1 hack.txt for Citadel/UX
2 written by Art Cancro (ajc@uncnsrd.mt-kisco.ny.us)
4 Much of this document is borrowed from the original hack.doc from
5 Citadel-CP/M and Citadel-86, because many of the concepts are the same. Hats
6 off to whoever wrote the original, for a fine document that inspired the
7 implementation of Citadel for Unix.
9 Note that this document is really out of date. It doesn't cover anything
10 about the threaded server architecture or any of the network stuff. What is
11 covered here is the basic architecture of the databases.
13 But enough of the preamble. Here's how Citadel/UX works :)
15 Here are the major databases to be discussed:
17 msgmain The big circular file that contains message text
18 quickroom Contains room info such as room names, stats, etc.
19 fullroom One fullrm file per room: message numbers and pointers.
20 usersupp Contains info for each user on the system.
22 The fundamental structure of the system differs greatly from the way
23 Citadels used to work. Citadel now depends on a record manager or database
24 manager of some sort. Thanks to the API which is in place for connecting to
25 a data store, any record manager may be used as long as it supports the
26 storage and retrieval of large binary objects (blobs) indexed by unique keys.
27 Please see database.c for more information on data store primitives.
29 The message base (MSGMAIN) is a big file of messages indexed by the message
30 number. Messages are numbered consecutively and start with an FF (hex)
31 byte. Except for this FF start-of-message byte, all bytes in the message
32 file have the high bit set to 0. This means that in principle it is
33 trivial to scan through the message file and locate message N if it
34 exists, or return error. (Complexities, as usual, crop up when we
35 try for efficiency...)
37 Each room is basically just a list of message numbers. Each time
38 we enter a new message in a room, its message number is appended to the end
39 of the list. If an old message is to be expired, we must delete it from the
40 message base. Reading a room is just a matter of looking up the messages
41 one by one and sending them to the client for display, printing, or whatever.
43 Implementing the "new message" function is also trivial in principle:
44 we just keep track, for each caller in the userlog, of the highest-numbered
45 message which existed on the *last* call. (Remember, message numbers are
46 simply assigned sequentially each time a message is created. This
47 sequence is global to the entire system, not local within a room.) If
48 we ignore all message-numbers in the room less than this, only new messages
49 will be printed. Voila!
52 Message format on disk (MSGMAIN)
54 As discussed above, each message begins with an FF byte.
56 The next byte denotes whether this is an anonymous message. The codes
57 available are MES_NORMAL, MES_ANON, or MES_AN2 (defined in citadel.h).
59 The third byte is a "message type" code. The following codes are defined:
60 0 - "Traditional" Citadel format. Message is to be displayed "formatted."
61 1 - Plain pre-formatted ASCII text (otherwise known as text/plain)
62 4 - MIME formatted message. The text of the message which follows is
63 expected to begin with a "Content-type:" header.
65 After these three opening bytes, the remainder of
66 the message consists of a sequence of character strings. Each string
67 begins with a type byte indicating the meaning of the string and is
68 ended with a null. All strings are printable ASCII: in particular,
69 all numbers are in ASCII rather than binary. This is for simplicity,
70 both in implementing the system and in implementing other code to
71 work with the system. For instance, a database driven off Citadel archives
72 can do wildcard matching without worrying about unpacking binary data such
73 as message ID's first. To provide later downward compatability
74 all software should be written to IGNORE fields not currently defined.
76 The type bytes currently defined are:
78 BYTE Mnemonic Comments
80 A Author Name of originator of message.
81 B Phone number The dialup number of the system this message
82 originated on. This is optional, and is only
83 defined for helping implement C86Net gateways.
84 D Destination Contains name of the system this message should
85 be sent to, for mail routing (private mail only).
86 G Gateway domain This field is provided solely for the implementation
87 of C86Net gateways, and holds the C86Net domain of
88 the system this message originated on. Unless you're
89 implementing such a gateway, there's no need to even
90 bother with this field.
91 H HumanNodeName Human-readable name of system message originated on.
92 I Original ID A 32-bit integer containing the message ID on the
93 system the message *originated* on.
94 M Message Text Normal ASCII, newlines seperated by CR's or LF's,
95 null terminated as always.
96 N Nodename Contains node name of system message originated on.
97 O Room Room of origin.
98 P Path Complete path of message, as in the UseNet news
99 standard. A user should be able to send Internet mail
100 to this path. (Note that your system name will not be
101 tacked onto this until you're sending the message to
103 R Recipient Only present in Mail messages.
104 S Special field Only meaningful for messages being spooled over a
105 network. Usually means that the message isn't really
106 a message, but rather some other network function:
107 -> "S" followed by "FILE" (followed by a null, of
108 course) means that the message text is actually an
109 IGnet/Open file transfer.
110 T Date/Time A 32-bit integer containing the date and time of
111 the message in standard UNIX format (the number
112 of seconds since January 1, 1970 GMT).
113 U Subject Optional. Developers may choose whether they wish to
114 generate or display subject fields. Citadel/UX does
115 not generate them, but it does print them when found.
119 Let <FF> be a 0xFF byte, and <0> be a null (0x00) byte. Then a message
122 Apr 12, 1988 23:16 From Test User In Network Test> @lifesys (Life BBS)
125 might be stored as...
126 <FF><40><0>I12345<0>Pneighbor!lifesys!test_user<0>T576918988<0> (continued)
127 -----------|Mesg ID#|--Message Path---------------|--Date------
129 AThe Test User<0>ONetwork Test<0>Nlifesys<0>HLife BBS<0>MHave a nice day!<0>
130 |-----Author-----|-Room name-----|-nodename-|Human Name-|--Message text-----
132 Weird things can happen if fields are missing, especially if you use the
133 networker. But basically, the date, author, room, and nodename may be in any
134 order. But the leading fields and the message text must remain in the same
135 place. The H field looks better when it is placed immediately after the N
140 Citadel nodes network by sharing one or more rooms. Any Citadel node
141 can choose to share messages with any other Citadel node, through the sending
142 of spool files. The sending system takes all messages it hasn't sent yet, and
143 spools them to the recieving system, which posts them in the rooms.
145 Complexities arise primarily from the possibility of densely connected
146 networks: one does not wish to accumulate multiple copies of a given
147 message, which can easily happen. Nor does one want to see old messages
148 percolating indefinitely through the system.
150 This problem is handled by keeping track of the path a message has taken over
151 the network, like the UseNet news system does. When a system sends out a
152 message, it adds its own name to the bang-path in the <P> field of the
153 message. If no path field is present, it generates one.
155 With the path present, all the networker has to do to assure that it doesn't
156 send another system a message it's already received is check the <P>ath field
157 for that system's name somewhere in the bang path. If it's present, the system
158 has already seen the message, so we don't send it. (Note that the current
159 implementation does not allow for "loops" in the network -- if you build your
160 net this way you will see lots of duplicate messages.)
162 The above discussion should make the function of the fields reasonably clear:
164 o Travelling messages need to carry original message-id, system of origin,
165 date of origin, author, and path with them, to keep reproduction and
166 cycling under control.
168 (Uncoincidentally) the format used to transmit messages for networking
169 purposes is precisely that used on disk, except that there may be any amount
170 of garbage between the null ending a message and the <FF> starting the next
171 one. This allows greater compatibility if slight problems crop up. The current
172 distribution includes netproc.c, which is basically a database replicator;
173 please see network.txt on its operation and functionality (if any).
177 At this point, all hardware-dependent stuff has been removed from the
178 system. On the server side, most of the OS-dependent stuff has been isolated
179 into the sysdep.c source module. The server should compile on any POSIX
180 compliant system with a full pthreads implementation and TCP/IP support. In
181 the future, we may try to port it to non-POSIX systems as well.
183 On the client side, it's also POSIX compliant. The client even seems to
184 build ok on non-POSIX systems with porting libraries (such as the Cygnus
188 "Room" records (quickroom)
190 The rooms are basically indices into msgmain, the message database.
191 As noted in the overview, each is essentially an array of pointers into
192 the message file. The pointers consist of a 32-bit message ID number
193 (we will wrap around at 32 bits for these purposes).
195 Since messages are numbered sequentially, the
196 set of messages existing in msgmain will always form a continuous
197 sequence at any given time.
199 That should be enough background to tackle a full-scale room. From citadel.h:
202 char QRname[20]; /* Max. len is 19, plus null term */
203 char QRpasswd[10]; /* Only valid if it's a private rm */
204 long QRroomaide; /* User number of room aide */
205 long QRhighest; /* Highest message NUMBER in room */
206 long QRgen; /* Generation number of room */
207 unsigned QRflags; /* See flag values below */
208 char QRdirname[15]; /* Directory name, if applicable */
209 char QRfloor; /* (not yet implemented) */
212 #define QR_BUSY 1 /* Room is being updated, WAIT */
213 #define QR_INUSE 2 /* Set if in use, clear if avail */
214 #define QR_PRIVATE 4 /* Set for any type of private room */
215 #define QR_PASSWORDED 8 /* Set if there's a password too */
216 #define QR_GUESSNAME 16 /* Set if it's a guessname room */
217 #define QR_DIRECTORY 32 /* Directory room */
218 #define QR_UPLOAD 64 /* Allowed to upload */
219 #define QR_DOWNLOAD 128 /* Allowed to download */
220 #define QR_VISDIR 256 /* Visible directory */
221 #define QR_ANONONLY 512 /* Anonymous-Only room */
222 #define QR_ANON2 1024 /* Anonymous-Option room */
223 #define QR_NETWORK 2048 /* Shared network room */
224 #define QR_PREFONLY 4096 /* Preferred users only */
226 [Note that all components start with "QR" for quickroom, to make sure we
227 don't accidentally use an offset in the wrong structure. Be very careful
228 also to get a meaningful sequence of components --
229 some C compilers don't check this sort of stuff either.]
231 QRgen handles the problem of rooms which have died and been reborn
232 under another name. This will be clearer when we get to the userlog.
233 For now, just note that each room has a generation number which is
234 bumped by one each time it is recycled.
236 QRflags is just a bag of bits recording the status of the room. The
239 QR_BUSY This is to insure that two processes don't update the same
240 record at the same time, even though this hasn't been
242 QR_INUSE 1 if the room is valid, 0 if it is free for re-assignment.
243 QR_PRIVATE 1 if the room is not visible by default, 0 for public.
244 QR_PASSWORDED 1 if entry to the room requires a password.
245 QR_GUESSNAME 1 if the room can be reached by guessing the name.
246 QR_DIRECTORY 1 if the room is a window onto some disk/userspace, else 0.
247 QR_UPLOAD 1 if users can upload into this room, else 0.
248 QR_DOWNLOAD 1 if users can download from this room, else 0.
249 QR_VISDIR 1 if users are allowed to read the directory, else 0.
250 QR_ANONONLY 1 if all messages are to recieve the "****" anon header.
251 QR_ANON2 1 if the user will be asked if he/she wants an anon message.
252 QR_NETWORK 1 if this room is shared on a network, else 0.
253 QR_PREFONLY 1 if the room is only accessible to preferred users, else 0.
255 QRname is just an ASCII string (null-terminated, like all strings)
256 giving the name of the room.
258 QRdirname is meaningful only in QR_DIRECTORY rooms, in which case
259 it gives the directory name to window.
261 QRpasswd is the room's password, if it's a QR_PASSWORDED room. Note that
262 if QR_PASSWORDED or QR_GUESSNAME are set, you MUST also set QR_PRIVATE.
263 QR_PRIVATE by itself designates invitation-only. Do not EVER set all three
264 flags at the same time.
266 QRroomaide is the user number of the room's room-aide (or zero if the room
267 doesn't have a room aide). Note that if a user is deleted, his/her user number
268 is never used again, so you don't have to worry about a new user getting the
269 same user number and accidentally becoming a room-aide of one or more rooms.
271 The only field new to us in quickroom is QRhighest, recording the
272 most recent message in the room. When we are searching for rooms with
273 messages a given caller hasn't seen, we can check this number
274 and avoid a whole lot of extra disk accesses.
276 There used to also be a structure called "fullroom" which resided in one
277 file for each room on the system. This has been abandoned in favour of
278 "message lists" which are variable sized and simply contain zero or more
279 message numbers. The message numbers, in turn, point to messages on disk.
281 User records (usersupp)
283 This is the fun one. Get some fresh air and plug in your thinking cap
284 first. (Time, space and complexity are the eternal software rivals.
285 We've got lots of log entries times lots of messages spread over up to nnn
286 rooms to worry about, and with multitasking, disk access time is important...
287 so perforce, we opt for complexity to keep time and space in bounds.)
289 To understand what is happening in the log code takes a little persistence.
290 You also have to disentangle the different activities going on and
291 tackle them one by one.
293 o We want to remember some random things such as terminal screen
294 size, and automatically set them up for each caller at login.
296 o We want to be able to locate all new messages, and only new
297 messages, efficiently. Messages should stay new even if it
298 takes a caller a couple of calls to get around to them.
300 o We want to remember which private rooms a given caller knows
301 about, and treat them as normal rooms. This means mostly
302 automatically seeking out those with new messages. (Obviously,
303 we >don't< want to do this for unknown private rooms!) This
304 has to be secure against the periodic recycling of rooms
307 o We want to support private mail to a caller.
309 o We want to provide some protection of this information (via
310 passwords at login) and some assurance that messages are from
311 who they purport to be from (within the system -- one shouldn't
312 be able to forge messages from established users).
314 Lifting another page from citadel.h gives us:
316 struct usersupp { /* User record */
317 int USuid; /* uid account is logged in under */
318 char password[20]; /* password */
319 long lastseen[MAXROOMS]; /* Last message seen in each room */
320 char generation[MAXROOMS]; /* Generation # (for private rooms) */
321 char forget[MAXROOMS]; /* Forgotten generation number */
322 unsigned flags; /* See US_ flags below */
323 int screenwidth; /* For formatting messages */
324 int timescalled; /* Total number of logins */
325 int posted; /* Number of messages posted (ever) */
326 char fullname[26]; /* Bulletin Board name for messages */
327 char axlevel; /* Access level */
328 long usernum; /* Eternal user number */
329 long lastcall; /* Last time the user called */
332 #define US_PERM 1 /* Permanent user; don't scroll off */
333 #define US_LASTOLD 16 /* Print last old message with new */
334 #define US_EXPERT 32 /* Experienced user */
335 #define US_UNLISTED 64 /* Unlisted userlog entry */
336 #define US_NOPROMPT 128 /* Don't prompt after each message */
337 #define US_PREF 1024 /* Preferred user */
339 Looks simple enough, doesn't it? One topic at a time:
341 Random configuration parameters:
342 -screenwidth is the caller's screen width. We format all messages to this
343 width, as best we can. flags is another bit-bag, recording whether we want
344 prompts, people who want to suppress the little automatic hints all through
347 Attachments, names & numbers:
348 -USuid is the uid the account was established under. For most users it will
349 be the same as BBSUID, but it won't be for users that logged in from the shell.
350 -fullname is the user's full login name.
351 -usernum is the user's ID number. It is unique to the entire system:
352 once someone has a user number, it is never used again after the user is
353 deleted. This allows an easy way to numerically represent people.
354 -password is the user's password.
355 -axlevel is the user's access level, so we know who's an Aide, who's a problem
356 user, etc. These are defined and listed in the system.
359 -timescalled is the number of times the user has called.
360 -posted is the number of messages the user has posted, public or private.
363 -lastcall holds the date and time (standard Unix format) the user called, so
364 we can purge people who haven't called in a given amount of time.
366 Finding new messages:
367 This is the most important. Thus, it winds up being the most
368 elaborate. Conceptually, what we would like to do is mark each
369 message with a bit after our caller has read it, so we can avoid
370 printing it out again next call. Unfortunately, with lots of user
371 entries this would require adding lots of bits to each message... and
372 we'd wind up reading off disk lots of messages which would never
373 get printed. So we resort to approximation and a small table.
375 The approximation comes in doing things at the granularity of
376 rooms rather than messages. Messages in a given room are "new"
377 until we visit it, and "old" after we leave the room... whether
378 we read any of them or not. This can actually be defended: anyone
379 who passes through a room without reading the contents probably just
380 isn't interested in the topic, and would just as soon not be dragged
381 back every visit and forced to read them. Given that messages are
382 numbered sequentially, we can simply record the most recent message ID#
383 of each room as of the last time we visited it. Very simple.
385 Putting it all together, we can now compute whether a given room
386 has new messages for our current caller without going to the message base
387 index (fullroom) at all:
389 > We get the usersupp.lastseen[] for the room in question
390 > We compare this with the room's quickroom.QRhighest, which tells us
391 what the most recent message in the room is currently.
394 REMEMBERING WHICH PRIVATE ROOMS TO VISIT
396 This looks trivial at first glance -- just record one bit per room per
397 caller in the log records. The problem is that rooms get recycled
398 periodically, and we'd rather not run through all the log entries each
399 time we do it. So we adopt a kludge which should work 99% of the time.
401 As previously noted, each room has a generation number, which is bumped
402 by one each time it is recycled. As not noted, this generation number
403 runs from 0 -> 127 (and then wraps around and starts over).
404 When someone visits a room, we set usersupp.generation for the room
405 equal to that of the room. This flags the room as being available.
406 If the room gets recycled, on our next visit the two generation numbers
407 will no longer match, and the room will no longer be available -- just
408 the result we're looking for. (Naturally, if a room is public,
409 all this stuff is irrelevant.)
411 This leaves only the problem of an accidental matchup between the two
412 numbers giving someone access to a Forbidden Room. We can't eliminate
413 this danger completely, but it can be reduced to insignificance for
414 most purposes. (Just don't bet megabucks on the security of this system!)
415 Each time someone logs in, we set all "wrong" generation numbers to -1.
416 So the room must be recycled 127 times before an accidental matchup
417 can be achieved. (We do this for all rooms, INUSE or dead, public
418 or private, since any of them may be reincarnated as a Forbidden Room.)
420 Thus, for someone to accidentally be led to a Forbidden Room, they
421 must establish an account on the system, then not call until some room
422 has been recycled 127 to 128 times, which room must be
423 reincarnated as a Forbidden Room, which someone must now call back
424 (having not scrolled off the userlog in the mean time) and read new
425 messages. The last clause is about the only probable one in the sequence.
426 The danger of this is much less than the danger that someone will
427 simply guess the name of the room outright (if it's a guess-name room)
428 or some other human loophole.
432 This is exactly the opposite of private rooms. When a user chooses to
433 forget a room, we put the room's generation number in usersupp.forget for
434 that room. When doing a <K>nown rooms list or a <G>oto, any matchups cause
435 the room to be skipped. Very simple.
437 SUPPORTING PRIVATE MAIL
439 Can one have an elegant kludge? This must come pretty close.
441 Private mail is sent and recieved in the Mail> room, which otherwise
442 behaves pretty much as any other room. To make this work, we have a
443 separate Mail> room for each user behind the scenes. The actual room name
444 in the database looks like "0000001234.Mail" (where '1234' is the user
445 number) and it's flagged with the QR_MAILBOX flag. The user number is
446 stripped off by the server before the name is presented to the client.
448 This requires a little fiddling to get things just right. For example,
449 make_message() has to be kludged to ask for the name of the recipient
450 of the message whenever a message is entered in Mail>. But basically
451 it works pretty well, keeping the code and user interface simple and
455 PASSWORDS AND NAME VALIDATION
457 This has changed a couple of times over the course of Citadel's history. At
458 this point it's very simple, again due to the fact that record managers are
459 used for everything. The user file (usersupp) is indexed using the user's
460 name, converted to all lower-case. Searching for a user, then, is easy. We
461 just lowercase the name we're looking for and query the database. If no
462 match is found, it is assumed that the user does not exist.
464 This makes it difficult to forge messages from an existing user. (Fine
465 point: nonprinting characters are converted to printing characters, and
466 leading, trailing, and double blanks are deleted.)