IGC File Name Decoding Made Easy

[ContestID67]

I had a pile of log files of my flights this summer. I typically
transfer files from my flight recorder to my PC via my IPAQ's removable
compact flash (CF) card.

Unfortunately, the IPAQ had the wrong date (year, time, everything) so
it was a little difficult figuring out which log file was which based
on the date stamp of the file.

However, there is a way out of this dilemma. The IGC file is named by
the flight recorder in such a way that it "encodes" the date when the
flight was made along with other information.

Below is your personal decoder ring for IGC file names.

Enjoy, John
email: john<at>derosaweb.com

=======================================================

YMDCXXXF.IGC
||||\ /|
|||| | |
|||| | +--- F=Flight # of day 1-9 then A-Z (35 flights max)
|||| +----- XXX=Serial number of logger (3 alpha-numberic digits)
|||+------- C=Manufacturer of logger (see note below)
||+-------- D=Day of month (see note below)
|+--------- M=Month of year (see note below)
+---------- Y=Last digit of year

Example
=======
5AGC3WK1.IGC

October, 16, 2005
Cambridge recorder serial number 3WK
Flight #1

Numbering Breakdown
===================
Month of year
-------------
1 = January
2 = February
3 = March
4 = April
5 = May
6 = June
7 = July
8 = August
9 = September
A = October
B = November
C = December

Day of Month
------------
1 = 1
2 = 2
3 = 3
4 = 4
5 = 5
6 = 6
7 = 7
8 = 8
9 = 9
A = 10
B = 11
C = 12
D = 13
E = 14
F = 15
G = 16
H = 17
I = 18
J = 19
K = 20
L = 21
M = 22
N = 23
O = 24
P = 25
Q = 26
R = 27
S = 28
T = 29
U = 30
V = 31

Manufacturer of Data logger
---------------------------
A = Garrecht
C = Cambridge
E = EW Avionics
F = Filser
L = LX Navigation
S = Streamline data instruments
W = Westerboer
X = All other manufacturers

Full technical specs for the IGC files can be found at:
http://www.fai.org/gliding/gnss/tech_spec_gnss.pdf

[David Kinsell]

Post by ContestID67
I had a pile of log files of my flights this summer. I typically
transfer files from my flight recorder to my PC via my IPAQ's removable
compact flash (CF) card.
Unfortunately, the IPAQ had the wrong date (year, time, everything) so
it was a little difficult figuring out which log file was which based
on the date stamp of the file.
However, there is a way out of this dilemma. The IGC file is named by
the flight recorder in such a way that it "encodes" the date when the
flight was made along with other information.
Below is your personal decoder ring for IGC file names.
Enjoy, John
email: john<at>derosaweb.com
=======================================================
YMDCXXXF.IGC
||||\ /|
|||| | |
|||| | +--- F=Flight # of day 1-9 then A-Z (35 flights max)
|||| +----- XXX=Serial number of logger (3 alpha-numberic digits)
|||+------- C=Manufacturer of logger (see note below)
||+-------- D=Day of month (see note below)
|+--------- M=Month of year (see note below)
+---------- Y=Last digit of year
Example
=======
5AGC3WK1.IGC
October, 16, 2005
Cambridge recorder serial number 3WK
Flight #1

Or, could be 1995. Or 2015. With all the stink about Y2K,
it's a little surprising they came up with a scheme that has
a 10 year rollover problem.

-Dave

[hannu]

Post by David Kinsell
Or, could be 1995. Or 2015. With all the stink about Y2K,
it's a little surprising they came up with a scheme that has
a 10 year rollover problem.

NOW it would be very easy decode the year by base 36 number to push the
problem until 2035... ;)

Only they decided on format in the last millenium.

hannu

[David Kinsell]

Post by hannu

Post by David Kinsell
Or, could be 1995. Or 2015. With all the stink about Y2K,
it's a little surprising they came up with a scheme that has
a 10 year rollover problem.

NOW it would be very easy decode the year by base 36 number to push the
problem until 2035... ;)

That's the weird thing. They used digits and characters for the date,
they used digits and characters for the month, and just used digits
for the year.

Post by hannu
Only they decided on format in the last millenium.

Yep. I guess some people just can't think ahead :-)

Post by hannu
hannu

[Tim Newport-Peace]

X-no-archive: yes

Post by David Kinsell

Post by hannu

Post by David Kinsell
Or, could be 1995. Or 2015. With all the stink about Y2K,
it's a little surprising they came up with a scheme that has
a 10 year rollover problem.

NOW it would be very easy decode the year by base 36 number to push the
problem until 2035... ;)

That's the weird thing. They used digits and characters for the date,
they used digits and characters for the month, and just used digits
for the year.

Post by hannu
Only they decided on format in the last millenium.

Yep. I guess some people just can't think ahead :-)

As I recall, IGC adopted a file naming system which was used on the
early Cambridge Recorders, which pre-date the specification. But any
flight that is 10 years old or more is history.

The 8-digit filename came about when DOS and Windows where limited to
8-digits. The specification now allows for Long Filenames (see
Specification Appendix 1 Paragraph 2.5.2) but no-one seems to use them.

In the unlikely event of a clash of filenames being problematical,
rename them to the long format. Problem Solved.

Tim Newport-Peace ***@spsys.demon.co.uk

"Indecision is the Key to Flexibility."

[Martin Gregorie]

Post by Tim Newport-Peace
As I recall, IGC adopted a file naming system which was used on the
early Cambridge Recorders, which pre-date the specification. But any
flight that is 10 years old or more is history.
The 8-digit filename came about when DOS and Windows where limited to
8-digits. The specification now allows for Long Filenames (see
Specification Appendix 1 Paragraph 2.5.2) but no-one seems to use them.

There's one unexplained item in the GNSS file specification: some loggers,
e.g. the EW model D, have a serial number of the form Dnnnn which does not
convert into the three alphanumeric characters in any obvious way. There's
not a hint in the file specification as to how this feat might be
performed.

The serial number is specified as three alphanumeric characters in both
short and long form file names.

[Tim Newport-Peace]

X-no-archive: yes

Post by Martin Gregorie

Post by Tim Newport-Peace
As I recall, IGC adopted a file naming system which was used on the
early Cambridge Recorders, which pre-date the specification. But any
flight that is 10 years old or more is history.
The 8-digit filename came about when DOS and Windows where limited to
8-digits. The specification now allows for Long Filenames (see
Specification Appendix 1 Paragraph 2.5.2) but no-one seems to use them.

There's one unexplained item in the GNSS file specification: some loggers,
e.g. the EW model D, have a serial number of the form Dnnnn which does not
convert into the three alphanumeric characters in any obvious way. There's
not a hint in the file specification as to how this feat might be
performed.
The serial number is specified as three alphanumeric characters in both
short and long form file names.

Dnnnn is not the serial number. The serial number is held in the
Firmware and is used in the IGC filename (but not the *.EWT filename.

On some older recorders the serial number is expressed to Base-10. This
has to be converted to Base-36 for use in IGC filenames.

Tim Newport-Peace ***@spsys.demon.co.uk

"Indecision is the Key to Flexibility."

[Martin Gregorie]

Post by Tim Newport-Peace
X-no-archive: yes

....

Post by Tim Newport-Peace

Post by Martin Gregorie
There's one unexplained item in the GNSS file specification: some
loggers, e.g. the EW model D, have a serial number of the form Dnnnn
which does not convert into the three alphanumeric characters in any
obvious way. There's not a hint in the file specification as to how this
feat might be performed.
The serial number is specified as three alphanumeric characters in both
short and long form file names.

Dnnnn is not the serial number. The serial number is held in the Firmware
and is used in the IGC filename (but not the *.EWT filename.

I've just looked again at the GNSS FR Specification. Section 2.5.3 allows
for a 5 digit internal serial number to be converted to the 3 character
external group but:

- the 5 digit serial number must appear in the A record
- it doesn't say how the conversion is to be done
- the s/n on the case must match the XXX form in the file name
- existing loggers in 1995 could "grandfather" their existing rules
but new loggers must obey section 2.5.3 exactly

Let's see how this fits my EW model D:

The A line from a trace: AEWAD0430 99429850
The serial number on the case: D0430
The serial number in a filename: 79Y

Post by Tim Newport-Peace
On some older recorders the serial number is expressed to Base-10. This
has to be converted to Base-36 for use in IGC filenames.

79Y in base 36 is 9430 - so base 36 is certainly possible if EW
number their model D loggers from 9000 in the firmware.

So, either exact implementation this part of the GNSS spec would appear to
be optional or the EW model D is an older design than I thought it was,
i.e. pre 1997. In that case, just how old are the EW models A, B and C ???

[Tim Newport-Peace]

X-no-archive: yes

Post by Martin Gregorie

Post by Tim Newport-Peace
X-no-archive: yes

....

Post by Tim Newport-Peace

Post by Martin Gregorie
There's one unexplained item in the GNSS file specification: some
loggers, e.g. the EW model D, have a serial number of the form Dnnnn
which does not convert into the three alphanumeric characters in any
obvious way. There's not a hint in the file specification as to how this
feat might be performed.
The serial number is specified as three alphanumeric characters in both
short and long form file names.

Dnnnn is not the serial number. The serial number is held in the Firmware
and is used in the IGC filename (but not the *.EWT filename.

I've just looked again at the GNSS FR Specification. Section 2.5.3 allows
for a 5 digit internal serial number to be converted to the 3 character
- the 5 digit serial number must appear in the A record
- it doesn't say how the conversion is to be done
- the s/n on the case must match the XXX form in the file name
- existing loggers in 1995 could "grandfather" their existing rules
but new loggers must obey section 2.5.3 exactly
The A line from a trace: AEWAD0430 99429850
The serial number on the case: D0430
The serial number in a filename: 79Y

Post by Tim Newport-Peace
On some older recorders the serial number is expressed to Base-10. This
has to be converted to Base-36 for use in IGC filenames.

79Y in base 36 is 9430 - so base 36 is certainly possible if EW
number their model D loggers from 9000 in the firmware.
So, either exact implementation this part of the GNSS spec would appear to
be optional or the EW model D is an older design than I thought it was,
i.e. pre 1997. In that case, just how old are the EW models A, B and C ???

All present EW models have grandfather's rights.

Tim Newport-Peace ***@spsys.demon.co.uk

"Indecision is the Key to Flexibility."

[309]

Bet you the cost of a tow that they come out with a new igc "standard"
before 2015...

Hey, wait...I'm not going to bet a tow, that could be $100 or
more...based on the other postings.

Maybe IGC will require USB connections...as they get phased out for ???

FWIW, I'm still using a VolksLogger with DOS, having sought out an
"old" laptop that has a com port, and thrashed the computer into a dual
boot system (using System Commander software) so that I can use the
Garrecht "secure" DOS download software ("vali.exe"), then reboot to
WinXP and get useful information from the trace using SeeYou.

Those that know me know I do EVERYTHING the hard way (e.g., fly a
1-26).

Thanks for the decoder ring; it sure is nice to understand the
gibberish naming of the file, I think.

-Pete

[Tim Newport-Peace]

X-no-archive: yes

Post by 309
Bet you the cost of a tow that they come out with a new igc "standard"
before 2015...

Two Years ago actually.

Post by 309
Hey, wait...I'm not going to bet a tow, that could be $100 or
more...based on the other postings.
Maybe IGC will require USB connections...as they get phased out for ???
FWIW, I'm still using a VolksLogger with DOS, having sought out an
"old" laptop that has a com port, and thrashed the computer into a dual
boot system (using System Commander software) so that I can use the
Garrecht "secure" DOS download software ("vali.exe"), then reboot to
WinXP and get useful information from the trace using SeeYou.
Those that know me know I do EVERYTHING the hard way (e.g., fly a
1-26).

You must do. There is a Free Windows shell program and Volkslogger DLL
available from http://www.fai.org/gliding/gnss/freeware.asp

Tim Newport-Peace ***@spsys.demon.co.uk

"Indecision is the Key to Flexibility."

[Derrick Steed]

The American marines did it before the IGC - they had a system which hit
the decade rollover in 1980

Rgds,

Derrick Steed

[309]

To IGC's defence, some of that issue could be traced to DOS, with it's
file name limitation (8 characters maximum).

Admittedly, hex/alpha coding year would extend the rollover period, but
make it yet more difficult for us to track our flight files...

And there was a GPS clock rollover back in the '90's, wasn't there? Or
did I sleep through the alarm clock?

-Pete

[Peter Wyld]

On Wed, 07 Sep 2005 14:53:10 +0100, Tim Newport-Peace

Post by Tim Newport-Peace
X-no-archive: yes
In article , Martin
Gregorie writes

.....

Post by Tim Newport-Peace

Post by Martin Gregorie
some
loggers, e.g. the EW model D, have a serial number
of the form Dnnnn
which does not convert into the three alphanumeric
characters in any
obvious way. There's not a hint in the file specification
as to how this
feat might be performed.
The serial number is specified as three alphanumeric
characters in both
short and long form file names.

Dnnnn is not the serial number. The serial number
is held in the Firmware
and is used in the IGC filename (but not the *.EWT
filename.

(snip)

Post by Tim Newport-Peace
On some older recorders the serial number is expressed
to Base-10. This
has to be converted to Base-36 for use in IGC filenames.

79Y in base 36 is 9430 - so base 36 is certainly possible
if EW
number their model D loggers from 9000 in the firmware.
So, either exact implementation this part of the GNSS
spec would appear to
be optional or the EW model D is an older design than
I thought it was,
i.e. pre 1997. In that case, just how old are the EW
models A, B and C ???
--
gregorie. | Essex, UK
org |

EWA Serial numbers convert directly to base36
EWB add 2000 to the serial number and convert to base36
EWD add 9000 to the serial number and convert to base36

Cambridge, Garrecht (Volkslogger), Zander all use the
external serial.

Filser,LX,Colibri,PrintTechnik convert external serial
directly to base36

[ContestID67]

Please see my other post for the continuation of this thread. Please
don't add to this thread.

"FAI/IGC/CAI File Name Decoding - Automated"

I've got a head of steam up on getting this right.

Enjoy, John
email: john<at>derosaweb.com