I always thought alpha, beta and so on only referred to perceived code/feature-set maturity of whoever is doing the version numbering. For me, 'alpha' = "works well enough and long enough to potentially be useful" while 'beta' = "works as expected most of the time".
Many freeware apps have public alphas and other such so alpha/beta/etc. do not imply any privacy in most cases. When I really mean for a build to be private, I add the 'private' tag to the version string to remind people I send stuff to not to redistribute.
In case you havent noticed - the ST franchise is self-sustaining.;)
Not for much longer if Paramount starts a tradition of starting a new ST show and abort it half-way through. Questionable commitment is bad for both fans and network/advertiser support.
If Paramount starts another ST show only to cancel it again because it did not beat UNreality shows, they should quit the ST business.
My friend works AT MS... so what the world now knowns as beta2 which was released four days ago is most likely a few iterations behind what is currently deployed internally. That would explain why he could not think of any major bugs when I asked about how usable VS2005 was... because the major beta1 bugs had already been fixed by the time he started using it, ~6 months ago.
I have a friend who happens to be working at M$. He told me they started to use VS2005 internally late last summer, that would be beta-1 I think. Beta-2 has been around for a few months, apparently a must-have from what I was told.
Some betas are worth other's releases and vice-versa.
If I successfully forged a 128KB data piece that has the same hash value as the original block, the 128KB block hash becomes useless in determining which block was bad: the whole-file hash would fail but all block hashes would check out ok.
As for the BT case, the INFO hash is a hash of hashes and would not even be in any way useful in detecting a file containing forged data.
There are many ways to work around this, the simplest one being to use two hash sets, another would be altering initial hash conditions.
Fonts are somewhat like a document's core library.
GCC and its core libraries are also GPL'd but GCC's terms explicitly limit the GPL's scope to exclude derived works (third-party code compiled and linked executables) from GPL coverage. Without this clause, writing closed-source commercial software for Linux would be practically illegal.
If using GPL'd fonts effectively GPL'd documents, GPL'd fonts would get ditched pretty quick since they would represent unnecessary liabilities.
With grey areas like these surrounding the GNU acronym, it is no wonder most companies approach anything related to it with extra caution. I generally keep my own projects GPL-free because I personally think the GPL is excessively restrictive... if I had to pick a license, I would go with a slightly stricter form of BSD's.
For optimum results, this should be two DIMMs per channel, two channels per CPU, four CPUs, 16 DIMMs total.
In this optimum (generally) case, a CPU must be fitted in the socket to which the DRAM banks are connected - the Opteron has built-in DRAM controllers, no CPU = no controller = dead slots.
Also, probably only one or two of the CPUs are connected to the chipset so these slot(s) have to be used first to provide an IO bridge for the others - signals do not mysteriously jump across sockets when the CPUs use point-to-point busses, HyperTransport in AMD's Opteron/A64 case.
If advertisers want to have more obtrusive ads, fine with me. Today's TV is mostly mediocre shows with even worse adds so I have almost completely given up on it. For online stuff, I do use *-block extensions since flashy flash ads are often unbearably irritating... I used to resize and scroll my browser's window to hide them before getting flashblock. I would have preferred not installing Flash altogether but some sites I go to are flash-only.
After withstanding years of online ads abuse, I barely (or often not at all) notice banners and any other form of adds both online and in real life, the more ads are within my field of vision, the less I notice them.
Like someone else in some other thread here said, I too make mental notes to specifically NOT buy anything from companies from which I see annoying (flashy/irritating/stupid/etc.) ads.
I will do the same with HD as I did with DVD... stay as far away from it as I can until someone restores fair use... as in wait until something like DeAACS becomes available.
I fundamentally boycotted all things backed by the MPAA/RIAA ever since they started making asses of themselves about file sharing, this is just one more reason for me to keep doing so.
Yes, my sweet-spot desired retail prices are roughly equal to current OEM pricing - that would probably take care of most of the grey-market OEMs and make many people (except MS bean counters) much happier.
As for the actual cost of bulk OEM licenses, last time I read about these, Home was around $45. I do not remember seeing quotes for bulk OEM Pro licensing though.
Without market share for an OS/hardware platform, getting programmer mindshare for it is much harder... why bother tuning software for a platform that has only 5% market share, focusing on the 90% platform is much more likely to pay off.
More programmers brings more software to a platform. More software gets the platform more visibility and potential install base. A larger platform deployment increases the platform's visibility on developers' radars, increasing the likelyhood of more software becoming available, etc.
Regular users may not submit patches but they still count as install base. Install base is many platforms' most valuable marketable asset. Who would code for a gaming platform with 5% market share? What commercial software vendor would bother with Linux's 10% market share? Other than high-end, extremely expensive software, very few.
Things should become really insteresting a few years after Linux reaches 10% home desktop market share.
Time to dig out all those forgotten Win98SEs... they might not have the more stable NT core but at least they suffer none of the nonsense restrictions and it also has more functionnal file sharing than XP Home.
Of course, with Win9x, a firewall (at the very least) is pretty much mandatory.
I almost go berserk when I have to deal with XP Home because stuff I use all the time is either "misplaced" or disabled... if I had to deal with XP Starter, the temptation to simply throw the whole PC out the window could be dangerously strong.
I hope competition will eventually force MS to drop XP Pro pricing to a reasonable level... like $100 retail-boxed - but I will not be holding my breath. In the meantime, I love free, campus-wide-licensed MSDNAA stuff.
Anyway, the way Microsoft is selling such outrageously crippled Windows XPs is... outrageous. If it were not for programs requiring Win2k or higher being increasingly more common, I would still prefer Win98SE over XP Home/Starter.
Yes, Starter is not worth using. An XP Starter CD belongs pretty much to the same value category as AOL CDs. An OS that cannot be used to do anything useful is not worth the CD it is distributed on or the bandwidth used to download it.
At least we can get some form of consolation from the fact that XP Starter asian launches so far have been practically absolute failures. Let's hope this bulk rejection trend will continue and that MS will eventually make the right choice: kill Starter, slash Home and Pro prices... to something like $60 for Home and $120 for Pro.
They already mention that they hope fans will be looking forwards to the next chapter of trek.
Well, unless some sort of miracle happens before this next chapter starts airing, I predict a new tradition of cancelled treks.
If they start another series and quit half-way though again, even the most loyal of trek fans will start turning their backs on Paramount and trek.
My not following through with Enterprise, Paramount has slapped the franchise's fans in the face, which is highly unlikely to help the next series' marketability.
The fan base is Trek's main marketing asset. Killing series will quickly decimate the fan base. Finish Enterprise or quit the Trek business altogether - I feel Trek is unlikely to survive another cancellation. (I mean after Enterprise)
Only when you have an AMD64 chip though - in most benchmarks I have seen so far, AMD64 usually at least as fast and often indecently faster than EMT64, I guess Intel deserves that for somewhat half-heartedly implementing it as a "me too" feature to avoid losing too much market share.
I used to be an Intel-only kind of person... but when Intel started doing last-minute half-baked moves to preserve their marked share, I pretty much lost all respect I had for them. (The fact that fully-featured A64 laptops cost almost the same as mid-range Celeron-based models also helped me break my 'tradidion'... and getting a castrated P4 would not bring me any new capabilities anyway.)
There is more to x86-64 than addressing...
on
Windows XP X64 Goes Gold
·
· Score: 2, Interesting
I find it somewhat irritating that Intel is promoting only the addressing part of x86-64's benefits.
Extended addressing might sound nice but in the real world, it translates to no performance improvement unless you have >4GB in your PC while gains from recompiling to use the extra registers (and some rewriting to combine high/low parts into int64s, reducing initial register usage) are often in the 20%-40% range - though this can vary wildly depending on GCC options and across GCC versions.
Well, it is all marketing so Intel's EMT64 campaign does not need to make any technical sense as long as it sells.
The only way to do it in a civilized manner would be to get a "people's party" to power with overwhelming majority and start anew.
Since elections are a publicity stunt and getting listed already costs quite a bit, getting >300 people listed, known and elected would cost a fortune. In a world of mostly narrow-minded idiots, it is highly unlikely that a party of "average" people ("average" because truly average people mostly do not care and would not get involved) would get to power.
Linux is even more extremely cheap yet nearly every Linux distro can potentially be used to setup computers for newbies, professionals, industrial servers, data-centers and supercomputers/clusters.
I wish Linux was a more immediate threat to Microsoft's practical monopoly... if it was, Microsoft could not afford abusing the market as much as it appears to intend.
To me, there are three basic XP versions... - Standard (Pro) - Crippled (Home) - More crippled (Starter)
I'd rather use Linux than either of the Windows Crippled Edition... and I'd also rather use Linux than pay full-price for XP Standard Edition. (Currently, I have access to MSDNAA and a friend who works at MS so I can get legit Pros dirt cheap.)
Every single time I had to use XP-Home I ended up bald and red from frustration as I was quickly reminded of how much artificially crippled XP-Home is compared to "Pro".
A few years ago, Micro$oft was saying they wanted to move everyone to the NT kernel so they would have only one architecture to support... but now that they have started forking/crippling/customizing it every which way, support nightmare must not be too far down the road.
To me, XP-Home is psychological cruelty and I dare not imagine what would happen if I got stuck having to use XP-Starter... (I'd probably use it to download a random Linux distro, burn that and junk Starter ASAP - artificially crippled products is one of the top things that really piss me off.)
My first mouse had three buttons and I had an almighty 8088 PC to use it with - I still have that PC's MoBo in a box somewhere.
I wonder what's the big deal with Apple "innovating" a two button mouse when nearly every other manufacturer's have at least three.
Yup, Apple's two-button innovation will be a flop beyond Apple's own OEM market since all people currently owning Apple kit who care about 2+ buttons almost certainly already bought a normal non-Apple mouse.
How much of that hypothetical 1TB/month does your site use on average? How does that compare to the data center's average?
The trend would probably look similar to ISPs: less than 10% of customers account for more than 90% of all bandwidth usage so the data center (or ISP) can afford to pass down its bandwidth at lower-than-cost rates.
Similar thing for me... but $7 CDN per extra GB rather than USD.
Since bulk bandwidth costs under $2/GB, anything over $5 seems quite unreasonable. Thankfully, I rarely need to exceed my service plans' base 20GB/month and when I do, it rarely is by more than 100-200MBs.
Slashdot Mirror
Oh?
I always thought alpha, beta and so on only referred to perceived code/feature-set maturity of whoever is doing the version numbering. For me, 'alpha' = "works well enough and long enough to potentially be useful" while 'beta' = "works as expected most of the time".
Many freeware apps have public alphas and other such so alpha/beta/etc. do not imply any privacy in most cases. When I really mean for a build to be private, I add the 'private' tag to the version string to remind people I send stuff to not to redistribute.
In case you havent noticed - the ST franchise is self-sustaining. ;)
Not for much longer if Paramount starts a tradition of starting a new ST show and abort it half-way through. Questionable commitment is bad for both fans and network/advertiser support.
If Paramount starts another ST show only to cancel it again because it did not beat UNreality shows, they should quit the ST business.
My friend works AT MS... so what the world now knowns as beta2 which was released four days ago is most likely a few iterations behind what is currently deployed internally. That would explain why he could not think of any major bugs when I asked about how usable VS2005 was... because the major beta1 bugs had already been fixed by the time he started using it, ~6 months ago.
I have a friend who happens to be working at M$. He told me they started to use VS2005 internally late last summer, that would be beta-1 I think. Beta-2 has been around for a few months, apparently a must-have from what I was told.
Some betas are worth other's releases and vice-versa.
If I successfully forged a 128KB data piece that has the same hash value as the original block, the 128KB block hash becomes useless in determining which block was bad: the whole-file hash would fail but all block hashes would check out ok.
As for the BT case, the INFO hash is a hash of hashes and would not even be in any way useful in detecting a file containing forged data.
There are many ways to work around this, the simplest one being to use two hash sets, another would be altering initial hash conditions.
Fonts are somewhat like a document's core library.
GCC and its core libraries are also GPL'd but GCC's terms explicitly limit the GPL's scope to exclude derived works (third-party code compiled and linked executables) from GPL coverage. Without this clause, writing closed-source commercial software for Linux would be practically illegal.
If using GPL'd fonts effectively GPL'd documents, GPL'd fonts would get ditched pretty quick since they would represent unnecessary liabilities.
With grey areas like these surrounding the GNU acronym, it is no wonder most companies approach anything related to it with extra caution. I generally keep my own projects GPL-free because I personally think the GPL is excessively restrictive... if I had to pick a license, I would go with a slightly stricter form of BSD's.
For optimum results, this should be two DIMMs per channel, two channels per CPU, four CPUs, 16 DIMMs total.
In this optimum (generally) case, a CPU must be fitted in the socket to which the DRAM banks are connected - the Opteron has built-in DRAM controllers, no CPU = no controller = dead slots.
Also, probably only one or two of the CPUs are connected to the chipset so these slot(s) have to be used first to provide an IO bridge for the others - signals do not mysteriously jump across sockets when the CPUs use point-to-point busses, HyperTransport in AMD's Opteron/A64 case.
I thought we were ~28M not so long ago and that 30M was an optimistic figure... then again, I have never been any good with 'details'.
Canada?
I've been there... actually, I'm still there and so are ~30M other people!
If advertisers want to have more obtrusive ads, fine with me. Today's TV is mostly mediocre shows with even worse adds so I have almost completely given up on it. For online stuff, I do use *-block extensions since flashy flash ads are often unbearably irritating... I used to resize and scroll my browser's window to hide them before getting flashblock. I would have preferred not installing Flash altogether but some sites I go to are flash-only.
After withstanding years of online ads abuse, I barely (or often not at all) notice banners and any other form of adds both online and in real life, the more ads are within my field of vision, the less I notice them.
Like someone else in some other thread here said, I too make mental notes to specifically NOT buy anything from companies from which I see annoying (flashy/irritating/stupid/etc.) ads.
Yup.
I will do the same with HD as I did with DVD... stay as far away from it as I can until someone restores fair use... as in wait until something like DeAACS becomes available.
I fundamentally boycotted all things backed by the MPAA/RIAA ever since they started making asses of themselves about file sharing, this is just one more reason for me to keep doing so.
Yes, my sweet-spot desired retail prices are roughly equal to current OEM pricing - that would probably take care of most of the grey-market OEMs and make many people (except MS bean counters) much happier.
As for the actual cost of bulk OEM licenses, last time I read about these, Home was around $45. I do not remember seeing quotes for bulk OEM Pro licensing though.
This is a chicken&egg situation.
Without market share for an OS/hardware platform, getting programmer mindshare for it is much harder... why bother tuning software for a platform that has only 5% market share, focusing on the 90% platform is much more likely to pay off.
More programmers brings more software to a platform. More software gets the platform more visibility and potential install base. A larger platform deployment increases the platform's visibility on developers' radars, increasing the likelyhood of more software becoming available, etc.
Regular users may not submit patches but they still count as install base. Install base is many platforms' most valuable marketable asset. Who would code for a gaming platform with 5% market share? What commercial software vendor would bother with Linux's 10% market share? Other than high-end, extremely expensive software, very few.
Things should become really insteresting a few years after Linux reaches 10% home desktop market share.
Time to dig out all those forgotten Win98SEs... they might not have the more stable NT core but at least they suffer none of the nonsense restrictions and it also has more functionnal file sharing than XP Home.
Of course, with Win9x, a firewall (at the very least) is pretty much mandatory.
I almost go berserk when I have to deal with XP Home because stuff I use all the time is either "misplaced" or disabled... if I had to deal with XP Starter, the temptation to simply throw the whole PC out the window could be dangerously strong.
I hope competition will eventually force MS to drop XP Pro pricing to a reasonable level... like $100 retail-boxed - but I will not be holding my breath. In the meantime, I love free, campus-wide-licensed MSDNAA stuff.
Anyway, the way Microsoft is selling such outrageously crippled Windows XPs is... outrageous. If it were not for programs requiring Win2k or higher being increasingly more common, I would still prefer Win98SE over XP Home/Starter.
Yes, Starter is not worth using. An XP Starter CD belongs pretty much to the same value category as AOL CDs. An OS that cannot be used to do anything useful is not worth the CD it is distributed on or the bandwidth used to download it.
At least we can get some form of consolation from the fact that XP Starter asian launches so far have been practically absolute failures. Let's hope this bulk rejection trend will continue and that MS will eventually make the right choice: kill Starter, slash Home and Pro prices... to something like $60 for Home and $120 for Pro.
They already mention that they hope fans will be looking forwards to the next chapter of trek.
Well, unless some sort of miracle happens before this next chapter starts airing, I predict a new tradition of cancelled treks.
If they start another series and quit half-way though again, even the most loyal of trek fans will start turning their backs on Paramount and trek.
My not following through with Enterprise, Paramount has slapped the franchise's fans in the face, which is highly unlikely to help the next series' marketability.
The fan base is Trek's main marketing asset. Killing series will quickly decimate the fan base. Finish Enterprise or quit the Trek business altogether - I feel Trek is unlikely to survive another cancellation. (I mean after Enterprise)
Only when you have an AMD64 chip though - in most benchmarks I have seen so far, AMD64 usually at least as fast and often indecently faster than EMT64, I guess Intel deserves that for somewhat half-heartedly implementing it as a "me too" feature to avoid losing too much market share.
I used to be an Intel-only kind of person... but when Intel started doing last-minute half-baked moves to preserve their marked share, I pretty much lost all respect I had for them. (The fact that fully-featured A64 laptops cost almost the same as mid-range Celeron-based models also helped me break my 'tradidion'... and getting a castrated P4 would not bring me any new capabilities anyway.)
I find it somewhat irritating that Intel is promoting only the addressing part of x86-64's benefits.
Extended addressing might sound nice but in the real world, it translates to no performance improvement unless you have >4GB in your PC while gains from recompiling to use the extra registers (and some rewriting to combine high/low parts into int64s, reducing initial register usage) are often in the 20%-40% range - though this can vary wildly depending on GCC options and across GCC versions.
Well, it is all marketing so Intel's EMT64 campaign does not need to make any technical sense as long as it sells.
It is not going to change any time soon.
The only way to do it in a civilized manner would be to get a "people's party" to power with overwhelming majority and start anew.
Since elections are a publicity stunt and getting listed already costs quite a bit, getting >300 people listed, known and elected would cost a fortune. In a world of mostly narrow-minded idiots, it is highly unlikely that a party of "average" people ("average" because truly average people mostly do not care and would not get involved) would get to power.
IDKJ about regex but this does appear to be somewhat unfair to people who host their eMail on a .info domain.
Linux is even more extremely cheap yet nearly every Linux distro can potentially be used to setup computers for newbies, professionals, industrial servers, data-centers and supercomputers/clusters.
I wish Linux was a more immediate threat to Microsoft's practical monopoly... if it was, Microsoft could not afford abusing the market as much as it appears to intend.
To me, there are three basic XP versions...
- Standard (Pro)
- Crippled (Home)
- More crippled (Starter)
I'd rather use Linux than either of the Windows Crippled Edition... and I'd also rather use Linux than pay full-price for XP Standard Edition. (Currently, I have access to MSDNAA and a friend who works at MS so I can get legit Pros dirt cheap.)
Every single time I had to use XP-Home I ended up bald and red from frustration as I was quickly reminded of how much artificially crippled XP-Home is compared to "Pro".
A few years ago, Micro$oft was saying they wanted to move everyone to the NT kernel so they would have only one architecture to support... but now that they have started forking/crippling/customizing it every which way, support nightmare must not be too far down the road.
To me, XP-Home is psychological cruelty and I dare not imagine what would happen if I got stuck having to use XP-Starter... (I'd probably use it to download a random Linux distro, burn that and junk Starter ASAP - artificially crippled products is one of the top things that really piss me off.)
I'm just too goddamn tired of seeing companies sticking the "innovation" label on minor and obvious steps in evolution's staircase.
That and I haven't slept for >30h.
My first mouse had three buttons and I had an almighty 8088 PC to use it with - I still have that PC's MoBo in a box somewhere.
I wonder what's the big deal with Apple "innovating" a two button mouse when nearly every other manufacturer's have at least three.
Yup, Apple's two-button innovation will be a flop beyond Apple's own OEM market since all people currently owning Apple kit who care about 2+ buttons almost certainly already bought a normal non-Apple mouse.
It is the same averaging game as everybody does.
How much of that hypothetical 1TB/month does your site use on average? How does that compare to the data center's average?
The trend would probably look similar to ISPs: less than 10% of customers account for more than 90% of all bandwidth usage so the data center (or ISP) can afford to pass down its bandwidth at lower-than-cost rates.
Similar thing for me... but $7 CDN per extra GB rather than USD.
Since bulk bandwidth costs under $2/GB, anything over $5 seems quite unreasonable. Thankfully, I rarely need to exceed my service plans' base 20GB/month and when I do, it rarely is by more than 100-200MBs.