home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!paladin.american.edu!auvm!INNOSOFT.COM!DAN
- Errors-to: epmdf@YMIR.BITNET
- X-Envelope-to: PMDF-L@IRLEARN.BITNET
- X-VMS-To: IN%"MAC@gvg34.gvg.tek.com"
- X-VMS-Cc: IPMDF
- MIME-version: 1.0
- Content-type: TEXT/PLAIN; CHARSET=US-ASCII
- Content-transfer-encoding: 7BIT
- Message-ID: <01GTEA14QRWI8WW94D@YMIR.CLAREMONT.EDU>
- Date: Mon, 11 Jan 93 23:26:43 GMT
- Sender: PMDF Distribution List <PMDF-L@IRLEARN.BITNET>
- From: "Daniel C. Newman" <DAN@INNOSOFT.COM>
- Subject: RE: How is estimated time to send FAX calculated?
- Newsgroups: bit.listserv.pmdf-l
- Lines: 101
-
- > One of my users would like to know how
- >
- > Estimated duration of the phone call:
- >
- > is calculated. I assume that a good guess is made from the size of the file
- > to be sent to the modem?
-
- The time is indeed a guess-estimate. However, I cannot derive it from the file
- size for two reasons. First, the FAX devices spend time on the phone line
- doing negotiations at the start of the FAX transmission, before each page of
- the transmission, and at the end of the transmission. I have no idea how long
- these negotiations may take. In the first set of negotiations the machines
- are, among other things, exploring the possible capabilities of each other
- trying to see how best to transmit the data. This in itself adds a lot a
- variability to the length of the initial set of negotiations. Secondly, the FAX
- devices can dynamically change the rate of data transmission depending upon the
- quality of the phone line (as determined by the frequency of detected
- transmission errors). So, the first observation means that any estimate I will
- make will be too low (as much as 30 seconds per page). The second observation
- also indicates that my estimates will be low, but in this case may be off by a
- non-minor multiplicative factor. (The FAX modem will never increase the speed,
- only decrease it.)
-
-
- So, below is a description on how I estimate the call duration. This
- description takes the form of explaining a sample log entry. (Read: "I
- doctored a response I sent someone else so as to answer your question".)
-
- > 14-JUL-1992 07:04:25 g3_to_fax 070135070304 S 2 HAL9000@ACME.COM
- > 4551656 DI-070301 TS-070301 CO-070301 PS-070324 PS-070424 NT-070424
- > RD-070424
-
- Here's a sample entry from one of the log files. Let's dissect it:
-
- (1) 14-JUL-1992 07:04:25 -- This is the time at which the log entry was made
-
- (2) g3_to_fax -- Name of the PMDF channel making the log entry
-
- (3) 070135070304 -- Estimated start and end times for the FAX transmission:
- Estimated start time was 07:01:35
- Estimated end time was 07:03:04
- ---------------------------------
- Estimated elapsed time 2:31
-
- The estimated start time is the time at which we send the "DIAL this phone
- number" command to the FAX modem and then begin sending the actual FAX image
- data. Now, the modem doesn't actually place the phone call for about 6 - 10
- seconds after this time. Then, there's another delay while we wait for the
- remote FAX machine to answer the call.
-
- The estimated end time is when we stop sending data to the modem. The
- modem, however, will still be transmitting data over the phone line for
- some unknown length of time
-
- (4) S -- Call was successful.
-
- (5) 2 -- 2 pages sent
-
- (6) HAL9000@ACME.CMD -- Person sending the FAX
-
- (7) 4551656 -- Destination FAX telephone number
-
- (8) DI-070301 -- PMDF-FAX received the "dial" response from the FAX modem
- at the time 07:03:01.
-
- TS-070301 -- PMDF-FAX received the TSI data from the remote FAX machine
- at 07:03:01.
-
- CO-070301 -- PMDF-FAX received the "Connected" response from the FAX
- modem at 07:03:01.
-
- Now, note that these responses were all received at the same time by
- PMDF-FAX. This is because the FAX modem batches up these things before
- sending them to the host. These were most likely sent after the cover
- page was transmitted. So, from this about all we can tell is that we
- started sending data to the modem at 07:01:35 [item (3)] and at 07:03:01
- the first page had finished transmitting.
-
- (9) PS-070324 -- At 07:03:24, PMDF-FAX received an acknowledgment that the
- first page was sent successfully.
-
- PS-070424 -- At 07:04:24, PMDF-FAX received an acknowledgment that the
- second page was sent successfully.
-
- NT-070424 -- PMDF-FAX received from the FAX modem a "Call termination"
- response at 07:04:24.
-
- RD-070424 -- PMDF-FAX received from the FAX modem a "Ready" response at
- 07:04:24.
-
-
- Now, from the DI-070301 [(8)] and NT-070424 [(9)] we could estimate the call to
- have had a duration of 1:24. The estimate from (3), which is what was reported
- in the status report mailed back to the sender, was 2:31. Which estimate is
- correct? Hard too say. The estimated from (8) and (9) is usually too small as
- these responses are batched up by the modem and not sent at the time the event
- they report occurred. However, is what I've found to be the best estimate.
- It's sometimes too low, sometimes too high, and never exact to within the
- second.
-
- Dan
-