boinc/vda/examples.txt

scenario 1:

0: REC
    0.0: PRES
        0.0.0: PRES
        0.0.1: PRES
        0.0.2: UNREC
    0.1: REC
        0.1.0: REC
        0.1.1: REC
        0.1.2: REC
    0.2: REC
        0.2.0: REC
        0.2.1: REC
        0.2.2: REC

- decode 0.0
- encode 0.0
- delete 0.0
- delete 0.0.0, 0.0.1

after decide:
0.0: need_recon

---------------------------------
scenario 2:

0: REC
    0.0: REC
        0.0.0: REC, upload in progress
        0.0.1: REC
        0.0.2: REC
    0.1: PRES
        0.1.0: PRES, download in progress
        0.1.1: PRES
        0.1.2: UNREC
    0.2: UNREC
        0.2.0: REC
        0.2.1: UNREC
        0.2.2: UNREC

- decode 0.1
- encode 0.1
- delete 0.1
- delete 0.1.1
- start upload of 0.0.1 or 0.0.2
- start download of 0.1.2

after decide:
0: need_present
    0.0: need_present
        0.0.0: need_present
        0.0.1: need_present
    0.1: keep_present, need_recon
        0.1.0: keep_present
        0.1.1: keep_present

---------------------------------
scenario 3:

0: PRES
    0.0: PRES
        0.0.0: PRES
        0.0.1: PRES
        0.0.2: REC
    0.1: PRES
        0.1.0: PRES
        0.1.1: PRES
        0.1.2: UNREC
    0.2: UNREC
        0.2.0: REC
        0.2.1: UNREC
        0.2.2: UNREC

- decode 0.0
- delete 0.0.0, 0.0.1, 0.0.2
- decode 0.1
- encode 0.1
- delete 0.1.0, 0.1.1
- decode 0
- encode 0
- delete 0.0, 0.1
- encode 0.2
- delete 0.2.0
- delete 0.2
- start download of 0.1.2, 0.2.1, 0.2.2

after decide:
0: need_recon
    0.0: needed_by_parent, need_recon
    0.1: needed_by_parent, need_recon
---------------------------------
scenario 4:
    same as 3, but 0.2.1 is REC
    In this case we don't reconstruct 0;
    we just start uploads of 0.2.0 and 0.2.1
    (minimize disk usage)