I can say that the developers I work with take things like
subselects and stored procedures for granted
Me too (developing with Oracle and PostgreSQL). I had PostgreSQL
and SAP DB in mind when I wrote that reply, not MySQL.
IMHO, adding features to MySQL has been and is a bad idea: it takes
away the only reason one would use it in the first place (the
simplicity of an SQL-like query language combined with Berkeley
DB-like speeds), and makes it into what probably is the worst of all
the RDBMS out there (just as links is one of the best textual
browsers, it is also the worst graphical browser when compiled with
graphic support).
Nonetheless, MySQL has an apparently large portion of its huge
userbase demanding the features needed to cope with all the
scalability and reliability issues that promptly rise when the number
of connected (ab)users grows... so the KISS principle is hard to apply
in this case.
Agreed. OTOH, you should ask yourself in how many places you really
need nothing less than the state of the art. Probably less than the
50% of the total.
My guess is that such percentual is going to shrink as more people
become aware of free RDBMS (I'm thinking more about PostgreSQL or SAP
DB than MySQL), but Ellison has nothing to worry, since absolute
numbers of RDBMS users will go up as well, and some of them are going
to need Oracle sooner or later.
In other words, a small percentual of a large number may still be a
large number. I'd say there's plenty of space for everyone (everyone
that does not stand still, obviously).
OTOH, computers usually are noisy and don't fit too well in the living room. Plus, I'm too lazy to
Look for an IrDA remote control
Develope a unified front-end to mplayer, ogg123, mpg321 and an image viewer.
I was getting tired to plug/unplug my laptop into the TV set, so I just bought that player, which is less versatile than my laptop, but it is quiet and it does the job.
I purchased one three weeks ago (a KiSS DP-450), but AFAIK it has been on the european market since late 2002.
Basically, the DP-450 it is a VCR-sized box with a 150Mhz StrongARM running Linux 2.4.x + busybox + custom software + custom hardware helping MPEG2 and MPEG4 decoding + a (Toshiba?) DVD drive + remote control. No ethernet on the DP-450 (but it is there on the DP-500). No fans:-)
Just insert a CD/DVD and it starts playing what's on it (but press the load button: just pushing the loading bay is not enough):
if it's a DVD, well, it plays the DVD, just like every other DVD player
if it's a CD full of MP3/OGG files, it is mounted and you can browse the content with the remote control and play the file. Of course it is really Linux under the hood, so it understands also symlinks. Apparently it ignores ID3 tags and similars. No playlists. No fast-forward/rewind while playing.
if it's a CD full of JPEG images, is starts a full-screen slideshow (and you can navigate, zoom and rotate with the remote control). Not exactly fast if your average image is 1MB, but acceptable.
if it's a CD with DivX files on it, you can browse the content, select and play
Briefly said: this is an MPEG2 and MPEG4 player (hence DivX 4 and 5; old DivX 3 is out of question), and as of now just MPEG4 Simple Profile features are supported (thus it won't play everything out there, as of now: be warned).
Image quality is nice, but not excellent (blacks aren't so... black). Firmware upgrades on the DP-450 are performed by downlowading an iso image (of a couple of megabytes) from the manifacturer website, and then booting the player with it.
All in all, a nice piece of hardware, easy to use, somewhat expensive (I purchased mine for 400 Euros). But it sits there beside your TV set and it just works.
Too heavy for a 56k dialup connection? I didn't think it was any worse than Citrix there, but I could be wrong about that.
Some reasons why X is not a good idea in this case:
Plain X protocol doesn't work well with high-latency connections (like a 56k dialup connection): it's the best thing since sliced bread in a LAN, it employs several tecniques to save bandwidth, but unfortunately apps have to continuosly talk with the X server and latencies play a big role. IMHO this is expecially true with modern (read: relatively young) toolkits like Qt or GTK+, which probably could do better at keeping the number of requests to the X server to a minimum (note: the number, not the size). Of course, there is the Low Bandwidth extension to X (LBX, to be used via lbxproxy, and the Differential X Protocol Compressor dxpc which, aside from compression, employs some caching to avoid contacting the X server on the other side when it is not absolutely necessary (thus making latency less of a problem).
The X session is fragile: dialup connections tend to go down unexpectedly, and in such cases your X session is abruptly closed. For each connected client (application on the application server), the X server keeps a lot of state which isn't saved. If the link goes down, you can't reconnect later and find your session again (or, for the same reason, transparently move an app from one X server to another - even if there are some tricks about the subject)
For these two reasons, VNC seems to be a better idea in this case.
Vim / emacs (Please delete as appropriate) is a good text editor, but that's all.
Emacs? You know, if people used to say "Emacs is a nice OS, but I prefer UNIX", there is a reason.:-)
GNU Emacs / XEmacs is to text editors what Mozilla is to web browsers (well, to "mere HTML renderers with some buttons attached" would be more appropriate). I'd go further and say that Mozilla is definitively the 21st century Emacs (a proof? Komodo).
Example 1: a.zip archive is to a file contained in it roughly as an
Ogg file is to a Vorbis stream inside it. Vorbis is the codec, Vorbis
is the format, Ogg is the container.
Example 2: AVI is a container for audio/video streams encoded with
several codecs. Your typical DivX;-) 4/5 file is an MPEG4 video
stream + MP3 audio stream packed together inside an AVI file, where
DivX;-) is the video codec, MPEG4 is the video format, lame/whatever
is the audio codec, MPEG2 layer 3 is the audio format and AVI is the
box containing them.
The same is true for Quicktime: technically it is just a container
for audio/video streams encoded with several codecs (usually Sorenson
for video).
Example 3: AFAIK,.doc is just a container for a set of OLE
serialized objecs. Accessing single OLE objects inside the.doc file
should be easy. Understanding the exact meaning of each serialized
object is the hard part.
If I were to send you a document in Microsoft Word format, nothing at all would happen to any of the other documents on your computer.
Feed the troll...
Coders giving out code under GPL basically say that if you dont'want to share your sources with others (but only the binaries), they don't want to share their sources with you. You have the choice to deal with that or not. Nobody is forcing GPL on your code. Stallman is not pointing a gun to your head. Even if you decide it is convenient to use that GPL code, it is still your own choice.
People using proprietary document formats basically say that if you want to read their documents you have to use a proprietary application. Nobody is forcing you to use such proprietary applications. Bill Gates is not poining a gun to your head. Even if you then decide it is convenient to use that format too, it is still your own choice.
Coders that release their derived works under the GPL just in order to use other GPL components do that uniquely because of convenience.
People that adopt the same proprietary tools as the ones used by people sending them documents in proprietary formats are doing that uniquely because of convenience.
GPL coders mainly want the sources. BSD coders mainly want the credits. Proprietary coders mainly want what's in the wallet. Presto, decide where's the convenience for you.
Re:A bzip2 version would be nice ...
on
Gzip on a PCI card
·
· Score: 4, Interesting
gzip works with streams, producing input as it gets output. OTOH bzip2 treats the input as blocks.
Gzip works with blocks of data too, but the block size is 32KB instead of nearly 1MB and it is not nearly as CPU intensive as bzip2, so this is why it appears to produce a continuous stream of compressed data (even if, strictly speaking, it doesn't).
Gzip just seems to be a well-balanced compromise between resources and resulting compression ratio, plus it is Free Software (hint: bzip2 is Free Software too, but Rar isn't).
Why isn't it possible to produce incremental binary patches containing just the diffs?
Better said: why not provide also rsync access to packages?
After all, it is possibile that foo-x.y.z-N+1.arch.rpm is
mostly the same as foo-x.y.z-N.arch.rpm (same for.deb
packages).
Or it could be as well that this is far from being true.
Assuming (for the sake of the argument) it is true, unfortunately,
rsync is NOT the right tool, since diffs are done on the fly
and this would put huge workloads on the server side.
But a tool specially crafted for this that makes a local copy of
the package to update, retrieves via http all the needed patchsets
(precomputed static files on the server-side), applies them in the
proper order and does a bit of MD5summing to check if the result is
right would be something nice to have.
Probably there is something out there that already implements most
of the needed functionalities (unfortunately I don't know where),
since the idea is so simple.
Opterons are Optimized for Optimum performance, suggesting that it's the best of the best. The Latin optimus, from which the prefix is ultimately derived, simply means "best".
To me (and my newphew), "opteron" mainly evokes the image of a pterodactyl of the jurassic age cross-breeded with an eigth-legged monster of some sort (which is relevant, since the opteron can do 8-way SMP).
My mother tongue is italian, should I worry if I didn't catch the "optimum" pun at all until I read your message?:-)
That the idea to use it as a platform to develope portable applications (using ECMAScript + XUL) is catching on slower than some people would expect. This is a pity, because ungodly amounts of effort goes in making this possible, and still people see it just as a web browser (a large one).
Other than that, Mozilla-the-web-browser is fine, Mozilla-the-messaging suite is at least good enough, and Mozilla-the-javascript-debugger shows lots of promises.
I don't include Mozilla-the-IDE (Komodo) in the list, since it deviates too much from the usual distribution (even if it is Gecko Inside(TM)).
Now waiting for Mozilla-the-organizer (thru Calendar, planned for 1.4 ~ 1.5). Perhaps a Mozilla-the-file-manager would be something worth implementing (but Meow seems definitively dead).
A united GUI is the best chance Linux has for a respectable marketshare on the desktop.
And of course it has to be done the way I like it, which incidentally is the opposite as you like it, right? Of course you will adopt my way of doing things. Or perhaps I should adopt yours. Better, let's both adopt a way neither I nor you like, so we will do nothing in the end but at least none will have an unfair advantage.:-)
If you really knew anything you would know that the "inconsistent" program is Emacs, which takes Ctrl+Y
For the humor-impaired, this requires a little explanation: Emacs has been using C-w, M-w and C-y respectively to "kill" and "yank" to/from the "kill ring" well before the "watered down" notion of "cut", "copy", "paste" and "clipboard" surfaced the screens.
I'm saying "watered down" because Emacs allows also to merge a "kill" (cut or copy) with a previous kill (M-C-w) and yank (paste) previous values of the kill ring (with one or more M-y after a yank - the default is to keep the last 16 values, but of course you can increase that).
So, if we look at what comes first, it can be as well said that it is the current usage of C-x, C-c and C-v which is inconsistent (and less powerful)... (ducks).
Someone wants a simple feature that other commercial offerings have had for years, and they are told to go look for an abandoned sourceforge page somewhere
man xclipboard.
It has been part of the standard X11 distribution (thus, also XFree86) for ages.
IE for Mac OS X could be called a "clunky port", maybe (of IE for Mac OS 9, which was an elegant port of IE for Windows).
AFAIK, IE for MacOS is special in that it is not a port of IE for Windows, but instead it has been rewritten for the most part by a separate development team, offering good support for W3C recommendations (it should be at least on par with Mozilla/Gecko). The only things in common with its Window cousin would be the name and the logo.
With GNU Emacs you can set named bookmarks referring to a specific position in specific files (i.e. with M-x bookmark-set).
Then, in a different session, jumping to a bookmark (M-x bookmark-jump) automatically opens that file again and positions the text where you set the bookmark (even if the text changed in the meantime: Emacs bookmarks keep some context lines with them).
OTOH, the thing I really miss is a nice program from the Amiga days called "muchmore", which provided a sort of full-screen equivalent of the well known "less", but with smooth scrolling, autoscrolling, and both scrolling speed and direction could be changed by simply moving the mouse pointer towards or away from the screen borders. IIRC, it too had bookmarks on texts.
If you are going to need that 24MB for other apps, they are paged out on the swap file anyways, (it can also happen incrementally, while the system is idle), and can be paged in when needed (which is faster than allocating and constructing ex novo everything).
Re:Doom (and Doom II) is probably the best game ev
on
25 Best Linux Games
·
· Score: 1
Duke Nukem 3D was fab as well but I guess it's now dead because the engine wasn't GPL'ed:-(
Well, the Build Engine on which Duke Nukem 3D is based is (to some degree) open source, and has been ported to Windows and Linux, but apparently it isn't in a great shape...
No, Linux has its own implementation of networking code rewritten from scratch.
This is why problems affecting the traditional *BSD implementation of TCP/IP (which is used pratically everywhere except for Linux) don't usually affect Linux. Of course, the opposite is also true.
That said, the FreeBSD kernel is known (or, at least, it has been known) for being able to handle high load/low resource conditions far more gracefully than Linux.
Slashdot Mirror
Me too (developing with Oracle and PostgreSQL). I had PostgreSQL and SAP DB in mind when I wrote that reply, not MySQL.
IMHO, adding features to MySQL has been and is a bad idea: it takes away the only reason one would use it in the first place (the simplicity of an SQL-like query language combined with Berkeley DB-like speeds), and makes it into what probably is the worst of all the RDBMS out there (just as links is one of the best textual browsers, it is also the worst graphical browser when compiled with graphic support).
Nonetheless, MySQL has an apparently large portion of its huge userbase demanding the features needed to cope with all the scalability and reliability issues that promptly rise when the number of connected (ab)users grows... so the KISS principle is hard to apply in this case.
Agreed. OTOH, you should ask yourself in how many places you really need nothing less than the state of the art. Probably less than the 50% of the total.
My guess is that such percentual is going to shrink as more people become aware of free RDBMS (I'm thinking more about PostgreSQL or SAP DB than MySQL), but Ellison has nothing to worry, since absolute numbers of RDBMS users will go up as well, and some of them are going to need Oracle sooner or later.
In other words, a small percentual of a large number may still be a large number. I'd say there's plenty of space for everyone (everyone that does not stand still, obviously).
OTOH, computers usually are noisy and don't fit too well in the living room. Plus, I'm too lazy to
I was getting tired to plug/unplug my laptop into the TV set, so I just bought that player, which is less versatile than my laptop, but it is quiet and it does the job.
Basically, the DP-450 it is a VCR-sized box with a 150Mhz StrongARM running Linux 2.4.x + busybox + custom software + custom hardware helping MPEG2 and MPEG4 decoding + a (Toshiba?) DVD drive + remote control. No ethernet on the DP-450 (but it is there on the DP-500). No fans :-)
Just insert a CD/DVD and it starts playing what's on it (but press the load button: just pushing the loading bay is not enough):
Briefly said: this is an MPEG2 and MPEG4 player (hence DivX 4 and 5; old DivX 3 is out of question), and as of now just MPEG4 Simple Profile features are supported (thus it won't play everything out there, as of now: be warned).
Image quality is nice, but not excellent (blacks aren't so... black). Firmware upgrades on the DP-450 are performed by downlowading an iso image (of a couple of megabytes) from the manifacturer website, and then booting the player with it.
All in all, a nice piece of hardware, easy to use, somewhat expensive (I purchased mine for 400 Euros). But it sits there beside your TV set and it just works.
Some reasons why X is not a good idea in this case:
For these two reasons, VNC seems to be a better idea in this case.
Emacs? You know, if people used to say "Emacs is a nice OS, but I prefer UNIX", there is a reason. :-)
GNU Emacs / XEmacs is to text editors what Mozilla is to web browsers (well, to "mere HTML renderers with some buttons attached" would be more appropriate). I'd go further and say that Mozilla is definitively the 21st century Emacs (a proof? Komodo).
Example 2: AVI is a container for audio/video streams encoded with several codecs. Your typical DivX ;-) 4/5 file is an MPEG4 video
stream + MP3 audio stream packed together inside an AVI file, where
DivX ;-) is the video codec, MPEG4 is the video format, lame/whatever
is the audio codec, MPEG2 layer 3 is the audio format and AVI is the
box containing them.
The same is true for Quicktime: technically it is just a container for audio/video streams encoded with several codecs (usually Sorenson for video).
Example 3: AFAIK, .doc is just a container for a set of OLE
serialized objecs. Accessing single OLE objects inside the .doc file
should be easy. Understanding the exact meaning of each serialized
object is the hard part.
Feed the troll...
Coders giving out code under GPL basically say that if you dont'want to share your sources with others (but only the binaries), they don't want to share their sources with you. You have the choice to deal with that or not. Nobody is forcing GPL on your code. Stallman is not pointing a gun to your head. Even if you decide it is convenient to use that GPL code, it is still your own choice.
People using proprietary document formats basically say that if you want to read their documents you have to use a proprietary application. Nobody is forcing you to use such proprietary applications. Bill Gates is not poining a gun to your head. Even if you then decide it is convenient to use that format too, it is still your own choice.
Coders that release their derived works under the GPL just in order to use other GPL components do that uniquely because of convenience.
People that adopt the same proprietary tools as the ones used by people sending them documents in proprietary formats are doing that uniquely because of convenience.
GPL coders mainly want the sources. BSD coders mainly want the credits. Proprietary coders mainly want what's in the wallet. Presto, decide where's the convenience for you.
Gzip works with blocks of data too, but the block size is 32KB instead of nearly 1MB and it is not nearly as CPU intensive as bzip2, so this is why it appears to produce a continuous stream of compressed data (even if, strictly speaking, it doesn't).
Gzip just seems to be a well-balanced compromise between resources and resulting compression ratio, plus it is Free Software (hint: bzip2 is Free Software too, but Rar isn't).
Better said: why not provide also rsync access to packages?
After all, it is possibile that foo-x.y.z-N+1.arch.rpm is mostly the same as foo-x.y.z-N.arch.rpm (same for .deb
packages).
Or it could be as well that this is far from being true.
Assuming (for the sake of the argument) it is true, unfortunately, rsync is NOT the right tool, since diffs are done on the fly and this would put huge workloads on the server side.
But a tool specially crafted for this that makes a local copy of the package to update, retrieves via http all the needed patchsets (precomputed static files on the server-side), applies them in the proper order and does a bit of MD5summing to check if the result is right would be something nice to have.
Probably there is something out there that already implements most of the needed functionalities (unfortunately I don't know where), since the idea is so simple.
To me (and my newphew), "opteron" mainly evokes the image of a pterodactyl of the jurassic age cross-breeded with an eigth-legged monster of some sort (which is relevant, since the opteron can do 8-way SMP).
My mother tongue is italian, should I worry if I didn't catch the "optimum" pun at all until I read your message? :-)
That the idea to use it as a platform to develope portable applications (using ECMAScript + XUL) is catching on slower than some people would expect. This is a pity, because ungodly amounts of effort goes in making this possible, and still people see it just as a web browser (a large one).
Other than that, Mozilla-the-web-browser is fine, Mozilla-the-messaging suite is at least good enough, and Mozilla-the-javascript-debugger shows lots of promises.
I don't include Mozilla-the-IDE (Komodo) in the list, since it deviates too much from the usual distribution (even if it is Gecko Inside(TM)).
Now waiting for Mozilla-the-organizer (thru Calendar, planned for 1.4 ~ 1.5). Perhaps a Mozilla-the-file-manager would be something worth implementing (but Meow seems definitively dead).
And of course it has to be done the way I like it, which incidentally is the opposite as you like it, right? Of course you will adopt my way of doing things. Or perhaps I should adopt yours. Better, let's both adopt a way neither I nor you like, so we will do nothing in the end but at least none will have an unfair advantage. :-)
Hell is others...
For the humor-impaired, this requires a little explanation: Emacs has been using C-w, M-w and C-y respectively to "kill" and "yank" to/from the "kill ring" well before the "watered down" notion of "cut", "copy", "paste" and "clipboard" surfaced the screens.
I'm saying "watered down" because Emacs allows also to merge a "kill" (cut or copy) with a previous kill (M-C-w) and yank (paste) previous values of the kill ring (with one or more M-y after a yank - the default is to keep the last 16 values, but of course you can increase that).
So, if we look at what comes first, it can be as well said that it is the current usage of C-x, C-c and C-v which is inconsistent (and less powerful)... (ducks).
man xclipboard.
It has been part of the standard X11 distribution (thus, also XFree86) for ages.
"600 MB" which takes just a couple of KB on your HD.
AFAIK, IE for MacOS is special in that it is not a port of IE for Windows, but instead it has been rewritten for the most part by a separate development team, offering good support for W3C recommendations (it should be at least on par with Mozilla/Gecko). The only things in common with its Window cousin would be the name and the logo.
OTOH, IE for Solaris and HP-UX were just ports using a proprietary equivalent of Winelib (by Mainsoft) providing a Win32 API on such systems.
Then, in a different session, jumping to a bookmark (M-x bookmark-jump) automatically opens that file again and positions the text where you set the bookmark (even if the text changed in the meantime: Emacs bookmarks keep some context lines with them).
OTOH, the thing I really miss is a nice program from the Amiga days called "muchmore", which provided a sort of full-screen equivalent of the well known "less", but with smooth scrolling, autoscrolling, and both scrolling speed and direction could be changed by simply moving the mouse pointer towards or away from the screen borders. IIRC, it too had bookmarks on texts.
Of what use is memory if it stays unused?
If you are going to need that 24MB for other apps, they are paged out on the swap file anyways, (it can also happen incrementally, while the system is idle), and can be paged in when needed (which is faster than allocating and constructing ex novo everything).
Well, the Build Engine on which Duke Nukem 3D is based is (to some degree) open source, and has been ported to Windows and Linux, but apparently it isn't in a great shape...
It was a Novell Netware file server, administered remotely: see The Register
A system can be solid at most as much its administration is?
No, Linux has its own implementation of networking code rewritten from scratch.
This is why problems affecting the traditional *BSD implementation of TCP/IP (which is used pratically everywhere except for Linux) don't usually affect Linux. Of course, the opposite is also true.
That said, the FreeBSD kernel is known (or, at least, it has been known) for being able to handle high load/low resource conditions far more gracefully than Linux.
FWIW, I clearly remember an italian version called "Grillo Parlante" (Talking Cricket).
According to what's here, there were also a French and a German version.
Well, some patents were filed from the very start, for example the patent on the SUID bit, owned (at the time) by Bell Labs, inventor: Dennis Ritchie.