Project plan
From W
global links:
- new NPD admin system (test version)
- new NPD user system (test version)
- Reports - weekly reports. setup and test reports, and others
- Meetings - information on past and future meetings
- Project plan - the original, and current updates
- the completed dissertation
- Original NPD - info and links
- Web services links
This page details the plan that is being worked to at present, and will be updated as changes are required, with short notes on why the change is occurring. Fuller details of such changes should be available in the Weekly reports.
week 1 to 3
Due to problems with original NPD accessibility in week 1, and its failure in week 2, some changes need to be made for the intentions to create the NPD admin system. Current plans are to complete a working prototype of the admin system by week 3, after which time Wendy can test it and suggest changes. it is now essential to get this working, given that the original NPD admin system no longer operates at the HGU.
As the full NPD could not be accessed, it will be impossible to refactor that system to reflect changes to the database schema. I have identified areas where the schema could have been altered, so these will be documented but unimplemented. The admin system will read and write to the tables as they currenly exist on the mysql server at the HGU.
A meeting should be held at the end of week 3 to analyse the status of the admin system and compare to the tasks declared in the original plan. Also, the testing carried out will be formally documented.
A first draft of the dissertation should also be prepared for the end of week 3.
week 4
Get ogsa-dai running, compare its functionality with web services available from other relevant databases, work on updating the prototype admin system based on these findings, look into forwarding info from the NPD to the gene ontology database. Get a response from Marco.
week 5 to 7
Meet with Bob and Jano on Tuesday of Week 5 to go over project progress, and get some feedback in dissertation draft 1. There will then be a lecture on what sort of info we need to include in our dissertations.
Have another meeting with Wendy in week 6 to confirm required updates to the admin system.
Apply the new querying tools, with the intention to have a version available by the end of week 6. This should include ogsa-dai and traditional web service functionality, both for the admin mechanism and for users. Examples will hopefully include passing data to the gene ontology database and pulling data automatically from EBI and other services, and showing a user interaction via Taverna with the web services offered, based on input from Marco Roos.
Prepare dissertation draft 2, with more content for the intro and NPD refactoring sections, and at least a plan of the layout of the web services section.
week 8 to 9
Work on the user testing with Marco, to produce a working model of a user interaction with the data and an evaluation of its applicability. Submit dissertation draft 3, which will require all of the layout in place, and alot of the content ready. Have Wendy try a test of the admin system, and answer a short questionnaire on its suitability as compared to the old system.
week 10
Receive feedback on dissertation draft 3, make any quick changes that I can submit back to Jano for comment before the end of the week. Work on updating installation notes and code comments. Complete any user testing.
week 11 and 12
Continue dissertation write-up, installation notes and comments. Prepare all the required software for installation on a live server, so that an archive is ready for submission with the dissertation.
Original plan
The original plan as prepared for the Research Project Preparation class:
Feedback
I have written some paragraphs paraphrasing the points made during the marking of the research project planning course. Things to keep in mind from the feedback on the original plan:
Development work is described very well and is structured in an excellent manner, but is limited to the 'job at hand'. A good in-depth probe is shown into the current database structure and what are the technologies proposed, but answer Why is grid enabling important/useful/... anyway?
Clarify what assessment will be conducted, what and who is involved and when it will be done. Ask: "How am I going to declare the success of my project?" Then find a way of measuring 'success' and devise an experiment to do so. Incorporate into the workplan effort dedicated to writing the dissertation. The dissertation will make up most of the final mark.
A good case was made as to why the project is of importance to the HGU staff directly associated with the NPD but more discussion of how publishing the NPD through OGSA-DAI should help expand or extend its use within the community is required: mention is made of Marco Roos as a potential new user, so maybe more could be made of that - discussing how the current system limits what he can do with the data, for example.
This could end up looking like a software implementation task, rather than a research project, so keep in mind those aspects which concern the assessment of the usefulness of e-science technologies in this domain, not just to this specific application. A bit more information on the benefits of the GRID to the project, more background on OGSA-DAI and the GRID and a stronger case for the benefits of grid computing to research in general would be good. How generalisable is the proposed work: could the lessons learnt in this project (and, maybe, even some of the software) be re-used for enhancing other, similar databases?
A simpler test client than Taverna for verifying the proper functioning of the OGSA-DAI installation, and a more explicit presentation of the test suites that will be used would be beneficial.
