Crashes/assertions when having alignments without related entries

[epruesse]

/home/epruesse/arbsilva/arb/lib/libARBDB.so(GBK_dump_backtrace+0x3c)[0xb7f674fc]
/home/epruesse/arbsilva/arb/lib/libARBDB.so(sigsegv_handler_dump+0x43)[0xb7f6a3b3]
[0xb800e400]
arb_edit4[0x807afa5]
arb_edit4[0x807b01c]
arb_edit4[0x807b01c]
arb_edit4[0x807b01c]
arb_edit4[0x808a07d]
arb_edit4[0x807398e]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe5)[0xb780f775]
arb_edit4[0x8058231]

Unangenehmer Weise funktioniert es mit DEBUG=1 …

[epruesse]

Hmm, hier noch ein paar Zeilen davor …

Action: 'arb_edit4 -c default_configuration &']
No '.arb_prop/edit4_ali_5s.arb' found
No '.arb_prop/edit4_rna.arb' found
Using '.arb_prop/edit4.arb'
Note: SILVA Aligner disabled (galign not found)
Note: SILVA Aligner disabled (galign not found)

-------------------- ARB-backtrace for 'received signal 11':

[fog]

From my point of view no need to fix this - seems to be specific for demo.arb because I never encountered problems with other databases.

[baderk]

There is an alignment entry ali_16s in the demo.arb database, but there are no related sequence datasets.

Generally: when creating a new alignment entry without having related datasets (yet), some ARB functionalities (like EDIT4 or the database optimization) may throw segfaults or assertions.

[westram]

Meanwhile most components are quite stable if species do not contain data in the used alignment. Just tried ali_16s of demo.arb with editor, parsimony, dist and phylo: no crashes, no completely weird error messages.