(NB: A previous release of this code is available at Li Heng's MAQ site; maq.sf.net)
"tgap" is an experimental curses based (Unix) sequence viewer. It's a completely new program designed to explore new algorithms for handling the huge scale of data from modern sequencing instruments (Illumina/Solexa, ABI SOLiD, etc).
It's not a robust and finished product, and nor is it likely to remain stable in terms of file formats. For now though it'll probably remain text based as this happens to be fast and easy for development/debugging. My advise would be if you want to experiment with it keep a copy of the source and binary along side your databases you produce. I'm not likely to be keeping binary compatibility between releases until much further down the development process.
The long term plan is to improve the fundamental indexing and querying algorithsm a bit more and then start adding more Gap4 code to it (editing, contig joining/breaking, etc) before finally adding a proper Gap4 style GUI to it. At that stage it'll probably become Gap5.
I cannot really commit to much support on how to get the data in the correct format to use this, so instead I'll just plug MAQ instead. It's well worth using that in conjunction with the -m option of tgap.