Oracle has always been just Oracle's DB. A bunch of young programmers saw the work being done at IBM (SQL/DS, System R) and decided they wanted to join the party. They created a working implementation of the standard based on the white paper for the VAX. Intent was to get the DB to work on many of the lower-cost minicomputers of the time, IIRC, which was possible once they re-wrote it in C for version 3.
Considering that PostGres and SQL Server have the same Berkeley roots, I'm going to have to disagree there.
SQL Server and PostGres are at about the same level as far as sophistication goes. SQL Server wins in the ease of administration department, but lags flexibility and openness.
As far as "just being an SQL 92 RDBMS" running on a SMP (but single image system), all three systems are at the same level. (That means any ignoring any specific clustering/dev frameworks/warehousing stuff)
Do you mean "double fault"? That's when the system experiences a (new, unexpected) page fault in the kernel while handling a page fault for user space. It indicates some kind of bug in the VM, as the page handler of any x86 OS must be specifically designed to avoid that problem.
Only on BSD and Linux... you want readlink.
on
Vista To Get Symlinks?
·
· Score: 2, Informative
The readlink() function shall place the contents of the symbolic link referred to by path in the buffer buf
which has size bufsize. If the number of bytes in the symbolic link is less than bufsize, the contents of the
remainder of buf are unspecified. If the buf argument is not large enough to contain the link content, the first
bufsize bytes shall be placed in buf.
If the value of bufsize is greater than {SSIZE_MAX}, the result is implementation-defined.
RETURN VALUE
Upon successful completion, readlink() shall return the count of bytes placed in the buffer. Otherwise, it shall return a
value of -1, leave the buffer unchanged, and set errno to indicate the error.
You know, if all the root servers went down, it wouldn't be the end of the internet. Most ISPs could reconfigure their name servers in short order to take over the functionality of the root servers and operate without peering for however many months it'd take to get new root servers up and running for a TLD.
So... I don't see that as being how the conversation went. World leaders are not CIOs...
Checkout scanners don't read the serial # off the barcode, just the 10-digit UPC label. Ever scan the "wrong" barcode with a self-checkout system? Sometimes retailers add the UPC label on the box after the fact with a built-in theft-deterrent device.
In any case, when you checkout, the POS system doesn't know the serial. The only time the serial "enters the system" is when you try to get service, register your warranty... or if you bought your equipment online from the vendor directly, or something like that.
The system could slew phase until the desired cancelation is achieved at some arbitrarily high frequency. It is the rate of change of the parameters of the system (whether in time or frequency space) to which it must adapt that ultimately governs its abilities. Fan noise is fairly stationary... so.
However there are all sorts of other issues which can not be easily solved (microphone/speaker placement, limited affected area, etc.).
There is a "cleaner" that is on the TODO list. Of course, you can delete files and re-use the space. But the performance slows down greatly once you start filling in "holes" left in the log after wrapping to the end of the allocated area. (A similar situation to database where you might want to compact, vacuum, condense, etc. a table).
If you can't do basic analysis (which might require some rudamentary calc) then you won't be able to model (and thus justify) any algorithmic choices you make as a project manager.
Well, if performance isn't your main objective then it might not matter so much... you're probably one of those get-it-to-work-then-shove-it-out-the-door kinda project managers.
You vote, right? Well, having a sense of persepctive and knowing "Hmmm... this political situation seems familiar" is something that you gain invaluable insight too through history.
Everyone should be familiar with history and geography. We live in a connected, opinionated world.
Simply hook the firewire up to your PC. Run any application that can capture MPEG2-TS (like dvgrab on linux, or windows video editing apps, or quicktime on macosx).
You use MythTV to control an IR device or serial input to the cable box to remotely change the channel at the appropriate time. Note that the MPEG2-TS that the firewire gives you contains no program information... Also I'm noting the MythTV 0.18.1 changes includes the following: "Internal channel-change over firewire support for DCT-6200 series cable boxes - no external program required like before."
So it seems MythTV has the infrastructure already set up.
If your cable box has the firewire features of the firmware disabled, it may still support external firewire HD attachment. If this is the case, you can record to the HD with the cable DVR features, then load it up on your PC. You should see.TSP files (which are just the file-based MPEG2-TS mentioned earlier) on FAT or ext2 partitions.
If you find things behaving in ways where your inverse-power-X laws don't always apply, then if you add other dimensions and distribute your interacting elements through it you can control the distribution and control which X the power law exhibits at different scales. Or you can translate things into these microscopic dimensions and explain how this "translation" has differing effects on normal matter due to the dimensional propensities of various force interactions.
And if these theoretical "rotations" of reality gel with string theory and stuff... BONUS.
From the application programmer's point of view, it can be treated like a FS-layer object. A.NET application will be able to open files or return a file set based on a SQL query, in effect. The explorer shell will be able to implement a bevy of virtual folders... etc.
Which is the whole point.
It seems the "WinFS"-ness of WinFS extends into the kernel such that a WinFS filesystem will be considered a seperate type from NTFS and FAT. I think they'll probably take of the reserved MFT entries to point at the dedicated relational metadata store to be used by the system-level Yukon engine. So it'll be "part" of the file system, the database using some of the space that you lose to "formatting". Of course, it's still using NTFS internally, and if you mount it as NTFS, it'll behave the way you expect (minus the metadata system).
Sort of like ext3 vs. ext2 (ext3 is a reserved inode pointing to a journal that the ext2 driver ignores)
You still need an indexing service. All that metadata isn't just going to poof out of the thin air. Metadata where it gets entered (save dialog in office, ID3 tags, thumbnails on pictures, etc.) needs to find it's way into this API, or it needs to be programmatically extracted by background processes.
I actually like the latter, it takes the burden off the applications.
Also, it'd be nice if concepts like the "Recently Used Files" and stuff like that gets rolled into it (that is, recently used is just a metadata field and the RUF directory is a "view" or "Select" against the database with appropriate criteria).
It's too bad WinFS isn't a real database-backed file store. Because then you could do all sorts of weird stuff... (and it's easy enough to provide a compatibility layer for a hierarchical-filesystem-assuming Win32 API)
The "Background" layer in GIMP is special if it doesn't contain transparency. In that case it tries to keep you from putting foreground layers (which might be transparent) behind it which would not be good if you tried to Flatten Image or something.
Just duplicate the resized layer, then do a canvas fill on the Background layer. That's it.
Slashdot Mirror
Oracle has always been just Oracle's DB. A bunch of young programmers saw the work being done at IBM (SQL/DS, System R) and decided they wanted to join the party. They created a working implementation of the standard based on the white paper for the VAX. Intent was to get the DB to work on many of the lower-cost minicomputers of the time, IIRC, which was possible once they re-wrote it in C for version 3.
Considering that PostGres and SQL Server have the same Berkeley roots, I'm going to have to disagree there.
SQL Server and PostGres are at about the same level as far as sophistication goes. SQL Server wins in the ease of administration department, but lags flexibility and openness.
As far as "just being an SQL 92 RDBMS" running on a SMP (but single image system), all three systems are at the same level. (That means any ignoring any specific clustering/dev frameworks/warehousing stuff)
Do you mean "double fault"?
That's when the system experiences a (new, unexpected) page fault in the kernel while handling a page fault for user space.
It indicates some kind of bug in the VM, as the page handler of any x86 OS must be specifically designed to avoid that problem.
NAME SYNOPSIS DESCRIPTION RETURN VALUE
My local Best Buy/Circuit City/Costco don't do that.
I don't shop at Walmart, so I can't comment on that.
My point is that it's a seperate step at POS, and apparently optional.
Mail order is a different story.
Let's hope it doesn't continue too far down that slippery slope.
You know, if all the root servers went down, it wouldn't be the end of the internet. Most ISPs could reconfigure their name servers in short order to take over the functionality of the root servers and operate without peering for however many months it'd take to get new root servers up and running for a TLD.
So... I don't see that as being how the conversation went. World leaders are not CIOs...
Checkout scanners don't read the serial # off the barcode, just the 10-digit UPC label.
Ever scan the "wrong" barcode with a self-checkout system? Sometimes retailers add the UPC label on the box after the fact with a built-in theft-deterrent device.
In any case, when you checkout, the POS system doesn't know the serial. The only time the serial "enters the system" is when you try to get service, register your warranty... or if you bought your equipment online from the vendor directly, or something like that.
The system could slew phase until the desired cancelation is achieved at some arbitrarily high frequency.
It is the rate of change of the parameters of the system (whether in time or frequency space) to which it must adapt that ultimately governs its abilities.
Fan noise is fairly stationary... so.
However there are all sorts of other issues which can not be easily solved (microphone/speaker placement, limited affected area, etc.).
nt
nt
There is a "cleaner" that is on the TODO list.
Of course, you can delete files and re-use the space. But the performance slows down greatly once you start filling in "holes" left in the log after wrapping to the end of the allocated area. (A similar situation to database where you might want to compact, vacuum, condense, etc. a table).
You can.k ing/120/= SunStore&cmdStartWebConfig_CP&familyCode=RACK90036 N&baseSelected=1
http://www.sun.com/products-n-solutions/hw/networ
http://store.sun.com/CMTemplate/CEServlet?process
Expensive as hell though.
Anytime you say: take the limit (as some parameter goes to 0, goes to inf.) you need calc.
L'Hopital's rules ARE calc.
If you can't do basic analysis (which might require some rudamentary calc) then you won't be able to model (and thus justify) any algorithmic choices you make as a project manager.
Well, if performance isn't your main objective then it might not matter so much... you're probably one of those get-it-to-work-then-shove-it-out-the-door kinda project managers.
You vote, right?
Well, having a sense of persepctive and knowing "Hmmm... this political situation seems familiar" is something that you gain invaluable insight too through history.
Everyone should be familiar with history and geography. We live in a connected, opinionated world.
That's how I used to do it too... it was hard to convince people why it was a good idea, though,.
And it acheives the same result as this book (polynomials for most analysis) although through different means.
Yes you can.
.TSP files (which are just the file-based MPEG2-TS mentioned earlier) on FAT or ext2 partitions.
Simply hook the firewire up to your PC. Run any application that can capture MPEG2-TS (like dvgrab on linux, or windows video editing apps, or quicktime on macosx).
You use MythTV to control an IR device or serial input to the cable box to remotely change the channel at the appropriate time. Note that the MPEG2-TS that the firewire gives you contains no program information...
Also I'm noting the MythTV 0.18.1 changes includes the following: "Internal channel-change over firewire support for DCT-6200 series cable boxes - no external program required like before."
So it seems MythTV has the infrastructure already set up.
If your cable box has the firewire features of the firmware disabled, it may still support external firewire HD attachment. If this is the case, you can record to the HD with the cable DVR features, then load it up on your PC. You should see
If you find things behaving in ways where your inverse-power-X laws don't always apply, then if you add other dimensions and distribute your interacting elements through it you can control the distribution and control which X the power law exhibits at different scales. Or you can translate things into these microscopic dimensions and explain how this "translation" has differing effects on normal matter due to the dimensional propensities of various force interactions.
And if these theoretical "rotations" of reality gel with string theory and stuff... BONUS.
Amirite?
So I don't know how much credit they really deserve.
Was the really good stuff Sybase's work or their own?
From the application programmer's point of view, it can be treated like a FS-layer object. A .NET application will be able to open files or return a file set based on a SQL query, in effect. The explorer shell will be able to implement a bevy of virtual folders... etc.
Which is the whole point.
It seems the "WinFS"-ness of WinFS extends into the kernel such that a WinFS filesystem will be considered a seperate type from NTFS and FAT. I think they'll probably take of the reserved MFT entries to point at the dedicated relational metadata store to be used by the system-level Yukon engine. So it'll be "part" of the file system, the database using some of the space that you lose to "formatting". Of course, it's still using NTFS internally, and if you mount it as NTFS, it'll behave the way you expect (minus the metadata system).
Sort of like ext3 vs. ext2 (ext3 is a reserved inode pointing to a journal that the ext2 driver ignores)
You still need an indexing service.
All that metadata isn't just going to poof out of the thin air. Metadata where it gets entered (save dialog in office, ID3 tags, thumbnails on pictures, etc.) needs to find it's way into this API, or it needs to be programmatically extracted by background processes.
I actually like the latter, it takes the burden off the applications.
Also, it'd be nice if concepts like the "Recently Used Files" and stuff like that gets rolled into it (that is, recently used is just a metadata field and the RUF directory is a "view" or "Select" against the database with appropriate criteria).
It's too bad WinFS isn't a real database-backed file store. Because then you could do all sorts of weird stuff... (and it's easy enough to provide a compatibility layer for a hierarchical-filesystem-assuming Win32 API)
The "Background" layer in GIMP is special if it doesn't contain transparency. In that case it tries to keep you from putting foreground layers (which might be transparent) behind it which would not be good if you tried to Flatten Image or something.
Just duplicate the resized layer, then do a canvas fill on the Background layer.
That's it.
http://www.mathworks.com/products/symbolic/
Download "Network" SP1a
Microsoft Webpage on deploying SP1