keep around old cassette recorders, VCR's and such.
Sound producers have been mumbling/rumbling about new CD formats with better sound (~megabit range)...even now you hear audiophiles bemoan bad audio CD players that simply play the current frequency for the entire 22.73 microseconds between sample as opposed to better decoders that interpolate the difference between each sample and gradually change the frequencies between intervals constantly over the 22.73 microsecond interval. More than one audiophile claims to be able to hear the difference.
Nevertheless, if they push 1Mbit DVD-Audio out as a new standard, it's real easy to also push out a new DRM scheme. Then a scratch on a CD will lose you minutes of music instead of seconds! Such a deal!
Anway....recording off streams will be even easier to control copying of since you'll be going from digital to digital, so less chance of loss of embedded watermarks. So record away while you can...:-(!
I have 5 year old HW in the other room. It has 1G of main memory and 2x1G CPU's. The disks range from 10K SCSI's to 7.2K IDE's w/8m write cache.
But if I am writing anything under 900M from my client to the server, that can all be buffered in memory -- and will be in the timeframes we are talking about. 2.6 will be very likely to buffer the entire free-memory area before dumping to disk and when it does, it will probably have everything arranged in memory to do maximal contiguous writes with no extraneous head movements.
The server has no graphics console and only runs server type SW.
TODAY -- if I was outfitting a similar server it might have 2-4G main memory -- I *might* even have RAID working (not that I haven't tried with the older server, but RAID combos I've tried have run slower than single disk options, so I obviously don't have raid configuration down to an exact science yet (not high enough priority).
Fact is, with the clients I run, they can spit ripped CD's to a network drive faster than to a local drive (bad dell laptop design putting internal CD and internal drive both on same IDE bus). Now let's talk about a 8x DVD ripper running over a 400Mb firewire. 400Mb is 4x faster than 100Mb ethernet -- so I'm likely to be able to rip DVD images 4x faster than my ethernet can take. I may start running into server limitations, but it's easy to see a 400Mb firewire overwhelming a 100Mb ethernet real easily. USB2.0 isn't a problem as it is only around 100-110Mb throughput in my measurements (multiple devices/ multiple HD's). It runs about 10x USB 1.1 due to sloppy USB 2 implementations (mentioned on Tom's).
So 100Mb ethernet is already inadequate if you are coming close to maxing 400Mb firewire -- and when 800Mb firewire arrives...100M ethernet will be dust and your disks better be able to handle near 80Mb/s write to disk speed (or you better have a very large system memory on your target system).
We're on the "edge" now. 1G is quickly going to become a necessity. There have already been articles about how 8x DVD's need to be internal vs. USB 2.0 (don't know about FW). Ack!:-)
I can see smoke coming out of my ethernet cables with the bits being pushed too fast....:-)...
Trackball -- yes. Try this on for size. Buy two (yes expensive) Turbo Pro's -- Wireless variety. Set them both for the same channel. It has 4 main buttons and a wheel that does dual duty as a middle mouse button. I set lower right button for select and drag, and lower left for double click, upper left and right as normal. It has 6 programmable buttons for anything you want.
The secret is the wireless option. Since both mice are on the same channel, you can switch hands to whichever is more convenient -- seamlessly. As long as you don't try to use both at the same time, they don't send signals. Battery life is pretty good -- though I use a steady stream of rechargables.
I use 1 receiver (it's USB, but they include a USB-PS/2 adapter). Because my carpal tunnel symptoms are worse in my right hand, I've become predominatly a left hand mouzer, but the ability to mouse with either hand w/o the hassle of moving the mouse -- but which ever is more convenient is a major bonus. Sometimes I need to primarly use keyboard keys on one hand or the other in combination with a mouse press. Depending on which hand has to be on the keyboard, I can automatically mouse with the other hand.
Becoming bi-mousual, is a major help for overuse on one-hand mousing problems. Also...kensington support has been *GREAT*. Lifetime warantee on the mice. If one malfunctions or I just wear it out -- I call and give them the serial #, and they send out a new one (and I return the bad unit or not depending on random audits). I've had them for about 3 years and maybe gotten 2-3 replacements because I've worn down the wheel's such that the ball starts rubbing against the walls of the inside of the mouse and no longer moves. No problem...they send out a new one. I've never had a company (besides Dell with next business day on-site service) be so good with support. Compare that type of support to HP "Depot" service where you are without your part for 1.5 months while they ship the unit to India to be repaired. I don't think so!
Seriously. It's your lively hood. Lawyers don't think twice about spending 1000's of dollars a year to keep up on the law, and electricians don't think twice about shelling out for a high quality Fluke over a cheaper Radio Shack model.
It's no wonder so many programmers get so damaged, when they work with bear skins and stone knives to carve out their programs. It's your hands, your eyes, your back, your spine (ergo seating, proper posture, monitors with print that doesn't cause you to squint or have to bend forward to read). Think about it. Whether your employer pays for it or not -- if they won't, I'd still claim it as a tax deduction as a necessary tool of your trade (IANATC).
Oh -- I forgot -- corporations were declared to be people.
Um...but what about "equal representation"? Ah...well...the makers set government up to be run by the richest individuals -- with the rich individuals able to command the most influence.
It was a system that sorta worked....with the dream of lower class becoming upperclass, and with 'equal' votes --- but without the votes, directly, really counting -- we elect some "special" group that I'm, a bit fuzzy on -- an "electorate", that makes the real decisions as to who gets elected with individual states getting to choose how individual votes are counted (or discarded) and which electorate entities finally get sent to cast the "representative" [sic] votes of the people.
Classic example: through illegal vote negation, Florida was able to swing the last election -- not Florida, but 1 person, in the family of the to-be elected president. It's the same as the original system -- the rich are the real ones deciding the election and making the laws to keep themselves in power.
The main hiccup that's wreaking the most havoc is this bit about corporations having the status of being "people". There can never be the equality of "all men being created equal" when you have corporations that can outlive any single "man" and can become a super "meta-person" with abilities and resources no single "man" could ever have.
Until that "waving of hands" is undone and "corporations" are stripped of the right of "personhood" with property only being owned by real people who have limited lifespans, we, the people will never have equality and will continue to find ourselves oppressed, more and more, by large corporate "people".
Must be that piracy leads to increased pirated (stolen/shoplifted CD's)...
Certainly people couldn't be using downloaded CD's to see if they want to *buy* CD's....we know those downloaders are criminals...we just have re readjust the facts to fit our story....it must be CD' thefts that have increased!:-/ -l
Having seen perfectly good stories rejected, and seen other, questionably newsworthy stories posted, with comments at the end like "go off and discuss among yourselves", I've felt, sometimes, like the gatekeepers view slashdottee's like their children with articles posted as gifts to us. I've felt (at times) they hand out the articles like like "toys" to children where they are then told to go play with their toys for a while. I think it may be the case that sometimes they are rushed or overbusy and just through out "whatever" -- where they may not realize that either they are being played by a submitter they normally trust or they might be unconsciously creating their own spin on an article.
I often feel that the slashdot.org system has one central flaw that distorts the community feel of the board: users are not allowed to moderate initial postings. While comments can be moderated by the user community, allowing good opinions to be highlighted and things that are deemed "not good" to be rated down, I feel that for slashdot to be a community resource, the community should be allowed to moderate the initial submissions as well as responses.
I "wish" that could be implemented and the initial submitters wouldn't take it personally any more than we are told to take frequent initial submission rejects personally, but I don't see the "powers that be" being very motivated to add such a feedback mechanism as there isn't really a competitive board to drive such a change.
They are having "trouble" installing linux in Munich because it is more expensive? What is more expensive? What is the "trouble"?
Oh!...The last time the installed, they had to install via floppies and they had to upgrade the systems with CD or DVD readers? Is that the source of the "expense"...?
The news article is not very "enlightening"...probably even sourced by MS as press fodder. I mean, what is the "trouble"? You can copy one image to the network, then use autoinstall to setup 100's of PC's. Now that is all included in my one copy of $100 SuSE. I cannot even buy one copy of Windows for use in a network for that, let alone be provided with autoinstall SW.
The article feels more like FUD than substance...does anyone know what part of the upgrade cost / workstation is costing more?
Besides, why don't these cities just setup the conversion as an open-source project on source-forge -- they could likely get all the software conversion done for near free and hire someone locally to maintain it? What, exactly, does a city need in a distro to "run itself?" If multi-cities are converting over, why not just develop an open-source city-distro. Seems like a specialized class of SuSE's...it or Mandrake has the best multi-lang support. Red Hat seems aimed at commercial, but why not have cities turn to its citizens, or citizens of the net to develop a "by-the-people" software distro?
What software do you want to run your government tomorrow? Perhaps these cities are used to operating in a close environment. But why not have an open process development to develop a distro to run a city? The citizens themselves could make sure it's secure before handing over the "keys" to the mayor (so to speak).
It seems like the problem might be not using open source processes to develop and deploy the solution(s). It seems there is a need to buy the distro from a retailer or from the manufacturer, and have something ready off-the-shelf or made for them.
But such customization would be necessary on MS software -- and would *have* to be proprietary, since not everyone has the source to all the pieces used and such customization could likely only be done on site or by a closed group. However, with open source SW, such customization doesn't have to be done by traditional closed-source methods. If fact, it's better done via open-source methods and the cities to form, create and become part of the open source community to make sure they are on top of fixes as they come out. Otherwise, you can have the same problems of "unupdated" SW as MS has. One of the positive aspects of the open-source system is that SW fixes, even fixes in customized software, can be done rapidly and deployed just as rapidly.
With MS, security and customization of OS components can only come from closed source and closed development sources: often contracts with the one closed source vendor, MS.
Perhaps the cost and problem with conversion to an open source base for running a city government is that they are running the project as they would a closed source project. I wonder if anyone in the Munich government even knows what Slashdot is, let alone reads it or if someone there in their IS department is up to date on the LKML action or any of the distro support groups?
Anyway...given the lack of details in the original article almost implies that they are treating open-source exactly as they would closed-source and have no process for doing an open-source project.
Oh come on, everyone knows how easy it is to convert teaspoons to gallons and drums....:-/
Besides...it fuels the calculator industry...its amazing the number of specialized calculators now sold to building contractors...and 12 gauge wire is how many inches in thickness?....:-/
Software is built in a craft/guild model? What alternate dimension are you living in? I've only been in the SW industry for 20+ years or so, never seen this process at any company I've been at and I've been at startup to 10+K employee places. Maybe some.
Software engineering will become a true engineering distance when "cp" fails. Engineering is about applying known plans, layouts, formula, blue prints to reproduce widgets. If software has been built before, then there isn't a reason to reproduce it other than by "cp".
Software crafting would be the equivalent in building, of constantly adding on features... sorta like the Winchester Mystery House where the widow who got her fortune from then gun with the same name felt guilt over those killed by her family's invention and felt the only way to stay ahead of them was to keep building onto her mansion year after year after year...different builders, no plan -- stairways that end in doors that open out to open space 10 feet up. I.e -- a building-bug. It was never fixed. Patch never applied. The alternative to adhoc continuing to grow and grow a program is to restart with "Hello World" and go a new direction designing a building with completely new functions covered under any previous "building code"....
In engineering, change happens slowly -- such that standards and building codes and laws are written to govern minimum standards, such that you can have a craft/guild model or schools that teach standards. Like automotive engineering. Same basic engine for 100 or so years -- fossil fuel, internal combustion, 4 tires, steering wheel. Seats that are build for price or style but not spinal health of people, engine usually in front, drive usually in back until recent years.
In software, its not usually worth doing unless it hasn't been before. Competition, in some ways, is actually bad, as it has often had customer lock-in as a cost. Multiple software documents standards, in some ways are like multiple Video tape formats. Except that "engineering" practices with software allow creating entire new copies of a widget ith a simple "cp".
You can't make buildings so easily, cars, TV's etc....nothing in the physical engineering world compares to the act of making 1000 units of linux or MS Office. MS's materials cost and time to make 1000 or 10,000 units is the real act of engineering.
But nothing that goes onto that first disk is engineering. It's creating a completely new "invention" that works uniquely from any other. It's like on the first rev, you release a radio, then a TV, then a VCR, then a DVD, etc... except that after the first successful radio, completely new decices get put out at the rate of 3-10 a year.
Now -- how do you develop safety standards for making radios, TV's, etc...when completely new appliances are being created by the 100's every year?
The answer is you can't -- unless you slow down the rate and build testability, testing and redundancy into each and every program. What would be the cost of having multiple groups develop code segments with duplicate functionality that are run as checks against other groups' answers? Even if you cut the rate of sofware development in half (doubling costs) you'd still have change coming faster out of the computer software industry considerably faster than any previous technology.
Until accountability is built into the software down to the project level, project practices like extending development time and cutting test time to 1/5th the development time). Testing is also usually one of the least glamorized jobs -- but maybe things would change -- for each bug in an engineer's code that was found, the engineer is doc'd $100 of pay. If his salary falls below miniumum wage, he's shown the door, and for each bug found by an engineer, $100 is added to their salary. Might bring new interest in people's desire to become a test engineer.:-)
Nope...at my last company, it was against company cul
Why not try this analogy on for size: "You are as capable of rewriting terms of an employment agreement (as an employee) as you are changing an inwall light switch to a dimmer switch". The alternative is to call an electrician to do it for you.
Me, I can change a switch. Some people don't know how. Some people don't know how to change a tire, some do. Some people know how to change an employment agreement, some don't. At my first job out of school, I listed a game I'd worked on before I came to Intel as exempt to their agreement. Two years later when the game had become popular among fellow employees, I was approached by marketing droids about integrating voice command tech into the game that they could use as a demo of their board. Did the work, worked fine. When it came down to actually using it the marketing droids were just going to pay me a small lump sum, which was fine w/me at the time, to use it. It was vetoed by management. M. tried to claim they owned my game and I would be paid nothing. I told them to check the employment agreement. M. pulled the plug on the deal. They didn't want to create precedence for paying me extra money for something that was my own work so the project and demo was killed. Nice of them. Under current California law it is illegal for an employer to have such an agreement in their employment contracts.
The deciding point, now, is whether or not the employee created the invention outside of work, on own time and equipment, is unrelated to employer's line of business and not created as a derivative work of work done by the employee for the employer (CA employment codes 2870-2872).
Section 2872 also says employee has to be made aware that any employment agreement they signed that claims rights to works other than those allowed by section 2870 is limited. Providing the employer has informed the employee, in writing, and has proof the the employee's notification (like their signature the revised legal requirements on file), the the burden of proof on proving ownership is on the employee. If the employer has only an "illegal" agreement signed and on file, then the burden of proof appears to be on the employer.
Please note. I'm not a lawyer, I'm just reading the legal code.
The full code is online on the california government webpages. If you want the hypertext linked version of the code, you have to pay about $50 more, per article of code, from LexisNexis (also includes a hard copy version).
If the police didn't catch her in the act, do they have witnesses that will testify that it was her in the bar and that she was naked? Did they check her ID? If not, whose to say who it really was and a bit of digital magic and you could have almost anyone nude in that bar.... Considering it's digital "art" and such is airbrushed regularly, I'd think it might be hard to _really_ prove she was nude in the bar.
If they did can they prove it the bar was open at the time? If it was closed, wouldn't it be a private party/establishment? I'm sure being nude in a private space isn't illegal in Lincoln, or it must issue alot of tickets for nudity for taking showers at home....yeah, right.
Dell has put it's bet on the +r/+rw standard for DVD's for over a year now. If you read on their HW boards or talk to their tech support/sales geeks (if you get one in America, anyway), have held off on any rewriteable or writeable DVD's until the + models were available. They thought the + standard was superior and didn't want to be stuck, like HP, with issuing free upgrades to everyone who bought a -DVD.
You read their user forums and they'd be bitching about no writeable DVD support and saying they didn't care if + wasn't ready, they'd rather have "-" and worry about compat later. Dell held firm and said "+" only. Now they are coming through on it and even throwing them in for free, sounds like, to make up, perhaps, for the long wait.
Dell is in the driver seat now. They no longer have to go out on limbs on sketchy technology -- and they don't. They can also afford to alienate the home market by dividing the business and home markets with feature sets that make top end inspirons unsuitable for business and make business models more expensive and less technologically attractive for bleeding edge tech lovers.
They don't have alot of credible competition in the direct-to-consumer market so they are getting more in a position of being able to dictate terms and be downright consumer hostile.
Had a horrible experience with them in purchasing a new laptop where they mucked up the order -- they refused to allow a replacement to be sent until the old one was sent back and my account was fully credited which didn't make life easy (since I'd already moved all my stuff over and reconfig'ed to the new system). Then, by the time I went back, they took all compatible laptops off the market and replaced their high end laptops with wide screen units that sported a digital interface for a then, (and now only high priced), external monitor. Since I'd just purchased a 1600x1200 LCD only a year earlier I didn't feel like dumping my investment even if there were monitors that were 1900x1080 (or whatever the res is). The first laptop that comes with a standard 4:3, (1600:1200) interface has no DVI out nor a docking station -- the only that was offered was a USB port expander with 10Mbit ethernet (oh so state of the art!...as I ponder upgrading my server-"workstation" (laptop) link to 1Gbit because 100Mb is too slow).
I asked their sales people who made their system design decisions, and I was told customers -- I said, I'm a customer -- how do I make my input known? No one knew. Sales, Tech, India, Texas, Alabama...even wrote a note to pres Michael...no response.
Seems they don't need customer input to make their design decisions...it was great toying with 1 of their sales drones when they answered the phone and asked me "Hi, what can we build for you today?"....
That sales motto doesn't apply anymore, as the sales person quickly realized.
It's the one thing they don't "get". We don't want a feature, or a frill, or any particular thrill -- just the source, the full source, unrestricted for all to use or see.
So in any future era, like the present one now at hand, when MS says goodbye to Windows [X] (current, 98), anyone can make the decision to transition based not on fear of no more security fixes or fear of being unable to keep up with new hardware -- but based on need or want for the new things in whatever new Win OS delights!
That's the enduring process of open-source. The maker goes out of business and you aren't left stranded -- you have all the source code for your product. How many pieces of junk CD's have I tossed because the no longer ran on newer OS's and the companies no longer support them (assuming the company is still around). You get to a point when you want to say "STOP".
Even linux, I watch...was true with 2.4 and 2.6....the benchmarks showed how nicely linux scaled under load with multiple processors...but the bottom end response time seemed to increase in both (just looking at graphs). Yeah, 2.4 handles itself better under high load and might not lock up like 2.2, but under low load that my system is at 99.9% of the time...?
Does anyone still run linux on an i386? or even i586? aren't most at the Pentium-Pro/PII/PIII or better level? Does linux still run on a 386 or 486?
Anyway...seems like so much "growth" is by forced change...is that really growth -- the music industry thought the upspurt in CD sales in the 90's was. When it was just the public playing "catchup". Now they have to go and invent a new format (duh...it's not about piracy...)
One of the XFS authors said anyone who wants to undertake such a port -- "go for it".
Considering the difficulty in ensuring data integrity and support for B-tree arranged data, Microsoft would not look kindly upon XFS being ported to NT, since their next generation OS is supposed to include database like features to speed up indexing and accessing data like XFS already has built-in. It would really rain on their parade. Also, benchmarking shows NTFS is considerably slower than XFS (or FAT32 for that matter) for large files and NTFS has no support for Real-time I/O partitions or journals being located on separate disks.
NTFS also requires (according to ad-copy) constant defragmentation due to their primitive block allocation scheme while XFS does quite well even without the XFS FSR (File System Reorganizer). XFS's FSR was created for 1 specific customer who had a particular application that generated excessively fragmented disks. Before that, an FSR (/defragmenter) wasn't considered necessary because XFS is intelligent about how it lays out files when they are written and how it stores free space (with free space also stored in ordered B-tree's by powers-of-two size of the free space blocks.
The only benchmark I've seen XFS run noticeably slower on linux, on is deleting large numbers of small files -- something one doesn't notice on IRIX, since the space deallocation happens in background on IRIX, and only the inodes need be marked deleted before the user prompt returned. I seem to remember on Linux the space had to be deallocated synchronously for some reason or another.
Makes sense given the way free space is managed -- when files are deleted, free blocks are recursively combined with adjacent free blocks to create the largest possible 'free block size' (I think up to 128k blocks, default=4k block size) (my numbers may be a bit rusty). Free space blocks were combined asynchronously, under IRIX (as I understand it), in a system thread after the last reference to an inode was released. Linux, if I remember correctly, didn't support the facilities for such a background thread -- thus the block combining happens synchronously, explaining the performance hit for file tests that delete lots of small files: there are many small free blocks that are candidates for being merged with adjacent free space.
I'm not entirely sure why a special "XFS_del" process couldn't be started at system run time who's sole purpose was taking unreferenced inodes and doing the space combining in background, allowing foreground programs to continue asynchronously after simply marking the inode as unusable and enqueing it to the XFS "free space" combining process. It is quite possible some of this has been implemented and my information is dated. But free space combining on cleanup is one of the main reasons why, historically, XFS file systems, didn't need to have _continually_ running programs like Executive Software's, _DiskKeeper_, running, full time in background: because XFS had it's own built-in defragmentation every time a user did a file-delete.
For the degenerate case -- *one* customer was not getting sufficient speed for real-time, uncompressed video recording to disk (back in the early to mid 1990's when disks were much slower). The swat team, assigned to the problem, found that the customer's particular use kept many small files around while deleting some files in a way that prevented automatic space consolidation. This odd usage was just enough to slow down direct-to-disk video recording (something quite difficult on systems in the early to mid 90's when disks were not so fast and SCSI-2 was still state of the art). To solve this problem for *one* customer, the "xfs_fsr" util was written.
To make the most of the efforts spent on the one customer, SGI incorporated xfs_fsr into the general OS to be run occasionally to stave multi-month/year buildup of possible, similar degenerate cases. I.e. XFS customers considered fragmentation such an unlikely / non-issue, that the X
Slashdot Mirror
Alternatively, you could use a Radio-Shack Rechargable battery pack (couldn't find URL, but catalog no. _was_ 23-047). It's about the size of 4 cassette tapes, ~12.6oz (~350g), output voltage selectable from 3-9v output and rechargable by plugging it into a wall output or from a 12v400ma source. If you are a doit-yourselfer, buy a 4 "cell" holder and wire it in series. Radio Shack sells battery adapter extension cord and heads singlely. You could choose capacity and weight by cell size (though note, I've often seen "D" rechargable cells with same ratings as the "C" indicating they've just stuck a "C" cell in a larger container.
Externally powered, this _should_ slightly increase laptop runtime (i.e. active external cooling => less internal fan use).
It holds the laptop on rubber feet about 7mm above 2 fans sucking air from center of underside and venting out the back.
-l
I just went with the defaults. You are right. Perhaps, MS doesn't know the best defaults for their fs's...wouldn't be surprising...
-l
noatime and no 8.3 file name generation is off in reg for NTFS so it already has a benefit there. Originally the disk was a replacement disk (7200RPM laptop drive vs 5400). I was looking at perf figures before I installed it as a system disk with the FAT32 and NTFS fs's in the sam loc during after reformatting the same partition.
.5X by the end of the disk.
.4 as fast as FAT32, USB2.0 was topping out at ~.6 of the speed on FAT32, but about .5 the speed when using NTFS and that's where NTFS was slowest.
I'm aware of the outer sectors reading slowly -- I see it painfully obviously where on a clean disk a "dd" (I have linux on the sys too, dual boot) will start at rate X and almost be down to
I was running tests over USB2.0, Firewire and as an "extra" internal HD. Firewire and internal IDE HD w/NTFS were about
Now, I get some of the bene's I need for username security by using a small NTFS partition (basically to install their unix extensions on to play w/), while getting perf bene's for files that don't need separate username management.
Most of my files don't need username management because it is a _laptop_. You still need a username/pw to log into the computer. The fs isn't exported (2-way 3rd party fw, default block).
For a _personal_ computer, I haven't found the overhead of NTFS (and problems easily reading/writing it from Linux). Like the original poster, I'd like rely on something other than windows code to have r/w access to my files. Theoretically, doesn't matter if you can use the virtualized Windows driver for r/w access -- you still need to have bought an authorized copy of Windows to run the driver on your machine. No?
-l
Lots of comments on Apple vs. XP, but with Apple using a Unix core, I was slightly more interested on how it used run time tracing in background and tries to record caching signatures for apps and "fastlinks" for common libraries. Also boot and init.d tracing and moving all the startup programs into the same area of the disk (maybe near the beginning of a partition where read times are usually faster). Wouldn't have to pre-load a cache but having startup programs intelligently laid out would seem to help boot time. Using page-in behavior of programs traced in background to determine pre-fetch
strategies might help program startup times. I dunno if Gnu libs/gcc use common lib[c,etc] addr's to speed runtime linkage. Anyone know??
-l
The CPU wastes billions of CPU cycles -- doesn't mean "uses tons of CPU cycles". If I click on a program that isn't in cache, the CPU wastes tons of cycles unless you have it constantly running background programs. This usually isn't the case on a system designed for interactive user-use.
Anyone with a partial clue has been told to turn off as many as possible, or all background applications to speed up performance. So if there are no background applications, just how is the CPU *NOT* wasting time waiting on a READ? Writes are virtually free because of DMA while your program continues on, but READ's -- ya just gotta stop and wait.
As for the answer "just" being a faster HARD DISK -- will admit a 15K HD will feel nice for almost anyone, but the biggest delay these days -- internet latency. Bandwidth is continuing to be expanded, I just recently got mine upped by about 3x down/.75up at the cost of doubling the 10ms latency I had to the nearest router.
It isn't just internet delays -- but worse the latency in a connection to open and return info from a READ. Was just reading a slashdotted article on ARS.Technica...a standard template of the page was returned in under a second, but to display the article content took an additional 30-45 seconds.
The more widgets, the more time to fetch them. Certainly you can re-use a an HTTP connection sometimes, but then you're only getting 'serial' fetch performance.
It seems as more apps work with the network, the single app with worst latency is going to be web browswers.
While there is new HTML standards to allow for prefetch of data after the first read, it's seems it's rarely used. On some occasions, I've gotten noticably less frustrated spidering a site while fetching a cup of coffee, and then reviewing the site in my webcache that waiting 30-45 seconds between each page I want to read. Ug!
-l
Well NTFS also has it's "sucking" points as in "sucking performance". More than one disk read/write test (Sandra, and one used by Magix) has shown my NTFS partition on the same physical hard disk to be 1/4th (that's 25%) of the speed of my FAT32 partition.
Both partitions were formatted using WinXP defaults, FAT32 at 16k segment size, and NTFS at a 1K segment size. NTFS has 86% free space, FAT32 has 38% free space. Both are defragmented and the results were duplicated on multiple hard disk drives, two at 5400RPM and one at 7200RPM. The best I could get out of a NTFS drive was "only" a 66% performance hit (that's 1/3rd the speed of
FAT32), sized at same size and place on same disk, successive tests).
NTFS may be more reliable, but on Win2000 and WinXP, I submit that unplanned shutdowns are less frequent and only rarely do I have to run a chkdsk on bootup because a FAT32 partition wasn't unmounted cleanly.
It's not like in linux where the difference between ext2 and journaled filesystems 50-60% depending on the ops and the journaled FS's.
I was distinctly unimpressed. Of course your Mileage may vary.
-l
Even if you have a gig of memory and never use more memory for programs
than 300K, and even if you set the memory manager to give priority to running
programs over file blocks, you will still see NT use swap.
If you don't give it swap, you won't be able to get various stats related
to paging (as when it pages a program into memory) since the module that
shows those stats also tries to show swap stats and basically does a divide
by zero somewhere and crashes.
My linux server, on the other hand will use 'zero' swap (even if it is
allocated, provided I never try to use more memory than I have
physical). This was in the 2.4 era. Things may have changed in 2.6 to
make it more like Win XP. Dunno.
-l
Sorry, I've called Fry's to do a stockcheck. It doesn't take that long --
not even for them to do a physical stock check. It sounds more like the guy
was just not wanting to wait on the phone for a non-commision sale.
United States, 56.74%
Canada, 6.80%
China (including Hong Kong), 6.24%
But this "thespamweblog" article from November 2003 shows:
The same source report (along with sources of digital attacks was mentioned in theInquirer. I'm sure if I kept tracking links in google, I'd find other reports and other percentages.
Maybe recent US lawsuits have, within a few months, forced most spammers off shore or perhaps it's just part of the offshoring of US [SPAMMER] jobs.
;-/
In regards to my sig, a bit of dark thoughts: imagine if China became a strict Moslem nation. If stealing incurs the lost of a hand, I could see the penalty for sending spam to be the loss of one finger for each separate mailing. Of course the worker would be fined and their boss. Eventually that might make a dent in spam. >:=}>
-l
An update system that offers to overwrite /etc/passwd (and presumably every other security file) hardly seems like a safe or easy upgrade process.
:-( & :-)
I can't say my DoC (SuSE) has it better -- they don't ever seem able to upgrade my system in a sane or coherent manner. Last time around, it upgraded my squid 3.0 to squid2, tried, unsuccessfully to put my named in a basement mail (when it hadn't even been bad), but it was thrown in the basement w/o the root servers file and when the root servers all expired some large amount of time (~3-4 months) later, various TLD's started disappearing. It was bizzare watching large sections of internet just "go away" a few days before it completely consumed itself. Then I found the problem -- it hadn't copied in the root servers file from the previous upgrade (and/or didn't install a new copy). I tried grabbing some updates with their Yast Online solution, but it kept downloading copies of 8.2 binaries when I have 9.0 loaded. I never had 8.2 loaded -- I went straight from 8.1 to 9.0. Later, I found, buried in some paragraph of fine print somewhere that their updates only support updating from the immediately preceding version -- this was after it had removed all unknown packages fro the package database. At this point I had all the 8.1 packages installed, but no longer noted as "installed" in the database over which it automatically upgraded and installed about 10-15 packages out of the 100-150 it should have installed (I guess ~10-15 packages kept some same valid name). I'm always rather afraid to do an upgrade under SuSE as I know it will usually involve lots of pain.
On the flip side -- a fresh install of 9.0 for a never-used-linux user went real smooth -- they were able to navigate their way around after only one or two hiccups -- like buttons weren't where they used to be under Win, but I just told them they'd have to experiment a bit and find out how things were arranged differently. Once they experiemented some, they started finding what they needed surprisingly well.
-l
Old idea, new context and as others have mentioned, the tech isn't even equal to AM quality in many cases let alone FM stero.
:-(!
But...like yeah, users can turn to the FM band, record a few hours direct to disk, then go back and edit out the talk/advertisements.
You'll get better quality recordings (if you have good FM receiver, and can pull in a strong quality FM station). If you live close to a college many run low-wattage stations geared just to the "right" age group...
Legal? Remember Ted Turner comparing digital video recorders that could skip advertising as video theft! Why do you think video corps are trying to make everything go digital and embed watermarks and require copy control tech in anything that can copy -- and can detect the watermarks even when played over an analogue medium.
Remember the DMCA? It'll be illegal to circumvent any of the copy protection they put in your recording devices. Maybe not a bad idea to keep around old cassette recorders, VCR's and such.
Sound producers have been mumbling/rumbling about new CD formats with better sound (~megabit range)...even now you hear audiophiles bemoan bad audio CD players that simply play the current frequency for the entire 22.73 microseconds between sample as opposed to better decoders that interpolate the difference between each sample and gradually change the frequencies between intervals constantly over the 22.73 microsecond interval. More than one
audiophile claims to be able to hear the difference.
Nevertheless, if they push 1Mbit DVD-Audio out as a new standard, it's real easy to also push out a new DRM scheme. Then a scratch on a CD will lose you minutes of music instead of seconds! Such a deal!
Anway....recording off streams will be even easier to control copying of since you'll be going from digital to digital, so less chance of loss of embedded watermarks. So record away while you can...
You are forgetting a minor point.
:-)
I have 5 year old HW in the other room. It has 1G of main memory and 2x1G CPU's. The disks range from 10K SCSI's to 7.2K IDE's w/8m write cache.
But if I am writing anything under 900M from my client to the server, that can all be buffered in memory -- and will be in the timeframes we are talking about. 2.6 will be very likely to buffer the entire free-memory area before dumping to disk and when it does, it will probably have everything arranged in memory to do maximal contiguous writes with no extraneous head movements.
The server has no graphics console and only runs server type SW.
TODAY -- if I was outfitting a similar server it might have 2-4G main memory -- I *might* even have RAID working (not that I haven't tried with the older server, but RAID combos I've tried have run slower than single disk options, so I obviously don't have raid configuration down to an exact science yet (not high enough priority).
Fact is, with the clients I run, they can spit ripped CD's to a network drive faster than to a local drive (bad dell laptop design putting internal CD and internal drive both on same IDE bus). Now let's talk about a 8x DVD ripper running over a 400Mb firewire. 400Mb is 4x faster than 100Mb ethernet --
so I'm likely to be able to rip DVD images 4x faster than my ethernet can take. I may start running into server limitations, but it's easy to see a 400Mb firewire overwhelming a 100Mb ethernet real easily. USB2.0 isn't a problem as it is only around 100-110Mb throughput in my measurements (multiple devices/ multiple HD's). It runs about 10x USB 1.1 due to sloppy USB 2 implementations (mentioned on Tom's).
So 100Mb ethernet is already inadequate if you are coming close to maxing 400Mb firewire -- and when 800Mb firewire arrives...100M ethernet will be dust and your disks better be able to handle near 80Mb/s write to disk speed (or you better have a very large system memory on your target system).
We're on the "edge" now. 1G is quickly going to become a necessity. There have already been articles about how 8x DVD's need to be internal vs. USB 2.0 (don't know about FW). Ack!
I can see smoke coming out of my ethernet cables with the bits being pushed too fast....:-)...
-l
Trackball -- yes.
Try this on for size. Buy two (yes expensive) Turbo Pro's -- Wireless variety. Set them both for the same channel. It has 4 main buttons and a wheel that does dual duty as a middle mouse button. I set lower right button for select and drag, and lower left for double click, upper left and right as normal. It has 6 programmable buttons for anything you want.
The secret is the wireless option. Since both mice are on the same channel, you can switch hands to whichever is more convenient -- seamlessly. As long as you don't try to use both at the same time, they don't send signals. Battery life is pretty good -- though I use a steady stream of rechargables.
I use 1 receiver (it's USB, but they include a USB-PS/2 adapter). Because my carpal tunnel symptoms are worse in my right hand, I've become predominatly a left hand mouzer, but the ability to mouse with either hand w/o the hassle of moving the mouse -- but which ever is more convenient is a major bonus. Sometimes I need to primarly use keyboard keys on one hand or the other in combination with a mouse press. Depending on which hand has to be on the keyboard, I can automatically mouse with the other hand.
Becoming bi-mousual, is a major help for overuse on one-hand mousing problems.
Also...kensington support has been *GREAT*. Lifetime warantee on the mice. If one malfunctions or I just wear it out -- I call and give them the serial #, and they send out a new one (and I return the bad unit or not depending on random audits). I've had them for about 3 years and maybe gotten 2-3 replacements because I've worn down the wheel's such that the ball starts rubbing against the walls of the inside of the mouse and no longer moves. No problem...they send out a new one. I've never had a company (besides Dell with next business day on-site service) be so good with support. Compare that type of support to HP "Depot" service where you are without your part for 1.5 months while they ship the unit to India to be repaired. I don't think so!
Seriously. It's your lively hood. Lawyers don't think twice about spending 1000's of dollars a year to keep up on the law, and electricians don't think twice about shelling out for a high quality Fluke over a cheaper Radio Shack model.
It's no wonder so many programmers get so damaged, when they work with bear skins and stone knives to carve out their programs. It's your hands, your eyes, your back, your spine (ergo seating, proper posture, monitors with print that doesn't cause you to squint or have to bend forward to read). Think about it. Whether your employer pays for it or not -- if they won't, I'd still claim it as a tax deduction as a necessary tool of your trade (IANATC).
-l
Oh -- I forgot -- corporations were declared to be people.
Um...but what about "equal representation"? Ah...well...the makers set government up to be run by the richest individuals -- with the rich individuals able to command the most influence.
It was a system that sorta worked....with the dream of lower class becoming upperclass, and with 'equal' votes --- but without the votes, directly, really counting -- we elect some "special" group that I'm, a bit fuzzy on -- an "electorate", that makes the real decisions as to who gets elected with individual states getting to choose how individual votes are counted (or discarded) and which electorate entities finally get sent to cast the "representative" [sic] votes of the people.
Classic example: through illegal vote negation, Florida was able to swing the last election -- not Florida, but 1 person, in the family of the to-be elected president. It's the same as the original system -- the rich are the real ones deciding the election and making the laws to keep themselves in power.
The main hiccup that's wreaking the most havoc is this bit about corporations having the status of being "people". There can never be the equality of "all men being created equal" when you have corporations that can outlive any single "man" and can become a super "meta-person" with abilities and resources no single "man" could ever have.
Until that "waving of hands" is undone and "corporations" are stripped of the right of "personhood" with property only being owned by real people who have limited lifespans, we, the people will never have equality and will continue to find ourselves oppressed, more and more, by large corporate "people".
-l
Must be that piracy leads to increased pirated (stolen/shoplifted CD's)...
:-/
Certainly people couldn't be using downloaded CD's to see if they
want to *buy* CD's....we know those downloaders are criminals...we just
have re readjust the facts to fit our story....it must be CD' thefts that
have increased!
-l
Having seen perfectly good stories rejected, and seen other, questionably newsworthy stories posted, with comments at the end like "go off and discuss among yourselves", I've felt, sometimes, like the gatekeepers view slashdottee's like their children with articles posted as gifts to us. I've felt (at times) they hand out the articles like like "toys" to children where they are then told to go play with their toys for a while. I think it may be the case that sometimes they are rushed or overbusy and just through out "whatever" -- where they may not realize that either they are being played by a submitter they normally trust or they might be unconsciously creating their own spin on an article.
I often feel that the slashdot.org system has one central flaw that distorts the community feel of the board: users are not allowed to moderate initial postings. While comments can be moderated by the user community, allowing good opinions to be highlighted and things that are deemed "not good" to be rated down, I feel that for slashdot to be a community resource, the community should be allowed to moderate the initial submissions as well as responses.
I "wish" that could be implemented and the initial submitters wouldn't take it personally any more than we are told to take frequent initial submission rejects personally, but I don't see the "powers that be" being very motivated to add such a feedback mechanism as there isn't really a competitive board to drive such a change.
They are having "trouble" installing linux in Munich because it is more expensive? What is more expensive? What is the "trouble"?
Oh!...The last time the installed, they had to install via floppies and they had to upgrade the systems with CD or DVD readers? Is that the source of the "expense"...?
The news article is not very "enlightening"...probably even sourced by MS as press fodder. I mean, what is the "trouble"? You can copy one image to the network, then use autoinstall to setup 100's of PC's. Now that is all included in my one copy of $100 SuSE. I cannot even buy one copy of Windows for use in a network for that, let alone be provided with autoinstall SW.
The article feels more like FUD than substance...does anyone know what part of the upgrade cost / workstation is costing more?
Besides, why don't these cities just setup the conversion as an open-source project on source-forge -- they could likely get all the software conversion done for near free and hire someone locally to maintain it? What, exactly, does a city need in a distro to "run itself?" If multi-cities are converting over, why not just develop an open-source city-distro. Seems like a specialized class of SuSE's...it or Mandrake has the best multi-lang support. Red Hat seems aimed at commercial, but why not have cities turn to its citizens, or citizens of the net to develop a "by-the-people" software distro?
What software do you want to run your government tomorrow? Perhaps these cities are used to operating in a close environment. But why not have an open process development to develop a distro to run a city? The citizens themselves could make sure it's secure before handing over the "keys" to the mayor (so to speak).
It seems like the problem might be not using open source processes to develop and deploy the solution(s). It seems there is a need to buy the distro from a retailer or from the manufacturer, and have something ready off-the-shelf or made for them.
But such customization would be necessary on MS software -- and would *have* to be proprietary, since not everyone has the source to all the pieces used and such customization could likely only be done on site or by a closed group. However, with open source SW, such customization doesn't have to be done by traditional closed-source methods. If fact, it's better done via open-source methods and the cities to form, create and become part of the open source community to make sure they are on top of fixes as they come out. Otherwise, you can have the same problems of "unupdated" SW as MS has. One of the positive aspects of the open-source system is that SW fixes, even fixes in customized software, can be done rapidly and deployed just as rapidly.
With MS, security and customization of OS components can only come from closed source and closed development sources: often contracts with the one closed source vendor, MS.
Perhaps the cost and problem with conversion to an open source base for running a city government is that they are running the project as they would a closed source project. I wonder if anyone in the Munich government even knows what Slashdot is, let alone reads it or if someone there in their IS department is up to date on the LKML action or any of the distro support groups?
Anyway...given the lack of details in the original article almost implies that they are treating open-source exactly as they would closed-source and have no process for doing an open-source project.
-l
Oh come on, everyone knows how easy it is to convert teaspoons to gallons and drums....:-/
Besides...it fuels the calculator industry...its amazing the number of
specialized calculators now sold to building contractors...and 12 gauge wire is
how many inches in thickness?....:-/
Software is built in a craft/guild model? What alternate dimension are you living in? I've only been in the SW industry for 20+ years or so, never seen this process at any company I've been at and I've been at startup to 10+K employee places. Maybe some.
... sorta like the Winchester Mystery House where the widow who got her fortune from then gun with the same name felt guilt over those killed by her family's invention and felt the only way to stay ahead of them was to keep building onto her mansion year after year after year...different builders, no plan -- stairways that end in doors that open out to open space 10 feet up. I.e -- a building-bug. It was never fixed. Patch never applied. The alternative to adhoc continuing to grow and grow a program is to restart with "Hello World" and go a new direction designing a building with completely new functions covered under any previous "building code"....
:-)
Software engineering will become a true engineering distance when "cp" fails. Engineering is about applying known plans, layouts, formula, blue prints to reproduce widgets. If software has been built before, then there isn't a reason to reproduce it other than by "cp".
Software crafting would be the equivalent in building, of constantly adding on features
In engineering, change happens slowly -- such that standards and building codes and laws are written to govern minimum standards, such that you can have a craft/guild model or schools that teach standards. Like automotive engineering. Same basic engine for 100 or so years -- fossil fuel, internal combustion, 4 tires, steering wheel. Seats that are build for price or style but not spinal health of people, engine usually in front, drive usually in back until recent years.
In software, its not usually worth doing unless it hasn't been before. Competition, in some ways, is actually bad, as it has often had customer lock-in as a cost. Multiple software documents standards, in some ways are like multiple Video tape formats. Except that "engineering" practices with software allow creating entire new copies of a widget ith a simple "cp".
You can't make buildings so easily, cars, TV's etc....nothing in the physical engineering world compares to the act of making 1000 units of linux or MS Office. MS's materials cost and time to make 1000 or 10,000 units is the real act of engineering.
But nothing that goes onto that first disk is engineering. It's creating a completely new "invention" that works uniquely from any other. It's like on the first rev, you release a radio, then a TV, then a VCR, then a DVD, etc...
except that after the first successful radio, completely new decices get put out at the rate of 3-10 a year.
Now -- how do you develop safety standards for making radios, TV's, etc...when completely new appliances are being created by the 100's every year?
The answer is you can't -- unless you slow down the rate and build testability, testing and redundancy into each and every program. What would be the cost of having multiple groups develop code segments with duplicate functionality that are run as checks against other groups' answers? Even if you cut the rate of sofware development in half (doubling costs) you'd still have change coming faster out of the computer software industry considerably faster than any previous technology.
Until accountability is built into the software down to the project level, project practices like extending development time and cutting test time to 1/5th the development time). Testing is also usually one of the least glamorized jobs -- but maybe things would change -- for each bug in an engineer's code that was found, the engineer is doc'd $100 of pay. If his salary falls below miniumum wage, he's shown the door, and for each bug found by an engineer, $100 is added to their salary. Might bring new interest in people's desire to become a test engineer.
Nope...at my last company, it was against company cul
Why not try this analogy on for size: "You are as capable of rewriting terms of an employment agreement (as an employee) as you are changing an inwall light switch to a dimmer switch". The alternative is to call an electrician to do it for you.
Me, I can change a switch. Some people don't know how. Some people don't know how to change a tire, some do. Some people know how to change an employment agreement, some don't. At my first job out of school, I listed a game I'd worked on before I came to Intel as exempt to their agreement. Two years later when the game had become popular among fellow employees, I was approached by marketing droids about integrating voice command tech into the game that they could use as a demo of their board. Did the work, worked fine. When it came down to actually using it the marketing droids were just going to pay me a small lump sum, which was fine w/me at the time, to use it. It was vetoed by management. M. tried to claim they owned my game and I would be paid nothing. I told them to check the employment agreement. M. pulled the plug on the deal. They didn't want to create precedence for paying me extra money for something that was my own work so the project and demo was killed. Nice of them. Under current California law it is illegal for an employer to have such an agreement in their employment contracts.
The deciding point, now, is whether or not the employee created the invention outside of work, on own time and equipment, is unrelated to employer's line of business and not created as a derivative work of work done by the employee for the employer (CA employment codes 2870-2872).
Section 2872 also says employee has to be made aware that any employment agreement they signed that claims rights to works other than those allowed by section 2870 is limited. Providing the employer has informed the employee, in writing, and has proof the the employee's notification (like their signature the revised legal requirements on file), the the burden of proof on proving ownership is on the employee. If the employer has only an "illegal" agreement signed and on file, then the burden of proof appears to be on the employer.
Please note. I'm not a lawyer, I'm just reading the legal code.
The full code is online on the california government webpages. If you want the hypertext linked version of the code, you have to pay about $50 more, per article of code, from LexisNexis (also includes a hard copy version).
-l
If the Athlon 15 stage beat the p4-20, is it possible for Intel two do two competing pipes that could do both branchs of a True/False if?
In single decisions, wouldn't that be likely to improve performance dramatically?
Dunno if it is possible, but just a thought....
-l
If the police didn't catch her in the act, do they have witnesses that will testify that it was her in the bar and that she was naked? Did they check her ID? If not, whose to say who it really was and a bit of digital magic and
you could have almost anyone nude in that bar.... Considering it's digital "art" and such is airbrushed regularly, I'd think it might be hard to _really_ prove she was nude in the bar.
If they did can they prove it the bar was open at the time? If it was closed, wouldn't it be a private party/establishment? I'm sure being nude in a private space isn't illegal in Lincoln, or it must issue alot of tickets for nudity for taking showers at home....yeah, right.
Just some random things I'd think of off hand....
Dell has put it's bet on the +r/+rw standard for DVD's for over a year now. If you read on their HW boards or talk to their tech support/sales geeks (if you get one in America, anyway), have held off on any rewriteable or writeable DVD's until the + models were available. They thought the + standard was superior and didn't want to be stuck, like HP, with issuing free upgrades to everyone who bought a -DVD.
You read their user forums and they'd be bitching about no writeable DVD support and saying they didn't care if + wasn't ready, they'd rather have "-" and worry about compat later. Dell held firm and said "+" only. Now they are coming through on it and even throwing them in for free, sounds like, to make up, perhaps, for the long wait.
Dell is in the driver seat now. They no longer have to go out on limbs on sketchy technology -- and they don't. They can also afford to alienate the home market by dividing the business and home markets with feature sets that make top end inspirons unsuitable for business and make business models more expensive and less technologically attractive for bleeding edge tech lovers.
They don't have alot of credible competition in the direct-to-consumer market so they are getting more in a position of being able to dictate terms and be downright consumer hostile.
Had a horrible experience with them in purchasing a new laptop where they mucked up the order -- they refused to allow a replacement to be sent until the old one was sent back and my account was fully credited which didn't make life easy (since I'd already moved all my stuff over and reconfig'ed to the new system). Then, by the time I went back, they took all compatible laptops off the market and replaced their high end laptops with wide screen units that sported a digital interface for a then, (and now only high priced), external monitor. Since I'd just purchased a 1600x1200 LCD only a year earlier I didn't feel like dumping my investment even if there were monitors that were 1900x1080 (or whatever the res is). The first laptop that comes with a
standard 4:3, (1600:1200) interface has no DVI out nor a docking station -- the only that was offered was a USB port expander with 10Mbit ethernet (oh so state of the art!...as I ponder upgrading my server-"workstation" (laptop) link to 1Gbit because 100Mb is too slow).
I asked their sales people who made their system design decisions, and I was told customers -- I said, I'm a customer -- how do I make my input known? No one knew. Sales, Tech, India, Texas, Alabama...even wrote a note to pres Michael...no response.
Seems they don't need customer input to make their design decisions...it was great toying with 1 of their sales drones when they answered the phone and asked me "Hi, what can we build for you today?"....
That sales motto doesn't apply anymore, as the sales person quickly realized.
It's the one thing they don't "get". We don't want a feature, or a frill, or any particular thrill -- just the source, the full source, unrestricted for all to use or see.
So in any future era, like the present one now at hand, when MS says goodbye to Windows [X] (current, 98), anyone can make the decision to transition based not on fear of no more security fixes or fear of being unable to keep up with new hardware -- but based on need or want for the new things in whatever
new Win OS delights!
That's the enduring process of open-source. The maker goes out of business and you aren't left stranded -- you have all the source code for your product. How many pieces of junk CD's have I tossed because the no longer ran on newer OS's and the companies no longer support them (assuming the company is still around). You get to a point when you want to say "STOP".
Even linux, I watch...was true with 2.4 and 2.6....the benchmarks showed
how nicely linux scaled under load with multiple processors...but the bottom end response time seemed to increase in both (just looking at
graphs). Yeah, 2.4 handles itself better under high load and might not lock up like 2.2, but under low load that my system is at 99.9% of the time...?
Does anyone still run linux on an i386? or even i586? aren't most at the Pentium-Pro/PII/PIII or better level? Does linux still run on a 386 or 486?
Anyway...seems like so much "growth" is by forced change...is that really growth -- the music industry thought the upspurt in CD sales in the 90's was. When it was just the public playing "catchup". Now they have to go and invent a new format (duh...it's not about piracy...)
-l
One of the XFS authors said anyone who wants to undertake such a port -- "go for it".
Considering the difficulty in ensuring data integrity and support for B-tree arranged data, Microsoft would not look kindly upon XFS being ported to NT, since their next generation OS is supposed to include database like features to speed up indexing and accessing data like XFS already has built-in. It would really rain on their parade. Also, benchmarking shows NTFS is considerably slower than XFS (or FAT32 for that matter) for large files and NTFS has no support for Real-time I/O partitions or journals being located on separate disks.
NTFS also requires (according to ad-copy) constant defragmentation due to their primitive block allocation scheme while XFS does quite well even without the XFS FSR (File System Reorganizer). XFS's FSR was created for 1 specific customer who had a particular application that generated excessively fragmented disks. Before that, an FSR (/defragmenter) wasn't considered necessary because XFS is intelligent about how it lays out files when they are written and how it stores free space (with free space also stored in ordered B-tree's by powers-of-two size of the free space blocks.
The only benchmark I've seen XFS run noticeably slower on linux, on is deleting large numbers of small files -- something one doesn't notice on IRIX, since the space deallocation happens in background on IRIX, and only the inodes need be marked deleted before the user prompt returned. I seem to remember on Linux the space had to be deallocated synchronously for some reason or another.
Makes sense given the way free space is managed -- when files are deleted, free blocks are recursively combined with adjacent free blocks to create the largest possible 'free block size' (I think up to 128k blocks, default=4k block size) (my numbers may be a bit rusty). Free space blocks were combined asynchronously, under IRIX (as I understand it), in a system thread after the last reference to an inode was released. Linux, if I remember correctly, didn't support the facilities for such a background thread -- thus the block combining happens synchronously, explaining the performance hit for file tests that delete lots of small files: there are many small free blocks that are candidates for being merged with adjacent free space.
I'm not entirely sure why a special "XFS_del" process couldn't be started at system run time who's sole purpose was taking unreferenced inodes and doing the space combining in background, allowing foreground programs to continue asynchronously after simply marking the inode as unusable and enqueing it to the XFS "free space" combining process. It is quite possible some of this has been implemented and my information is dated. But free space combining on cleanup is one of the main reasons why, historically, XFS file systems, didn't need to have _continually_ running programs like Executive Software's, _DiskKeeper_, running, full time in background: because XFS had it's own built-in defragmentation every time a user did a file-delete.
For the degenerate case -- *one* customer was not getting sufficient speed for real-time, uncompressed video recording to disk (back in the early to mid 1990's when disks were much slower). The swat team, assigned to the problem, found that the customer's particular use kept many small files around while deleting some files in a way that prevented automatic space consolidation. This odd usage was just enough to slow down direct-to-disk video recording (something quite difficult on systems in the early to mid 90's when disks were not so fast and SCSI-2 was still state of the art). To solve this problem for *one* customer, the "xfs_fsr" util was written.
To make the most of the efforts spent on the one customer, SGI incorporated xfs_fsr into the general OS to be run occasionally to stave multi-month/year buildup of possible, similar degenerate cases. I.e. XFS customers considered fragmentation such an unlikely / non-issue, that the X