[Bioperl-pipeline] The latest

[Shawn Hoon]

Kiran will have more on this.

On Wednesday, March 19, 2003, at 12:01 AM, Brian Osborne wrote:

> pipeline developers,
> I'd written down a few questions relating to some work I know you'd 
> been
> doing, let me run these by you:
> There was a proposal to allow analysis id's to be text, as in <analysis
> id="blastall">. Did you decide to make this change?
>
This is will be a nice feature but we are not there yet. Kiran is 
working on this and it
will involve some work on the parser but it is not critical right now 
so I guessing it will
take a while.

> There was a question about whether <logic_name></logic_name> could be
> nullable. Is it required or not?

I forgot what we agreed from the hackathon :) its currently not 
required if I'm not wrong and takes
the name from the program  parameter if not specified. But I  vote we 
make it required..since
people should have a name for what they want run.


> There was an effort to remove the requirement for XML::SimpleObject. 
> What's
> the latest on that?

Kiran has implemented a pure perl solution to solve all the XML 
installation troubles. So the way it
works now is first it checks whether XML::Parser (uses Expat and is 
faster) is installed. If not, it will
go with XML::SAX::PurePerl which works just as well, since we only 
parse XML only once.
We still require XML::SimpleObject which is pure perl and we feed in 
either XML::Parser or XML::SAX::PurePerl depending
on your configuration.This  I think that will ameliorate most of the 
XML pains...


shawn

>
> Thanks again,
> Brian O.
>
>
>
> _______________________________________________
> bioperl-pipeline mailing list
> bioperl-pipeline at bioperl.org
> http://bioperl.org/mailman/listinfo/bioperl-pipeline
>