home *** CD-ROM | disk | FTP | other *** search
Text File | 1996-06-19 | 100.6 KB | 2,027 lines |
-
- 94/10/30
- NBETA000: The prototype beta envelope, never released.
-
- 94/10/31
- NBETA001: The first released beta. Several bugs reported within hours,
- including:
-
- - The MORE in displaying the file menu help (and I assume in
- many other places) wouldn't abort. This was due to a logic
- bug in the USEND routine. Fixed 94/11/01.
-
- - Someone got a Runtime 003 when trying to insert a new user in
- the user editor. Runtime 003 is a "path not found" error, and
- indicates that a path the BBS was trying to access does not
- exist. So the problem was chalked up to a lack of proper
- pathing after I managed to duplicate it. However, sysops
- who unpack the archive into a direcory other than C:\BBS would
- have a helluva time grappling with this same bug so I added code
- to: make sure the system and temporary paths exist each time the
- BBS is called (and create them if not), and: move all the
- files to C:\BBS (creating if needed) before proceeding with
- initial installation. I agree that this is a patchwork
- solution, and I actually have sketchy plans underway
- already to implement a full-featured installation program
- like those used by Apogee games some time prior to release,
- possibly in the marketability-testing phase between Nov.21
- and Nov.30.
-
- - An observation has been made that in local logons the last-8
- log showed a baud rate of "115200" instead of the proper
- "Local". I was actually aware of this for a long time, and
- never was bothered by it enough to do anything about it.
- Nevertheless, it's fixed.
-
- Thanks go to John Richards for pointing out all of these and
- several other things which I will address later.
-
- 94/11/01
- NBETA002:
-
- I have added support for read-only message bases and a
- [readpriv] UFO keyword which returns an error code if the user
- is NOT allowed to post in the current area (insufficient access
- or read-only area). Also I have added support for exit
- semaphores as a way to interface NanoBBS to external mail
- processors through creative batch hacking. I will document the
- black art of batch file writing in detail in the release docs,
- but the basic semaphores are listed in the docs.
-
- Also added to the doc file is a brief listing of keys you can
- hit while the user is online. Alt-C to chat and so forth.
-
- NBETA002 addresses all the bugs and "features" listed above in
- NBETA001.
-
- 94/11/02
- NBETA003:
-
- A comment from one of my users prompted me to do some cosmetic
- cleanup in flagging files during a search or list. I wasn't
- happy with the appearance there anyhow. I also noticed that
- flagging was impossible during new file scans and partial
- filename searches - seems the moretype variable wasn't being set
- properly before the high speed searches were invoked. This is
- fixed as of this beta.
-
- Also, I have added support for file-attach messages. Any area
- can be defined as a file-attach area and can have its own
- file-attach directory. File attach access requirements are also
- definable on an area by area basis. File attachment is
- super-easy in the local mode: all you do is specify the filename
- and path of the file you want to send the user, and the whole
- uploading process is skipped. For remote users using
- file-attach is just like uploading and downloading files only
- it's done in conjunction with a message, so that files can be
- sent privately from user to user. This is one of those creature
- comforts that users just _adore_...
-
- As always, let me know how it works for you. I have no path
- enforcement code yet so I suspect the unprepared may get a
- runtime 003 when the BBS tries to write or read a nonexistent
- path...
-
- 94/11/03
- NBETA004:
-
- In the echo, B.J. Verruyt has pointed out a few bugs; this
- release fixes them.
-
- I have also done some more cosmetic cleanup.
-
- 94/11/04
- NBETA005:
-
- On THC Freedom Online BBS, I set my adult cutoff at 18 years; I
- think this is the legal age to buy adult stuff in BC. Anyway I
- had an 18 year old user point out a problem in the age
- calculating code that prevented him from accessing the "fun"
- doors at certain times of the month. This was due to a logic
- bug and is fixed for NBETA005.ZIP.
-
- I have installed a vastly improved swap routine that will swap the
- BBS lock stock and barrel to EMS if available, or to a swap file
- on the defined RAM disk. It should leave you with LOTS more
- memory to run doors or external archivers (and external
- protocols when they are implemented). In theory it is supposed
- to leave you with only a few k resident, but I would be very
- interested to know exactly how much memory the resident part of
- NBBS occupies after it's been swapped. I would appreciate it if
- the OS/2 testing platoon would let me know how this works in a
- DOS box...
-
- I have also noticed that I'd forgotten to update the version
- number display with each beta. I have properly set this
- constant with this release and will try to be more vigilant from
- now on [sheepish look].
-
- One last note for this beta: I have tightened one of the units
- by over 10k. This is to shrink the size of the overlay and
- reduce overall memory requirements, and this will be repeated
- for all the units by release time. However it's possible that
- in some places efforts to tighten the code may introduce new
- bugs. Please let me know if you encounter any! The unit
- tightened today was the one concerned witb processing UFO
- keywords, so if you have been hacking in UFO keep it up!
-
- 94/11/05
- NBETA006:
-
- Sure enough the code-tightening campaign created some new bugs.
- String comparisons in UFO (in the IFEQ, IFNE, etc keywords)
- suddenly became case sensitive. RAID! These keywords (should)
- work again with this release.
-
- An earlier beta accidentally shut off displaying message text.
- This is fixed.
-
- I have finally tackled some very elusive, but very annoying bugs
- in the message reading code, particularly regarding the personal
- new scan at login and reading messages in reverse order. I
- -think- these functions are smooth now; before today they were
- useless. These bugs had been with me for a very long time.
-
- I have also fixed a bug in the full screen message editor where
- if you hit the end of a line and it wrapped, and the character
- you happened to be typing at the moment was a space, the space
- wouldn't be continued on the next line and your typing
- afterwards would be concatenated to the previous word. This bug
- has been fixed.
-
- I have added a hot key to invoke the user editor while the user
- is on. The key is Alt-E. If you edit the user who is online,
- his online parameters are also affected, so that the change
- takes effect immediately.
-
- At the request of a user, I have added options in the user
- [O]ptions off the main menu to set a default archiver and
- default protocol. If a default is set, the BBS doesn't ask for
- protocol or archiver when downloading, uploading, or assembling
- QWK packets. I will add UFO keywords for these later.
-
- There is now a relatively pain-free way to have file
- descriptions of virtually limitless length. Read Nanobbs.Doc
- for the details. I will likely add another mechanism for the
- same thing in order to have plug-and-play compatibility with a
- CD I happen to have (Night Owl 13 has FILES.BBS with a
- continuation mechanism different from mine).
-
- 94/11/08
- NBETA007:
-
- A couple of days hiatus from coding was due to some hardware
- problems here. All is temporarily OK, and some new code is now
- available.
-
- I have added another mechanism for continuing file descriptions.
- See NANOBBS.DOC.
-
- I have also added automatic FILE_ID.DIZ importation: if NanoBBS
- is able to extract a FILE_ID.DIZ from an uploaded file using any
- of the archivers defined in the config file, it will do so and
- will import it to the FILES.BBS description instead of asking
- the user for a manually entered description.
-
- This uses the same code as the QWK upload processor. If QWK
- works so should this. If this works, so should QWK. Let me
- know if either doesn't.
-
- 94/11/10
- NBETA008:
-
- I have added some automatic path detection code which should
- address problems reported by several testers. As long as there
- is a valid compiled config file and language file in the same
- directory as the executables, NanoBBS should run fine even with
- invalid paths specified.
-
- I am now working on adding a full RIPscrip suite to UFO (about
- 50 commands) and this will consume a great deal of time, so the
- new betas will be pretty sparse for the next week or so. I will
- continue to attempt to address bugs as they are reported but the
- next "new feature" will be a kickass RIP interface which is
- several days off. I wish there were more hours in a day.
-
- 94/11/15
- NBETA009:
-
- I am still working on adding RipScrip graphics to NanoBBS, which
- is why the long delay since the last beta. Unlike most BBSes,
- NanoBBS's RipScrip will not be limited to a built in detector
- and a bit to set to determine whether to use a .ANS or a .RIP
- file. NanoBBS will have full plaintext command support for the
- entire protocol built into both the language files and the UFO
- language. In fact the UFO RIP support allows the use of
- iterated commands and variables integrated into the RIP support
- to provide effects and interactivity of a calibre that is simply
- impossible under BBS systems that use static RIP files only! Of
- course, support for static RIP files still exists as it should.
- The protocol is now implemented but I am still working out some
- picky bugs and will document the protocol when the bugs I find
- are gone; this should be no more than a couple more days.
-
- Don St. Germain has reported trouble with the BBS hanging up at
- the end of a call; I have noticed this on THC also and I have
- added a 20 second timeout in the hangup code, and a slight
- modification to the sequence of the DTR drop which is supposed
- to cause the hangup. Hopefully this should fix the problem.
-
- I have also (in the process of adding RIP) made NCOMP several
- times faster - it compiles to memory instead of disk and saves
- the whole enchilada at once. Much faster! And easier on the
- hard disk. I think I'll add this feature to UFO.EXE and make it
- faster than Mecca - which is written in the allegedly-faster
- Watcom C!
-
- There are also a few new (documented) UFO verbs; INC, DEC, SHL1,
- SHR1, SHL and SHR. These perform increments, decrements and
- bitwise shifts left and right on variables in UFO.
-
- The debug code which printed out those C:\BBS\ type directories
- at user logon has been eliminated. It was never visible to
- users anyway.
-
- Oh yeah, for those of you running Rancho Nevada, I took a little
- time out this evening to fix a couple of bugs there too.
- $RN332.ZIP can now be file requested from me at 1:340/26. The
- FEATURES.TXT file in the archive explains the fixes.
-
- One last thing, I would remind everyone on the beta team that
- there is an echo whose purpose is reporting of bugs - and it's
- the deadest echo on my system, while my netmail area is crammed
- with bug reports! Please, use the echo to report bugs so that
- other testers experiencing similar problems can pipe up and
- offer more details. The quality of this BBS is equal to the
- quality of the problem reporting each of you gives! In order to
- make a top-quality BBS I need top-quality feedback, so please
- make this easier and use the echo!
-
- 94/11/16
- NBETA010:
-
- Some of the RIP stuff is now documented. Check out UFO.DOC. I
- have made improvements to NCOMP.EXE, and added totally
- transparent RIP language support - each language has an
- ASCII/ANSI language source file (ending in .TXL) and a RIPscrip
- source file (ending in .RXL). Very few other BBSes have this
- level of RIP integration available, let alone the RIP UFO
- scripting ability. If both are present NCOMP compiles both. If
- a user has a language selected and is using RIP and a RIP
- version is available the RIP version is used, otherwise the
- ANSI/ASCII version is used.
-
- It's worth noting that the [rip_pixel] keyword is considered a
- sort of "junk" command by the very people who designed the
- specification for RIP; this is because they expected RIP to be
- used in a static fashion as ANSI is today. One of the .UFO
- examples, RIPSTARF.UFO, uses this so-called "inefficient"
- command to very good effect, and I am sure the imagination of
- other sysops will give birth to much more spectacular results!
-
- I have included a couple of .UFO files which illustrate RIP UFO
- tokens. It's really cool (IMNSHO) to be able to do RIP in
- near-plain English without buying an expensive paint program,
- but I have already noticed several syntactic bugs and I assume
- that there are more. The chief symptom is that a command will
- simply not execute or will execute totally wrong. If this
- happens to you you may have found a command which is not parsing
- enough parameters; please let me know which command did this and
- include a copy of your UFO source if possible.
-
- I have also made some refinements to the RIP detection sequence.
- It still occasionally registers a false negative, but apparently
- the timing of Desqview is more to blame than anything else. I
- will experiment with some special timeslice forfeiture
- techniques RSN if this keeps up...
-
- Note: if RIP is detected NanoBBS will expect .RIP files in
- certain places!!! I have not yet built any .RIP files for these
- but if you copy the .ANS versions to .RIP the users will be able
- to use the system normally until I get some nice RIP. Or, you
- can build your own, whichever suits you after all it's YOUR BBS.
-
- The sample .UFO rip files are in the /misc directory...
-
- In other news, there are a couple of new UFO keywords; ADD and
- SUB will quickly and efficiently add and subtract nouns. This
- is more convenient than using the mathematical adjectives (which
- are still useful as they are more flexible).
-
- Also I have been informed that some users are getting no
- linefeeds. I still do not know if this is because they have
- turned off their linefeeds (and couldn't figure out how to turn
- them back on after the screen turned to mush) or due to a bug,
- but I have taken two steps to solve the problem anyway:
- Firstly, there is now an idiotproof subroutine which asks the
- user how many rows of dots they see, and if they only see one
- row they need linefeeds (and are told to respond that way). The
- other step I took was to add redundant code in several places to
- force linefeeds on until the user has actually logged on. This
- may screw up some non-IBM types but they are a tiny minority
- who can still turn the linefeeds off after they're logged in
- anyhow.
-
- A cosmetic bug in the wait-for-call has been fixed. It used to
- not properly reset the "rings" counter to zero when an incoming
- ring failed to produce a connection. This is expected fixed
- with this release.
-
- 94/11/19
- NBETA011:
-
- I should clarify that the INSTALL procedure is for NEW
- INSTALLATIONS ONLY. DON'T RUN INSTALL EVERY TIME OR IT WILL
- OVERWRITE YOUR EXISTING CONFIGURATION!
-
- The files PROGS.EXE and STUFF.EXE are self-extracting archives.
- It should be safe to run PROGS.EXE every time (and you should).
- You should run STUFF.EXE in a scratch directory and copy any new
- files you need from that - like this one.
-
- The RIP tokens in UFO are now fully documented - they can also
- all be used in language files exactly as they are in UFO except
- that Nouns cannot be used - all embedded language RIP commands
- must have constant parameters! The syntax for RIP in the
- language files is otherwise the same. I should mention that
- there are a few picky arcana in the RIP protocol which I was
- unable to decipher from the Telegrafix spec sheet; in these
- cases I have put in a comment like "please see the original
- RIPscrip spec from Telegrafix for more information" - which you
- should in those cases, as you likely will learn more than I know
- about RIP by doing so. I should mention that a week ago I
- knew virtually squat about the technical end of the RIP
- system... I have had to "grow up fast" on that one so there may
- be BUGS, let me know...
-
- The language files now come with two sources: one for ASCII/ANSI
- which ends in .TXL and one for RIP which ends in .RXL. The
- version of NCOMP released in NBETA010 compiles both in sequence.
-
- I have added support for a Time Bank to the BBS. You can set
- the maximum allowed withdrawal and deposit amounts in the
- general config file using the MAX DEPOSIT {amount} and MAX
- WITHDRAW {amount} keywords as illustrated in the example
- NODE1.CTL. It is fairly self explanatory for the users. I
- think I have it smooth, so that withdrawal limits are enforced
- and the user is properly docked his time when he deposits. Any
- bugs, you know where to post...
-
- I have also added the ability to pass a time-to-next-event
- parameter. The /T{minutes} command line parameter can be tied
- into your BBS batch files called up by Binkley or Frodo to force
- the user into a time limit that may be shorts than he's
- otherwise entitled to. This can be used, for example, to ensure
- that your mailer is available at the beginning of ZMH. I have
- not tested this feature in conjunction with the time bank at
- all, and I suspect that the user may be able to defeat this
- using the time bank if he has an existing balance. Let me know
- what happens.
-
- I have also fixed a long running bug in the code that
- compensates users time for their uploads. It now properly adds
- time to the current call and properly reports this time to the
- user when it is awarded.
-
- I have also added another security feature, the "barricade"
- file. This lets you selectively bar individual users from
- specific message bases without otherwise affecting their access,
- providing an unobnoxious way of complying with echomail
- moderators' enforcement directives. The NANOBBS.DOC file now
- contains what will likely by the final form of the chapter on
- this feature.
-
- I have improved the intelligence of the time warnings, so that a
- five-minute warning that occurs two minutes before the user is
- due to be axed does not tell the user falsely that s/he has five
- minutes left. There are many reasons (doors, external programs,
- downloads) why a user might get the first time warning late, but
- now it tells the user how many minutes s/he's REALLY got. Ditto
- the final warning; it no longer is fixed to 60 seconds (but
- still displays its warning in seconds).
-
- Last update I reported a cosmetic bug fixed where the number of
- rings received by the WFC screen was incorrectly reported. That
- bug was only made weirder and not fixed by NBETA010 but is fixed
- (and I have personally tested it) with this version.
-
- Lastly, now that the gargantuan task of implementing a RIP
- scripting language is substantially behind me, the updates
- should start coming on a nightly or almost-nightly basis again.
-
- 94/11/20
- NBETA012:
-
- What was to be whipped up in an hour ended up taking eleven
- hours today. But, the EVENTS are now coded! See the new
- NODE1.CTL for a set of examples; NanoBBS events are EASY to set
- up! Events can be forced or relaxed, and have great day/time
- flexibility!
-
- I have fixed a rather horrific bug in the Alt-E user editor - it
- would remain in local mode and the user would be unable to
- resume his/her session after you returned from the editor! This
- is now confirmed fixed (I checked).
-
- I have twiddled with the low level comm code some more in an
- effort to fix the problem where the carrier isn't dropped on
- hangup (sometimes). I am using an extremely straightforward
- FOSSIL call so this shouldn't be happening. However I have put
- in some debug code to help me identify the problem when it
- occurs. The little blue hangup window will be a bit ugly for a
- few days, sorry. There -is- a timeout in place now so that your
- whole system won't be brought to its knees when the program
- can't drop DTR.
-
- I have added a neat feature that I used to enjoy in my Maximus
- days: There are now 40 sysop-definable function keys. The keys
- F1 thru F10 will cause the compiled UFO files F1.BBS thru
- F10.BBS to be run according to the key pressed. This allows the
- sysop to display a UFO file to the user at almost any point
- within the user's call. The example files F1.UFO, F9.UFO and
- F10.UFO are included. F1 simply sets the user's time left to
- 60 minutes. F9 is my legendary(????) "Idiot User" key (grin)
- and F10 is a key I used to love way back in the dark ages when
- Blue Board had a function key to drop the user to a 5 minute
- warning no matter how much time actually remained...
- The other 30 keys are Ctrl-F1...Ctrl-F10 (filenames CF1.BBS to
- CF10.BBS), Shift-F1...Shift-F10 (SF1.BBS to SF10.BBS) and Alt-F1
- to Alt-F10 (AF1.BBS to AF10.BBS).
-
- I also found to my horror today that I had neglected to include
- a [timeleft] noun in UFO - this unspeakable situation has been
- ameliorated with the addition of a read/write [timeleft]. It is
- used in more than one of the example F-key files!
-
- 94/11/24
- NBETA013:
-
- Lucky 13! A bug that was only visible thanks to a faulty
- motherboard on my end is now fixed. I'll explain the problem.
- Every now and then, the tick interrupt would simply stop
- occurring, and with that so would the DesqView timeslices and
- the real time clock. Usually it would come back within a couple
- of seconds and if a user was online he could continue as if
- nothing happened (and in fact might not even notice at all if
- he's using a slow modem, thanks to buffer lag). But sometimes
- it would stay frozen long enough for the user to assume the BBS
- had crashed and hang up. When this happened in a file list, it
- would cause the BBS to lock up tight. I couldn't figure out why
- until I decided the problem might be in the code that actually
- writes out the file descriptions (since that's where the lockups
- occurred). Sure enough, there was NOTHING to prevent a loop
- which writes the spaces that tab out the formatted description
- from continuing should the user hang up. Worse, with no carrier
- it would wait on a dead com port for input, by now at a no
- longer valid baud rate, never to recover. Sometimes this loop
- would actually tight-lock the whole system although I still
- don't know why -that- happened. Anyway, I added some extra
- carrier checks to that part of the code and it should no longer
- be a personal problem on my own BBS, nor for any other sysop
- experiencing similar hardware trouble.
-
- This might possibly be related to a long-standing bug in Rancho
- Nevada which Allen Walker and Don St. Germain have long
- reported. I will keep you posted on the RN situation. RN and
- NanoBBS share a lot of code; much of Nano's code was originally
- developed for use in my doors, and many of Nano's code
- improvements have been re-integrated into Rancho. The two
- products are as intimately related as if they were sisters.
- That's why I am mentioning Rancho problems in the NanoBBS echo
- and in this file.
-
- A few logic bugs have been fixed, and I have re-organized memory
- usage such that you should have close to 80k more memory
- available, or better yet it can run in windows as much as 80k
- smaller than before!
-
- I have implemented a full internal TommCHAT in this version. I
- will be adding some UFO tools and canned main menu options to
- work with the chat before release, but for now if you enable the
- "multinode" keyword in the config file and specify a valid IPC
- path your users will be able to chat with each other (assuming
- you have 2 or more lines). There is a new "/" command at the
- main menu for the main chat system and a "W"ho's online command.
-
- 94/11/27
- NBETA014:
-
- Time is now frozen in chat mode! This is a rather important
- feature if, like me, you sometimes get into REALLY long
- protracted chats with users. I was getting fed up with having
- to manually give users time back, now you don't have to as time
- spent in chat "doesn't count".
-
- A bug reported a few weeks ago by Don St. Germain, that global
- newscans would go round in circles and never terminate, has also
- shown up in downloading .QWK packets. The cause is the same as
- that of the global newscan bug, and the fix is exactly the same
- as was the global newscan fix. I have also made the packet
- assembly procedure more informative for the user - if it turns
- out that there is still an endless loop in QWK (which is very
- unlikely now) at least users will know!
-
- I have also added the key code, node limits and the two
- unregistered limitations. Without a key, NanoBBS will only
- allow two copies of the BBS to run with the same system
- directory. It also limits all calls - even for the sysop - to a
- maximum of 60 minutes. You will each be receiving a USL
- (unlimited site licence) key within a day or so of receiving
- this beta. You'll also be getting a copy of the legalese that
- goes with the key; it applies to all keyholders including the
- beta team so read it!
-
- Another bug reported by Don is being investigated. What happens
- is that sometimes "K"ill defaults to an invalid message number.
- I have added code to read in a valid current message, but I
- somehow don't think this will solve the problem.
-
- There is now an option to blank out a user record in the user
- editor. As a safety feature, the user must be "deleted" with
- the [Del] key first! Use Alt-B to blank out users and remember
- blanked users are NOT recoverable!
-
- IMPORTANT: I have changed the syntax of the SYSOPKEYS and
- NEWKEYS keywords in the NODE1.CTL file!!! THE OLD SYNTAX IS
- GONE!
-
- Example of old syntax:
-
- ;This indicates that flags D, E and F must be set for sysop
- ;functions!
- SYSOPKEYS -------------XXX
-
- Examples of new syntax:
-
- ;Same thing!
- KEYS SYSOP DEF
-
- ;No keys set for new users
- KEYS NEW
-
-
-
- Also, the undocumented "plaque" command ($ at main menu) has
- been enhanced. It had been very buggy but was fixed around
- NBETA007 or so... I had forgotten to mention this...
-
- Overall, I am noticing fewer and fewer problems. As release
- approaches, NanoBBS is looking more and more like a finished,
- market-ready product. Keep the quality bug reports coming,
- guys!
-
- 94/11/28
- NBETA015:
-
- The product is still getting smoother and smoother. I'll be
- able to see my face in it pretty soon...
-
- I have updated some of the user booleans and completely
- documented their UFO syntax equivalents. All user booleans are
- now fully implemented.
-
- I have also fixed a few little problems I've noticed in NanoCHAT
- and added a generic QUIT, HELP and EXIT command (with no slash)
- so that NanoCHAT is more familiar to users who call Galacticomm
- systems. I think I may also add a super-k-neet0 /GO command
- feature as well; this would be incredibly powerful and would
- only take a little more extra code. I'll work on it after
- release... And maybe a clone of Galacticomm's Registry feature.
-
- I have fixed the problem where the screen is full of overwritten
- diarrhea after returning from a door. Seems it was
- repositioning the cursor at (1,1) on the nicely saved and
- swapped video screen. This fault is no longer possible.
-
- I have added UFO keywords to deal with frozen time - the
- [freezetime] and [unfreezetime] keywords do what their names
- imply, while the [iftimefrozen] keyword processes to the next
- mark if time is frozen.
-
- A problem with ANSi and RIP autodetection had been caused by
- placing the detect routines in a little-used overlay whose
- overlay overhead was too high for the timing required. This has
- been fixed by putting those two routines back where they came
- from, which unfortunately increases the .EXE size a bit. I have
- also increased the size of the overlay buffer by 4272 bytes in
- an effort to prevent future such problems.
-
- Don has noticed a few more bugs with the global new scans, and
- in my efforts to fix them I have also noticed a couple of logic
- bugs in the "S"etup Scans code. I am pretty sure that it is now
- impossible for the endless area scan loop to occur, and the bugs
- in the lastread code which caused some weirdness in repeated
- newscans are also fixed. The "S"etup Scans are also
- (apparently) fixed.
-
- NOTE: My BBS is 100% Squish based. I have not tested NanoBBS
- with *.MSG or JAM on any more than a trivial basis. I have been
- operating on the assumption that Mark May's MKMSG units are
- uniform over each format and I realize that this may not be a
- valid assumption. However, since I don't run *.MSG or JAM I am
- not able to exhaustively test these formats myself, therefore I
- must rely on the beta team to provide me with -quality- bug
- reports, meaning every little inconsistency between Squish and
- *.MSG and JAM that you can find needs to be reported as
- completely as you can. I have the source - I can fix May's bugs
- if need be. I just hope I don't have to...
-
- Several picky NanoCHAT bugs are fixed.
-
- 94/11/29
- NBETA016:
-
- No longer are "spaces" (actually nulls) displayed when using
- sysop keys in chat mode. Also it is no longer possible to
- recursively invoke chat when you're already in chat. No-one had
- reported problems with this but it could get annoying and may
- have caused a stack error if allowed to recurse too many times.
-
- There is now a freeze-time sysop key. Hit Alt-F to toggle
- frozen time.
-
- There is now a drop-to-DOS sysop key. Alt-J swaps out the BBS
- (giving you the maximum size window possible) and gives you a
- real live honest to Gh0d DOS prompt! Kool!
-
- I have also added the Alt-Z sysop key which switches between the
- three video modes (TTY, ANSi and RIP). It's still a bit buggy
- but I'm working on it...
-
- When you edit a user online, if you change his access level, his
- time limit is now immediately updated to reflect the change.
-
- I have added a %c parameter to the doors. You can use %c in
- place of the commmand.com comspec. A use for this is if you
- want to call up a door using a batch file (as is required by BRE
- and some others). Your COMMAND parameter would look like this:
-
- command %c /c bre.bat
-
- This is documented in the DOORS.CTL file that comes with the
- distribution archive.
-
- NanoBBS now attempts to auto-detect ANSi -before- the [Hit
- Enter] prompt whose purpose is to swallow spurious connect
- strings before they can fuck up the ANSi detector. If ANSi is
- not detected the BBS tries again after the [Hit Enter] and
- before the licence banner... It is at this point that the line
- is assumed to have settled down, and an attempt is made to
- auto-detect RIPscrip.
-
- 94/11/30
- NBETA017:
-
- I have added some improvements to Doors. For starters, you can
- now specify a "bogus" last name to be used when a user is using
- a one-word pseudo. You can also make NanoBBS use a user's real
- name when calling doors instead. I have also fixed a bug which
- wrote a "local baud" instead of the proper "0 baud" to
- DORINFO1.DEF.
-
- 94/12/04
- NBETA018:
-
- Now working full time on the docs. I will fix any more bugs
- that are reported but all "new" features are finished!
-
- I have noticed that users can post to a real name message base
- using a pseudonym via the QWK channel. This is no longer
- possible, doubtless to the delight of Scott James. QWK uploaded
- messages will now be posted with the real name or username as
- apropos.
-
- I think I have fixed an undefined variable in the wait for
- caller code which forced Nano to dump the user as soon as the
- connection was complete.
-
- Runtime errors are now written to the NANOBBS.LOG file.
-
- NANOBBS.LOG now reports node numbers for multi-node installations.
-
- It is now logged when you enter and exit chat with a user.
-
- You can now log all text seen in SysOp chat. Just hit Alt-L at
- any time during chat to toggle SysOp chat logging. This can be
- useful if a user has been abusive or threatening in chat, or if
- you want to save a phone number or some other piece of
- information the user gives you in chat. You could even capture
- a text file this way although it'd be slow and inefficient.
-
- I have also fixed a bug in the NIDX / BBS/F code which would
- abort with a runtime error after a certain number of file areas
- had been processed.
-
- 94/12/04
- NBETA019:
-
- I have added code to check for the presence of the IPC directory
- and try to create it if it doesn't exist. This is for when
- you've specified a subdirectory of a RAM disk - which would not
- exist at system startup and needs to be created. I think this
- is the cause of a Runtime 003 reported by John Richards.
-
- I have also located an undefined boolean (LOCALMODE) which caused
- NanoBBS to write a baud rate of 0 to DORINFO1.DEF and DOOR.SYS
- no matter what the user's baud rate was. I run mostly my own
- doors (and Allen Walker's) which aren't affected by this so I
- hadn't noticed, but Don St. Germain pointed out that this causes
- some doors to just sit there forever. I am sure that it is
- fixed now.
-
- The UFO syntax docs are finally complete! This beta comes with
- a comprehensive description of the UFO lexicon in UFO.DOC.
-
- 94/12/06
- NBETA020:
-
- I have made major changes to the installation system. It will
- now -update- your language files, not overwrite them, preserving
- your custom strings while adding in new ones. If your
- NODE1.CTL, MSGAREA.CTL, or FILEAREA.CTL files already exists it
- will not overwrite ANY config files. If any of the WELCOME.*,
- BYE.* or LOGO.* files exist, it will not overwrite any of the
- MISC directory files. It will now always overwrite all of the
- executables and overlays - this is important as a mismatched
- version between the executable and overlay would cause an
- immediate crash and probably a system lockup.
-
- It should be -safe- to install updated betas using INSTALL now!
-
- Some improvements have been made to the message base code. I
- have reduced the memory size for the message objects to just
- above the maximum amount that Squish will toss. JAM is
- unaffected by this as it was already more memory efficient than
- Squish or *.MSG. It should be possible to get more message
- objects into memory at once if need be, thus reducing the
- possibility of an out-of-memory error...
-
- I have fixed a cosmetic bug in the sysop bar (it'd display
- "ANSI" even if the user had RIP or TTY until an Alt-S).
-
- I have also fixed a couple of places in NanoCHAT where a user
- who was not in chat could have actions and messages sent to them
- - this is no longer possible.
-
- Until such time as I find a truly bulletproof way to handle the
- .NNS (Nanobbs Node Stat) files in the event of a crash, I have
- disabled rigid licence enforcement. Better business to err on
- the side of trust, and all that. I will actually likely release
- the software without fixing this.
-
- When will the software be released? My "bug board" is empty and
- I have no new features to code. Barring any catastrophic
- failure reports from any of you guys, the BBS will be released
- to the public as soon as I get the docs done. That means all I
- have to do is finish NANOBBS.DOC as the UFO.DOC file is now
- complete. I should also create a shareware licence doc and
- order form, I spoze. Maybe a NanoBBS RIP icon too. I am going
- to release NanoBBS, and THEN I am going to hoist a Christmas
- tree. That's my general (revised) schedule.
-
- 94/12/08
- NBETA021:
-
- Some long-standing logic bugs in the message reading code that
- have plagued me since I first coded Nano back in February are
- now FINALLY fixed! I tested it using various usernames and
- access levels and it is now smooth and stable and secure. This
- should address those message reading bugs reported by Don some
- time ago.
-
- I have also noticed that File Attach messages weren't working -
- it would let you upload the file but wouldn't attach it. This
- is also fixed.
-
- I have also fixed a logic bug in the ANSi detection - if ANSi is
- not detected on the first go, the user is no longer asked TWICE
- if s/he wants to proceed with ANSi anyway.
-
- 94/12/09
- NBETA022:
-
- Several people have reported a runtime 158 copying various $I
- files during INSTALL... err... MPATH shouldn't have been copying
- those in the first place! I have fixed a string compare which
- caused this. I am just amazed no-one lost all their BBS files
- as a result, as that's what happened to me when I tried to
- duplicate the scenario! Good thing I was only using a test
- scratch dirctory...
-
- I have also improved the intelligence and appearance of the
- UPDLNG program which updates the language files.
-
- There is some debug code in the file search code. If you're
- running in debug mode, I expect file lists and such will be very
- ugly for a little while. I'll take this out in a release or
- two, if it bothers you simply don't use debug mode.
-
- Most of today's effort has been in docs writing. NANOBBS.DOC
- is much improved with extended descriptions of all of the
- NODE1.CTL parameters and about half the MSGAREA.CTL parameters
- likewise better documented (with detailed examples) then those
- found in the file itself.
-
- 94/12/12
- NBETA023:
-
- I have been struggling with a bug in the code that gets a file
- description. Sometimes it would just lock right up! This
- occurred in file searches and new-file scans, and the fault was
- damned intermittent. But take heart; I have FINALLY pinpointed
- this most irritating bug (which has caused me to wake up to a
- frozen BBS many times) and the fix is included in this release!
-
- A temporary bug in which the 5 minute warning would tell you you
- had 300 minutes is fixed (i.e. it's back to displaying it in
- minutes, not seconds).
-
- John Richards has reported that /? in NanoCHAT doesn't offer a
- More prompt as it should - this is due to earlier efforts at
- removing such prompts from the automatic /m commercials. It is
- now fixed!
-
- The docs are getting there...
-
- 94/12/15
- NBETA024:
-
- This will likely be the last beta before public release.
-
- I have noticed maybe two times in the past eleven months, and
- again last night, a sharing violation in the code that updates
- the NANOBBS.SYS file. Although this is a near-freak occurrence,
- I have added code that actually checks to see it it's alright to
- open NANOBBS.SYS before updating it. If not it retries up to 25
- times. This is code originally developed for TommCHAT/NanoCHAT
- and is well tested, so the error should not occur any more.
-
- John Richards pointed out a problem where the User Editor screen
- binary is not shown. Additional path checking has hopefully
- fixed this.
-
- 94/12/15
- NANO100B/
- NBETA025:
-
- This beta is issued concurrent to the public release. Last
- night you got NANO100A; it's a non-beta and a non-release. Nuke
- it and use this.
-
- I have performed a number of cosmetic cleanups and added a
- feature to compensate for the lack of .RIP files: if a .RIP is
- not found, NanoBBS will try to fall back on a .ANS and then a
- .ASC before giving up. I'll make some nice RIP stuff in
- January.
-
- The docs are complete. They print fine on my Dataproducts 8070
- printer. Users of some HP printers may need to force 62 line
- (or longer) pages. Pages are 60 lines long.
-
- This is the last official beta until January. If any emergency
- bugs that cause your BBS to not work pop up I will fix them, but
- otherwise I am taking two weeks off for the holidays, and to
- write the THC Xmas Demo III. Feliz Navidad!
-
- 94/12/17
- NBETA026:
-
- Jerry Driscoll reports that his key no longer works.
- Interesting, since I haven't changed the key encryption
- algorithm at all. However, the ubiquitous path error leapt to
- mind as a likely culprit, so I added code to first check the
- SYSPATH path, then the path where the executable actually runs,
- then the current directory, and finally the root directory
- before giving up and declaring the key nonexistent. Jerry, let
- me know if your key works now.
-
- 94/12/18
- NANO100C/
- NBETA027:
-
- Although I am not officially developing right now, the bug
- reported by Ryan Clifford (not a tester) of a runtime 003 was
- deemed today to be too detrimental to post-release marketability
- to leave to an information report, therefore an emergency
- maintenance release has been issued. The runtime in question
- was caused by NanoBBS trying to write to a non-existent IPCPATH
- directory. Additional checking is now in place to disable all
- multi-line communications in the event the IPC path does not
- exist and cannot be created (as in the case of a non-existent
- drive).
-
- 94/12/27
- NBETA028:
-
- A problem reported by Don St. Germain has been fixed. Seems
- NanoBBS wasn't checking to see if the caller is local before
- trying to upload or download. The protocol engine has horrible
- DCD checking (which I cannot fix without violating copyright) so
- I added code to disallow uploads in all protocols, and downloads
- in protocols other than ASCII, in local-mode logons. Your
- remote callers will be unaffected. I should point out that this
- is how Maximus behaves.
-
- An extremely unimportant cosmetic fix: when you go into
- NanoCHAT, it now says "¡Hola!" (properly punctuated Spanish)
- instead of "Hola!" (poorly anglicised). This will only take
- effect if you manually update your language files. If you want
- to do it yourself, the upside-down exclamation point is ASCII
- 173.
-
- RIPscrip parsing has been improved. NanoBBS will try to keep
- all lines of RIPscrip code to 65 characters or less, where
- practical, and there is now only one "!" per line. Should be a
- bit faster on the user's end, and there should be less problems
- with less-than-perfect emulations of RIPscrip on the terminal
- end.
-
- I now have a NANOBBS_.ICN file - it's a really slick looking RIP
- Icon of the NanoBBS graphic logo...
-
- A [rip_fill_pattern] example in UFO.DOC had several incorrect
- decimal representations of the binary fill pattern (the sort-of
- hexagonal pattern in particular). This documentation error has
- been corrected.
-
- Several RIP colour parameters were not being properly parsed by
- NCOMP.EXE. This is now fixed.
-
- Don reports that the BBS has frozen a couple of times at
- the get-filenames prompt when downloading. This prompt uses the
- same input routine used throughout the BBS for string entry, so
- I find it incredible that a bug should cause it to only freeze
- at that one prompt. I suspect it may be due to a user hanging
- up at that prompt, possibly not knowing what to type there, and
- the input routine may not be handling all carrier-loss
- scenarios. I have added additional DCD checks and a line of
- code which prevents FOSSIL I/O calls when the carrier has been
- lost. If this clears the problem up, great, but I still
- haven't duplicated it here.
-
- I have also added code to explain two common runtime errors and
- how the sysop can resolve them without calling on me for
- support. Runtime 003 is a path not found error - this is
- usually because of pathing problems, and although I have added
- much code recently to compensate for sysop confusion over paths,
- the possibility of some arcane pathing causing a problem will
- always exist, so a reminder has been added when this error
- occurs to check the paths in the config. Similarly, Runtime 100
- is an attempt to read past the end of a file, usually caused by
- running a new version without running CONFIG first. A reminder
- is also displayed (and logged) when this happens.
-
- 95/01/02
- NBETA029:
-
- I have added a new feature: "File Groups". It is a convenient
- way to classify groups of file areas by general fields of
- interest, to make it easier for users to find a particular file.
- This feature has been added in response to criticism from some
- of my BBS's users that THC's 250+ file areas can be confusing to
- wade through in search of files of interest.
-
- I have also sped up the behind-the-scenes login code quite a
- bit. By that, I mean the code that updates the user, updates
- the logs, loads up the user's limits definitions, and a slew of
- other necessary things that both user and sysop are blissfully
- unaware of. There is still a way to go on this, but I should
- have truly fast logons by the end of January.
-
- 95/01/03
- NBETA030:
-
- A cosmetic blemish in the language files has been fixed. SysMsg
- 687, which informs the user that the BBS is searching for a file
- (in a find-file search) now has a couple of carriage returns
- after it, giving a neater appearance.
-
- Another item of concern was that the backspace key didn't go up
- to the next line if it occurred at the beginning of the line, as
- is expected in a screen editor. This little featurette has been
- added for your convenience...
-
- 95/01/06
- NBETA031:
-
- I have added a debug line which displays when the hot-keys
- handler has been invoked. I would like to find out if a
- rarely-occurring flaw in the hotkey handler (or routines that
- call it) might be responsible for occasional lockups such as
- those experienced by both Don and myself. No other new items
- are in this release. If the BBS locks on your user, see if the
- last thing the BBS wrote to the screen was "Hotkey Handler
- Called" or something like that...
-
- 95/01/09
- NBETA032:
-
- I have added much code to make it easier to handle the external
- ASCII, ANSI, RIP and compiled UFO files that NanoBBS uses. In
- addition I have expanded the random welcome screen concept
- sixfold to allow random ANSI, ASCII and RIP in both the logo and
- post-logon welcome screens. I have added a section to the
- NANOBBS.DOC documentation file explaining all the files in the
- /MISC directory. Hopefully this will help... And, the
- practical upshot of all this is that the problem mentioned earlier
- in this file (for NBETA010) about missing RIP files causing an
- error and user confusion, can not exist anymore if an ANS or ASC
- file is available!
-
- I have greatly streamlined the code which sets up NanoCHAT upon
- a user's logon. Once a valid password is received, users should
- be logged to the system in about 1 second, not 10. Much faster!
-
- If a .RIP file is being displayed, it is now "non-stop" so that
- spurious "more" prompts will not spoil long displays.
-
- I have created BYE.RIP. I think it looks pretty cool.
-
- 95/01/15
- NBETA033:
-
- Most of this beta's accomplishments are infrastructural, to
- increase speed, stability and sysop convenience.
-
- Some more RIP improvements include code to always start a .RIP
- file on a new line, and two new languagefile entries for the RIP
- languagefile only which reset the windows at strategic places.
-
- I have streamlined the efficiency of the ASCII/ANSI/RIP/UFO file
- output by about 50% on heavily-ANSI-laden files, and by much
- more than that on files with little or no ANSI. It should be
- noticeably faster now, especially in local mode!
-
- I have modified the same routine so that when it is aborted with
- a hotkey, the colour is always reset to light gray. Of course,
- subsequent language stuff will probably change this right away
- but at least you won't be stuck with an unwanted background
- colour if you should happen to abort an ANSi with a background
- other than black.
-
- I have also added code to automatically select a memory-based
- rather than a disk-based text file output wherever possible.
- This will further speed things along and add UFO syntax
- capability to many more files.
-
- Fixed a runtime that could occur if the "Use FileGroups" option
- is enabled but no file groups dat file exists... it will now
- instead do a normal file area list instead.
-
- Fixed a bug that would cause a runtime if FILES.BBS exists but
- has a length of zero.
-
- Added the following Sysop file-area functions: Adopt/Redescribe
- (with auto-DIZ extraction), Global adopt, hide, raw directory,
- override path.
-
- If you have sysop access you can now download any file on the
- system no matter where it is, so long as it exists. It doesn't
- even have to be in files.BBS. Just use the override. I'll add
- move and kill later.
-
- I have completely re-written the QWK scanner to be tighter, and
- to leave the message base "open" during the scan instead of
- opening and closing it for each message. QWK packing should be
- MUCH faster now, even faster than most Fidonet message packers
- on a per-message basis. By the way, the dots you see during QWK
- packing use a dot to represent four messages. So if you see
- five dots, 20 messages have been packed. This has been reduced
- from 10 messages per dot to increase the psychological illusion
- of speed, further enhancing the real speed improvements.
-
- A little while ago, in an effort to improve speed, I created a
- LIMITS.DAT file which is derived from LIMITS.CTL and compiled
- when CONFIG.EXE is run. Unfortunately, it's not intuitive that
- changes to LIMITS.CTL are not automatic so I have added code
- which will detect when LIMITS.CTL has a newer filedate than
- LIMITS.DAT, and will recompile it automatically when the user
- logs on if this is the case. Prettycoolhuh!
-
- I'm late for my January 15th release. If you notice any new
- bugs PLEASE let me know RIGHT AWAY so I can get them fixed and
- get 1.01a out to the public SAP! If I don't receive any
- non-trivial bug reports by the 19th I will place 1.01a (which
- will be this beta plus any minor bugs fixed in the interim)
- on THC and on several FTP sites on Friday, the 20th.
-
- 95/01/16
- NBETA034:
-
- OK, Don has pointed out a very nasty bug, which I have noticed
- on my own system, i.e. it took no effort at all to duplicate. A
- new feature I had added, code named "Gerg Yur Up" (from a
- message generated by an external program by Greg Shergold which
- does the same for QuickBBS) would cause the BBS to go into
- ignore-carrier and out of dump-the-user-cause-he's-out-of-time
- mode, putting him back at the main menu, with a ridiculous time
- left (6:25:and change) and unable to function because the BBS
- was in local mode, as it should be after it thinks it has turfed
- the user. However, GergYurUp wasn't working properly either,
- and as a result it was leaving the BBS in local mode where it
- wouldn't hang up the phone. Big problem. FIXED! I hope.
-
- BTW, it's not documented yet, but Alt-G toggles GergYurUp. I
- will probably call it "Sysop Waiting". What it does is, after
- the user has logged off, the BBS will play a PagePlay tune to
- signal that the BBS is available, and it will wait for the sysop
- to hit a key, or for 5 minutes, whichever is sooner, before
- re-cycling. This is to give the sysop a chance to log on or do
- important maintenance on a busy day without having to commit the
- social faux pas of throwing a user off... When Sysop Waiting is
- active, a flashing black colon ":" on a cyan background appears
- in the lower left corner of the screen.
-
- Also, I am pretty sure that if, for any reason, a user ends up
- with negative time left, the BBS will always display 00:00:00
- instead of the negative time. This used to cause serious
- cosmetic problems as the resultant time remaining would scroll
- off the screen, which in the sysop bar is one line long, and,
- well, if it's happened to you, you know what it looks like.
-
- 95/01/19
- NBETA035:
-
- This will most likely be NANO101A.
-
- As requested by Don, a NEWS.* file has been added. It is
- displayed immediately following the WELCOME.* file. The docs
- (NANOBBS.DOC) have been updated to explain this. It's intended
- as a general system news file, and it follows the same general
- rules as most other system textfiles: first it looks for .BBS,
- then (.RIP/.ANS/.ASC) as apropos, .TXT, no extension, and so
- forth.
-
- The GergYurUp problem wasn't fixed after all. Now it is.
-
- I have greatly streamined the process of checking for new email
- at logon. I used to get mystery lockups during this, I think
- these are gone now.
-
- Additional new system textfiles: ULPOLICY.* and DLPOLICY.* are
- displayed prior to uploads and downloads, respectively. Sample
- .TXT files are included. They are used to indicate system
- policy regarding uploads and downloads, for the user's
- re-edification. In terms of encouraging adherence to the BBS's
- rules, these work well in WildCat and other BBSes which use
- them. So now we've got 'em.
-
- I am still chasing a really mysterious bug where the BBS
- sometimes won't drop DTR to hang up. It should be; I have debug
- code in this beta which shows that the INT 14 drop DTR call -is-
- being executed. I have temporarily killed the non-FOSSIL
- version of the DTR drop and the if-then which decides which
- method to use; I suspect I could solve the problem better the
- other way around, by not using the FOSSIL at all...
-
- 95/01/23
- NBETA036/
- NANO101A:
-
- This finally fixes the bug where the BBS would go into local
- mode and stay there forever (until the sysop logs the user off
- manually) after a user runs out of time. This is the
- "mid-January" official release.
-
- There's also a new "S"earch command off the files menu. It is
- like "F"ind but it searches description strings, at a bit of a
- premium in speed.
-
- There have also been a couple of minor bug fixes in NEd. Sysop
- keys are now properly trapped without generating phantom
- characters, and ^Y will no longer let you delete a line when you
- are already on the last one.
-
- 95/02/05
- NBETA037:
-
- I have made some improvements to the Account Expiry code, in
- order to facilitate some policy changes on my own BBS.
-
- The Alt-G SysopNext feature now works properly.
-
- I have modified logic somewhat such that "hidden" files
- (actually, files with no FILES.BBS listing including .QWK
- packets) are "free" - ratio is not updated, time is not
- affected, and download limits are ignored. This should fix the
- problem where users exhaust their ratio downloading QWK packets,
- plus allow sysops to have free hidden files, for whatever
- nefarious purposes this will serve.
-
- 95/02/18
- NBETA038:
-
- I have added much to this beta, and I will give plenty of time
- to work the bugs out before I release.
-
- Firstly, Nano now has a "Chat Prompt" mode for multinode use, in
- which most prompts in the BBS will allow display of
- current-channel chat traffic (silencable) and even participation
- in chat from a prompt! The /M auto messages can now be silenced
- by users of a specific access level, so that you can make this a
- "registered users only" feature if you want. I'll release this
- feature to Joe Q. Public when I have it integrated with the UFO
- syntax...
-
- Secondly, there is now a "Review New Users" mode. Hit Alt-R at
- the User Editor screen and any user the BBS quasi-intelligently
- deems non-validated is displayed for you to consider, nuke,
- edit, or quickly promote to a preset access level.
-
- 95/03/15
- NBETA039:
-
- Some users reported new-email check not finding their email.
- This is fixed.
-
- I think I have also fixed some sticky mystery crashes. Hotkeys
- oughta work better now. They DON'T work at the logon prompt,
- and this is deliberate - I want any extra connection characters
- to be swallowed before the user is expected to input anything!
-
- In addition, abortability of lists and other output with respect
- to NanoCHAT prompt messages is now considerably improved.
-
- 95/03/20
- NBETA040:
-
- A really weird lockup has been isolated and, we hope, fixed. It
- seems to only affect multi-node installations.
-
- I am going to attempt to refine and debug the transfer
- protocols. Please report in detail any problems you or your
- users have with file transfers that cannot be explained by user
- dummkopf error ;-) In particular I need to know which terminal
- programs are giving users the most problems, especially in
- Zmodem, so that I can modify Nano's Zmodem to be more friendly
- with not-quite-perfectly-compliant terminals... And I am the
- first to acknowledge that the Zmodem in Nano itself may be
- deficient; this is what I most want to address and fix at all
- costs!
-
- 95/03/22
- NBETA041:
-
- There is extensive debug messaging in the logon procedure so I
- can ensure that I have really gotten rid of that lockup bug at
- logon - it'll be gone in another beta or two.
-
- I have also added two UFO verbs and a UFO adjective. See
- UFO.DOC for updates - the verbs are [ghostpost] and
- [ghostpostp], and the adjective is [creditcard]. It is now
- possible to write a UFO script that accepts credit cards,
- upgrades the user's access based on this, logs the fact, and
- informs the sysop BY EMAIL! Try THAT with an external Visa
- door!!! I will be writing such a script for my own use before
- next beta and will include it in the distribution archive.
-
- 95/03/30
- NBETA042:
-
- A crash on my machine forced me to fall back on a backup of
- source from before [ghostpost] therefore it will not work as
- documented for a beta or two. This has also given me the chance
- to identify some inefficiencies in code and clean them up, so
- all is not disaster and chaos. (on a personal note, I know what
- caused my crash. I'd like to tell you where NOT to buy an IDE
- multi-I/O card but I don't feel like being sued today...)
-
- I have added a little code to the chat system code which will
- prevent a rare runtime error from occuring when two nodes try to
- write to the same temporary filename simultaneously. If a
- collision is detected the later task (meaning: the one which
- would have crashed before this mod) will try again with a
- different filename, preventing the BBS from booting the user off
- with a Runtime Error 5.
-
- A new "post rewards" system has been implemented. When a user
- enters a message, a variable amount of time is deposited to the
- user's time bank balance. There are two tiers so you can reward
- posters of long messages more than for trivial posts, and you
- can also set the cutoff of what counts as a "long" message.
- This does not count for QWK uploads (yet).
-
- Another new feature, related to the above, is the capacity to
- "freeze" time during message posts.
-
- 95/04/30
- NBETA043:
-
- After a month of poring over the Zmodem code (which I did not
- write) I believe that I have successfully identified and
- eliminated the problems some users have been experiencing in
- downloading with Zmodem. With that, this Public Beta is
- released.
-
- The UFO [creditcard] token now recognizes Discover cards.
-
- 95/05/02
- NBETA044:
-
- NEd (NanoBBS Full Screen Editor) now has a Ctrl-A key which
- inserts a space into the current line at the cursor position.
-
- A runtime PagePlay and a bunch of page tunes are now included.
- PAGEPLAY.DOC explains how to create the tunes and use
- PAGEPLAY.EXE to test them without having to load up the whole
- BBS. PagePlay itself is a stand-alone utility which, in its
- full implementation, can also play and convert tunes for other
- BBS software. The two hardcoded tunes are PAGE.PPL (played when
- a user pages) and SOS.PPL (played when a fatal error occurs).
- Any other PagePlay tune can be played most anywhere through the
- use of the [pageplay] UFO keyword. TUNES.ZIP contains the tunes
- that come with PagePlay. Give them a try, some are very good,
- others are quite short and designed specifically to be used as
- sysop pages or error/warning messages.
-
- The UFO [ghostpost] and [ghostpostp] keywords are back!
-
- 95/05/09
- NBETA045:
-
- There was a place in the QWK packer where it could possibly hang
- if the user got impatient and dropped carrier while the packet
- was assembling. (I hope) this is now fixed...
-
- A new CONFIG token has been added - SHOWUPLOADER, when active,
- will tell NanoBBS that, after a user uploads and describes a
- file, the user's username should be added to the end of the
- description, e.g. "Uploaded by: {UserName}".
-
- 95/05/13
- NBETA046/
- NANO110:
-
- This is a public general release. A freak hangup bug has been
- corrected. A new config token "HIDESYSOP" which refrains from
- including sysop logons in user-viewable logs, has been added.
- The docs have been updated. Some default textfiles have been
- changed (un-THC'd). The NNS files and bogus licence violations
- are toast - I've gone to a completely different, much more
- stable licence enforcement system! Also all of the cosmetic
- features (such as the $ info) which had been lost in the crash
- last month are now restored!
-
- 95/05/27
- NBETA047:
-
- MFB now supports an INCLUDE parameter, making it much easier to
- rotate CD-ROMs. MMB and CONFIG also include this parameter.
- Several include files are, well, included as examples. This
- should greatly simplify configuring multi-node systems.
-
- Formerly, when you went /ks in NanoCHAT, it would report -1
- bytes free on the IPC drive. This has been corrected. This
- problem also existed in the TommCHAT product and has been
- corrected there too.
-
- Four new UFO tokens have been added which will enhance
- flexibility for sysops who want to have multiple logons per
- online session. The [logoff], [bye_nohangup], [donthanguponbye]
- and [hanguponbye] tokens let you control whether the BBS
- actually hangs up after a logoff.
-
- 95/06/03
- NBETA048:
-
- I have improved the intelligence of the line counter quite a bit
- - it now detects unintentional line wraps and bumps the counter
- so that More prompts occur like clockwork even when some lines
- are longer than the screen! Likewise, the message headers now
- stay on the screen 'til the first More prompt, instead of the
- BBS displaying the full 24 lines of a message's text and shoving
- the header into limbo...
-
- I have also finally solved the last of the problems with
- searching new e-mail at logon. Note that this feature only
- searches the private email area; to scan all areas users must do
- a global new scan (which can be time consuming).
-
- 95/08/07
- NBETA058:
-
- There have been a great number of fixes since NBETA048, so many in
- fact that there hasn't been time to type them all out here. The
- bug where the last message area defined would not show up in a
- global scan or QWK scan is now eliminated. The bogus "Missing"
- files are gone. A number of very silly defaults in the
- distribution archive have been corrected. The BBS will no longer
- crash on account of your not having run BBS /F beforehand. There
- are now 32 user keys. And yes, Don, you can now select whether you
- want new-personal-mail scans to include all areas in the user's
- scan list or just the defined private area.
-
- 95/08/08
- NBETA059(b): (Some testers may have 059, no letter, which lacks a few of
- the release 1.12 features)
-
- There is now true support for non-disclosed gender. Assuming you
- have set gender as an optional user field, users may choose, during
- the new-user application process, to protect the privacy of that
- information by not disclosing the gender. Non-disclosure is
- treated by NanoBBS as a third gender. Thus, undisclosed users
- cannot access message areas that are restricted to women only or to
- men only. You may want to apprise your callers of this fact before
- they apply for access.
-
- The docs have also been overhauled. To be different (NOT to be
- politically correct!), all personal pronouns are now in the female
- gender (grin). I did a series of search-replace macros, so it's
- unlikely I missed anything, but if you notice any serious
- grammatical or spelling errors (NOT stylistic warts) let me know.
-
- An inconsistency in the JAM code (versus the Squish and *.MSG code)
- has been detected that would cause, in JAM areas only, the BBS to
- fail to display the last line of each message. This is now fixed.
- Also apparently affecting JAM areas only was that it was possible
- to create a message -20 if it was the first message. Only problem
- was, NanoBBS chokes on negative numbered messages. Both the
- illness (the negative numbered messages) and the symptom (Nano's
- tendency to choke on them) have been cured.
-
- The non-FOSSIL serial support has finally been added! So far, beta
- testing has been done only with a USR Sportster V.32bis internal
- modem, but the results are very solid. NanoBBS supports up to
- COM16 by itself; my alpha testing so far has encompassed COM1 and
- COM2 only. You can set non-standard addresses and IRQs for any of
- the 16 ports. If no port config has been set (i.e. if you're using
- a NODE1.CTL that came with an earlier version) NanoBBS will default
- to using settings that will work with virtually every COM1 and COM2
- in the world and with many internal modems on COM3 and COM4
- (assuming their IRQs mirror those of COM1 and COM2). See the docs
- and NODE1.CTL for configuration examples and information.
-
- New code has been added to accommodate a few non-standard FILES.BBS
- files I have encountered on several CD-ROMs. The new tokens in
- FILEAREA.CTL will allow you to run many CD-ROMs of questionable
- compatibility virtually straight out of the box! There are now
- five example include files illustrating the use of these tokens,
- and with only some paths to be changed you could use the examples
- directly with the CD's they are designed for (Night Owl 15 and 16,
- Monster Media 10, PC-Sig 14 and The Hacker Chronicles ][).
-
- 95/08/21
- NBETA060:
-
- This beta includes the first of the new features for 1.2: it now
- supports UFO labels up to 16 characters in length. This is a big
- improvement over [label BF] and [goto BF] - now you can put
- [label bigfat] or [goto bullfudge]. Labels are case insensitive.
-
- NanoBBS no longer updates lastreads when checking for personal
- email.
-
- The BBS now handles "free" files much better. It no longer adds
- the byte counts for "free" files to the user's download totals,
- precluding later downloading of non-free files. Before, you could
- download all the free files you wanted but if you went over your
- limit on them, tough. The BBS now also hides the "/b" used to
- indicate a free file.
-
- The BBS no longer skips ratio checking on CD-ROM downloads. This
- one drove me spare until I nailed it.
-
- The BBS can now handle even more kinky versions of FILES.BBS. For
- example, on some really wacky CDs, FILES.BBS includes file dates
- and sizes, with the file size working backward until it actually
- touches the filename, like this:
-
-
- SMALFILE.ZIP 50065 94-01-01 This is a smaller file.
- FILENAME.ZIP1678082 95-01-01 This is a file.
-
- Previously, Nano would deem FILENAME.ZIP1678082 a "missing" file.
- Now, it detects this occurrence and splits it up, neatly displaying
- the file size (or not, if you have {correctly} specified the
- PCBSTYLE keyword in your filearea.ctl).
-
- A number of minor cosmetics have been cleaned up, mostly in the
- form of refinements to the file listing bug fixes of the 1.12a
- release.
-
- There are two more CD-ROMs included in the example control files:
- Night Owl 17 and Night Owl BBS Shareware (from 1994). I have a
- source for a great many more and plan to include all of them if I
- can.
-
- 95/08/31
- NBETA60B:
-
- This, along with NBETA60A.ZIP (not written up here) is a sort of
- "emergency beta" for John Barnard. The high speed file searches
- are totally, totally rewritten. The file area file has a new
- structure so the file areas MUST be recompiled using the new
- MFB.EXE. Also run BBS /F after MFB has been run! The NANOBBS.HDX
- and NANOBBS.HXA files are obsolete and can be deleted now. The new
- MASTER.FIL generated by BBS /F is larger but it is a plain ASCII
- file and delivers comparable speed on new-file searches and
- filename searches, and VASTLY GREATER speed on description
- searches! This rewrite eliminates the remaining extended
- description display bugs and all runtime errors associated with
- description searches! And it's a whole lot smaller in code size to
- boot!
-
- A note to the programmers on the beta team - Pascal purists will
- tell you that the GOTO statement is a travesty, an abomination in
- the face of God and J&W. BULLSHIT. In rewriting the file
- searches, the use of just three GOTO statements eliminated nearly
- 20k of redundant source, plus at least 10 indices and booleans in
- the remaining code, enabling me to use a single universal search
- routine where I had needed three. The source is more readable and
- the resulting program is faster and far more solid and airtight.
- There are some instances in programming where a GOTO is the
- absolute best way possible to get things done. Never let anyone
- tell you otherwise.
-
- Also if you experience weird wrapping problems with the extended
- descriptions in the file lists, add a ~, embedded keyword to the
- beginning of [sysmsg 603] and [sysmsg 604] in all your language
- files and then recompile them with NCOMP. ~, forces a carriage
- return if the cursor is not on the leftmost column already. This
- is documented in NanoBBS technical bulletin #2.
-
- Once I add multinode paging, I shall release NanoBBS 1.14.
-
- 95/09/09
- NBETA61C:
-
- This is the last beta before the 1.14 release (barring a
- catastrophe).
-
- I have added Paging and Telegrams to the NanoCHAT features and
- added many of the TommCHAT bug fixes to NanoCHAT (in fact I ported
- the entire SmartActions code from the enhanced TommCHAT 1.13b
- product into NanoCHAT lock, stock, and barrel!). If you have
- multinode enabled, I think your callers will really appreciate
- these new features - mine do!
-
- I have also strengthened the WAIT4IT function that waits for a
- locked file to come unlocked before attempting to open it. Once in
- a while a user may experience a delay of up to 2 seconds in a
- multi-user function - this is WAIT4IT doing its job: the BBS would
- have exited with a runtime error otherwise.
-
- File Flagging is now much faster thanks to a last-200-files list
- buffer. Most of the time, users will no longer have to wait for
- NanoBBS to search the whole index file when flagging a file or
- adding it to the batch from the batch menu. The BBS will first
- search an in-memory list of up to 200 most recently listed files,
- and only if that fails to turn up anything will the BBS attempt a
- global search of MASTER.FIL.
-
- A new bug, where the Flagging option was absent from the More
- prompt in file lists and searches, is now fixed.
-
- 95/09/11
- NANO114:
-
- Owing to a lack of feedback (and rotgut) from Don St. Germain, I
- dub Nano 1.14 ready for prime time!
-
- Last minute tweaks:
-
- I noticed that if a caller enters a message and then aborts it
- without saving, the BBS wouldn't resume the clock if you have
- POSTFREEZETIME active. Now it does.
-
- I've noticed a *possible* place where NanoCHAT might drop a message
- every now and then - under certain circumstances it could write the
- IPC itself but not update the receiver's .STC (stack) file. You
- may or may not notice an improvement in NanoCHAT's reliability,
- depending whether or not you were experiencing such a problem
- before.
-
- I have also added PAGE and TELEGRAM as internal NanoCHAT commands,
- so they are accessible from within NanoCHAT as well as from the
- canned who's on menu. They're also available in UFO as
- [canned_pageuser] and [canned_telegram]. I may never add un-canned
- components of these two tokens as they are a bit complicated. Then
- again, I might yet. Time will tell.
-
- 95/10/14
- NBETA062:
-
- I have spent the last few weeks taking a breather from serious
- programming after the release of 1.14, and I am now working on
- 1.16. Fitting with the new version numbering scheme, all 1.16
- betas will be version 1.15.
-
- In this beta, I have completely rewritten the global newscan code,
- which should eliminate many nagging problems in navigating through
- global newscans, email checks, and even general message reading.
- I've alpha tested this a lot today and found that all the old
- familiar quirks are blissfully absent... It also turns out that
- there was a logic bug in some of the code that was not rewritten,
- which had occasionally caused an elusive runtime 004. This is now
- fixed.
-
- A new feature has been added: you can now type the message number
- of a message you want to read directly at any read prompt. This
- way you can skip past large numbers of unwanted messages or quickly
- jump back or forward to a message whose number you know. Most
- other BBS software already had this; Nano should've had it in its
- first release, it has it now.
-
- There's a new sysmsg 356 to service this feature.
-
- In addition, when you flag a file from a file listing, you get the
- "More" prompt again. I will do some cosmetic work on this to
- minimize the number of lines scrolled out during a file flag; I
- might even implement some sort of a screen capture/redraw for a
- zero-disruption flag prompt!
-
- 95/10/28
- NBETA063:
-
- A couple of bugs in the new newscan code are fixed: it no longer
- shows private messages you shouldn't be seeing, and it no longer
- gets stuck in a prompt loop without going to the next base.
-
- I have streamlined a number of things resulting in 28k less heap
- consumption. This means NanoBBS will run in tighter memory
- quarters than in the last beta.
-
- I have also fixed a cosmetic bug in the Telegrams where, when you
- received a telegram, you'd see your own name and node as those of
- the person who'd sent you the telegram.
-
- Quite some time ago, I added a new PagePlay kernal to the BBS for
- its chat page music, but neglected to add in hooks to the modem so
- that the user could abort the paging. Up until this beta, only the
- sysop could abort a page. Now, the user can too, as was originally
- intended.
-
- Additionally, three new UFO keywords have been added, [writebang],
- [noabort] and [abortok], and the UFO script compiler itself has
- been updated slightly as well, with a very rare infinite loop bug
- discovered and fixed.
-
- And, I have at last located the line of code that sent a carriage
- return but failed to increment the line counter (in file lists
- only) resulting in file lists that often scrolled 30 or 40 lines
- (when you're set up for a 24 line page) before you'd get a More
- prompt. This is now fixed. At the same time, I have added extra
- intelligence to the file description wraparound code: it will not
- auto-wrap an extended description that has a high-bit character or
- certain regular ASCII characters as the LAST character of the first
- line of the extended description. If it detects such a character,
- it will assume that the extended description has extended ASCII art
- or a box or some other large column-sensitive graphic, and will
- honour its line breaks.
-
- 95/11/01
- NBETA064:
-
- There was an error in UFO.EXE where it would compile a GHOSTPOSTP
- as a GHOSTPOST and vice versa, resulting in public messages when
- you intended private. This is now fixed.
-
- I have further refined the user editor's user flags editor. It's
- much nicer now.
-
- I noticed that occasionally, when a user encountered a
- [harddownload] UFO keyword, it would fail to give the user the
- file, either because it was not in a FILES.BBS or because the
- user's ratio was unfavourable. Since the most common use of
- [harddownload] is for files that ought to be free anyway (such as
- local FAQs, registration forms, must-have files like RIP icons,
- etc) I have added extra code that forces "free" mode when the
- keyword is invoked, so that files thus downloaded do not count
- towards the user's download totals and the ratio is not checked.
-
- 95/11/04
- NBETA64E:
-
- This is a response to a serious bug reported by Don St. Germain, in
- which the Chat Prompt would generate a runtime 104. I was never
- able to duplicate the problem but I strongly suspect an aberrant
- FILEMODE caused the error, so I went over every FILEMODE directive
- in the source by hand and made a few changes which should fix any
- further "not open for input" problems (i.e. runtime 104).
-
- I have also enhanced the Sysop "file directory override" and "raw
- directory" options at the file menu. The raw directory now lists
- directories and the override lets you use . and .. as well as
- partial paths.
-
- 95/11/11
- NBETA065:
-
- Some problems with flagging files and downloading files that are in
- the last-100-listed buffer have been fixed. The last of the
- cosmetic fixes for file lists have been applied.
-
- 96/01/06
- NBETA066:
-
- OK, NanoBBS development is underway once again after a busy
- Christmas season which left me with no time for programming...
-
- The runtime errors which commonly threw users out of NanoCHAT have
- been obliterated! I have put in beta code which prints a warning
- to the sysop screen only whenever the BBS thinks a runtime error
- would have occurred if this fix were not in place. After a short
- time, you will notice how bad the problem was.
-
- I have added two more levels of indexing to the master file index.
- MASTER.PRE is a pre-index which is exactly 729 records in length
- and points to offsets in MASTER.IDX. MASTER.IDX contains
- filenames, file area numbers and offsets into MASTER.FIL. Although
- I have not yet incorporated the code that searches these new
- indices, the preliminary code (still in testbed) has shown
- incredible results; any file in the system can be located (even
- with wildcards) in less than 1 second on my 486-33. That's in a
- MASTER.FIL of 2.3 megs in size, indexing 39,000 files totalling
- over 2 gigs. Not bad! Next beta will have the new searching
- capability in file flagging and in the Download command and
- File-Search commands. The new-file scan and description scan will
- necessarily continue to use the flat-file technology currently
- employed in all file searching, which takes a noticeable time to
- find a file. Although, the flat-file system itself may be improved
- before 1.16.
-
- There is now code in place to free disused heap blocks (such as
- in-memory FILES.BBS lists) prior to memory-critical operations such
- as message entry and saving. This reduces overall real-world
- memory requirements and makes out-of-memory runtimes much less
- likely.
-
- This is the first beta of 1996! Note new copyright dates.
-
- 96/01/06
- NBETA067:
-
- OK, the new searching code is in place in the file area. The
- "D"ownload command, the flagging from within a file list, and the
- Batch Menu are affected. All three of these functions are now much
- faster. Next beta, the "F"ile Search (by filename/wildcard) will
- also have the same technology.
-
- 96/01/07
- NBETA068:
-
- Oops! The new file searching code doesn't search the current
- directory. This isn't a problem if you're in an indexed file
- section (as all files sections are indexed) but if you're in an
- override directory, that directory may not be indexed, in which
- case a download would fail to find your file. I have added code to
- use the old method when adding files to the batch from an override
- directory.
-
- As promised, the "F" command at the file menu now uses the new
- search code. It's now an FFFF (Fucking Fast File Finder)!
-
- One thing, the new search code has a weakness: because it derives
- its speed from indexing the first two letters of each filename
- (effectively resulting in a tree with 729 branches at the first
- level) it must search more files if less than two letters of the
- base filename are supplied by the user. For example, if a user
- wants to download EX*.*, the BBS positions its index pointer at the
- beginning of the files beginning with EX and finds them all very
- fast. But, if the user requests E*.DAT, the program must search
- through all the files beginning with E. This may not be so bad
- because all the E's are in one block in the index, but in a large
- BBS it may be noticeably slower. The worst case is when a user
- supplies only an extension, e.g. *.PHK. Then, NanoBBS must search
- the entire database for files with a .PHK extension, and is only
- marginally faster than the flat-file code.
-
- The flat-file search engine, BTW, is still in place because it is
- still the fastest way to do a description search without creating
- another huge index file of description words. The same applies to
- searches by date.
-
- Override keys like the ones you can program for the canned main
- menu are now also available in the File Section. They work the
- same way, except that in the file section you can only have up to
- 30 of them, and the keyword in the config file is FOVRSTR instead
- of OVRSTR.
-
- Doors can now be addressed by their number instead of just the key.
- The prior limitation used to effectively restrict NanoBBS to just a
- few dozen doors; this was unsuitable for a dedicated gaming BBS.
- The canned door menu still uses single keys but this should still
- be more than enough for most BBSes; those that need a huge number
- of doors can now (and should!) build a UFO script to present an
- extended doors menu. The single-key syntax still works with the
- same UFO keyword, so that existing scripts are not broken.
-
- I have also taken out some very ugly debugging messages,
- particularly in the file flagging and QWK building code. This and
- a few rewrites of various inefficient functions has saved nearly
- 10k of memory usage.
-
- It's been a long and productive weekend. I smell a public release
- coming up!
-
- 96/01/16
- NBETA069:
-
- Mostly cosmetic cleanups. Only John Barnard got this one.
-
- I have added NanoWATCH, which is an IPC watcher. Type BBS /A on an
- unused task on a busy multiline system and you'll soon see its
- usefulness. It provides an at-a-glance status report on the
- current activity of up to 20 nodes.
-
- 96/01/20
- NBETA070:
-
- I have added a "divider" feature to message and file area lists.
- When you do a file area list or a message area list (i.e. use the J
- and ? keys) you can now use dividers to break up the list, perhaps
- in categories of topic or of mail network affiliation. The syntax
- is explained in the sample MSGAREA.CTL and FILEAREA.CTL files. For
- faster speed you can use a preset sysmsg from the language file, or
- for greater flexibility you can specify an external filename for
- each divider.
-
- I have also added an "automatic area number" feature to MFB.EXE and
- MMB.EXE. By using the @NEXT keyword instead of an area number
- (e.g. filearea @next instead of filearea 650) you can have MFB or
- MMB automatically assign the area the number that comes after the
- area number of the previous area. You can also instruct MMB or MFB
- to automatically increment areas by amounts other than one.
-
- I have fixed a problem with the full screen editor, in which quoted
- text no longer appears. This was a new bug introduced in the last
- code-tightening and memory-efficiency campaign.
-
- I have made many critical NanoCHAT messages non-abortable. The
- abortability of messages had caused some people to lose /p's and
- the such.
-
- I have updated the More prompts to be more informative and improved
- the appearance (and verbosity) of the flagging-file lists (lists of
- files that are being flagged from within a file list). Eventually
- I intend to have a really slick full-ANSI file-tagging system such
- as the one in PC-Board but that's probably not going to happen
- until version 2.0. Which may come out sooner than anyone expects,
- if progress on NanoBBS continues at this pace.
-
- 96/01/27
- NBETA071:
-
- This is the final beta before the release of 1.16.
-
- I have *FINALLY* nailed the bug where Zmodem transfer rates would
- sometimes really suck. I had been unable to duplicate it because
- it only manifests for those who've locked their comports at 57600
- or faster, which I don't. An uninitialized variable in Mark
- Dignam's protocol engine was the source - apparently Dignam did not
- forsee locked baud rates faster than 38400! I have expanded that
- part of his code to take into account any baud rate up to 655360
- bps...
-
- I have also taken out the Zmodem beta text which was pretty ugly
- and has now served its purpose. While I was at it, I improved the
- progress bar in the download screen - it uses a ▌ character to give
- twice as many gradations in the progress bar, so it advances twice
- as often now. Overall it looks better that way.
-
- 96/01/31
- NANO116:
-
- This is the release version of 1.16.
-
- In addition to all the new features and function improvements and
- bug fixes since the release of 1.14 in September, 1.16 also
- includes a utility to convert the userbase, message area
- definitions, and file area definitions from a RemoteAccess 2.0
- system.
-
- New features have also been added: You can now prohibit one word
- pseudonyms or one word real names. And, if you're in sysop chat
- you can squelch the user by hitting Alt-Q. This is useful when
- you're busy and you just want to say one thing and get out of chat
- and the user won't stop typing.
-
- The docs are also clarified in a few places.
-
- The structures are now included.
-
- 96/04/14
- NBETA075:
-
- Fixed /KT in nanochat: erroneously changed node on user, now
- correctly changes channel.
-
- /S in NanoCHAT no longer counts nodes that are "waiting for a call"
- in the "number of users online".
-
- With the purchase of two 7-tray CD changers on THC, NanoBBS has
- been upgraded to support changers that require an external command
- to change the disc. Through the use of a CD_ID filearea field, you
- can minimize unneccessary use of external CD change shells. Your
- CD changer may come with a tray-swap program, I had to put mine on
- a dedicated server on the LAN and assign each tray's drive letter
- to a network directory name, which was then mapped on demand to a
- single common drive letter across the LAN. Using this technique it
- is also possible to map multiple hard disk volumes to a single
- drive letter, providing a virtually unlimited file area
- addressability under DOS.
-
- I have fixed a long standing problem where internet e-mail
- addresses sometimes got truncated at 32 characters. Email
- addresses can now be up to 80 characters.
-
- 96/06/16
- NBETA077:
-
- There has been so much added that I didn't have any time to write
- up NBETA076! Where do I start? For starters, the gender dysphoria
- in NanoCHAT is now finally fixed. NanoCHAT also now supports up to
- 1000 actions and can index them far faster than before; this is in
- preparation for the upcoming user-definable actions which should be
- ready next beta. Also, >name commands which cause directed public
- speech in NanoCHAT are now implemented and have significant
- intelligence improvements over the one in Glop. Also, you can now
- use /handle to send a /p to a user without really knowing his node.
-
- Not all the new stuff is in NanoCHAT. I have also fixed ALL of the
- time usage problems by totally reworking the time usage calculation
- code. To test it, I made the /i8 file in NanoCHAT and also added
- an option to display the same thing at logoff.
-
- I have fixed quite a few things in NEd, especially dealing with the
- Ctrl-Y line delete.
-
- There is now code in place to support miniature uses of the
- internal message editors for things like file descriptions, chat
- entry/exit messages, sig files, "about me" files and much more!
- This is preliminary to the user base structure integration which
- places the entry/exit messages in a Squish message base along with
- other message-like user data. The same file will also hold an
- "about me" message which will be part of the upcoming Registry.
- Note: we will have to call the Registry something other than that -
- we're already stomping on Galacticomm's toes enough as it is. Any
- suggestions?
-
- Now here's the big one: I have made a MAJOR change to the user
- record structure! In order to assist sysops running 1.12, 1.14 and
- 1.16 and associated betas, I have included the CONVUSER program
- which does an elegant job of turning the old userlog into the new
- one.
-
- The first benefit you will notice from the new structure is that
- user logons (and lookups) are way, WAY faster. Prior to this beta,
- user #3000 took about 10 seconds for the BBS to find before even
- asking the password; now this lookup time is imperceivable. The
- new userlog structure is indexed and the indices are updated on the
- fly as required. There is also a UHASH.EXE program to rebuild the
- indices from scratch is they ever become corrupted.
-
- Users can now have up to 6 lines in their /u1 and /u2 NanoCHAT
- entry/exit messages. They are now entered in a mini version of the
- message editor, AND they can now be "edited" instead of totally
- re-doing them!
-
- User passwords are now SECURE. NanoBBS now stores a CRC32 of the
- user's actual password instead of the password itself. The CRC, if
- known, cannot be used to logon. I will shortly add an even more
- secure Base64 encrypted password system to the system as well. At
- that point, you will be able to have plaintext, CRC32 and Base64
- encrypted passwords in the same user file, used at random. At this
- writing, the user change-password facility and the new-user
- password entry both store the CRC, and the CONVUSER program turns
- 1.16's plaintext passwords into CRCs.
-
- I have also started putting in code to make NanoCHAT friendly with
- "linked" chat systems. I am sure I will have to write an entire
- chapter on links before I release 2.00. When a user logs into an
- account that has been flagged as a "link" account (there is a key
- in the user editor to do this) they will be thrust directly into
- the chat system, their own messages will not echo back to them, and
- /q will disconnect the call. In addition, ANSI is disabled. At
- this writing, a remote system with dialout ASCII link capability
- should be able to call a NanoCHAT, log in, and comfortably be
- linked until its link account runs out of time. Needless to say, a
- special account should be created for each system you expect to
- link in. I will very soon write outgoing link code as well, with
- an internal dialer and script processor to make links a matter of
- using a /k command from NanoCHAT.
-
- Among the many other benefits of the new structures, is the keeping
- of a separate "Internet Name" field, which is a 9-character
- username suitable for use as the basis of an internet address. The
- User Editor has had several new fields added to it, and among these
- is the Internet Name. If a user doesn't have an Internet Name, you
- can give her one in the User Editor by simply hitting Alt-I. This
- will make NanoBBS quasi-intelligently synthesize an Internet name
- out of the user's logon handle.
-
- I have also fixed 3 or 4 annoying bugs in the fast file searching
- code. Chief among these was that when *.ext (where .ext is any
- extension) was specified in a filename search, NanoBBS would quite
- happily display any file it found that matched the wildcard,
- without regard for whether the user is allowed to download the file
- or not. This, of course, was inconsistent with Nano's basic design
- philosophy that: areas the user may not access are invisible to
- her. This is now fixed, and thank Ghod it never did let the user
- flag these errant files...
-
- There will be many amazing and innovative new features added in the
- coming weeks as support for the new userlog fields is added one
- feature at a time.
-
-