Well, generally yes. But remember that they are light-as-a-feather and can move easily. Mine are about in the middle of my desk but their tripod stand deal is offset so I can push my keyboard back to give me more than enough room to put a large text or reference book up there along with a notebook if you want to take notes. This is without moving the LCDs. If I need gobs of space I can easily scoot them back.
I, too, used to think flat-panel LCD monitors were useless in the home desktop mixed-application (office apps, coding, gaming, etc.). I do a lot of outside contracting so my home computer occasionally doubles as an after-hours workstation. I long ago discovered the joys of multiple monitor programming and so for several years had two 19" Sony Trinitron-tube monitors running - it still allowed for high-color gaming and I was able to split my code windows up over the two displays (it is nice to have code on one screen and the output and references on the other).
As luck would have it, I also own a wonderful Dell Inspiron 8000 laptop with the 15" 1600x1200 display. For the longest time I was running Windows 2000 on it, but when I heard about ClearType (sub-pixel font rendering) on Steve Gibson's site (GRC.com) I thought I'd give it a try. I was pretty impressed and decided to try the full-blown item with Windows XP Professional a try.
It was, in a word, mind-blowing!
Never before had text looked so sharp and clear - and after using it for several multi-hour coding sessions (aside from only the single display) it didn't give me headaches like my two CRTs did (even though I'd tried virtually every refresh rate setting from 60 to 120Hz).
That made the difference for me, so I slowly replaced my two 19" CRTs with two 17" Iiyama black LCDs (retail around $640 US for the digital versions, $610 for analog). Remember that CRTs cheat and don't really measure the actual diagonal, so that the 19" CRT is much more like 17.something (mine were 17.8"), whereas the LCDs are ACTUAL diagonal. So those who are saying, "Not until they have affordable 19" displays!" are missing the boat - for all intents and purposes a 17" LCD *is* equal to a 19" CRT in terms of usable display size. The LCD goes from bezel edge to bezel edge with ZERO loss in quality in the corners, something my CRTs could not claim. So in practice, I would guestimate the actual screen real estate are about equal.
As I didn't have the cash right away to drop on two, I ran for quite a while on one LCD (digital through my GeForce4 Ti4400) and one CRT. The differences were astounding - the CRT was noticeably not as bright as the LCD, and text was HORRIBLE to read. It was like trying to focus on a blurry photograph - your eyes keep trying to find the right focus and could never really adjust. It game me a TERRIBLE headache after only a few minutes, so I sold the CRT and waited for my next paycheck and got the second one. While both displays are digital (along with secondary analog inputs) my Ti4400 only has a single DVI output, so my secondary monitor is in *analog* mode until I can fix that (maybe a PCI card with digital?). I won't lie and say they look identical, but it is not nearly as dramatic as the CRT comparison.
Resolution-wise 1280x1024 is not a big deal to me. I enjoy my laptop's 1600x1200 and would hope we start seeing 17" affordable desktop LCDs in that range soon, but it isn't killing me. XP's ClearType provides for amazingly sharp text as it is, so the slightly larger fonts look pretty anyway.:) In short, you won't notice lack of clarity due to resolution size difference.
Gaming, or LCD's *supposed* Achilles' heel. I play a couple games - Half-Life Counter-Strike, UT, Medal of Honor: Allied Assault, Stronghold Crusader - and have noticed ZERO blurring problems. None. Nada. Zip. Zilch. Zero. Get it? It's a NON-ISSUE. Perhaps my system (Athlon 1.333Ghz, 512MB PC2100, ATA100 HDD, 128MB GeForce4 Ti4400, etc.) is not fast enough to get the frame rate above the pixel response rate, but it is not a problem with my Iiyama digital LCD displays. Certainly it can be an issue with older or slower displays, but anything in the sub 30ms rate should be fine. The calculation (from Tom's Hardware) to turn response into FPS is: 1 / (rate * 0.001). So 30 ms = 1 / 0.030 = 33.3 images displayed on the screen per second. 25ms = 40fps. Your eye notices things at 25fps or higher to be continuous 'full-motion' and at 30 it seems to be the 'magic number'. Remember to read the fine print on each monitor since some companies will list 'average' display rate or some other random numbers. Find the worst-case percentage and see if you can live with it. The only problem I've noticed is that when the games exit and the video mode switches the LCD panel never 'wakes up'. I am sure everyone knows what I am talking about - the screen flickers, the green indicator turns amber for a while, and then it dumps you back to your desktop. Well, the LCD (or my Beta 4 Detonator drivers) must give up early on the video card and goes to a 'You have nothing hooked up' screen. I have to turn it on and off to re-cycle the display.
I have noticed that certain extremely light web-page backgrounds (the old background pattern on http://www.angsttechnology.com for example) the colors appear to be washed out (you could only see a little bit of the pattern) but it is generally not an issue for anything you are *actively* looking at (games, photos, etc.). Contrast seems to be better than my CRT as well, and the images appear to be more saturated on the LCD, somewhat startlingly so. When playing something like Command and Conquer: Red Alert 2 (or any other real-time strategy game) the colors are MANY TIMES more vivid - it is hard to explain but a really cool feeling the first time you fire up a game. First-person shooters look different as well; generally I have to turn the gamma up a notch or two since they all appear BLACK in the dark areas. I am not sure if that is how it *should* be and my CRT was inadequate or it is an LCD artifact, but I generally noticed LCD improvement over the CRT images.
I also notice that certain images online are not as smooth looking on the LCD. Again on InkTank (http://www.angsttechnology.com/AT/index.cfm, a really good geeky comic; if you have time check out the archives!) you can see 'jaggies' or anti-aliasing artifacts far more clearly on the LCD. It appears to happen on thin black antialiased lines more than any other. The black outline on PennyArcade looks (I think) how it should be, but the lines on Ink Tank were the #1 first thing I noticed after I got the displays. Now, of course, my eyes have gotten adjusted and so I no longer see them any more.
Anyways, to make an entirely too long post short, I have no misgivings about the LCD purchase. You often get what you pay for, and a cheap panel will probably yield a poor display picture. Mine is pretty much perfect. Give it a try, you might like it!:)
Right, but I think the assumption (the one I made at least) is that most hard-core gamers would be working with few large-files, right? I am thinking of the FPS which traditionally have HUGE (several hundred MB to a GB!).pak files with all the junk in it.
I would think the large-file benefit would outweigh the many small-file hit. I used to recall that Adaptec Ez-CD Creator (when it was called that) would do benchmarks on small and large file disk I/O. The actuall performance of the system was somewhere in-between. I would suspect it would be the same with your typical gaming system. Faster than a single drive but not as fast as if you had all large files.
I would never use RAID0 because as I stated a bad drive would hose the entire system. RAID1 is the way to go for low-budget RAID.
RAID 5 (10/01 et al) requires too many disks to be a consumer-grade RAID solution, really.
No, RAID0 is faster than a single drive. It can stripe the reads and writes to spread the I/O over each disk. The problem is that it provides no redundancy in case of a disk failure (you've lost everything at that point!).
Another advantage is that it makes your two single drives of size N each look like a single drive of size N * 2. This was a big advantage back in the days of the all-common 9GB SCSI. If you had a 40GB movie file you were working with it was inconvenient (or impossible) to partition it up. Merging 5 x 9GB drives into one logical device was great, but with single-drive capacities reaching several hundred GB RAID0's single logical drive is not quie as useful.
Provided it is done via hardware and not software, of course.:)
I would agree wholeheartedly and get up on the horse with you if not for two things:
MySQL AB have indicated quite clearly that MySQL is the world's most popular Open Source Database, designed for speed, power and precision in mission critical, heavy load use.
And in their docs, press releases, interviews etc. they do not hide the fact that they are taking aim at Oracle, DB2, MS SQL, Sybase ASE, et al.
Secondly, whenever anything on/. talks about RDBMS's the armchair DBAs come out of the woodwork and seem to think their tiny website all of a sudden is 'Enterprise' and that their usage of MySQL is revolutionary and shows the world that MySQL is the *one true RDBMS*.
Those who know, do. Those who don't sit on/. and try and pretend MySQL is a cut above the rest.
Agreed. The difference is that our Sybase ASE servers in over 2.5 years running have NEVER failed. The only downtime came when installing routine patches/EBFs which is easily handled with Sybase's insanely great Replication Server product (end users noticed NOTHING).
Replication in MySQL is neat but should not try and make up for RDBMS shortcomings (no hot backup in MySQL) or instability issues. It should be there when you machines BLOW UP.
Again, we're talking different classes of software here.
I really like Postgres. Had it not really, really not sucked 3 years ago when we moved form MySQL we'd be on it now. We chose Sybase ASE which is I think a far better RDBMS than MySQL (and Postgres as well) but if we had the choice today it would've been Postgres (GREAT value for the money!!;) ).
If you were a DBA you would know that Oracle has everything *including* the kitchen sink installed. You can't say the 9i 2GB install has nothing more, can you? I smell a troll.
Cool, I stand corrected then. IMHO it should be added to the docs TEXT datatype because that was the first place I looked to come up with my on/off row analysis: http://www.postgresql.org/idocs/index.p hp?datatype -character.html
Maybe just a short blurb stating it is TOASTed.:)
And the docs here: http://www.postgresql.org/idocs/index.php?l argeobj ects.html
Is what pointed to the alternate syntax for the large object support. So does the pg_class silently use the large_object syntax to stuff TEXT columns?
Also this: Tip: There are no performance differences between these three types, apart from the increased storage size when using the blank-padded type.
Led me to beleive that the TEXT was stored on-row, because there will be a hair bit more I/O to find the TEXT data, which does not seem to be indicated.
DROPing a column is under the covers simply a SELECT INTO followed by a drop of the old table and a rename of the new. It is just packaged up a bit nicer by ALTER TABLE DROP bob
Yes, we all know virtually every RDBMS can store gobs of data. It's the whole 'doing stuff with it' wherein lie the rub.
I'm not saying that MySQL can't 'do stuff' to large amounts of data, but one can go to vBulletin.com and read about customers who routinely have to 'prune' their database of posts/threads (think HardOCP). We're not talking 100 million posts but maybe a mil or two. Now, certainly there are parts of the vBulletin application that just don't scale to 10 million posts. But, it seems MySQL has a bit of a problem dealing with complex JOINs, DISTINCTs, COUNT's (not the vanilla SELECT COUNT( * ) FROM table;) which other RDBMSs do not. It is not terribly hard to see why. In your traditional web apps most queries are of the type "SELECT * FROM myTable WHERE myAutoIncrement = CONSTANT" and so the more 'goofy' queries haven't been optimized as much. This is not idle conjecture but something I have witnessed 'in the field' and have done some rudimentary benchmarks in support of.
So, it is not unthinkable that someone doing a lot of heavy lifting with some moderate amount of data (couple million rows) will hit some brick walls with MySQL. As web applications start to 'grow up' MySQL will address the growing demand as they have started to already (with FK support, row-level locking, replication). The hope is, of course, that MySQL will remain just as easy to use and fast after the new features!
ASE 12.5+ (maybe 12) has the ability to go up to 16K row sizes (with the new larger page sizes) although you would be hard pressed to find a practical instance in which you *need* a row that big -- provided your relations are properly normalized.
The main problem that *existed* in PostreSQL was that TEXT (and other unstructured datatypes) live on-row in PostgreSQL (and MySQL, but I'll get to that later;)). So you could easily hit the row cap with gobs of text (or a single image!). Now that the cap is gone you shouldn't see any problems. Aside from TEXT you should never hit the limit anyway unless you design horrible tables.:)
Note that TEXT (IMAGE, etc.) data lives off-row in Sybase and MS SQL (and Oracle allows you to specify both I think!). This is very nice since you can keep your rows smaller (better cache hit ratio) and it also allows for a much higher probability that your TEXT data will be in a contiguous block saving expensive disk seeks.
Re: MySQL the Gemini table type has (IIRC) an easily hit max row size if you use text columns. Of course, no one uses Gemini over InnoDB so it is a non-issue.
InnoDB rows live on a HUGE 16KB pages (recompile MySQL to change; yuck!) which means the max row size in older versions is a function of page size (PAGE_SIZE / 2 - STUFF). In later versions this was removed, although limits still exist: http://www.mysql.com/doc/en/InnoDB_restrictions.ht ml But in short, know your data and you should have no problems.
Sybase IQ also runs VLDBs into the 100's of GB range. A very, very capablo product, although Terradata in pretty much the market leader ar you point out.
Is Sybase SQL Anywhere. It offers scalable, bidirectional synchronization of information between your handheld and your Sybase Adaptive Server Enterprise, Oracle, Microsoft SQL Server and IBM DB2 RDBMS's. I would even think it might poke your MySQL/PostgreSQL DB via ODBC or some other method.
The RIAA and all the lawyers in the world will never be able to completely stop pirating. Look at how much money the feds throw at drugs and the number of addicts on the street. If enough people want something, they'll get it.
I know one of my chief frustrations is to search for a song and either have it incomplete, or be of poor quality (e.g. pops or other defects) or to simply have it not be the same song that I downloaded. If I could search for a song, pay $SOME_SMALL_AMOUNT (e.g. $1US) for it and download a 'known perfect' copy at my choice of bitrates (e.g. 128, 160, etc.) then sure as heck I'd do it.
Distributing these poisoned files would take an enormous amount of bandwidth, so they'd have to have some sort of agreement worked out with ISPs and a mass-content provider, say Akamai. Akamai has tens of thousands of servers located in hundreds (if not more) of ISPs throughout the nation. I think on peak usage they're pushing out 100 GB/sec. in the US (if not more). Simply say "Ok Akamai, can we buy 10GB on each of your servers and push all these MP3s out?". Then you write a gnutella client for each box which offers all the MP3s up for distribution.
I can't remember how the gnutella protocol works but I think it broadcasts search requests to the nodes that store a cache of what they have and what their neighbors offer and then can pass the request off. Have your client log all the requests (so you can tell the record companies which songs were requested more) and of course offer up your files when requested. If you do this with 10,000 boxes full of identical content chances are you're going to drown out any signal out there.
If you're really tricky, you can even have the client 'fake' files so you don't actually need to have the file on the box; you could send a pre-existing obfuscated file, or even dynamically build and stream the poisoned MP3.
Of course, all of this is moot if you still don't have a very easy, cheap method of offering MP3s online for the mass public. You could pitch it like this "Yeah, so you won't make much money off of offering $SOME_SMALL_AMOUNT for each MP3. But you're a fool if you think simply shutting Morpheus off will result in even 10% of the Morpheus users buying the actual CD or using a painful, userUNfriendly pay-per-MP3 system. However, what if we have a method to net you 20 or 30% of users who wouldn't pay you anyway?" So the pitch would be "We can't get you all of them, but our method would give you more than you're getting now!". Frankly the people who post on SlashDot (from the very negative response to the Subscription model) are not a good cross-section of the vast majority of internet users out there:).
So in your obfuscated file you have it play maybe 20 seconds of the file and then say "Sorry, this is a copywrited file. Pirating files costs artists money. If you want to buy this MP3 for $SOME_SMALL_AMOUNT, please visit http://www.somestore.com. 80% of $SOME_SMALL_AMOUNT earned will go directly to the artist."
It gives them a reason to buy it - not only do you have SomeStore.com very easily accepting payment, but you ACTUALLY PAY THE ARTISTS A MAJORITY OF THE MONIES EARNED! So it can quell the naysayers who say "Well the artist wouldn't receive anything anyway!" (rant: but who are you hurting more, the billion dollar-industry or the Artist who NEEDS even the small cut they receive from each CD sold?).
Some drawbacks could be of course that someone writes a 'detector' to find and ignore the invalid MP3s, or they block the IP addresses of the servers, etc. but that is easily fixed. Most non-power users (e.g. the great and huddled masses of the internet) don't want to update their Morpheus client every time a new version is released. Heck, even programs which offer hassle-free updating (e.g. antivirus, windowsupdate.com) very rarely are by the majority of internet users. Also, you'd work out the server IP settings with the ISP so that they would rotate to a random IP in their pool - since most of the servers are located in most ISPs you couldn't ban the single IP but perhaps a subnet. But since the IPs are in the ISP, you have now banned a large chunk of users. If they are in every ISP, you will have to ban every ISP (see the problem in banning IPs?).
So, to boil it down to a sentence: Have very easy-to-use, hassle-free, cheap, reliable, etc. method for users to buy MP3s and they WILL
Slashdot Mirror
Well, generally yes. But remember that they are light-as-a-feather and can move easily. Mine are about in the middle of my desk but their tripod stand deal is offset so I can push my keyboard back to give me more than enough room to put a large text or reference book up there along with a notebook if you want to take notes. This is without moving the LCDs. If I need gobs of space I can easily scoot them back.
I, too, used to think flat-panel LCD monitors were useless in the home desktop mixed-application (office apps, coding, gaming, etc.). I do a lot of outside contracting so my home computer occasionally doubles as an after-hours workstation. I long ago discovered the joys of multiple monitor programming and so for several years had two 19" Sony Trinitron-tube monitors running - it still allowed for high-color gaming and I was able to split my code windows up over the two displays (it is nice to have code on one screen and the output and references on the other).
:) In short, you won't notice lack of clarity due to resolution size difference.
:)
:)
As luck would have it, I also own a wonderful Dell Inspiron 8000 laptop with the 15" 1600x1200 display. For the longest time I was running Windows 2000 on it, but when I heard about ClearType (sub-pixel font rendering) on Steve Gibson's site (GRC.com) I thought I'd give it a try. I was pretty impressed and decided to try the full-blown item with Windows XP Professional a try.
It was, in a word, mind-blowing!
Never before had text looked so sharp and clear - and after using it for several multi-hour coding sessions (aside from only the single display) it didn't give me headaches like my two CRTs did (even though I'd tried virtually every refresh rate setting from 60 to 120Hz).
That made the difference for me, so I slowly replaced my two 19" CRTs with two 17" Iiyama black LCDs (retail around $640 US for the digital versions, $610 for analog). Remember that CRTs cheat and don't really measure the actual diagonal, so that the 19" CRT is much more like 17.something (mine were 17.8"), whereas the LCDs are ACTUAL diagonal. So those who are saying, "Not until they have affordable 19" displays!" are missing the boat - for all intents and purposes a 17" LCD *is* equal to a 19" CRT in terms of usable display size. The LCD goes from bezel edge to bezel edge with ZERO loss in quality in the corners, something my CRTs could not claim. So in practice, I would guestimate the actual screen real estate are about equal.
As I didn't have the cash right away to drop on two, I ran for quite a while on one LCD (digital through my GeForce4 Ti4400) and one CRT. The differences were astounding - the CRT was noticeably not as bright as the LCD, and text was HORRIBLE to read. It was like trying to focus on a blurry photograph - your eyes keep trying to find the right focus and could never really adjust. It game me a TERRIBLE headache after only a few minutes, so I sold the CRT and waited for my next paycheck and got the second one. While both displays are digital (along with secondary analog inputs) my Ti4400 only has a single DVI output, so my secondary monitor is in *analog* mode until I can fix that (maybe a PCI card with digital?). I won't lie and say they look identical, but it is not nearly as dramatic as the CRT comparison.
Resolution-wise 1280x1024 is not a big deal to me. I enjoy my laptop's 1600x1200 and would hope we start seeing 17" affordable desktop LCDs in that range soon, but it isn't killing me. XP's ClearType provides for amazingly sharp text as it is, so the slightly larger fonts look pretty anyway.
Gaming, or LCD's *supposed* Achilles' heel. I play a couple games - Half-Life Counter-Strike, UT, Medal of Honor: Allied Assault, Stronghold Crusader - and have noticed ZERO blurring problems. None. Nada. Zip. Zilch. Zero. Get it? It's a NON-ISSUE. Perhaps my system (Athlon 1.333Ghz, 512MB PC2100, ATA100 HDD, 128MB GeForce4 Ti4400, etc.) is not fast enough to get the frame rate above the pixel response rate, but it is not a problem with my Iiyama digital LCD displays. Certainly it can be an issue with older or slower displays, but anything in the sub 30ms rate should be fine. The calculation (from Tom's Hardware) to turn response into FPS is: 1 / (rate * 0.001). So 30 ms = 1 / 0.030 = 33.3 images displayed on the screen per second. 25ms = 40fps. Your eye notices things at 25fps or higher to be continuous 'full-motion' and at 30 it seems to be the 'magic number'. Remember to read the fine print on each monitor since some companies will list 'average' display rate or some other random numbers. Find the worst-case percentage and see if you can live with it. The only problem I've noticed is that when the games exit and the video mode switches the LCD panel never 'wakes up'. I am sure everyone knows what I am talking about - the screen flickers, the green indicator turns amber for a while, and then it dumps you back to your desktop. Well, the LCD (or my Beta 4 Detonator drivers) must give up early on the video card and goes to a 'You have nothing hooked up' screen. I have to turn it on and off to re-cycle the display.
I have noticed that certain extremely light web-page backgrounds (the old background pattern on http://www.angsttechnology.com for example) the colors appear to be washed out (you could only see a little bit of the pattern) but it is generally not an issue for anything you are *actively* looking at (games, photos, etc.). Contrast seems to be better than my CRT as well, and the images appear to be more saturated on the LCD, somewhat startlingly so. When playing something like Command and Conquer: Red Alert 2 (or any other real-time strategy game) the colors are MANY TIMES more vivid - it is hard to explain but a really cool feeling the first time you fire up a game. First-person shooters look different as well; generally I have to turn the gamma up a notch or two since they all appear BLACK in the dark areas. I am not sure if that is how it *should* be and my CRT was inadequate or it is an LCD artifact, but I generally noticed LCD improvement over the CRT images.
I also notice that certain images online are not as smooth looking on the LCD. Again on InkTank (http://www.angsttechnology.com/AT/index.cfm, a really good geeky comic; if you have time check out the archives!) you can see 'jaggies' or anti-aliasing artifacts far more clearly on the LCD. It appears to happen on thin black antialiased lines more than any other. The black outline on PennyArcade looks (I think) how it should be, but the lines on Ink Tank were the #1 first thing I noticed after I got the displays. Now, of course, my eyes have gotten adjusted and so I no longer see them any more.
Anyways, to make an entirely too long post short, I have no misgivings about the LCD purchase. You often get what you pay for, and a cheap panel will probably yield a poor display picture. Mine is pretty much perfect. Give it a try, you might like it!
www.LCD.com/switch
Thanks for modding redundant even though mine was posted 9 minutes before the last link. :(
Here is an example with all sorts of definitions you can read.
t ml
http://mathworld.wolfram.com/PoincareConjecture.h
Yeah, like living next to North Korea.
Sybase ASE is already on OS X.
Right, but I think the assumption (the one I made at least) is that most hard-core gamers would be working with few large-files, right? I am thinking of the FPS which traditionally have HUGE (several hundred MB to a GB!) .pak files with all the junk in it.
I would think the large-file benefit would outweigh the many small-file hit. I used to recall that Adaptec Ez-CD Creator (when it was called that) would do benchmarks on small and large file disk I/O. The actuall performance of the system was somewhere in-between. I would suspect it would be the same with your typical gaming system. Faster than a single drive but not as fast as if you had all large files.
I would never use RAID0 because as I stated a bad drive would hose the entire system. RAID1 is the way to go for low-budget RAID.
RAID 5 (10/01 et al) requires too many disks to be a consumer-grade RAID solution, really.
No, RAID0 is faster than a single drive. It can stripe the reads and writes to spread the I/O over each disk. The problem is that it provides no redundancy in case of a disk failure (you've lost everything at that point!).
:)
Another advantage is that it makes your two single drives of size N each look like a single drive of size N * 2. This was a big advantage back in the days of the all-common 9GB SCSI. If you had a 40GB movie file you were working with it was inconvenient (or impossible) to partition it up. Merging 5 x 9GB drives into one logical device was great, but with single-drive capacities reaching several hundred GB RAID0's single logical drive is not quie as useful.
Provided it is done via hardware and not software, of course.
I have no clue, but I *do* know if I catch anyone using that garbage I won't have to fake my boot in their ass! :)
No, not at all. But there is NO way PosgreSQL contains everything Oracle does as the parent stated.
I would agree wholeheartedly and get up on the horse with you if not for two things:
/. talks about RDBMS's the armchair DBAs come out of the woodwork and seem to think their tiny website all of a sudden is 'Enterprise' and that their usage of MySQL is revolutionary and shows the world that MySQL is the *one true RDBMS*.
/. and try and pretend MySQL is a cut above the rest.
MySQL AB have indicated quite clearly that MySQL is the world's most popular Open Source Database, designed for speed, power and precision in mission critical, heavy load use.
And in their docs, press releases, interviews etc. they do not hide the fact that they are taking aim at Oracle, DB2, MS SQL, Sybase ASE, et al.
Secondly, whenever anything on
Those who know, do. Those who don't sit on
Agreed. The difference is that our Sybase ASE servers in over 2.5 years running have NEVER failed. The only downtime came when installing routine patches/EBFs which is easily handled with Sybase's insanely great Replication Server product (end users noticed NOTHING).
Replication in MySQL is neat but should not try and make up for RDBMS shortcomings (no hot backup in MySQL) or instability issues. It should be there when you machines BLOW UP.
Again, we're talking different classes of software here.
I really like Postgres. Had it not really, really not sucked 3 years ago when we moved form MySQL we'd be on it now. We chose Sybase ASE which is I think a far better RDBMS than MySQL (and Postgres as well) but if we had the choice today it would've been Postgres (GREAT value for the money!! ;) ).
If you were a DBA you would know that Oracle has everything *including* the kitchen sink installed. You can't say the 9i 2GB install has nothing more, can you? I smell a troll.
Cool, I stand corrected then. IMHO it should be added to the docs TEXT datatype because that was the first place I looked to come up with my on/off row analysis:p hp?datatype -character.html
:)
l argeobj ects.html
http://www.postgresql.org/idocs/index.
Maybe just a short blurb stating it is TOASTed.
And the docs here:
http://www.postgresql.org/idocs/index.php?
Is what pointed to the alternate syntax for the large object support. So does the pg_class silently use the large_object syntax to stuff TEXT columns?
Also this:
Tip: There are no performance differences between these three types, apart from the increased storage size when using the blank-padded type.
Led me to beleive that the TEXT was stored on-row, because there will be a hair bit more I/O to find the TEXT data, which does not seem to be indicated.
I see that now for the 'large object' although you are forced to use non-standard SQL in order to facilitate. It would be nice if it was invisible.
It does not work with TEXT data though.
Copying what data over? Adding a new column?
:(
That does not seem like a very good way to do it. You will have to explain it a bit further since it doesn't seem to make sense to me.
DROPing a column is under the covers simply a SELECT INTO followed by a drop of the old table and a rename of the new. It is just packaged up a bit nicer by ALTER TABLE DROP bob
Yes, we all know virtually every RDBMS can store gobs of data. It's the whole 'doing stuff with it' wherein lie the rub.
I'm not saying that MySQL can't 'do stuff' to large amounts of data, but one can go to vBulletin.com and read about customers who routinely have to 'prune' their database of posts/threads (think HardOCP). We're not talking 100 million posts but maybe a mil or two. Now, certainly there are parts of the vBulletin application that just don't scale to 10 million posts. But, it seems MySQL has a bit of a problem dealing with complex JOINs, DISTINCTs, COUNT's (not the vanilla SELECT COUNT( * ) FROM table;) which other RDBMSs do not. It is not terribly hard to see why. In your traditional web apps most queries are of the type "SELECT * FROM myTable WHERE myAutoIncrement = CONSTANT" and so the more 'goofy' queries haven't been optimized as much. This is not idle conjecture but something I have witnessed 'in the field' and have done some rudimentary benchmarks in support of.
So, it is not unthinkable that someone doing a lot of heavy lifting with some moderate amount of data (couple million rows) will hit some brick walls with MySQL. As web applications start to 'grow up' MySQL will address the growing demand as they have started to already (with FK support, row-level locking, replication). The hope is, of course, that MySQL will remain just as easy to use and fast after the new features!
ASE 12.5+ (maybe 12) has the ability to go up to 16K row sizes (with the new larger page sizes) although you would be hard pressed to find a practical instance in which you *need* a row that big -- provided your relations are properly normalized.
;)). So you could easily hit the row cap with gobs of text (or a single image!). Now that the cap is gone you shouldn't see any problems. Aside from TEXT you should never hit the limit anyway unless you design horrible tables. :)
t ml But in short, know your data and you should have no problems.
The main problem that *existed* in PostreSQL was that TEXT (and other unstructured datatypes) live on-row in PostgreSQL (and MySQL, but I'll get to that later
Note that TEXT (IMAGE, etc.) data lives off-row in Sybase and MS SQL (and Oracle allows you to specify both I think!). This is very nice since you can keep your rows smaller (better cache hit ratio) and it also allows for a much higher probability that your TEXT data will be in a contiguous block saving expensive disk seeks.
Re: MySQL the Gemini table type has (IIRC) an easily hit max row size if you use text columns. Of course, no one uses Gemini over InnoDB so it is a non-issue.
InnoDB rows live on a HUGE 16KB pages (recompile MySQL to change; yuck!) which means the max row size in older versions is a function of page size (PAGE_SIZE / 2 - STUFF). In later versions this was removed, although limits still exist: http://www.mysql.com/doc/en/InnoDB_restrictions.h
Argh I mis-typed GB meant TB. Sybase ASE does 100GB (1TB probably practical limit). Sybase IQ *does* do 100 TB.
Sybase IQ also runs VLDBs into the 100's of GB range. A very, very capablo product, although Terradata in pretty much the market leader ar you point out.
Is Sybase SQL Anywhere. It offers scalable, bidirectional synchronization of information between your handheld and your Sybase Adaptive Server Enterprise, Oracle, Microsoft SQL Server and IBM DB2 RDBMS's. I would even think it might poke your MySQL/PostgreSQL DB via ODBC or some other method.
m l
A whitepaper is here: http://www.sybase.com/detail/1,6904,1016232,00.ht
The RIAA and all the lawyers in the world will never be able to completely stop pirating. Look at how much money the feds throw at drugs and the number of addicts on the street. If enough people want something, they'll get it.
:).
I know one of my chief frustrations is to search for a song and either have it incomplete, or be of poor quality (e.g. pops or other defects) or to simply have it not be the same song that I downloaded. If I could search for a song, pay $SOME_SMALL_AMOUNT (e.g. $1US) for it and download a 'known perfect' copy at my choice of bitrates (e.g. 128, 160, etc.) then sure as heck I'd do it.
Distributing these poisoned files would take an enormous amount of bandwidth, so they'd have to have some sort of agreement worked out with ISPs and a mass-content provider, say Akamai. Akamai has tens of thousands of servers located in hundreds (if not more) of ISPs throughout the nation. I think on peak usage they're pushing out 100 GB/sec. in the US (if not more). Simply say "Ok Akamai, can we buy 10GB on each of your servers and push all these MP3s out?". Then you write a gnutella client for each box which offers all the MP3s up for distribution.
I can't remember how the gnutella protocol works but I think it broadcasts search requests to the nodes that store a cache of what they have and what their neighbors offer and then can pass the request off. Have your client log all the requests (so you can tell the record companies which songs were requested more) and of course offer up your files when requested. If you do this with 10,000 boxes full of identical content chances are you're going to drown out any signal out there.
If you're really tricky, you can even have the client 'fake' files so you don't actually need to have the file on the box; you could send a pre-existing obfuscated file, or even dynamically build and stream the poisoned MP3.
Of course, all of this is moot if you still don't have a very easy, cheap method of offering MP3s online for the mass public. You could pitch it like this "Yeah, so you won't make much money off of offering $SOME_SMALL_AMOUNT for each MP3. But you're a fool if you think simply shutting Morpheus off will result in even 10% of the Morpheus users buying the actual CD or using a painful, userUNfriendly pay-per-MP3 system. However, what if we have a method to net you 20 or 30% of users who wouldn't pay you anyway?" So the pitch would be "We can't get you all of them, but our method would give you more than you're getting now!". Frankly the people who post on SlashDot (from the very negative response to the Subscription model) are not a good cross-section of the vast majority of internet users out there
So in your obfuscated file you have it play maybe 20 seconds of the file and then say "Sorry, this is a copywrited file. Pirating files costs artists money. If you want to buy this MP3 for $SOME_SMALL_AMOUNT, please visit http://www.somestore.com. 80% of $SOME_SMALL_AMOUNT earned will go directly to the artist."
It gives them a reason to buy it - not only do you have SomeStore.com very easily accepting payment, but you ACTUALLY PAY THE ARTISTS A MAJORITY OF THE MONIES EARNED! So it can quell the naysayers who say "Well the artist wouldn't receive anything anyway!" (rant: but who are you hurting more, the billion dollar-industry or the Artist who NEEDS even the small cut they receive from each CD sold?).
Some drawbacks could be of course that someone writes a 'detector' to find and ignore the invalid MP3s, or they block the IP addresses of the servers, etc. but that is easily fixed. Most non-power users (e.g. the great and huddled masses of the internet) don't want to update their Morpheus client every time a new version is released. Heck, even programs which offer hassle-free updating (e.g. antivirus, windowsupdate.com) very rarely are by the majority of internet users. Also, you'd work out the server IP settings with the ISP so that they would rotate to a random IP in their pool - since most of the servers are located in most ISPs you couldn't ban the single IP but perhaps a subnet. But since the IPs are in the ISP, you have now banned a large chunk of users. If they are in every ISP, you will have to ban every ISP (see the problem in banning IPs?).
So, to boil it down to a sentence:
Have very easy-to-use, hassle-free, cheap, reliable, etc. method for users to buy MP3s and they WILL
What RDBMS are you using? Oracle, Sybase ASE, and MS SQL Server do not behave in such a manner.
You're pobably thinking of a generic data cache. While it does store table data it does not store query results. Very, very different.