The concept IBM is going for is to treat IT as another utility. Instead of some small company having to keep an expensive IT staff and maintain their own computers/network/storage, IBM says that it will do this for you.
This is exactly what management at the 60K+ financial firm I'm at would love to hear. The only problem is that IT and everything that comes with it needs to be customized to the business requirements. Are these commodity IT staff going to install every fruitcake application and know that you need INET.DLL from Win3.51 or that to restore files you use these tapes labeled that way, etc, etc. When the suits realize that they have to make investments in individuals and that IT (especially programmers) are not a commodity the recession will be over.
I copied over MS TTF fonts and X acknowledged them but looking at a webpage in Moz with Verdana shows huge disproportionate characters. Is this a problem with Mozilla or just a problem with raw TTF/xfs?
Aren't we past the point were Linux needs to be "adopted"? Is Linux an orphan that needs a home? Is Linux the red headed step child of operating systems? I think this shoud read "Linux Takes Over Chrysler (for vehicle simulations)".
X Windows Tip 151: Bind F3 to "previous workspace" and F4 to "next workspace". This works with most window managers (certainly does with WindowMaker). Then have all your apps and terminal windows maximized or tiled vertically. It would be hard to find a more efficient use of space. Why limit yourself to only one when you can many depending on how much video memory you have? Do an e-mail, hit F4, F4 to move over 2 workspaces, edit a source file, hit F4 again and compile and run in a separate terminal, F3 to go back and edit, F4, F3, F3 F4,...etc.
It is extremely difficult, in fact, to breed populations whose solutions *do* make sense. They find "organic", bizarre, complicated, twisted, fragile solutions much more often than something simple and straight-forward.
Without Rendezvous, you have several options-- all of them unappealing. You might start with some kind of application specific configuration file format
Why does it have to be application specific? Like using Rendezvous is less "specific".
and a cobbled-together system for distributing such data around to all the hosts in the system.
What's wrong with TFTP or NFS or any other network filesystem?
You might instead store such network configuration in some directory server on the network, like DNS
DNS is a directory server?
or LDAP or some kind of custom front end
So Rendezvous has a non-custom front end? That's not good. If my processes are specific to the particular problem domain (probably unique if it demands a "distributed system") then I would rather have a "front end" to match.
to some kind of evil database backend. Either way, you quickly encounter what we veterans like to call "the cache coherency problem,"
Then what do you call it? Because using a database sounds like a very good way to maintain cache coherency. You would rather state be strewn across the network?
i.e. the information in the directory must be consistent at all times with the actual state of the processes in the system.
So Rendezvous is going to guarantee state consistency between processes better than a database backend? It was compared to NetBIOS for cripes sake. I think you've made a pretty stong case against using Rendezvous for a "distributed system". My impression is that other than negotiating IP and multicast addresses a Rendezvous client is otherwise application specific.
Rendezvous is used to setup printers and wireless devices, not orchestrate large scale distributed processes.
Funny, any link for the draft you might find refers to draft-ietf-zeroconf-reqts-10.txt which just doesn't exist on the ietf site. So being real clever I flipped that 10 to an 11 and behold theres a new version dated the 26th.
To add to the irony the Windows executable was created using Samba source code. You can tell by the dialects it submits (one of them is "Samba") and by the fact that the tree connect was not batched with the session setup.
One thing that annoys me to no end is when I install the latest version and by default some new silly screwball "IMprovement" is activated.
Like the time line numbers (:set nu) have lines under them. You have to turn that off with (hi LineNr term=bold).
In the latest release (from RH actually) they put visual line breaks (not real newlines) in lines that are longer than the terminal is wide so that it doesn't print text in the left margin. Supposed to look nice I guess. Not only is that just a silly worthless feature but now if I select a line in X with the mouse and paste it somewhere I get stairstepping text.
I think their stuggling for new ideas. I wish they would clean up that morass and make a nice X widget instead.
I just setup Tomcat and Resin to develop a Filter that exposes our library to containers. Results: Tomcat is pathetically slow compared to Resin. No contest.
I may have been pretty horrible at math but I would feel very confident about walking over and plunking these guys over the head with a big rubber bat for publishing these nonsense plots. Three data points?! Are you kidding me? They obviously freehanded those dotted lines which could have been in the perpendicular direction and maintain the same statistical significance. All these graphs prove is that there are a lot more "gnulinux-users" in 1998 than there were in 1992 and 1993.
So what is this for? Would this be used to implement a Display Postscript widget or be used directly by drawing programs or something? What's the point?
First, this was not a USENET post. It's a message from the linux kernel mailing list that google is pumping into their groups.google.com database. Second, Linus is not saying he thinks Hammer is a better architecture. What he said in this message was that the current Linux page table implementation is not ideal for use with IA64 and therefore, for the sake of Linux servers everywhere, it would be better for Hammer to provail in the near to medium term future. I don't know his real position, but I would be very surprised if Linus though Hammer was a "better" architecture. X86 is an awkward instruction set that has been perpetuated by software designed to run on it. The core of these chips like Pentiums are really RISC chips with hardware wrappers to implement the X86 instructions. So it's just a waste if die space. IA64 is purer and a much better long term choice. Don't over analize a simple e-mail message from someone on lkml. These are not markedroid approved public service announcements.
The Standard C Library,P.J. Plauger I use this book *all* the time. It's very informative to see how a C library implementation works, and it has the important sections of the C standard reprinted with discussion. My #1 pick.
Advanced Programming in the UNIX Environment,W. Richard Stevens The comp.unix.programmer bible. A must have for UNIX programmers.
UNIX Network Programming Networking APIs: Sockets and XTI, W. Richard Stevens Another classic from Stevens. I have not done a tremendous amount of socket programming in C but I have extensively in Java and I am confident that I can in UNIX after having read select portions of this obviously high quaility book.
Lex & Yacc, John R. Levine I think yacc (the GNU version is called bison) is an extreemly usefull tool. I started writing yacc grammers for all sorts of configuration files on my system while reading this the first time. Great fun, and potentially very usefull. A good yacc grammer can save tremendous amounts of time and is often the proper way to handle a problem. And I'm not recommending the book just because it's the only of it's kind. It's genuinely good.
Algorithms on Strings, Trees, and Sequences Computer Science and Computational Biology, Dan Gusfield This is a tough book. Be prepared to ponder the presentation of each algorithm and skip the proofs (I believe you Dan). If you find yourself getting stuck in spegetti string manipulation, implement a few of these algorithms like calculating the edit distance of two sequences or Boyer-Moore. You'll be a better programmer after and there are very practical application hidden in this book. Again, the presentation is of the highest quality.
Introduction to Algorithms, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest I have three algorithms books. This one's the best and that's not a compermise. It's a great book.
The C Programming Language, Brian W. Kernighan, Dennis M. Ritchie Of course.
The Art of Computer Programming (Vols 1-3), Donald E. Knuth
The C++ Programming Language (Third Edition and Special Edition), Bjarne Stroustrup
Xlib Programming Manual, Adrian Nye
Applied Cryptography: Protocols, Algorithms, and Source Code in C, 2nd Edition, Bruce Schneier
X Window System Toolkit, A Complete Programmer's Guide and Specification, Paul J. Asente, Donna Converse, Ralph R. Swick, Paul Assente
Expert C Programming, Peter van der Linden
Life with UNIX, Don Libes and Sandy Ressler
Programming Pearls, Jon Bentley
Structured Computer Organization, Andrew Tanenbaum
The Code Book, Simon Singh
Etudes for Programmers, Charles Wetherell
Secrets and Lies, Bruce Schneier
The Pragmatic Programmer, Andrew Hunt and David Thomas
TCP/IP Illustrated, Richard Stevens
Exploring Expect, Don Libes
UNIX System Administration Handbook, Evi Nemeth et al If you use UNIX, you need this book. Forget Running Linux. This is much better and the latest edition covers Red Hat as well as BSD, Solaris, and HP/UX.
Digital Woes, Lauren Wiener
Surely You're Joking, Mr. Feynman!, Richard P. Feynman This was a pretty funny book. Very readable for anyone. A NYT Bestseller.
Structure and Interpretation of Computer Programs, Harold Abelson, Gerald Jay Sussman, and Julie Sussman
Essentials of Programming Languages, Daniel P. Friedman, Mitchell Wand, and Christopher T. Haynes
Software Fundamentals: Collected papers, David L. Parnas, Daniel M. Hoffman and David M. Weiss
The Invisible Computer, Donald A. Norman
Lion's Commentary on UNIX 6th Edition, with Source Code, John Lions
Envisioning Information, Edward R. Tufte
The Visual Display of Quantitative Information, Edward R. Tufte, I think I have this book. I was disappointed.
Visual Explanations, Edward E. Tufte
The Design of the UNIX Operating System, Maurice J Bach
Algorithms in C, Robert Sedgewick Not that great. Having lot's of code samples is nice but the comments are unbelieveable. Their HUGE. Really ruins and otherwise good book. I'd rather go with the Cormen book on Algorithms and just write c from there (done just that many times actually).
Introduction to Computer Theory, Daniel I A Cohen
Compilers, Principles, Techniques and Tools (The Dragon Book), Alfred V Aho, Ravi Sethi, and Jeffrey D Ullman
The Practice of Programming Kernighan, Brian W., and Rob Pike
The Mythical Man-Month, Anniversary Edition : Essays on Software Engineering, Frederick P. Brooks
Well thanks. Now I'm ticked. The./ post made this sound very interesting taking about framing and low level down to the metal "network programming". So I followed the links. I get 5 minutes into the spec and what do I find? It's another telnet based protocol that is largely XML. I'm going to puke if I see another stupid XML networking protocol. I can't believe I'm the only one who sees how stupid it is to format all network traffic in a verbose text format like XML. Here's 448 byte "hello" example.
RedHatis aparently shipping their latest LIMBO release with UTF-8 locales as the default. This is actually pretty impressive and somewhat scary. A lot of software supports UTF-8 but it's not been proven unlike the region specific encodings like ISO-8859's. I think the big hold up was Bash but apparently that's been worked out. I don't know if this helps with Chinese though. A suspect a lot of programs meant to be used with Chinese probably do so by using a locale specific to Chinese (and not UTF-8) but I'm straining my knowledge of the topic.
Forget metric time. How about standardizing on a paper size? Everyone but the US uses A4. Does anyone realize what sort of headaches that creates? I can appreciate it and I'm from the states. We should have switched looooong ago just like we should switch to the metric system. Don't you think we look silly with our "feet" and "gallons"? Ha ha. Let's just start selling A4 printers already and phase out Letter. That's my vote.
Slashdot Mirror
The concept IBM is going for is to treat IT as another utility. Instead of some small company having to keep an expensive IT staff and maintain their own computers/network/storage, IBM says that it will do this for you.
This is exactly what management at the 60K+ financial firm I'm at would love to hear. The only problem is that IT and everything that comes with it needs to be customized to the business requirements. Are these commodity IT staff going to install every fruitcake application and know that you need INET.DLL from Win3.51 or that to restore files you use these tapes labeled that way, etc, etc. When the suits realize that they have to make investments in individuals and that IT (especially programmers) are not a commodity the recession will be over.
The fact that perl lets you create a mess may be open to debate, but it certainly doesn't mean it will be a mess.
The Second Law of Themodynamics suggest that you will be proven wrong.
I copied over MS TTF fonts and X acknowledged them but looking at a webpage in Moz with Verdana shows huge disproportionate characters. Is this a problem with Mozilla or just a problem with raw TTF/xfs?
it's only worth a 1000 words.
What would our nation's forefathers be saying today?
Nothing. They'd be too busy pissing in their pants thinking about suitcase nukes and jumbo jets crashing into buildings.
Aren't we past the point were Linux needs to be "adopted"? Is Linux an orphan that needs a home? Is Linux the red headed step child of operating systems? I think this shoud read "Linux Takes Over Chrysler (for vehicle simulations)".
X Windows Tip 151: ...etc.
Bind F3 to "previous workspace" and F4 to "next workspace". This works with most window managers (certainly does with WindowMaker). Then have all your apps and terminal windows maximized or tiled vertically. It would be hard to find a more efficient use of space. Why limit yourself to only one when you can many depending on how much video memory you have? Do an e-mail, hit F4, F4 to move over 2 workspaces, edit a source file, hit F4 again and compile and run in a separate terminal, F3 to go back and edit, F4, F3, F3 F4,
Sun StarOffice 6.0* can view and edit Microsoft Excel®, Word®, and PowerPoint® documents - and save them in their native format.
Really? If you're writing a grocery list maybe. This kind of hair-splitting is the source of "irrational exuberance".
It is extremely difficult, in fact, to breed populations whose solutions *do* make sense. They find "organic", bizarre, complicated, twisted, fragile solutions much more often than something simple and straight-forward.
Sounds just like software.
Without Rendezvous, you have several options-- all of them unappealing. You might start with some kind of application specific configuration file format
Why does it have to be application specific? Like using Rendezvous is less "specific".
and a cobbled-together system for distributing such data around to all the hosts in the system.
What's wrong with TFTP or NFS or any other network filesystem?
You might instead store such network configuration in some directory server on the network, like DNS
DNS is a directory server?
or LDAP or some kind of custom front end
So Rendezvous has a non-custom front end? That's not good. If my processes are specific to the particular problem domain (probably unique if it demands a "distributed system") then I would rather have a "front end" to match.
to some kind of evil database backend. Either way, you quickly encounter what we veterans like to call "the cache coherency problem,"
Then what do you call it? Because using a database sounds like a very good way to maintain cache coherency. You would rather state be strewn across the network?
i.e. the information in the directory must be consistent at all times with the actual state of the processes in the system.
So Rendezvous is going to guarantee state consistency between processes better than a database backend? It was compared to NetBIOS for cripes sake. I think you've made a pretty stong case against using Rendezvous for a "distributed system". My impression is that other than negotiating IP and multicast addresses a Rendezvous client is otherwise application specific.
Rendezvous is used to setup printers and wireless devices, not orchestrate large scale distributed processes.
But all I have is U.S. dollars. Oh, well.
Funny, any link for the draft you might find refers to draft-ietf-zeroconf-reqts-10.txt which just doesn't exist on the ietf site. So being real clever I flipped that 10 to an 11 and behold theres a new version dated the 26th.
r oconf-reqts-11.txt
http://www.ietf.org/internet-drafts/draft-ietf-ze
To add to the irony the Windows executable was created using Samba source code. You can tell by the dialects it submits (one of them is "Samba") and by the fact that the tree connect was not batched with the session setup.
<select name="pie">
<option value="Apple" selected>
<option value="Cherry">
<option value="Blueberry">
</select>
Mmm, blueberry.
One thing that annoys me to no end is when I install the latest version and by default some new silly screwball "IMprovement" is activated.
Like the time line numbers (:set nu) have lines under them. You have to turn that off with (hi LineNr term=bold).
In the latest release (from RH actually) they put visual line breaks (not real newlines) in lines that are longer than the terminal is wide so that it doesn't print text in the left margin. Supposed to look nice I guess. Not only is that just a silly worthless feature but now if I select a line in X with the mouse and paste it somewhere I get stairstepping text.
I think their stuggling for new ideas. I wish they would clean up that morass and make a nice X widget instead.
Every connection is stateless? What's the point in using a contianer then? You might as well us CGI.
I just setup Tomcat and Resin to develop a Filter that exposes our library to containers. Results: Tomcat is pathetically slow compared to Resin. No contest.
I may have been pretty horrible at math but I would feel very confident about walking over and plunking these guys over the head with a big rubber bat for publishing these nonsense plots. Three data points?! Are you kidding me? They obviously freehanded those dotted lines which could have been in the perpendicular direction and maintain the same statistical significance. All these graphs prove is that there are a lot more "gnulinux-users" in 1998 than there were in 1992 and 1993.
So what is this for? Would this be used to implement a Display Postscript widget or be used directly by drawing programs or something? What's the point?
First, this was not a USENET post. It's a message from the linux kernel mailing list that google is pumping into their groups.google.com database. Second, Linus is not saying he thinks Hammer is a better architecture. What he said in this message was that the current Linux page table implementation is not ideal for use with IA64 and therefore, for the sake of Linux servers everywhere, it would be better for Hammer to provail in the near to medium term future. I don't know his real position, but I would be very surprised if Linus though Hammer was a "better" architecture. X86 is an awkward instruction set that has been perpetuated by software designed to run on it. The core of these chips like Pentiums are really RISC chips with hardware wrappers to implement the X86 instructions. So it's just a waste if die space. IA64 is purer and a much better long term choice. Don't over analize a simple e-mail message from someone on lkml. These are not markedroid approved public service announcements.
Apache The Difinitive Guide from O'Rielly is a brick.
The Standard C Library,P.J. Plauger
I use this book *all* the time. It's very informative to see how a C library implementation works, and it has the important sections of the C standard reprinted with discussion. My #1 pick.
Advanced Programming in the UNIX Environment,W. Richard Stevens
The comp.unix.programmer bible. A must have for UNIX programmers.
UNIX Network Programming Networking APIs: Sockets and XTI, W. Richard Stevens
Another classic from Stevens. I have not done a tremendous amount of socket programming in C but I have extensively in Java and I am confident that I can in UNIX after having read select portions of this obviously high quaility book.
Lex & Yacc, John R. Levine
I think yacc (the GNU version is called bison) is an extreemly usefull tool. I started writing yacc grammers for all sorts of configuration files on my system while reading this the first time. Great fun, and potentially very usefull. A good yacc grammer can save tremendous amounts of time and is often the proper way to handle a problem. And I'm not recommending the book just because it's the only of it's kind. It's genuinely good.
Algorithms on Strings, Trees, and Sequences Computer Science and Computational Biology, Dan Gusfield
This is a tough book. Be prepared to ponder the presentation of each algorithm and skip the proofs (I believe you Dan). If you find yourself getting stuck in spegetti string manipulation, implement a few of these algorithms like calculating the edit distance of two sequences or Boyer-Moore. You'll be a better programmer after and there are very practical application hidden in this book. Again, the presentation is of the highest quality.
Introduction to Algorithms, Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest
I have three algorithms books. This one's the best and that's not a compermise. It's a great book.
The C Programming Language, Brian W. Kernighan, Dennis M. Ritchie
Of course.
The Art of Computer Programming (Vols 1-3), Donald E. Knuth
The C++ Programming Language (Third Edition and Special Edition), Bjarne Stroustrup
Xlib Programming Manual, Adrian Nye
Applied Cryptography: Protocols, Algorithms, and Source Code in C, 2nd Edition, Bruce Schneier
X Window System Toolkit, A Complete Programmer's Guide and Specification, Paul J. Asente, Donna Converse, Ralph R. Swick, Paul Assente
Expert C Programming, Peter van der Linden
Life with UNIX, Don Libes and Sandy Ressler
Programming Pearls, Jon Bentley
Structured Computer Organization, Andrew Tanenbaum
The Code Book, Simon Singh
Etudes for Programmers, Charles Wetherell
Secrets and Lies, Bruce Schneier
The Pragmatic Programmer, Andrew Hunt and David Thomas
TCP/IP Illustrated, Richard Stevens
Exploring Expect, Don Libes
UNIX System Administration Handbook, Evi Nemeth et al
If you use UNIX, you need this book. Forget Running Linux. This is much better and the latest edition covers Red Hat as well as BSD, Solaris, and HP/UX.
Digital Woes, Lauren Wiener
Surely You're Joking, Mr. Feynman!, Richard P. Feynman
This was a pretty funny book. Very readable for anyone. A NYT Bestseller.
Structure and Interpretation of Computer Programs, Harold Abelson, Gerald Jay Sussman, and Julie Sussman
Essentials of Programming Languages, Daniel P. Friedman, Mitchell Wand, and Christopher T. Haynes
Software Fundamentals: Collected papers, David L. Parnas, Daniel M. Hoffman and David M. Weiss
The Invisible Computer, Donald A. Norman
Lion's Commentary on UNIX 6th Edition, with Source Code, John Lions
Envisioning Information, Edward R. Tufte
The Visual Display of Quantitative Information, Edward R. Tufte,
I think I have this book. I was disappointed.
Visual Explanations, Edward E. Tufte
The Design of the UNIX Operating System, Maurice J Bach
Algorithms in C, Robert Sedgewick
Not that great. Having lot's of code samples is nice but the comments are unbelieveable. Their HUGE. Really ruins and otherwise good book. I'd rather go with the Cormen book on Algorithms and just write c from there (done just that many times actually).
Introduction to Computer Theory, Daniel I A Cohen
Compilers, Principles, Techniques and Tools (The Dragon Book), Alfred V Aho, Ravi Sethi, and Jeffrey D Ullman
The Practice of Programming
Kernighan, Brian W., and Rob Pike
The Mythical Man-Month, Anniversary Edition : Essays on Software Engineering, Frederick P. Brooks
Well thanks. Now I'm ticked. The ./ post made this sound very interesting taking about framing and low level down to the metal "network programming". So I followed the links. I get 5 minutes into the spec and what do I find? It's another telnet based protocol that is largely XML. I'm going to puke if I see another stupid XML networking protocol. I can't believe I'm the only one who sees how stupid it is to format all network traffic in a verbose text format like XML. Here's 448 byte "hello" example.
/>]]> />]]>
C: MSG 0 1 . 52 158
C: Content-Type: application/beep+xml
C:
C: <start number='1'>
C: <profile uri='http://iana.org/beep/TLS'>
C: <![CDATA[<ready
C: </profile>
C: </start>
C: END
S: RPY 0 1 . 110 121
S: Content-Type: application/beep+xml
S:
S: <profile uri='http://iana.org/beep/TLS'>
S: <![CDATA[<proceed
S: </profile>
S: END
RedHatis aparently shipping their latest LIMBO release with UTF-8 locales as the default. This is actually pretty impressive and somewhat scary. A lot of software supports UTF-8 but it's not been proven unlike the region specific encodings like ISO-8859's. I think the big hold up was Bash but apparently that's been worked out. I don't know if this helps with Chinese though. A suspect a lot of programs meant to be used with Chinese probably do so by using a locale specific to Chinese (and not UTF-8) but I'm straining my knowledge of the topic.
Forget metric time. How about standardizing on a paper size? Everyone but the US uses A4. Does anyone realize what sort of headaches that creates? I can appreciate it and I'm from the states. We should have switched looooong ago just like we should switch to the metric system. Don't you think we look silly with our "feet" and "gallons"? Ha ha. Let's just start selling A4 printers already and phase out Letter. That's my vote.