Chempound-Quixote

[Peter Murray-Rust]

I have blogged in medium detail aboput Chempound-Quixote (note that we also have Chempound-Crystaleye) at http://blogs.ch.cam.ac.uk/pmr/2011/11/15/anyone-can-run-their-own-semantic-repository-chempound-quixote/

The position is that we need to document and deploy-test the system. Jorge did a great job of documenting and I suggest he coordinates effort on this. (A typical buglet was that he used UNIX classpaths (with ":") and Windows needs ";". It took me a little while of hair-tearing to solve this, then I forgot my solution and had to resort to urgent mails.

We need to do the following:
 * mend NWChem2CML conversion. I suspect that has to be me.
 * test the deployment (maybe with Gaussian as that works)
 * start a new phase of enthusiasm (e.g. deposit more data)

i think that deposition from the command line is fine, but it could be useful to have a GUI. I don't imagine that is too difficult. At present the conversion seems to be done separately from the server - i.e. on the machine where the commandline runs rather than on the server. Is that correct? If so maybe we need a server-side implementation where we submit the file to the server for conversion. Correct me if I am wrong:

P

--
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069

[Sam Adams]

Hi Peter,

I'll have a read of your blog in due course, but just to quickly answer your question:

Yes conversion is currently done client side, partly because it was simpler to set up, and also it distributes the load off the server.  

Sam

[Peter Murray-Rust]

On Tue, Nov 15, 2011 at 3:29 PM, Sam Adams <s.e....@gmail.com> wrote:

Hi Peter,

I'll have a read of your blog in due course, but just to quickly answer your question:

Yes conversion is currently done client side, partly because it was simpler to set up, and also it distributes the load off the server.  

Great. At this stage we probablyshould restrict users to those who know what a commandline is!

P.
 

Sam

On 15 November 2011 12:35, Peter Murray-Rust <pm...@cam.ac.uk> wrote:

I have blogged in medium detail aboput Chempound-Quixote (note that we also have Chempound-Crystaleye) at http://blogs.ch.cam.ac.uk/pmr/2011/11/15/anyone-can-run-their-own-semantic-repository-chempound-quixote/

The position is that we need to document and deploy-test the system. Jorge did a great job of documenting and I suggest he coordinates effort on this. (A typical buglet was that he used UNIX classpaths (with ":") and Windows needs ";". It took me a little while of hair-tearing to solve this, then I forgot my solution and had to resort to urgent mails.

We need to do the following:
 * mend NWChem2CML conversion. I suspect that has to be me.
 * test the deployment (maybe with Gaussian as that works)
 * start a new phase of enthusiasm (e.g. deposit more data)

i think that deposition from the command line is fine, but it could be useful to have a GUI. I don't imagine that is too difficult. At present the conversion seems to be done separately from the server - i.e. on the machine where the commandline runs rather than on the server. Is that correct? If so maybe we need a server-side implementation where we submit the file to the server for conversion. Correct me if I am wrong:

P

--
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069

[Jorge Estrada]

For convenience, I have posted the documentation Peter mentioned in a
previous email. You can see it at:

https://bitbucket.org/jestrada/quixote-docs/wiki/Home

Thanks to Peter, Sam and Pablo E. for their help while preparing these
documents (and I still have to include some of their last comments).

The source documents are available at:
https://bitbucket.org/jestrada/quixote-docs

Everybody is welcomed to contribute or build on them.

Jorge

[Sam Adams]

That looks great!

Thanks,

Sam

[Peter Murray-Rust]

On Fri, Nov 18, 2011 at 12:21 PM, Sam Adams <s.e....@gmail.com> wrote:

That looks great!

Agreed
 

And if you have followed, the system is now installable without requiring emails.

Jorge - the diagrams are exactly the sort of thing we need. I am going through dolcumenting the packages as far as I can.

Sam,
Any thoughts on the duplicated packages - I am tempted to rename.

P.

[pca...@db.uc3m.es]

Dear all,

Best wishes for a happy Christmas time and a fruitful new year from La
Mancha, home place for Quixote and seasonly disguised fellow rider
Sancho :)

P.

----
Pablo de Castro
SONEX Workgroup for Scholarly Output Notification and Exchange
http://sonexworkgroup.blogspot.com/
e-mail: pca...@db.uc3m.es

[Marcus D. Hanwell]

On Sun, Dec 25, 2011 at 3:13 PM, <pca...@db.uc3m.es> wrote:
> Dear all,
>
> Best wishes for a happy Christmas time and a fruitful new year from La
> Mancha, home place for Quixote and seasonly disguised fellow rider Sancho :)
>

Merry Christmas and Happy New Year from an unseasonably warm upstate
New York. The snowboard has yet to have an outing this season! Here's
hoping for a very fruitful new year with lots of exciting developments
already on the horizon.

Best wishes,

Marcus

[Peter Murray-Rust]

and from me. Settling down in the next day or so to refactor JUMBO-converters for the workshop