August 12, 2003 -
We're still 'live', I swear!
Several small updates (in no particular order):
- I've done a bit of minor updating to this website. The news on the front page is now generated 'dynamically' rather than being one big, sloppy file, so it ought to be more readable now, as well as easier to update. The 'Other Links' page has also been updated. Among other things, Serge Gregorio's GenePHP site and Nico Stuurman's PHPLabware projects are prominently listed.
- The "merging" of the codebases here and at GenePHP is on hold for the moment. Serge is mostly unavailable for another couple of months due to his college schedule, and the GenePHP CVS at sourceforge still appears to be having problems. For the moment at least, I will be keeping the new and updated code that I write here at bioinformatics.org.
- Serge has also mentioned that he was moving towards a more "quick and dirty" sort of approach, so while he may not be generating many 'formal-style' BioPHP modules (which is more where I'm hoping to focus) in the near future, he WILL likely be churning out quite a lot of usable 'standalone' code which people will find useful, so I recommend checking the GenePHP site for code examples, as well has here at bioinformatics.org.
- The current holder of the BioPHP.org domain name has contacted me to offer a transfer of the domain name, as he's been kept from making much use of it for his project due to circumstances beyond his control. I'm working on determining what it would take to get the transfer to happen - I don't know if it will be feasible or not.
- After Nico brought up BioSQL support in his last post, I started re-evaluating my priorities a bit. I am now going over the BioSQL specifications, with the intention of using them to guide a re-implementation of the sequence object, incorporating capabilities from both Serge's seq object and my own "nuc_sequence_class", with the internal data structure redesigned to be easier to 'map' to BioSQL (as well as other projects' data, e.g. BioJava's sequence objects). This should also make the implementation of sequence EFetch easier. Expect this alternative sequence object code to be ready for posting to CVS "Real Soon Now".
- I am also hoping to start making CVS snapshots available for download, to make it easier for people to examine, test, and comment on the code that's currently written.
- Finally, developing some 'design documentation' is also on my metaphorical to-do list.
July 6, 2003 -
CVS seems to be live, though Sourceforge's webCVS interface is still grossly behind (sourceforge seems to have been having problems lately...) I am endeavoring to keep the CVS on bioinformatics.org up-to-date as well, and their webCVS seems to work fine, so you can follow changes there as well.
In brief, my contributions to current CVS updates include a working interface to NCBI's online BLAST search and a parser for the results, a minor update to the import module (now attempts to guess file format by its name/URL before resorting to reading the file directly), updated core XML parser (now works with both filehandles and plain text), minor updates to ESearch and ESummary modules to use the updated XML parser, and, finally, the first working draft of the export module design. The Export module can be found in CVS in my "point-n-laugh" subdirectory (but will be renamed and moved into the official tree shortly), and the NCBI BLAST modules are in my "semi-official" section, pending consensus on where to place it in the "official" tree.
Coming soon (hopefully):interface methods in the seq object and support for ambiguity codes, additional export formats, support for import and export of phylip format files (including support for fastDNAml options), sequence EFetch, and (pending comment on My thoughts for BioPHP "URLS") possibly extensions of the NCBI BLAST results and sequence EFetch modules that work within the existing import framework.
June 23, 2003 -
Just a quick update: this web page has been very quiet, obviously, but development on BioPHP has not been. We are in the process of merging the features of my own work here and the code in the GenePHP project. My existing file parsers have already been rewritten and incorporated into the file import module, and basic ESummary and ESearch modules are now operating. I'm working on the file export framework now, as well as preparing for merging of my own sequence object code's features with the GenePHP "seq" object.
There are no code changes currently going into CVS, as it is being moved to Sourceforge (it will continue to be mirrored here, and the previously-mentioned mailing list will remain here as well.) Once the sourceforge CVS is ready, expect a variety of code updates. Details will be available on the mailing list...
Next 3 News updates