Finder has never froze on me when I have lost an NTFS share.
Presumably by "NTFS share" you mean "SMB share" - "share" generally refers to file systems accessed over the network, and OS X's NTFS (like the ones on other BSDs, and the one on Linux, and the one in Windows) accesses files on a local disk. You can access NTFS file systems over a network, but that'd be done with SMB (or NFS or AFP or Netware NCP, if you have server software for them on the machine with the NTFS file system).
The issues he's seeing might be Finder issues, or they might be smbfs issues.
And for uploading to a FTP, have you RTFM? Works fine for me if you have permission to write to a directory.
Yes - if you, for example, use the command-line FTP client, you can upload files. He was talking about doing it in the finder - which would mean doing it with OS X's ftpfs, which is implemented with a program (mount_ftp) that acts as a user-mode NFS server and FTP client. If, for example, you use Connect To Server with an ftp: URL, you're using ftpfs, and the FTP server looks like a mounted file system.
That's very convenient (and not just from the Finder; I've used it from the command line to grep through a bunch of log files somebody'd put up on FTP for me). However, ftpfs is read-only, not read-write; you can't use it to upload - you'd have to use the ftp command, or some GUI FTP client.
That's not a Finder issue, though, unless you consider it a Finder issue that the Finder only handles file systems accessible with standard UN*X APIs - unlike Konqueror, which supports file systems accessible through kioslaves. I think that's how Konqueror handles FTP access, for example; that means it acts more like a GUI FTP client. That makes it easier to implement write access, but it means that it's accessible only through KDE applications that support kioslaves; you can't just cd to a directory on the FTP server and run arbitrary UN*X programs (including, but not limited to, KDE programs) on those files.
I remember when seeing talk of either a GNOME VFS or a KDE kioslave for SMB, I wondered why this was necessary, at least on Linux and some BSDs, given that SMB client file systems are available, so you could just access SMB file systems through the standard UN*X APIs.
My personal inclination is towards such a scheme, with an automounter used for browsing (and that was my inclination even before I started using a Mac) - it just comes across as "more UNIXy" (and that's not just an aesthetic prejudice - it was Really Cool to be able to do the grepping to which I referred above without having to explicitly download the files (in effect, ftpfs did the downloading for me, and provided access to the downloaded local copies via NFS on the same host). It does mean you have to do a better job on the client file systems and the automounter (for example, ftpfs would need to be based on something other than NFS, so that it can find out when you close a file to which you're writing and thus can infer when it's time to upload the file to the server).
(Yes, it's perhaps a bit ironic that the OS X way is more command-line friendly than the UNIX+X free-software way of handling FTP from the file manager....)
Who wants/Applications,/System,/Users etc. CAPS!!! yes CAPS!!! in a UNIX OS! Evil.
Yeah, God forbid that KDE or GNOME, for example, put your desktop in a subdirectory of your home directory called "Desktop"....
Not to mention some dirs have spaces *gasp* very annoying at the command line.
Bash's file name completion - which I tend to use when typing pathnames, as it's more convenient than typing the full pathname, even for pathnames I create myself at the command line, at times - escapes spaces for you. (The Korn shell puts things in quotes, instead, and screws up if you continue doing completion; I'll have to see if I can improve that, as, otherwise, I prefer the Korn shell.)
Apple should start using some of Microsoft's ideas, like binding the browser on a molecular level to the OS itself!
Yeah, they should, for example, make Safari run on top of a toolkit, with the toolkit's functionality available to other applications. Maybe they could call that toolkit "WebKit", or something clever like that.
No BSOD, no registry conflicts, no "flagrant system errors", nothing!
Which of those did you get when you tried dragging IE to the Recycle Bin on Windows?
Which leads into this nonsense of "hiding file extensions". THEY ARE NOT FREAKING EXTENSIONS: THE ARE SUFFIXES!
Yes, boys and girls, it's the height of idiocy to look at the NAME of a file to determine how to handle it when you can look in a file to see what sort of thing it actually is. One of the stupidest "features" of Microsoft Windows is it's inability to understand that a JPEG file is actually a JPEG file even thought it's named "Foo", or, gosh darn it, even maliciously renamed to "Foo.GIF".
Then you'd like KDE, as it determines a file's type by looking at the file's contents, when possible - and you wouldn't like OS X, as it determine's a file's type by looking at the name, in at least some places. (I just tried it, and 10.3.9's Preview, at least, gets confused if you rename a JPEG file to ".gif" and run open on that file.)
The only advantage I see to using the name is that less file system activity is necessary to determine the file's type.
How about those "Atomic Clocks" sold at Radio Shack and Wal-mart?
If the one sold at Radio Shack is this Atomic Digital Travel Clock, then it receives a signal from WWVB, and their digital signal includes a daylight savings time indication, so assuming they do the right thing the "atomic clock" will Just Work.
Really, people, this is probably a lot less complicated than you might think. Many UN*Xes can just deal with it with zoneinfo file updates, and several people have indicated that it's a registry change in Windows. Perhaps some applications have their own time zone rule files, and perhaps OS/360^H^H^H^H^H^HMVS and z/OS aren't as easy to fix, but a lot of machines and applications will require only a small file/registry tweak.
Mod parent up "+5 Clueful". Yes, many UN*Xes use either the "Arthur Olson" time zone code or something compatible; this includes, at minimum, Solaris, OS X, most if not all Linux distributions, and all the free-software BSDs. For those systems, the time zone files would be updated, and anything using the OS calls for converting times, or otherwise using the time zone files, will Just Work.
Well I don't know it just says netbios-ns (name server?)
Or "name service" - see RFC 1001 and RFC 1002 (NetBIOS Name Service for NetBIOS-over-TCP, a/k/a "NBT", which, the name nonwithstanding, uses UDP for some functions).
That's a NetBIOS Name Service query; I'd only expect to see that if you're doing SMB mounts from an SMB server (Windows, Samba, etc.) or perhaps if you're running Samba - but you said Windows file sharing is turned off, so, if that refers to OS X Macs, they're not running Samba.
You might want to use Ethereal or Tethereal to look at those packets - you'll get a bit more detail, including the name of the host it's looking for, which might give more information to help figure out what's sending those requests.
(Perhaps it's time to improve the tcpdump code a bit and show, when decoding the NetBIOS Name Service traffic, what the name being looked up is. A lot of the code could be taken from tcpdump's DNS decoder.)
Then you will, of course, see AFP "noise" on your network, although I wouldn't call it "noise" if you're intentionally using it. You might get broadcast or multicast announcements of service from the machines acting as servers.
I think the thing that keeps announcing itself as Netbios is something to do with Rendezvous or AFP.
Unlikely - Rendezvous^H^H^H^H^H^H^H^H^HBonjour is, among other things, a mechanism for announcing services, so Bonjour-related stuff wouldn't need to use NetBIOS to announce itself. Similarly, AFP uses either AppleTalk or Bonjour mechanisms to announce itself.
What are the details of the NetBIOS announcement packets?
Lots of network noise from my Apple boxen. AFP, Rendezvous, Netbios etc.
Presumably by "AFP" you mean "AppleTalk" - I wouldn't expect to see Apple File Protocol traffic unless some machine is accessing a file server. Perhaps the Macs are sending some sort of AppleTalk broadcast announcements.
What sort of NetBIOS traffic are you seeing?
Now can someone please tell me why tcpdump and tcpflow -c don't do the same thing. tcpflow seems to grab the entire data sans headers but missees most all of the lower level traffic (e.g ARP whohas etc), whilst tcpdump only grabs the headers no matter how big I make the snarfen -s thing or if I do -vv still only grabs the headers.
What tcpdump "grabs", in the sense of getting from the OS's packet capture mechanism, is controlled by "-s", and that does work; "-v" doesn't affect what it captures.
"-v" controls what it displays, and tcpdump displays a summary of the packet or, with "-v" or higher levels of verbosity, some amount of detail about some packets. You can get it to display raw packet data with "-x" or, with newer versions, "-X".
I think I remember it being funny that HP owned Tandem and then sold it only to aquire it again from Compaq?
I don't remember HP owning Tandem - perhaps you're thinking of Stratus, although HP didn't own them, they just were partners of theirs, at least to the extent that HP sold (and still sell) PA-RISC processors to Stratus.
Apple has said that developers should not depend on the BIOS system being used in a release model. Apple likely threw these dev boxes together quickly and inexpensively, so they haven't had time to design all the extra stuff a final version would have, like Open Firmware.
In the laptop space, Apple won't be replacing 64-bit processors with 32-bit processors; iBooks and PowerBooks all have 32-bit processors. The same applies to the Mac Mini - but not to the iMac.
So what do you consider Alpha to have been? (I consider it to have been a RISC processor.)
The PowerPC will disappear from computing in 2007.
So much for the pSeries and iSeries servers, I guess....
(Presumably by "computing" you're referring to personal computers, although I'm not sure why you're including SPARC in that category; from your reference to Sun workstations, perhaps you mean "desktop computers", or perhaps, at least with low-end Sun workstations, "computers ordinary people might be able to buy".)
Slashdot Mirror
Presumably by "NTFS share" you mean "SMB share" - "share" generally refers to file systems accessed over the network, and OS X's NTFS (like the ones on other BSDs, and the one on Linux, and the one in Windows) accesses files on a local disk. You can access NTFS file systems over a network, but that'd be done with SMB (or NFS or AFP or Netware NCP, if you have server software for them on the machine with the NTFS file system).
The issues he's seeing might be Finder issues, or they might be smbfs issues.
Yes - if you, for example, use the command-line FTP client, you can upload files. He was talking about doing it in the finder - which would mean doing it with OS X's ftpfs, which is implemented with a program (mount_ftp) that acts as a user-mode NFS server and FTP client. If, for example, you use Connect To Server with an ftp: URL, you're using ftpfs, and the FTP server looks like a mounted file system.
That's very convenient (and not just from the Finder; I've used it from the command line to grep through a bunch of log files somebody'd put up on FTP for me). However, ftpfs is read-only, not read-write; you can't use it to upload - you'd have to use the ftp command, or some GUI FTP client.
That's not a Finder issue, though, unless you consider it a Finder issue that the Finder only handles file systems accessible with standard UN*X APIs - unlike Konqueror, which supports file systems accessible through kioslaves. I think that's how Konqueror handles FTP access, for example; that means it acts more like a GUI FTP client. That makes it easier to implement write access, but it means that it's accessible only through KDE applications that support kioslaves; you can't just cd to a directory on the FTP server and run arbitrary UN*X programs (including, but not limited to, KDE programs) on those files.
I remember when seeing talk of either a GNOME VFS or a KDE kioslave for SMB, I wondered why this was necessary, at least on Linux and some BSDs, given that SMB client file systems are available, so you could just access SMB file systems through the standard UN*X APIs.
My personal inclination is towards such a scheme, with an automounter used for browsing (and that was my inclination even before I started using a Mac) - it just comes across as "more UNIXy" (and that's not just an aesthetic prejudice - it was Really Cool to be able to do the grepping to which I referred above without having to explicitly download the files (in effect, ftpfs did the downloading for me, and provided access to the downloaded local copies via NFS on the same host). It does mean you have to do a better job on the client file systems and the automounter (for example, ftpfs would need to be based on something other than NFS, so that it can find out when you close a file to which you're writing and thus can infer when it's time to upload the file to the server).
(Yes, it's perhaps a bit ironic that the OS X way is more command-line friendly than the UNIX+X free-software way of handling FTP from the file manager....)
What sort of changes would you suggest for /Network?
Yeah, God forbid that KDE or GNOME, for example, put your desktop in a subdirectory of your home directory called "Desktop"....
Bash's file name completion - which I tend to use when typing pathnames, as it's more convenient than typing the full pathname, even for pathnames I create myself at the command line, at times - escapes spaces for you. (The Korn shell puts things in quotes, instead, and screws up if you continue doing completion; I'll have to see if I can improve that, as, otherwise, I prefer the Korn shell.)
Yeah, they should, for example, make Safari run on top of a toolkit, with the toolkit's functionality available to other applications. Maybe they could call that toolkit "WebKit", or something clever like that.
Which of those did you get when you tried dragging IE to the Recycle Bin on Windows?
Then you'd like KDE, as it determines a file's type by looking at the file's contents, when possible - and you wouldn't like OS X, as it determine's a file's type by looking at the name, in at least some places. (I just tried it, and 10.3.9's Preview, at least, gets confused if you rename a JPEG file to ".gif" and run open on that file.)
The only advantage I see to using the name is that less file system activity is necessary to determine the file's type.
Or just use localtime() and mktime(), or Windows native equivalents, when possible?
If the one sold at Radio Shack is this Atomic Digital Travel Clock, then it receives a signal from WWVB, and their digital signal includes a daylight savings time indication, so assuming they do the right thing the "atomic clock" will Just Work.
Really, people, this is probably a lot less complicated than you might think. Many UN*Xes can just deal with it with zoneinfo file updates, and several people have indicated that it's a registry change in Windows. Perhaps some applications have their own time zone rule files, and perhaps OS/360^H^H^H^H^H^HMVS and z/OS aren't as easy to fix, but a lot of machines and applications will require only a small file/registry tweak.
Mod parent up "+5 Clueful". Yes, many UN*Xes use either the "Arthur Olson" time zone code or something compatible; this includes, at minimum, Solaris, OS X, most if not all Linux distributions, and all the free-software BSDs. For those systems, the time zone files would be updated, and anything using the OS calls for converting times, or otherwise using the time zone files, will Just Work.
Well, I heard it from Apple Computer. Do you consider them a valid source?
OK, that might be where the NBNS broadcasts are coming from.
It already does, but you need to run with "-vv".
Or "name service" - see RFC 1001 and RFC 1002 (NetBIOS Name Service for NetBIOS-over-TCP, a/k/a "NBT", which, the name nonwithstanding, uses UDP for some functions).
That's a NetBIOS Name Service query; I'd only expect to see that if you're doing SMB mounts from an SMB server (Windows, Samba, etc.) or perhaps if you're running Samba - but you said Windows file sharing is turned off, so, if that refers to OS X Macs, they're not running Samba.
You might want to use Ethereal or Tethereal to look at those packets - you'll get a bit more detail, including the name of the host it's looking for, which might give more information to help figure out what's sending those requests.
(Perhaps it's time to improve the tcpdump code a bit and show, when decoding the NetBIOS Name Service traffic, what the name being looked up is. A lot of the code could be taken from tcpdump's DNS decoder.)
Then you will, of course, see AFP "noise" on your network, although I wouldn't call it "noise" if you're intentionally using it. You might get broadcast or multicast announcements of service from the machines acting as servers.
Unlikely - Rendezvous^H^H^H^H^H^H^H^H^HBonjour is, among other things, a mechanism for announcing services, so Bonjour-related stuff wouldn't need to use NetBIOS to announce itself. Similarly, AFP uses either AppleTalk or Bonjour mechanisms to announce itself.
What are the details of the NetBIOS announcement packets?
Presumably by "AFP" you mean "AppleTalk" - I wouldn't expect to see Apple File Protocol traffic unless some machine is accessing a file server. Perhaps the Macs are sending some sort of AppleTalk broadcast announcements.
What sort of NetBIOS traffic are you seeing?
What tcpdump "grabs", in the sense of getting from the OS's packet capture mechanism, is controlled by "-s", and that does work; "-v" doesn't affect what it captures.
"-v" controls what it displays, and tcpdump displays a summary of the packet or, with "-v" or higher levels of verbosity, some amount of detail about some packets. You can get it to display raw packet data with "-x" or, with newer versions, "-X".
And what fusion power experiments are being done there?
Did you mean The Editing Room floor?
So his professorship at the University at Southampton will be conducted over the phone from the USA?
I don't remember HP owning Tandem - perhaps you're thinking of Stratus, although HP didn't own them, they just were partners of theirs, at least to the extent that HP sold (and still sell) PA-RISC processors to Stratus.
...who are owned by HP now. See the HP Integrity NonStop systems.
Did the ATM machine run Windows 2000, which is built on Windows NT technology?
Only if you're posting over a DSL line.
You probably meant to say "wrote the BeOS file system".
Apple have also said that "Macintosh computers that use an Intel microprocessor do not use Open Firmware."
Yeah - try one of these boxes, for example.
In the laptop space, Apple won't be replacing 64-bit processors with 32-bit processors; iBooks and PowerBooks all have 32-bit processors. The same applies to the Mac Mini - but not to the iMac.
So what do you consider Alpha to have been? (I consider it to have been a RISC processor.)
So much for the pSeries and iSeries servers, I guess....
(Presumably by "computing" you're referring to personal computers, although I'm not sure why you're including SPARC in that category; from your reference to Sun workstations, perhaps you mean "desktop computers", or perhaps, at least with low-end Sun workstations, "computers ordinary people might be able to buy".)
Of course, there are always IBM workstations if you want a POWER-family processor (a family that includes PowerPC). They're a lot more expensive than the cheapest Sun SPARC workstation, however.