Prior doc= Next doc= SORTING OUT THE APRL SYSTEM A Sorted Affair* 102160,ae,g, Eurasia; 11430,aeb,g, -Asia; s/b # 102200 102260,aee,c, --Asia, East 102320,aeh,c, --Asia, North 102380,aek,c, --Asia, South 102440,aen,c, --Asia, Southeast 102500,aeq,c, --Asia, West 102560,aet,c, --Asia, Central 99990,aezzz,0,;spacer 11480,af,g, -Australasia; s/b 102700 99990,afzzz,0,;spacer 17470,ag,g, -Europe; s/b # 103240 103280,agb,c, --North 17120,age,c, --East; s/b 103340 38290,agh,c, --West; s/b 103400 33520,agk,c, --South; s/b 103460 14200,agn,c, --Central; s/b 103520 103580,agq,c, --Southeast 103660,agu,c, --Southwest 99990,agzzz,0,;SPACER Out of the 17 listings above 7 already are currently in the databse. These are those with ID numbers under 50,000. Nore: To place a sub-region like the Balkan Peninsula enter the next Location_Code ; likewise for all that relate. e.g. under Europe, Southeast the location seguence is 'agq' so go up one: agr. APRL#,agr,s,Balkan Peninsula or e.g. 'Europe, South' has Location_Code agk then use: agl 10140,agl,i, Agean Islands APRL#,agl,i,Ionian Islands or another example: In the Example above - 100860,abp,c,Africa, South (the compass region): has abpa-ags free (note: using abpa-abpz provides 26 sub entrites under S. Africa) (as do abq, abr and abs subsets: total 104; +3 for abq, abr & abs) 33180,abq,1e,South Africa, Rep. of (the nation) 41320,abqa,1e,Bophuthatswana (1977-1994);enclave in So. Africa 41340,abqa,1e,Ciskei (1983-1994) " 41360,abqa,1e,Transkei (1976-1994) " 41380,abqa,1e,Venda (1979-1994) " a pattern worth noting is that all major geographical regions & topics are 2 letter codes only. e.g. Europe is 'ag', Asia is 'aeb' (lossing 2 entries being the first entry incorporated under a more General Healding (here 'Eurasia') - this is taken into acct as, further, the attemp has been made to assign numbers after 100k in a direct relationship to the location codes here: note that from aa to ab there are 26 possibilities (aa-az) thus if aa is assigned the # 100,000 (and it is) then az (allowing 20 APRL #'s between entries) = 26x20=520 +100,000 =100520. (ab=+20 or 105540). The easiest way to view it is 'add 540' between dual letter changes ; i.e. aa to ab ; and you get the 1st # of the next iteration. In order to keep the letter system in check with the APRL # system subsequent extenstions must have some rules: 1. Third level additions (e.g. aaa-aaz) divides the 520 # gap that occurs between 'aa' and 'ab' into 26 possibilities. Going from 'aaa' to 'aab' is a 20 # step.: The 3rd level rule: letters only; no numbers. 2. Forth level additions (i.e. aaaa-...) must be limited to a range that can only express the 19 intergers that remain between 101100 (aca) and 101120 (acb); i.e. 101101 to 101119. Thus forth level additions must be limited to the range 'a' through 's' AS the 4th character in order to be usable in the APRL # system for adding unique data within the ordered system provided by the letter codes. There are many examples of this already; to wit. APRL # 11430 (representing 'Asia', in the example above, is assigned against the 3 letter code 'aeb'. This would be 102200 if the numbers tracked the letters (a change that is recommened for dividing regional information from actual philatelic entities). Note: This does not mandate that a fourth level addition have a APRL # but that # would not be usable in the area where this code is used unless....(see 'The Fourth Level' below) 3. Fifth level additions (i.e. aaaaa-.....): Fifth level items must be duplicates being used differently than the original or items that already hove an APRL # or no # at all such as a 'space' or note. To The Fouth Level Example and Analasys: This example represents what a 'delimited' file, composed of a short section of the modified database, would look like. The first line describes the field names. EXAMPLE "APRL#","Location_Code","Menu_type","Entity1","Entity2","Entity2Descript", "Entity3","Comment" "99990","a","0",,,,,,,";SPACER" "100000","aa","H", ,,"The World","aka: Earth",,,"position hopefully centers it"x "99990","aaazz","0", ,,,,,,";SPACER"x "10330","ab","g", "Africa",,,,,,";s/b 100540 ; using 20 APRL#'s per 3 ltr_Code" "100580","abb","c", ,"- North",,,,,"4th level: b,c,d,e,f, (a-s) = 95*" "100680","abg","c", ,"- East",,,,,"g,h,i,j,k = 95*" "100780","abl","c", ,"- Central",,,,,"l.m.n.o = 76*" "100860","abp","c", ,"- South",,,,,"76 poss.* sub divisions in p,q,r,s" "100940",abt,"c", ,"- Southwest",,,,,"t,u,v = 57*" "101000","abu","c", ,"- West",,,,,"w,x,y,z = 76*" "99990","abzzz","0", ,,,,,,";SPACER"x *by appending a-s (19 possible letters) to make a 4 character code *this limitation allows assigningg APRL# in a possible range assuming that all prior entries have required a new APRL #.: Each change of a 3 letter code (e.g. abb to abc) is 20 APRL #'s and Each change of a 4 letter code (e.g. abba to abbb) is 1 APRL #. Simplified Example APRL# code APRL# 100580, abb abba = 100581 abbb = 100582 abbc = 100583 ... thruough abbs - 100599 (i.e. 81-99) 100600, abc Thus abb to abba goes from 0 to 1 (all three letter codes end in 0; abba to abbb is 1 to 2 and so forth to 'abc' which is 20 APRL #'s total. 't-z' can be used but has to be thought out - ; an APRL # can be used if one of the prior (a-s 4th characters) didn't use one but, note, it will be out of sequence numerically in relation to the absolute order obtainable at higher levels. (what this implies is that it can't be easity calculated, in a program for example, though they can be manually tracked or accounted for in a program with a specifically designed routine. The key term above is 'easily'. One thing that might help is when entries are made in the t-a range requiring an APRL # (I feel that most won't at this level) then choose the next unsued # that falls between the allowable numbers for the prior level: e.g. let's say you are entering a record wanting to use 'abht' as a fourth level sort code requiring an APRL # and you notice that the prior entries within abha-abhs range used no APRL #'s over 100000, then choose the next or most desirable number within the 20 # step that occurs between abb and abc or what abha ought to have been if it would have reqauired one. --- random (ae)Asia, West (aes) Near East (aet)Arabian Peninsula (aeu) South Arabia (aeul) Yemen, Rep. of (current name) made from YAR & below (aeul1) YAR (aeul2) Yeman, People's Dem. Rep of Southern here add 6th # (aeul21) Aden , (aeul22) S.Arabian Fed. (aeul23) Yenmen,prs-x? i.e. (aeul21)-aeul24 to describe chain FURTHER: could these codes be used to set up 'See Also' refs?: they may be best empoyed in a seperate field so they don't require an additional record! an old idea and the question sill remains: How large a field? i.e. how many references to allow? Another consideration: Another table? Could consist of 2 fields for: 1. APRL# of POV (Point Of View - the entity under primary consideration) 2. APRL# of the 'See Also' reference. e.g. re: 10630 Albanian Occupation - Epirus field 1 would have # of entity where 'Epirus' is/was at the time field 2 would have 10630 You could then have a single y/n logical field in the main database indicating the presense of a 'See Also' reference... ----- * the sort seguence in the first example is an 'index' based on: "Location_Code"+"Entity1"+"Entity2Description"+"Entity2"+"Entity3"+ "Start_Yr_Date"+"End_Yr_Date" (c)opyright 2000 A.J. Ward