I suspect that which application server you want depends a lot on the application you want to build. I've had a look at Zope, and found it a bit of a pain because the actual development of the web site was too web-oriented. (I want to keep everything in files and keep these under revision control.) As well, it wasn't really optimised for the programming side of building applications.
What works much better for me is Enhydra, which is very good in terms of developing Java code, and lets you leave plain stuff (html files, images) just as it is.
However, as I said, I suspect which will be better for you depends a lot on what you're trying to build.
The NetBSD folks have put together a FAQ about the Multia, which includes information on a part that needs to be replaced to help avoid heat death.
Note that this is a fairly slow machine, mostly due to the very slow memory controller in it; DRAM access is about one half to one third the speed of a Pentium at the same clock speed.
In terms of disk and stuff like that, if you get one just to play with cheaply, you can just net boot it (at least under NetBSD).
...we fucking wrote the software that Red Hat sells. We own the company in a far more real sense than any of the moneyed lords with sufficient "liquid net worth" to take part in the IPO. They're auctioning my software off on the New York Stock Exchange to the highest bidder, and I can't take part!
Sorry to be a bore here, but this is sort of comment makes me think that E-Trade is right in their rejections of people based on low net worth and lack of investment experience. After all, here we have a fellow who's been coding for years and quite consciously putting his code under the GPL, so that anybody can use it for any purpose (so long as they give away the source). Now he sees a situation where his code is being used to make a pile of money on an IPO (or so he thinks) and he can't get in on it. Though this is a natural (and should be an expected) consequence of his decision to make his code freely available, he's complaining vociferously after the fact about something that's entirely his own doing. If he didn't want this to happen, he didn't have to free his code! Red Hat, E-Trade, and the SEC don't want to avoid exactly this should the stock plunge after the IPO.
The problem is, all improvements these vendors made they can keep for themselves, and so the commom code base suffers
It suffers no more than GPL'd code does, for which the improvements aren't written in the first place because the company didn't want to give away the code. In both cases, there's no free code added. So what's the difference?
As for Darwin, Apple has contributed all sorts of code and fixes back in to the NetBSD tree. So we certainly benefitted from this. What has Linux got from Darwin? Nothing.
cjs
Re:Hot swap != big deal, multiproc & linux
on
SGI's Linux Server
·
· Score: 2
the hardware must handle it, I mean, you must be able to add/remove a drive to a IDE bus/SCSI chain without everything going mad. Most hardware I met doesn't care if you remove a (umount'ed) drive.
Well, if you want to do this reliably, you want a connector that's designed to deal with this properly. The standard SCSI SCA connector is a good example: It's not possible to misalign, there are no pins to bend, and the connectors mate in the appropriate order (ground first, then data, then power). This also requires some software support, since you really want to be doing this on a quiescent bus.
If you want really good performance/reliability, go for RAID-5. Hardware support is not required anymore.
Well, it's not required if you don't want really good performance and reliability with your RAID-5.:-) To do it properly, you need an external unit (such as a Baydel) that has a separate SCSI bus for each drive, so that a single drive can't lock the bus. Then you need two controllers on the Baydell, one for each of the two controllers on your host. Now you can lose any of the following and still stay up:
a drive
a RAID controller
a SCSI cable
a SCSI controller on the host
Of course, on many machines you're still prone to motherboard failure at this point, which is why you're probably hooking this up to a Sun E-series box. Make sure the two SCSI controllers that the RAID box is hooked up to are on different I/O boards, so that if one I/O board dies, the other is still there. And of course make sure you have at least two separate CPU/Memory boards, so if one of those dies, the other is available.
E-Trade probably doesn't have much choice in the matter; all of this is due to SEC regulations, not E-Trade. As a broker, all sorts of nasty things can happen to you if an inexperienced client loses his shirt.
...I'd appreciate it if someone who is more knowledgeable than I would tell me how often developments in FreeBSD are integrated into the development tree of NetBSD, for example.
All the time. There's a massive amount of code trading back and forth. And this extends to the userland, too; if we add new features to daemons or whatever, they tend to go back and forth. That's something that the Linux distros, since they don't share userland source to the same degree, don't get as much of.
The other main advantage that Linux has for me is the applications that are being ported to Linux more and more.
So just install a set of linux libs and whatnot under/emul/linux (a package is available, so it's as simple as a single pkg_add command) and run the Linux programs. I use Linux stuff all the time. My Netscape Communicator and RealAudio player are Linux versions, for example.
All respect aside for the guys who figured out the clean solution, but a 1542 is a worthless piece of shit anyway, and it certainly isn't worth waiting two YEARS to get a driver for.
It's a fine piece of ISA hardware, as ISA hardware goes. And the driver for it existed already, it just didn't deal with systems with more than 16 MB of RAM. And you may think it's worthless, but for some people it's better than buying a new SCSI controller, and perhaps a new motherboard. If you're going to go down that route, much of the vast array of hardware that NetBSD runs on is considered `worthless' by someone. I really like my Sun 3/60s, even though many people think of them as `worthless.'
Bounce buffers aren't really what you want on Alpha anyway. The question is, can you do DMA to arbitrary 64-bit memory to and from a 32-bit PCI device on an Alpha?
I never said we used bounce buffers on the alpha. Yes, we can do arbitrary DMA to anywhere in memory that hardware will support, on any machine that NetBSD runs on. If this capability isn't available, we have to use bounce buffers in some circumstances. The point is, the driver for that specific piece of hardware neither knows nor cares exactly how the DMA'd data gets to a particular point in memory. The OS takes care of it all, on all platforms.
I think you have to realize he is more interested in the freedom of _users_ than the freedom of programmers.
Hm. That makes what he does make a bit more sense, I suppose, though I find it difficult to understand why he's more concerned with users' freedom than programmers'. After all, the vast majority of users couldn't give a d--- whether or not the program source is available. And, in fact, sometimes users are better served by closed-source systems; it's better to have a closed-source piece of software that does what you want than no software at all. And there are situations where closed-source works better economically for the entity making the software.
BTW, i wonder if the problem of relocating real 32 bit adresser under a > 32 bit plataform (as requiered to talk with 32 bit pci devices taht do bus mastering) would be much of a problem. I wonder if the people that has worked on NetBSD/Alpha has got a look into it.Perhaps the "bounce buffers" from the ISA days might come back....
In fact, this was originally dealt with for the NetBSD/alpha port, and extended to all other ports in a machine-independent way. Drivers no longer have to worry about what memory is DMA-able, or whether a card can access the entire address space; underlying machine-dependent software takes care of that. (The underlying software may use special DMA registers, bounce buffers, or whatever it needs.) Therefore the same driver works for a card whether it's in an Alpha, an i386, an Amiga, or whatever. (NetBSD is, as far as I know, the only OS that does this.)
Have wou NETBSDers discusser if it would be worthwile to implement a PSE36 port...
I don't know what this piece of hardware is, unfortunately.
On a 32 bit plataform, it is doubtfull if it makes any sense to have more than 4 GB, as this forces the os and apps that mght want to use it to do some segment translation
This is not necessarially true. I'm not familiar with just how the P6 MMU works with this, but the programs themselves, so long as they're happy with a 32-bit address space, work entirely in a VM map and have no clue about the amount of physical RAM in the machine. The OS has to be able to provide 36-bit physical addresses to the MMU to map these physical pages into an address space, but given that the OS usually is happy in a fairly small chunk of address space (usually 256 MB to 1 GB at the `top' of every process's address space), it doesn't need to deal with segmentation for itself.
The implication of this, of course, is that more than 4 GB of memory will be useful only if you have several processes that want a lot of physical memory, as an individual process is never going to use more than 4 GB without segmentation support.
How about, `a damn nice thing from the point of view of software consumers'?
As an example, I used to use a Telebit router. It had TCL built in, and thus had a far better and more powerful scripting language than it would likely have had otherwise. This made my life a lot easier. Had TCL been GPL'd or not available at all, there's little chance that Telebit would have spent the kind of effort (and money) necessary to make a scripting language that powerful for just this one purpose.
'd say it matters to the end user whether I write the code or not. I'd not write it if the BSD license was my only protection.
Just out of curiousity, why wouldn't you write it? It seems to me that if you don't write it, it adds nothing to the world, but if you do write it, even if you don't force someone else's code public, you've added something to the world. Why should it be so bothersome that someone else might release a different version of your program without source when your original program is still out there with source?
I wasn't sure why anyone would want to work hard producing code just so somebody else could tweak it a bit and *not* release the source back for others to use.
Because, for some of us, we want to give the users of our software maximum freedom.
In some situations it's simply more convenient not to have to deal with the restrictions of the GPL. If I can help people more by keeping them out of licensing goo, I'll go that route. I write software to make the world a better place for everyone, not to push some political agenda on others.
What RMS is doing is often not good, and certainly not necessary. One of his explicit goals is to rid the world not only of proprietary software, but also free software that is not GPL'd. He comes right out and says this when he explains why the readline library is GPL'd rather than LGPL'd.
I guess people replying didn't really read what I wrote. I said ``[RAID 0 is] quite common in many shops that need high reliability, because you don't use software RAID or internal RAID controller cards in such systems. You use external RAID boxes.'' [Emphasis added.]
External hardware RAID is a heck of a lot more reliable, and usually faster than, software RAID and in-box RAID controller cards. (A typical setup has two SCSI controllers in the host, each hooked to one of the two controllers in the RAID box, so you can lose a host controller, a cable or a RAID box controller without going down.)
Care to look at the development history for the *BSD projects? *None* of them are anywhere near 20 years old.
The original BSD project was started in the late seventies. I started using 4.2BSD on a VAX 11/780 in 1982. Then I was told that System III (remember System III?) was the real thing, and BSD was going to vanish. Then when the extent of that failure became apparent, System V was the real thing, and BSD was going to vanish. Then NT came out, and all Unices were going to vanish. Then the Berkely project wound down, and BSD was going to vanish. Then Jolitz never got out another release, and BSD was going to vanish. Then Linux came along and BSD was going to vanish. You'll excuse me if I'm getting a little skeptical about these predictions at this point.
You should learn a little more about the history of Unix before you start talking about how old BSD is. Peter Salus has a good book on it.
But then they would have become censors. Censoring content just doesn't seem very Harvard to me.
Removal of a site is also censorship.
Uh...neither of these are censorship. Censorship is when someone is prevented from saying something. Saying `I'm not going to host your web site or this particular bit of content' in no way stops the content owner from posting his content elsewhere.
Censorship is a serious issue, and it still does exist in modern democracies today. (There's nothing like the threat of a huge libel suit to silence someone without the resources to defend himself.) Please don't trivialise it by applying it to things that are not.
FreeBSD has very old, clumsy code. Matthew Dillon (FreeBSD coder) put it so....
Having read substantial portions of both the NetBSD and Linux networking code, I have to say that, even though it's much older, the BSD code is much better designed. But feel free to quote code here and prove me wrong....
Solaris scales up to 64 processors (in reality). NT, the New Technology OS, scales up to 16 processors (in marketing). Solaris is so much "worse," because it wasn't designed with SMP in mind.
Huh?!
SunOS 5 (the OS component of Solaris 2.x) certainly was built from the ground up with SMP support in mind. They knew they'd need it; SunOS 4 had MP support before SunOS 5 was even thought of.
The need comes from differing needs in the user and developer groups. A quite typical example is `bounce buffers' to let you use an Adaptec 1542 in a machine with more than 16 MB of RAM. NetBSD lived without this capability for almost two years longer than Linux, until a clean, general, platform-independent solution was implemented. This has advantages (NetBSD is still the only system in which you can put a 1542 in an Alpha and have it work), but there are a lot of users out there who didn't care how ugly the hack was, they wanted that 1542 working right away. They got what they needed in Linux and FreeBSD; the purists got what they wanted in NetBSD. Now, in your proposed merger, which user group were you going to tell to get lost?
Slashdot Mirror
I suspect that which application server you want depends a lot on the application you want to build. I've had a look at Zope, and found it a bit of a pain because the actual development of the web site was too web-oriented. (I want to keep everything in files and keep these under revision control.) As well, it wasn't really optimised for the programming side of building applications.
What works much better for me is Enhydra, which is very good in terms of developing Java code, and lets you leave plain stuff (html files, images) just as it is.
However, as I said, I suspect which will be better for you depends a lot on what you're trying to build.
cjs
The NetBSD folks have put together a FAQ about the Multia, which includes information on a part that needs to be replaced to help avoid heat death.
Note that this is a fairly slow machine, mostly due to the very slow memory controller in it; DRAM access is about one half to one third the speed of a Pentium at the same clock speed.
In terms of disk and stuff like that, if you get one just to play with cheaply, you can just net boot it (at least under NetBSD).
cjs
cjs
As for Darwin, Apple has contributed all sorts of code and fixes back in to the NetBSD tree. So we certainly benefitted from this. What has Linux got from Darwin? Nothing.
cjs
- a drive
- a RAID controller
- a SCSI cable
- a SCSI controller on the host
Of course, on many machines you're still prone to motherboard failure at this point, which is why you're probably hooking this up to a Sun E-series box. Make sure the two SCSI controllers that the RAID box is hooked up to are on different I/O boards, so that if one I/O board dies, the other is still there. And of course make sure you have at least two separate CPU/Memory boards, so if one of those dies, the other is available.cjs
E-Trade probably doesn't have much choice in the matter; all of this is due to SEC regulations, not E-Trade. As a broker, all sorts of nasty things can happen to you if an inexperienced client loses his shirt.
cjs
Personally, I find that the more innocious sites that are blocked by CyberSitter, the better. That's the best way to discourage people from using it.
cjs
cjs
cjs
cjs
cjs
cjs
The implication of this, of course, is that more than 4 GB of memory will be useful only if you have several processes that want a lot of physical memory, as an individual process is never going to use more than 4 GB without segmentation support.
cjs
cjs
How about, `a damn nice thing from the point of view of software consumers'?
As an example, I used to use a Telebit router. It had TCL built in, and thus had a far better and more powerful scripting language than it would likely have had otherwise. This made my life a lot easier. Had TCL been GPL'd or not available at all, there's little chance that Telebit would have spent the kind of effort (and money) necessary to make a scripting language that powerful for just this one purpose.
cjs
cjs
In some situations it's simply more convenient not to have to deal with the restrictions of the GPL. If I can help people more by keeping them out of licensing goo, I'll go that route. I write software to make the world a better place for everyone, not to push some political agenda on others.
cjs
What RMS is doing is often not good, and certainly not necessary. One of his explicit goals is to rid the world not only of proprietary software, but also free software that is not GPL'd. He comes right out and says this when he explains why the readline library is GPL'd rather than LGPL'd.
cjs
I guess people replying didn't really read what I wrote. I said ``[RAID 0 is] quite common in many shops that need high reliability, because you don't use software RAID or internal RAID controller cards in such systems. You use external RAID boxes.'' [Emphasis added.]
External hardware RAID is a heck of a lot more reliable, and usually faster than, software RAID and in-box RAID controller cards. (A typical setup has two SCSI controllers in the host, each hooked to one of the two controllers in the RAID box, so you can lose a host controller, a cable or a RAID box controller without going down.)
cjs
You should learn a little more about the history of Unix before you start talking about how old BSD is. Peter Salus has a good book on it.
cjs
Censorship is a serious issue, and it still does exist in modern democracies today. (There's nothing like the threat of a huge libel suit to silence someone without the resources to defend himself.) Please don't trivialise it by applying it to things that are not.
cjs
cjs
SunOS 5 (the OS component of Solaris 2.x) certainly was built from the ground up with SMP support in mind. They knew they'd need it; SunOS 4 had MP support before SunOS 5 was even thought of.
cjs
cjs
The need comes from differing needs in the user and developer groups. A quite typical example is `bounce buffers' to let you use an Adaptec 1542 in a machine with more than 16 MB of RAM. NetBSD lived without this capability for almost two years longer than Linux, until a clean, general, platform-independent solution was implemented. This has advantages (NetBSD is still the only system in which you can put a 1542 in an Alpha and have it work), but there are a lot of users out there who didn't care how ugly the hack was, they wanted that 1542 working right away. They got what they needed in Linux and FreeBSD; the purists got what they wanted in NetBSD. Now, in your proposed merger, which user group were you going to tell to get lost?
cjs