home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Turnbull China Bikeride
/
Turnbull China Bikeride - Disc 2.iso
/
BARNET
/
FREENET
/
SKELTON
/
SMTPD
/
SMTP
/
Changes
< prev
next >
Wrap
Text File
|
1996-01-22
|
2KB
|
50 lines
Change Log for SMTP Daemon for RISC OS
======================================
1/6/95 » Version 0.01 released for Beta test «
2/6/95 • Fixed Internet_GetHostByAddr exception when InetDB not present
• Extended line buffer to 4k
• Fixed username@localhost bug
• Fixed delay in initial server response
• Remembered about multiple recipients, and implemented it
• Fixed multiple recipients logfile bug
• Modified logfile handling slightly
13/6/95 • Fixed alias bug
» Moved to version 0.02 beta «
• Re-wrote alias & recipient code
• Sorted out handling of premature connection-closing by client
14/6/95 • Sorted out problems with connections to other smtp implementations
• Added EXPN functionality
• Added the commands *smtpusers and *smtpstats
• Tidied up socket shutdown
15/6/95 • Finally fixed the 'missing first byte' mystery during connection!
• Tidied up the way smtp talks to other servers
• Added the command *smtpclose, for panics and aborts
• Made the outputs of the other commands more pleasing
• Updated the documentation
20/6/95 • Fixed an obscure socket bug (non-standard socket descriptor set range)
• Fixed a bizarre state transition, which mysteriously appeared from nowhere
• Added real user names to the config file
• Altered smtp EXPN command to use real user names
• Updated the documentation
15/11/95 Changed format of output file to terminate lines with LF only, after many, many
complaints about the old version terminating lines with CRLF.
17/11/95 Fixed the RSET bug.
22/1/96 Fixed the HELO bug when no hostname is given. Lots of people have been bitterly
complaining around the clock for this to be fixed, and I'm happy to finally
oblige.
Fixed a problem which a couple of people had noticed by being really pedantic
with the module. :) Really long lines got horribly mangled, as a result of me
forgetting to test a huge swathe of the pending socket read code. This has now
all been sorted, and hopefully a lot of people will be much happier.