Running Zope on top of Apache or using the ArsDigita Community System are probably the best options available to a business today. The ACS would need to be hacked a bit if you don't want to use Oracle as the database; Zope comes with its own object database, and has free-as-in-speech Products for calendaring, web mail, discussion forums (Squishdot is both a real live site and the distribution point for the software running it -- try it out!).
Zope is extensible in Python. The ACS is a large package of tcl code that accesses the AOLserver API (AOLserver is now also free as in speech). Both encourage a style of programming that is more maintainable than Perl. If you knew Perl already, I strongly doubt you'd have asked your question. That's actually a good thing -- the same things that make Perl great for simple one-shots make it tough for novices to maintain. Python (and to a lesser extent, tcl) is a great deal cleaner.
I didn't mention Java or Jserv -- there is a package called JetSpeed which the Java-Apache group has put out, but my initial reaction was that it was very slow. Don't take my word for it, though -- take a look and decide for yourself.
Don't be an idiot and lock yourself into Yet Another Uncaring Vendor. You can get support for Zope or the ACS direct from the developers (Digital Creations or ArsDigita respectively). If you choose to use mod_perl and postgres, you still can get professional support. With Lotus you can look forward to servers that don't write log files, proprietary APIs, flat file "databases", and other such niceties.
This thing is a tool of the gods. You need to be reasonably comfortable with Oracle and (somewhat) with Perl, but DAMN is it fly. About the only thing it won't let you do really easily is get the query planner to barf up its plan, but that could probably be hacked in easily.
There's something called "Orac" as well, but it's in Perl/Tk rather than with a web interface, so I prefer the oracletool, myself. Simple to use, even NT people grasp it immediately.
whereas it IS for Postgres. Postgres on Linux also can run on Alpha machines, blowing away many of the painful 32-bit architectural restrictions. 6.5.2 is much, much better than older versions, but I'm not 100% sure that I'd trust it for a mission-critical system yet.
Oracle, on the other hand, is developed on Solaris and then ported to other platforms. So if you're going to run a huge Oracle installation, do it on Sun hardware. Use raw I/O (can't do this yet under Linux) and stick the (sievelike) Solaris boxes behind a firewall, for god's sake. But please don't be a martyr and run a big Oracle installation on Linux... yet.
Don't forget that Oracle is and will remain a complex beast. My personal viewpoint is that Oracle is the C++ of databases -- highly flexible, can be outrageously fast, but you MUST know what you are doing. Hire a professional DBA if you don't... a big installation like you describe will rapidly spin out of control unless someone skilled is around to manage it.
OpenBSD is great for firewalls, DNS servers, and other applications where you *really* do NOT want any chance of an intrusion. It is heavily audited, tightly locked down by default, and perfect for these applications (both because of the security and because of the license).
It isn't meant to replace Linux or FreeBSD, it can't even run on multiprocessors yet (AFAIK)! But it does its stated job very well.
Contrast this with NT or Linux and you'll better understand this. You don't tell the CIO or CTO "I'm setting up a Linux firewall", you say "I'm setting up the most secure firewall on the market" and they gratefully fuck off.
Maybe the hypothetical ISP should have considered this. Most of IBM's internal network runs on free software because security and IGS can sift through the code.
Maybe ISPs are in a competitive enough environment that a bad decision like that is enough to kill one. What do you think?
IBM has simply chosen to follow a path of enlightened self-interest. They built Jikes on their own dime, built OpenDX on their own dime (yes, I know full well that it was mostly to sell SP2's back in 1994, but they opened it and freely licensed the patents on the code), built a Linux JVM on their own dime, and support the Perl DBI interface on their own dime.
GCC 2.95 (with libgcj) now has support for compiling (non-graphical) Java code to native binaries. If you had hoped Jikes would do this, well, look into GCJ at
"Absurdly complex" appears to be quantifiable when one OS has something like 20 million lines of code and the other something on the order of 2 million.
One advantage Linux has is that it is relatively easy for a competent user to configure it the way he/she wants to. This appears to be much more difficult under NT. The "lots of little tools" philosophy isn't there -- a complex aggregate which cannot be broken down into simpler pieces is harder to understand and analyze than one that can.
In any event, anything worth doing is usually pretty tough. There's no competitive advantage in offering a service Just Like Everyone Else's, and doing easy, fully understood things isn't much fun. This goes far beyond OSes and webservers.
Perl is a lot faster to develop in than Java. Really. I've done complicated stuff in both. When you really push the envelope, Java's standard API abstraction barriers don't go far enough.
JSPs are okay, but so is embedded Perl or VBasic.
Cocoon is very unwieldy right now, and probably not yet worth the effort unless you are a huge shop with lots of Java and XML devotees. The examples assume an Oracle backend. That should tell you something. (Yes, I've played with Cocoon as well. Cocoon 2 may inded be revolutionary...)
I have never heard of Midgard and I've been around for a while now. This indicates to me that it could be difficult to hire someone quickly and get them up to speed on it. OTOH, if it's just like PHP only better, I could be quite wrong.
Just my experiences and thoughts.
Personally, I don't mind servlets, mod_perl, OR mod_php. They're popular for different things, so for maximum code reuse (laziness) I choose whichever gets the job done fastest.
(Zope, incidentally, is quite cool and OO, and would be great for a new installation, but too much of a transition IMHO for a preexisting site)
somebody moderate this idiot down
on
SGI Releases IDE
·
· Score: 0
At least FreeBSD, OpenBSD, and NetBSD can compete on their merits, instead of by buying/crushing/smearing the competition.
I love Linux, and I'm setting up my firewall/NAT box at home on FreeBSD. Why? Because less hackers are familiar with it, I have loads of good tuning information on it, and it works great for most of our production servers at work. Not that we don't have about 100 of each FreeBSD and Linux. And not that I don't run Linux on my laptop. I just feel better about having a firewall at home that is built on an "old-school" type of distro.
Competition is the inevitable result of scarce resources. Dvorak is being an idiot; the BSDers aren't. They have a good NOS and it shows.
Some psychopath wrote a BFS module, and announced it. Whether it works or not is another question... BFS seems fairly logical and is quite well-documented, so it was only a matter of time.
Researchers at Cornell University have demonstrated an NT-based cluster computer, the NT-O-Wulf, which offers some of the performance and some of the stability of its Linux counterpart for institutions interested in purchasing expensive commercial ports of freely available message passing software.
Robert Constable, Cornell's new dean for Microsoft services (CMS) believes that AC3 and AC3 Velocity will be valuable assets in his strategic vision for whoring out undergraduate research assistants and underpaid faculty to offer free R&D to giant corporations. "Windows 2000 and Intel architecture-based cluster computing are the direction of the future," he said. "The Theory Center is an integral part of CIS and we're delighted that they are taking a leadership position in this arena."
If you're not interested in using NFS behind a firewall (just because it's slow, insecure, ugly, unreliable, fault-intolerant, and buggy doesn't make it bad, right?) you might be interested in PVFS. It sounds like what you are looking for, and under the "Files" link are the sources.
Personally, I like the "adventure" of Coda, but haven't tried setting it up in a few months. Now that my roommates have agreed to be guinea pigs for the Windows client, I figure I'll set it up behind my NAT box and play with it again. It's overkill for everything but a big installation, but I still think it's kind of fun. The thought that terrifies me is working with a multi-GB datafile or such over Coda -- but since my roommates will probably be more interested in playing Dopewars and moving around small files on a FE network, I'm going ahead with the grand master plan anyways. Besides, I have a laser printer and a burning desire to experience the frustrations of Samba...
Microsoft is legendary for their anticompetitive practices. In the situations I have seen, they are/delighted/ to leave researchers twisting in the wind once the benefits to Microsoft start to decrease. That's not charity. That's exploitation. And the administrative staff of Generic State University is typically a bunch of burnout slackers who actually believe in a free lunch, so they merrily march into this trap. "TCO through the roof? Too bad, so sad. Hire some MCSEs!"
Microsoft will continue to arouse suspicion in every aspect of its practices until there is substantial evidence that they are doing something besides raping the intellectual capital of the world.
I wouldn't wish an "education" from many of these universities on my worst enemy. I was extremely lucky to have recieved the exposure I did to UNIX without being a CS student, although even the professors who do cooperate with MS at Cornell routinely lambaste the company and its products.
I wonder if students get the message -- "sure, we'll take their money, but the product still sucks". You'd have to be pretty thick not to.
A friend asked me whether this was good or bad. I wasn't really sure -- maybe Microsoft wants to roll it into Office or Visual Whatever or something, maybe they'll fuck it up, who knows.
But one thing is for sure. Dia, for any Unix that can build GTK+, is a great tool and in combination with LyX can produce truly professional results (embedding the dia.eps output into typeset LaTeX for that FrameMaker look) for free with little hassle. The architecture is meant to allow Visio-like plugins for transforming SQL to ER diagrams and vice versa, although AFAIK no one has implemented it (I started to look into it and got distracted).
There are some problems, but overall, I like my TP600E a lot. It's fast, light, and works OK.
I'm running
1) oracle, postgres, mysql 2) apache, apache 2.0, zope 2.0 3) OpenDX, VTK, and VMD
on a puny little laptop. That's kind of cool. Next week, fun with vmWare (need to port the Dopewars client to NT). It's a bloody *laptop*... all I really want is SOUND, which *is* a bit of a bitch. But my coworker got sound on his TP570, and I'm pretty close, so I'm hoping this certification nonsense will help me play my CDs through my laptop.
Suspend is fucked up. apmd will hose up and force a dirty shutdown if you suspend and resume. Keep your windows/FAT partition around (well, at least enough to dump the contents of memory to, in my case 192MB + a little cushion for bookkeeping) to suspend to. Maybe IBM will fix this niggly too.
CardBus cards suck. Sell yours and get a Xircom RealPort type III card before your X-jack breaks off. The Xircom card I have works like a charm and you don't have to unplug/re-plug it after a suspend. If you insist on using the 3Com card (don't say I didn't warn you), go into/etc/pcmcia/config.opts and change this line:
include memory 0xc0000-0xfffff, memory 0xa0000000-0xa0ffffff
to
include memory 0xc0000-0xfffff, memory 0x60000000-0x60ffffff
and do an insmod 3c575_cb.o (or whatever you have) to force it. Then ping somebody. DHCP appears to be fuct for some reason with this card.
I can't think of anything else. X configuration was a drag, but if you read the NeoMagic README it all becomes clear. I assume if you're reading Slashdot, that you already discovered this.
I want 32-bit color (any resolution) on this thing and would gladly pay up to $250 for the upgrade.
I can't think of anything else at the moment, but when I get sound working on *my* laptop I will be happy to gloat about it;-).
Keep plugging away at it. I wouldn't trade my Thinkpad for anything now that it works for me.
Read the fucking Linux Kernel Mailing List FAQ before shooting your mouth off. And if you're going to dispute it, benchmark it in a reasonable context and prove your claim.
is too hard? I almost bought a copy of Code Fusion and then said "naaah, what for?" when I realized I could do everything but the source navigation just using GCC 2.95.1, GDB 4.18, DDD 3.16, and XEmacs 21.4. Seeing as to how I already use all of these in earlier versions, it wasn't real hard to transition to writing Java->native code, my only problem now is getting DDD to work right with the executables. But DDD also lets one debug Perl and Python scripts. Can't do that with Code Fusion!
I was unable to find any mention of what OS was used for these benchmarks. Realistically, only SCO or Linux seems probable (since the application in question is Oracle8), and I do not know whether more recent in-house tuning has allowed Linux to scale past 4 CPUs on x86 machines. But as of a few months ago, NT and Linux both went from near-linear scaling to a near-plateau beyond 4 CPUs, so I am wondering if they didn't use SCO.
Slashdot Mirror
Maybe now they'll use autoconf and automake. GRASS was a nightmare to install when I did it for my former boss.
Yes, yes, and yes. However, someone just picking up the language is unlikely to.
Python and Java encourage encapsulation. Tcl... well, let's not go there.
Zope is extensible in Python. The ACS is a large package of tcl code that accesses the AOLserver API (AOLserver is now also free as in speech). Both encourage a style of programming that is more maintainable than Perl. If you knew Perl already, I strongly doubt you'd have asked your question. That's actually a good thing -- the same things that make Perl great for simple one-shots make it tough for novices to maintain. Python (and to a lesser extent, tcl) is a great deal cleaner.
I didn't mention Java or Jserv -- there is a package called JetSpeed which the Java-Apache group has put out, but my initial reaction was that it was very slow. Don't take my word for it, though -- take a look and decide for yourself.
Don't be an idiot and lock yourself into Yet Another Uncaring Vendor. You can get support for Zope or the ACS direct from the developers (Digital Creations or ArsDigita respectively). If you choose to use mod_perl and postgres, you still can get professional support. With Lotus you can look forward to servers that don't write log files, proprietary APIs, flat file "databases", and other such niceties.
Don't buy into it.
http://www.oracletool.com/
Yeah, it's free as in speech and beer, too.
There's something called "Orac" as well, but it's in Perl/Tk rather than with a web interface, so I prefer the oracletool, myself. Simple to use, even NT people grasp it immediately.
Good stuff!
> Of course, IBM is just finally recognizing that
> ANY *nix is better than forcing customers to use
> AIX
I see you've never used a SCO flavor.
You'll pine for smitty if you ever do...
whereas it IS for Postgres. Postgres on Linux also can run on Alpha machines, blowing away many of the painful 32-bit architectural restrictions. 6.5.2 is much, much better than older versions, but I'm not 100% sure that I'd trust it for a mission-critical system yet.
Oracle, on the other hand, is developed on Solaris and then ported to other platforms. So if you're going to run a huge Oracle installation, do it on Sun hardware. Use raw I/O (can't do this yet under Linux) and stick the (sievelike) Solaris boxes behind a firewall, for god's sake. But please don't be a martyr and run a big Oracle installation on Linux... yet.
Don't forget that Oracle is and will remain a complex beast. My personal viewpoint is that Oracle is the C++ of databases -- highly flexible, can be outrageously fast, but you MUST know what you are doing. Hire a professional DBA if you don't... a big installation like you describe will rapidly spin out of control unless someone skilled is around to manage it.
OpenBSD is great for firewalls, DNS servers, and other applications where you *really* do NOT want any chance of an intrusion. It is heavily audited, tightly locked down by default, and perfect for these applications (both because of the security and because of the license).
It isn't meant to replace Linux or FreeBSD, it can't even run on multiprocessors yet (AFAIK)! But it does its stated job very well.
Contrast this with NT or Linux and you'll better understand this. You don't tell the CIO or CTO "I'm setting up a Linux firewall", you say "I'm setting up the most secure firewall on the market" and they gratefully fuck off.
It's sure hard to audit proprietary crap...
Maybe the hypothetical ISP should have considered this. Most of IBM's internal network runs on free software because security and IGS can sift through the code.
Maybe ISPs are in a competitive enough environment that a bad decision like that is enough to kill one. What do you think?
One of our dev servers blew up in the middle of that comment. Lost track of what I was saying. Oops.
IBM has simply chosen to follow a path of enlightened self-interest. They built Jikes on their own dime, built OpenDX on their own dime (yes, I know full well that it was mostly to sell SP2's back in 1994, but they opened it and freely licensed the patents on the code), built a Linux JVM on their own dime, and support the Perl DBI interface on their own dime.
It's the customers, stupid. They asked for this.
GCC 2.95 (with libgcj) now has support for compiling (non-graphical) Java code to native binaries. If you had hoped Jikes would do this, well, look into GCJ at
Cygnus
and try using it. The binaries are indeed much faster than loading up a JVM, parsing bytecode, running it...
Jikes, on the other hand, compiles Java to bytecode much faster than javac. On the order of 10 to 100 *TIMES* faster. Use it instead of javac.
but just more worthless speculation.
"Absurdly complex" appears to be quantifiable when one OS has something like 20 million lines of code and the other something on the order of 2 million.
One advantage Linux has is that it is relatively easy for a competent user to configure it the way he/she wants to. This appears to be much more difficult under NT. The "lots of little tools" philosophy isn't there -- a complex aggregate which cannot be broken down into simpler pieces is harder to understand and analyze than one that can.
In any event, anything worth doing is usually pretty tough. There's no competitive advantage in offering a service Just Like Everyone Else's, and doing easy, fully understood things isn't much fun. This goes far beyond OSes and webservers.
/Life/ is absurdly complex. Get used to it.
Perl is a lot faster to develop in than Java. Really. I've done complicated stuff in both. When you really push the envelope, Java's standard API abstraction barriers don't go far enough.
JSPs are okay, but so is embedded Perl or VBasic.
Cocoon is very unwieldy right now, and probably not yet worth the effort unless you are a huge shop with lots of Java and XML devotees. The examples assume an Oracle backend. That should tell you something. (Yes, I've played with Cocoon as well. Cocoon 2 may inded be revolutionary...)
I have never heard of Midgard and I've been around for a while now. This indicates to me that it could be difficult to hire someone quickly and get them up to speed on it. OTOH, if it's just like PHP only better, I could be quite wrong.
Just my experiences and thoughts.
Personally, I don't mind servlets, mod_perl, OR mod_php. They're popular for different things, so for maximum code reuse (laziness) I choose whichever gets the job done fastest.
(Zope, incidentally, is quite cool and OO, and would be great for a new installation, but too much of a transition IMHO for a preexisting site)
This is clearly flamebait.
At least FreeBSD, OpenBSD, and NetBSD can compete on their merits, instead of by buying/crushing/smearing the competition.
I love Linux, and I'm setting up my firewall/NAT box at home on FreeBSD. Why? Because less hackers are familiar with it, I have loads of good tuning information on it, and it works great for most of our production servers at work. Not that we don't have about 100 of each FreeBSD and Linux. And not that I don't run Linux on my laptop. I just feel better about having a firewall at home that is built on an "old-school" type of distro.
Competition is the inevitable result of scarce resources. Dvorak is being an idiot; the BSDers aren't. They have a good NOS and it shows.
Some psychopath wrote a BFS module, and announced it. Whether it works or not is another question... BFS seems fairly logical and is quite well-documented, so it was only a matter of time.
Cornell duplicates Los Alamos innovation after 3 years of brutal hacking
Researchers at Cornell University have demonstrated an NT-based cluster computer, the NT-O-Wulf, which offers some of the performance and some of the stability of its Linux counterpart for institutions interested in purchasing expensive commercial ports of freely available message passing software.
Robert Constable, Cornell's new dean for Microsoft services (CMS) believes that AC3 and AC3 Velocity will be valuable assets in his strategic vision for whoring out undergraduate research assistants and underpaid faculty to offer free R&D to giant corporations. "Windows 2000 and Intel architecture-based cluster computing are the direction of the future," he said. "The Theory Center is an integral part of CIS and we're delighted that they are taking a leadership position in this arena."
If you're not interested in using NFS behind a firewall (just because it's slow, insecure, ugly, unreliable, fault-intolerant, and buggy doesn't make it bad, right?) you might be interested in PVFS. It sounds like what you are looking for, and under the "Files" link are the sources.
Personally, I like the "adventure" of Coda, but haven't tried setting it up in a few months. Now that my roommates have agreed to be guinea pigs for the Windows client, I figure I'll set it up behind my NAT box and play with it again. It's overkill for everything but a big installation, but I still think it's kind of fun. The thought that terrifies me is working with a multi-GB datafile or such over Coda -- but since my roommates will probably be more interested in playing Dopewars and moving around small files on a FE network, I'm going ahead with the grand master plan anyways. Besides, I have a laser printer and a burning desire to experience the frustrations of Samba...
Microsoft is legendary for their anticompetitive practices. In the situations I have seen, they are /delighted/ to leave researchers twisting in the wind once the benefits to Microsoft start to decrease. That's not charity. That's exploitation. And the administrative staff of Generic State University is typically a bunch of burnout slackers who actually believe in a free lunch, so they merrily march into this trap. "TCO through the roof? Too bad, so sad. Hire some MCSEs!"
Microsoft will continue to arouse suspicion in every aspect of its practices until there is substantial evidence that they are doing something besides raping the intellectual capital of the world.
Don't hold your breath.
Yah, Solaris is a real fortress.
"Who's this guy Bob?"
I wouldn't wish an "education" from many of these universities on my worst enemy. I was extremely lucky to have recieved the exposure I did to UNIX without being a CS student, although even the professors who do cooperate with MS at Cornell routinely lambaste the company and its products.
I wonder if students get the message -- "sure, we'll take their money, but the product still sucks". You'd have to be pretty thick not to.
Solaris and Irix secure... god damn that's funny.
A friend asked me whether this was good or bad. I wasn't really sure -- maybe Microsoft wants to roll it into Office or Visual Whatever or something, maybe they'll fuck it up, who knows.
.eps output into typeset LaTeX for that FrameMaker look) for free with little hassle. The architecture is meant to allow Visio-like plugins for transforming SQL to ER diagrams and vice versa, although AFAIK no one has implemented it (I started to look into it and got distracted).
But one thing is for sure. Dia, for any Unix that can build GTK+, is a great tool and in combination with LyX can produce truly professional results (embedding the dia
Check it out:
Dia, a diagram creation program
There are some problems, but overall, I like my TP600E a lot. It's fast, light, and works OK.
/etc/pcmcia/config.opts and change this line:
;-).
I'm running
1) oracle, postgres, mysql
2) apache, apache 2.0, zope 2.0
3) OpenDX, VTK, and VMD
on a puny little laptop. That's kind of cool. Next week, fun with vmWare (need to port the Dopewars client to NT). It's a bloody *laptop*... all I really want is SOUND, which *is* a bit of a bitch. But my coworker got sound on his TP570, and I'm pretty close, so I'm hoping this certification nonsense will help me play my CDs through my laptop.
Suspend is fucked up. apmd will hose up and force a dirty shutdown if you suspend and resume. Keep your windows/FAT partition around (well, at least enough to dump the contents of memory to, in my case 192MB + a little cushion for bookkeeping) to suspend to. Maybe IBM will fix this niggly too.
CardBus cards suck. Sell yours and get a Xircom RealPort type III card before your X-jack breaks off. The Xircom card I have works like a charm and you don't have to unplug/re-plug it after a suspend. If you insist on using the 3Com card (don't say I didn't warn you), go into
include memory 0xc0000-0xfffff, memory 0xa0000000-0xa0ffffff
to
include memory 0xc0000-0xfffff, memory 0x60000000-0x60ffffff
and do an insmod 3c575_cb.o (or whatever you have) to force it. Then ping somebody. DHCP appears to be fuct for some reason with this card.
Read this article:
Installing Debian on a Thinkpad 600E
I can't think of anything else. X configuration was a drag, but if you read the NeoMagic README it all becomes clear. I assume if you're reading Slashdot, that you already discovered this.
I want 32-bit color (any resolution) on this thing and would gladly pay up to $250 for the upgrade.
I can't think of anything else at the moment, but when I get sound working on *my* laptop I will be happy to gloat about it
Keep plugging away at it. I wouldn't trade my Thinkpad for anything now that it works for me.
Read the fucking Linux Kernel Mailing List FAQ before shooting your mouth off. And if you're going to dispute it, benchmark it in a reasonable context and prove your claim.
What, changing DDD to fire up
gnuclient +@LINE@ @FILE@
instead of
vi +@LINE@ @FILE@
is too hard? I almost bought a copy of Code Fusion and then said "naaah, what for?" when I realized I could do everything but the source navigation just using GCC 2.95.1, GDB 4.18, DDD 3.16, and XEmacs 21.4. Seeing as to how I already use all of these in earlier versions, it wasn't real hard to transition to writing Java->native code, my only problem now is getting DDD to work right with the executables. But DDD also lets one debug Perl and Python scripts. Can't do that with Code Fusion!
I was unable to find any mention of what OS was used for these benchmarks. Realistically, only SCO or Linux seems probable (since the application in question is Oracle8), and I do not know whether more recent in-house tuning has allowed Linux to scale past 4 CPUs on x86 machines. But as of a few months ago, NT and Linux both went from near-linear scaling to a near-plateau beyond 4 CPUs, so I am wondering if they didn't use SCO.
Regardless, it's odd that they don't say.