ssujun02.arb was updated as ssu_jan03.arb with respect to the status of the EMBL sequence databases as it was in January 2003. 
Only sequences comprising at least 1000 nucleotides encoding rRNA were included.
All entries were evaluated against the current (Oct. 2003) versions of the EMBL, RDP and European rRNA Database (tagged as EBI, RDP, DEW).
Entries no longer present in EBI, RDP, or DEW are not maintained in ssu_jan03.arb. 
Furthermore, entries containing conflicting data (with repsect to EBI, RDP, DEW) or violating the 1000 nucleotide criterion were removed.
It is up to you to reimport them from previous releases.
The sequence version stored in 'ali_16s/data' is that of the EMBL databases (usually).
The 'full_name' entry was taken from RDP (usually), given that this database is closest to the current approved list of bacterial names.

The alignment of variable regions within eucaryotic sequences has still to be improved!

The alignment in ssu_jan03.arb differs from that in previous releases mainly by additional (eucaryotic) insertions.
To adapt your current ARB database version it is recommended to use the most recent release of the ARB software pakage (Linux Beta 030822).
In this version the merge tool allows to use SAIs for adaptation of the alignments.

In case you were using ssujun02.arb alignment as template for your additional data and din't insert additional alignment gaps, the following procedure is recommended:
Create 'gap_updt_jan03' as an SAI containing 'x' (lower case) at any alignment position (take care not ! to use the editorin compressed mode!) in your database 
or import it from ssujun02_merge.arb (it's an excerpt of the original ssujun02.arb providing in addition 'gap_updt_jan03') using the merge tool 
('ARB_INTRO' 'MERGE TWO DATABASES' ssujun02_merge.arb = 'Database I', your datbase = 'Database II') and transferring that SAI to your database ('Transfer SAIs ...').
Merge your ARB database (= 'Database I') to ssu_jan03.arb (= 'DatabaseII').

Before merging, take care that your datbase is formatted (in 'ARB_NT' window: 'Sequence' > 'Sequence/Alignment Admin' > 'Format'. The new ARB software reminds you any 
time you start it or can be customised to automatically format the database (my choice). Furhtermore, check for damaged sequence data. (Damaging my occure occasionally 
working with not formatted sequences or importing 'poisoned' sequences (never pass data through MS windows or similar word processors!!!). Checking for 'poisoned'
sequences can be done by searching (SEARCH and QUERY') for all 'species', 'More Functions' > 'Modify fields of listed ..' > 'Count poisonde characters ..' The searching
result is stored in the fiels 'poisoned' and should be '0'. If not, remove, correct or replace the respective entries!
Unfortunately, ssujun03.arb seems to contain 'poisoned' sequences.   

Take care to synchronise the 'name' entries ('Check Names ...') before searching or transferring any data!!! 
Serach for the sequences in 'Database I' you want to transfer. ! Empty 'Search strig' field means: compare the respective firled in 'Database I' and 'Database II' 
for identity or difference according to your selection in 'QUERY TYPE'.
To adopt the alignments check 'Preserve Alignment' (we know, it's misleading, term will be changed), press 'SELECT'.
In the window popping up press 'Find candidates', select 'gap_updt_jan03' and press 'Add'. Then press the respective 'Transfer ..' button of the 'TRANSFER SPECIES' window.

In case your database is not based upon a previous ARB database release or the alignmet has been substantially modified try to 'Add' a selection of 'species' present in both 
databases and 'related' to those you want to transfer.

See also the information accessible via the 'HELP' buttons of the respective windows.

WL, Oct03