Network CD 2

home *** CD-ROM | disk | FTP | other *** search

/ Network CD 2 / Network CD - Volume 2.iso / programs / internet / tcp / server / amitcp_fspd.lha / fspd / ChangeLog next >

Wrap

Text File | 1994-05-20 | 26.3 KB | 566 lines

Version 1.0 -- Dec 91. Original source created. Version 2.0 -- Dec 91. Added man pages. Server bug fix. In 1.0, files in a directory that is protected from deletion can still be written over. In 2.0, it is prohibited as well. Version 2.1 -- Jan 1, 92. 1) The maximum length of a name in pre-2.1 server is both limited by the 1k-byte udp packet size and, in some cases more severely, by the OS it is running in. On some systems, the OS imposes a limite of 14 chars. For those systems, the server can now be made to hash long names into short ones. The effect is transparent to the clients except when two long names gets hashed into the same value. In that case, they will refer to the same file during downloading, and they will displace each others during uploading. 2) Since directories, unlike files, have undisputable ownership, removal of a directory owned by a client should depends only on the ownership of the directory itself, and not on that of its parent directory. In pre-2.1 server, the client must have delete permission to the parent directory before it is allowed to delete the subdirectory. In 2.1, the permission is not needed. Version 2.2 -- Feb 15, 92. 1) The fput and fgetcmd utilities will now accept a list of files from its standward input. See their man pages for this expanded usage. This is useful if you don't want file names to show on 'ps' output. 2) The retransmit timer for client code has been changed to use an adaptive algorithm. In pre-2.2 versions of FSP, clients will retransmit if it has not received a response from the server after a delay of 3 seconds. In version-2.2, two separate delays are used. The "busy" delay is used when waiting for a reply for the first try and the second try. Any time a second try is necessary, the delay is increased by 50%. Every first try brings the delay 12.5% closer to the initial delay setting (3 seconds by default, settable by the FSP_DELAY variable). For the third try and beyond, the "idle" delay is used. For each try, the delay is doubled. 3) Makes available FSP request code space 0x80 - 0xff for future user extensions. Each such extension will begin with a code byte in that range, and followed by a control block, which is likely to be different from the existing one. The basic operations of FSP will not require the use of such extensions. Note: 2.2 does not involve changes to the server itself. The server is been checked out and added to by a few people on the net. I will wait until the modifications stabilize before I consolidate them. Version 2.3 -- Mar 9, 92 1) The bsd_src/glob.c file has a bug in blkfree function in that it tries to free something that is actually on the stack of the glob function. The last call to free has been commented out to fix the bug. 2) The fspd has a bug whoes only visible manifestation is the that protection can't be changed. The fix has required a moderate amount of code change in fspd. 3) .FSP_NO_DEL and .FSP_NO_ADD mechanism has been replaced by .FSP_OK_DEL and .FSP_OK_ADD mechanism. Thus by default, a directory is no delete and no add. 4) fprocmd is modified to relax option checking, allowing arbitrary option characters be passed to the fsp server. 5) A fls bug is fixed so that a "fls /" in a directory other than the root directory will not cause an error. 6) A -DDIRENT switch is added in Makefile to help support those machines that does not have /usr/include/dirent.h. Version 2.3a -- March 20, 1992 1) In a few places, non-critical errors of the form: if(errno = EINTR) .... was made. It obviously should have been == instead. The problem is fixed. 2) In the previous versions there is a potential problem such that if the first reply packet is dropped or is messed up, making a retransmit necessary, fspd will make the client utility wait for a 1 minute timeout. A potential loop hole for this happening has been closed by modifying server_lib.c where the packet key is checked. Version 2.4 -- March 27, 1992 1) fgrab/fgrabcmd client utility is added. It is like fget except it also deletes the named files in such a way that if there are multiple fgrabs for the same file, only one will succeed. One bug/mis-feature of fgrab is that when fgrab is terminated abnormally, the file is not restored. This command is useful for things like sharing bone files for modified versions of nethack. 2) FSP now allows multiple client programs to run at the same time if one of the three multiplexing mechanisms are chosen during compile time. It is likely that at least one of them will work for your system. The multiplexing is done at the client side, and does not allow the client machine to issue request messages at a higher rate compared to the previous versions. The FSP_LOCALPORT variable is no longer needed. However, if used, it will force serialization of client programs, which might be desireable if none of the multiplexing mechanisms are suitable for your machine. In order to accommodate this change, the server code has been changed to store temporary upload files with a different name from the previous version. The names now have to identify sender's port as well as sender's internet id. Change is necessary to keep it under 14-character limit of some OS. 3) Progress report mechanism enabled by FSP_TRACE variable now try to update the kbyte display more promptly. 4) fcat/fcatcmd will now turn off FSP_TRACE during its file transfer if its stdout is a tty. Version 2.5 -- April 8, 1992 1) A bug introduced during the last release --- fspd will not timeout and exit when running under inetd mode. This bug is now fixed. It was introduced because it is more convienent at one time for it not to exit when it is still being debugged. The patch to stop it from exiting was not removed before release time. 2) A very rarely seen bug in some OS is circumvented. UDP packet's headers contain a checksum field. If the checksum option is disabled, the OS will set the checksum field to 0 before sending, and will not check the checksum field when message is received. If the option is enabled, the OS will stuff the checksum into the checksum field. When a packet is received, the checksum field is checked. If the checksum field is 0, then the OS assume that the sender does not have option enabled, and therefore the checking is skipped. Problem arises when a packet's computed checksum is zero. Some OS will stuff 0xffff into the checksum field before sending such a message to distinguish it from the case for which checksum is disabled. Other OS couldn't recognize 0xffff as an indication that the actual checksum could be either 0 or 0xffff, and subsequently reject the message. The client can thus get stuck forever in retransmitting a request that happens to have 0 for the computed checksum, but 0xffff for the checksum field. To get around the problem, the sequence number field of the request message (which has meaning only to the client software), is partitioned such that 2 of the bits reflect the retry sequence, and the remaining 14 bits reflect the request sequence. This will ensure that if one request got ignored because it happens to have a checksum of 0, the next retry will result in a packet that does not have a checksum of 0. The only file affected here is client_lib.c. 3) In response to those who have severely limited UDP packet sizes, the environment variable FSP_BUF_SIZE is supplied to set the data size of file and directory information transfers. The default is 1024. You cannot set it to be bigger than 1024, but you can set it to be smaller if you find that your system cannot pass UDP packets that includes 1024 bytes of data. This requires changes both in the server and the client code. The first parameter field of get-file and get-directory requests contains the path of the file or directory. The second parameter field was unused, but it is now used to hold the number of bytes to read. This change is backward compatible in the sense that if the second field is empty (old client code), the new server code will default to 1024. Changes to the client code is confined to client_util.c. Version 2.6 -- July 6, 1992 1) client_lock.c is fixed so that perror is always followed by exit. Previously, if System-V shared memory mechanism is chosen for locking control in the Makefile, but the mechansim has not in fact been configured into the UNIX kernel, the client utilities would die of a segmentation fault error. The error message will now be more reasonable. 2) mechanism is added so FSP server can put directory cache files in a special directory other than the one for which the files are intended. Previously a .FSP_CONTENT file is created in each directory the fspd process controls. Through variables in the Makefile, provider of fsp service can now elect to put cache files together in one single directory, and put a cap on the total number of such files that can exist simultaneously. The fsp server will create a cache file in the cache directory when it couldn't find one already exist in the directory being listed by the client. Changed are minor and invovles the routines server_get_dir, server_get_dir_2 [new] in "server_file.c"; the variable declaration section of "server_main.c", and Makefile. 3) a line in util_cd is changed, and a line in get_dir_blk is added in the file "client_util.c" to prevent fls from sending some spurious requests to the fsp server. 4) Makefile is changed to allow the option of combining all client utilities into one actual binary files. This is necessary for space saving reasons on some versions of UNIX. The various small main C files for the utilities had to be changed as well to accomodate this change. 5) fver now will take an optional argument. If none is supplied, it prints the version string of the server. If one is supplied, it prints the versioin string of the client FSP software. Version 2.6.1 -- Aug 31, 1992 1) Small changes to Makefile. In 2.6, "make" will fail if you needed to define LIB, or if your command shell is not /bin/sh, and you want to merge client utilities. Version 2.6.1 fixed them. Only Makefile is changed, everything else is identical to 2.6. Version 2.6.2 -- Sep 2, 1992 1) bsd_src/cmp.c was changed so that flscmd can sort its output by date correctly on all systems. This particular version of the BSD source made assumptions about what the comparison functions should return for the qsort routine, but those assumptions were not true for all systems. Version 2.6.3 -- Sep 19, 1992 1) client_lock.c is modified to fix a compilation phase bug that occur when NOLOCKING option is used in the Makefile. This affects only the client code. 2) server_file.c is modified so that upload uses the directory cache directory instead of the top directory. This is needed for those sites that want the top directory be write protected. (good for ftp sites). Version 2.6.4 -- Dec 12, 1992. 1) Make main makefile pass CC and CFLAGS definitions to bsd_src's Makefile 2) Add util_cd2 to client_util and make bsd_src/ls.c use it. This remove some of the spurious GET_DIR requests. Also, there is a bug in client_util's util_cd routine in that it always requests 1K worth of directory info no matter what FSP_BUF_SIZE setting is used. Causeing fsp clients to hang under situations where FSP_BUF_SIZE is limited by the network (slip, usually). 3) function server_interrupt in server_lib.c was defined to be of type void to avoid complaint from some compilers. There are several other minor fixes in functin variable declarations in several files for the same reason. 3.5) The retry delay function for the client is tweaked somewhat. Wed Dec 16 12:31:29 PST 1992 4) A change in the server_up_load function in the file server_file.c is made to fix a subtle bug. When a file is uploaded, the offset to begin write was not used to seek to end of file. Instead, the file is only appended to. This cause error if two retry messages in an upload was divided by a server going down. The new server is fixed so that seek based on the pos parameter is done. This eliminates the possibility of such an error. Version 2.6.5jt -- Dec 17, 1992 1) Added support for .FSP_OK_READ and .FSP_OK_MKDIR. .FSP_OK_READ allows reading the contents of files. .FSP_OK_MKDIR allows creation of subdirectories. .FSP_OK_ADD no longer means anything for creation of subdirectories, only for creation of files. Version 2.6.5jt.1 -- Dec 18, 1992 1) Changed the direction of intent on the .FSP_OK_READ file and renamed it to .FSP_PRIVATE. Now prescense of file means directory is private. Done so that spurious files didn't need to be created everywhere just to grant read access. In concordance with this, the procmd is no {+-}p instead of {+-}r. Version 2.6.5jt.2 -- Dec 29, 1992 1) Fixed a core dump in any of the programs that used file globbing. If you passed a ~ to the command in this state, you would seg violate. Fixed by removing the parsing of ~ in the glob.c file. It now errors with an 'illegal character in path'. Version 2.6.5jt.3 -- Jan 27, 1993 1) Added in the Enhanced Trace (cps) from pi@cs.sun.ac.za 2) Added protection from clobbering files on fget/fgrab courtesy of A.E.J.Fellows (A.E.J.Fellows@bham.ac.uk) 3) Various updates and corrections to the man pages. 4) Merged in VMS changes. .com files and VMS makefiles as well as VMS changes files all exist in ./vms. VMS specific sourcefiles are in ./vms_src and VMS specific header files are in ./vms_h. VMS changes done by S.A.Pechler@bdk.tue.nl. All changes made by Sven are in ./vms/VMS.ChangeLog. Some of the changes he made there are incorporated into the unix version as well. Version 2.6.5jt.4 -- Feb 2, 1993 1) Made the initial check for logfile opening be done with mode "a" instead of mode "w+" so that it didn't erase the file every time through. 2) Reordered the includes in common_def.h so that they preserved the original 2.6.5 ordering which some machine types seemed dependant on. 3) Added in Sven's repatches to make VMS compilation work. Version 2.6.5jt.5 -- Feb 4, 1993 1) Added extended version information and loggin information to the reply from the fver command. 2) Changed the message type on fver replies to CC_VERSION instead of CC_ERR. This just makes more sense. 3) One more special case where SCO unix seems to require a certain order in it's include files.. sheesh. 4) Changed duplicate variable clobbertype in both fgetcmd.c and fgrabcmd.c to be unique. Moved the #defines of the clobbertypes into the appropriate .h file 5) Added a few more small fixes from Sven. 6) Added support for restricted mode. When server is run in restricted mode ONLY users from sites appearing in the .ROGUE_HOSTS file can connect. This flag basically reverses the meaning of the .ROGUE_HOSTS file, but by doing so provides a cheap way of doing priveledged connect. Version 2.6.5jt.6 -- Feb 8, 1993 1) Added a contrib directory for user submitted programs such as my admin scripts. Currently thats the only thing here but I'm sure other things will wind up here in time. 2) Removed the redirecting of stdin/stdout/stderr to /dev/null under normal operation because it was causing some true error conditions (like it telling you that homedir didn't begin with a / to get dumped into the bit bucket.) 3) Fixed a logic error in the priveledged site code. sheesh. NOTE: To a disallowed client, it will appear as if the server doesn't exist. This is as far as I can tell the behavior intended by Wen. 4) Modified makefile to not use any file names longer than 14 characters (POSIX compliant). This only affected the merge programs. Version 2.6.5jt.7 -- Feb 9, 1993 1) Fixed some of the POSIX compliant naming that I missed the first time around. 2) Added Ruben Olsons admin tools and fsp-shell to the contrib distribution. 3) Fixed a small bug in switch parsing in fspd. (if you did fspd -l -r it would think that -r was the filename to log into. this is wrong). 4) Added a small patch suggested by Ruben Olson. Now, if the Remote FSP server doesn't respond (and you're not in the middle of a file transfer) the client will just exit instead of hanging indefinately. It takes about 3-5 seconds for the client to exit. Basically, if pos is zero (as it is unless you are transferring), then when the client would first print an 'I', it will instead now exit. (enable this by compiling with -DCLIENT_TIMEOUT) Version 2.6.6 -- Mar 10, 1993 1) Added some patches supplied by Sven Pechlar to fix a few last niggling bugs with the VMS port. 2) File sizes are now logged as part of GET_FILE (courtesy of Sven). This might make the Sumalog script useless until fixed. 3) Added an append mode to fgetcmd and fgrabcmd. Under this mode, fget/fgrab will try to append data to the end of a short file. This patch was submitted by robert@diku.dk (Robert Martin-Legene) 4) Extended the timeout time on CLIENT_TIMEOUT as requested. 5) Added code supplied by Sven Pechlar to do password protect on private directories, and to do internal file cacheing. The password protect code looks extremely broken to me. But I added it anyway as if no passwds are set, it shouldn't break anything. 6) Signal handling code (for BSD only) to handle cleaning up .fsptemp.XXXX files on abort (courtesy of AEJFellows) 7) Move setuid call before log file is opened. (AEJFellows) 8) Added throughput control courtesy of Laurent Montaron (montaron@eurecom.fr). See INFO file for details. Version 2.6.6.1 -- Mar 11, 1993 1) Reduced the client timeout from 5 to 4. 2) Made client timeout code default in the makefile. 3) Made printing of 'E' on retrys conditional on trace being true. 4) Shortened temp file names so that they will always fit under 14 character limit. 5) Yet more small fixes from Sven to make 2.6.6 compile on VMS cleanly. (thanks Sven) 6) .README file support thanks to Pihl (phil.richards@prg.oxford.ac.uk) 7) More advanced Hostname security thanks to Phil as well. (Look in the INFO file for details) 8) Rewrote logging to use varargs based on ideas from Phil. 9) If no password is supplied, and the .OWN.XXXXXXXX file is empty, then you will still be able to read the directory. This reinstitutes backward compatability in the server with older clients that cannot send passwords. 10) Changes to support a local working directory for all remote operations courtesy of Michael Meskes <meskes@ulysses.informatik.rwth-aachen.de> Version 2.6.6.2 -- Apr 8, 1993 1) Added patches to make the client_timeout value be settable via an environment value. Default timeout value is 4. The change has no effect unless you compile the clients to time out if there is no response. (patches by Michael Meskes) 2) Added fducmd and fsethostcmd written by Michael Meskes. fducmd is a standalone (or merge) version of the du cmd that phil put into his fspclient. Fsethostcmd supports a fairly simple resource file for setting up hosts and local directories and all environment variables based on information in a file. 3) Fixed a bug in fgrab that caused you to not be able to grab a file even though you could get it and then rm it. 4) Fixed the IPHOST code. It was returning automatic vars up the stack, and assuming that assigning a structure to a structure performed copy. This is not always the case, and when compiled under ultrix cc this assumption broke. The fix was to malloc the structures down and pass pointers up the stack initially, and then store those pointers. The code has been tested now as well, and verified to work on at least my system. 5) The server will now log an 'Unknown command.' error if it gets a command request that it doesn't know how to handle, and will return this string to the requesting client. (requested by Robert Martin-Legene) 6) Fixed a bug in build_dir() reported by Phil Richards. A file could get a garbaged 'long_name'. 7) Added better logging code supplied by Phil Richards. 8) Added reverse hostname lookup and read only mods supplied by Phil Richards. 9) Removed a bug that caused logging to be turned on even when the -l option wasn't specified. 10) Changed the name of the fsethostcmd to fhostcmd so that POSIX machine nameing conventions were followed. 11) Renamed the .fsprc file used by fhostcmd to .fsp_prof Version 2.6.6.3 -- Apr 12, 1993 1) Added bitfields containing the version and protection information into the getpro and fver commands. Newer clients should rely on that information and thus be able to present it in any display format desired. 2) Logging code will now stat the logfile and if it's no longer there, it will recreate it after closing the old fd. This saves the system expense of opening and closing a logfile each time we log something, and still allows the deletion of logfiles during a run. 3) Modified clients to present information from the bitfields for ver and pro, and modified the server to give only minimal textual information for ver and pro. 4) Added a sanity check to FSP_DELAY variable to never let it go below some specified minimum value. This minimum value is set in common_def.h and defaults to 500 usec. 500 usec should remove the problem of the infinite resend client. 5) Fixed a small bug in logging where it would add one newline for every 1k uploaded to the logfile. Version 2.6.6.4 -- Apr 13, 1993 1) Fixed yet another small bug with readme files that I'd created. 2) Encased all printing of strings recieved from a remote site inside some sanity to prevent them sending a %s or something as part of a string and having it choke. 3) A core dump in the logging code where there was a case under which it did not return after an error is fixed. Version 2.6.6.5 -- Apr 19, 1993 1) Fixed a bug in the .IPHOST code. (Thanks Laurent and Phil) 2) Fixed a few typos and compile problems with the merged client. (Thanks to the numerous people who reported it) 3) Changed the code for fducmd slightly (Thanks Michael Meskes) (NOTE: Michael has also written an ffindcmd that will be in the next release, it's not included here) Version 2.7.0 -- May 3, 1993 1) reorganized source code completely. 2) Installed the source code under Gnu autoconf 3) Added ffind client submitted by Michael Meskes. 4) Rewrote portions of the documentation. 5) Moved all configuration out of the Makefile and into header files and an fspd.conf file 6) Added -r recursive option to fgetcmd and fcatcmd (thanks to Michael Meskes). Read man pages for details. Version 2.7.1 -- May 7, 1993 1) Ironed out portability problems. 2) Fixed the pointer to in cast in server/filecache.c (thanks to wen-king) 3) Added WenKing's patches for the Dec Alpha 4) Merged in Sven's patches for fixing 2.7.0 on a VMS system. 5) Added aclocal.m4 with some macros to start cleaning up the configure.in file a bit. Some macros needed changing to get the configure script to configure properly on a linux. 6) Fixed a bug in server_upload where a file could not be uploaded unless the server had a valid cache dir set. Thus, any sites that didn't use cache dirs couldn't have files uploaded. Version 2.7.1 AMIGA -- March 25, 1994 1) Started with (re-)configuring the tweak.h file to suit SAS/C 6.2 2) Disabled various not-amiga-supported functions. (fork, setuid) 3) Added needed include-files. March 29, 1994 4) Rewrote signal handlers. (^C breaks program, ^F dumps .HTAB) 5) Fixed up server/build_dir_file to deal with amiga pecularities. 6) server/checkpath now returns a (more) valid amiga-path. To make this fool-proof will need more reworking. 7) Inetd-mode supported. Doesn't time-out though March 30, 1994 8) Programs sets it's name correctly when running in inetd-mode. (when inetd doesn't) 9) Checkpath now disallows client-requested paths with ':' in them. May 15, 1994 9) Timeout under inetd works. May 20, 1994 10) Get a usable version together, and upload it to aminet. :) Near future -- Grabfile doesn't seem to work, could be the clients fault. -- Get round the softlink problem. -- Handle homedir ending in ':' correctly. (or did I do that already?) What needs to be done EVENTUALLY Somebody should make a tar equivalent that read and write remote fsp archives. This will allow transfer of whole directory in one command. Sources should be make to pass lint with as little problem as possible.