Need Help With Langford's Flag Cancel Encyclopedia

I am learning to use Frederick Langford's Flag Cancel Encyclopedia.I am confused about the meaning of notations in column 4. For example, 17-02,17-128, 17-219, etc. Can anyone explain or tell where to find them..

Fred was intense (and quirky) in his collecting.
As it turns out, the flag portion moved around the country from city to city as machines came back for maintenence, replacement, and parts reused, etc.
The "17-xxx" notation records the movement from city to city of the flag portion.
See explantion in the 3rd edition, pages 96-97, or more extensively in the 4th edition pages 98-103.
In short, I believe most flag/machine collectors ignore this data, including myself, and I collect several flags quite seriously.

Add: As for quirky, beyond his unusual alphabetizing of the states, his list of what creates a variety or not (4th ed, page 5) clearly contains some "not a variety" listings which a specialist would readily accept as a collectible type/variety. The same is true for his Doremus machine cancel volume, where for example, a change (or alternation) in the killer from 16 to 18 vertical bars is not a variety, yet a change of the flag portion in an Amrican machine would be.

Bottom line, Langford's works are incredible research over a lifetime of study, but one can still go farther.

Thank you for the guidance and direction.I can't believeI I overlooked pages 96 and 97. I am not sure when that information would become important enough for me to make a notation with a cover or card. As I gain more experience, maybe I will find one. I did think things like the upside down date on this card was important enough to note because I am not sure how often it happens. Maybe one day I will find a correct one to put with it. Again, thank you for the help.

Yes, human error makes for some interesting postmarks. Very rarely, the year slug is intentionally inserted inverted for the first few days of January *IF* the new year slug has not yet been received, but this is very hard to prove, and is certainly not the case by August.

Beyond the possible confusion of seeing all the time/day/month/year slugs in mirror image, newer clerks changing the internal dial slugs may also be confused by the fact that everything operates upside down in the machine. I.e., notice in a stack of different height mail that the machine cancels are all the same distance down from the top. In reality, the faced mail is fed through the cancelling machine up side down, sliding along on its top side and the cancel applied a short distance above the bottom.. With both the letter and town dial inserted up side down. The result is a stack of mail looking normal when we read it.

Scottsburg, Indiana has one of the more interesting slug arrangements, reading either up or down with the year slug in a normal position at the bottom - and switching back and forth every few weeks or months. It is a fairly scarce flag in either orientation.