axiope.com

The Axiope Blog

Wednesday, December 10. 2008

User Input Posted by Rory Macneil at 11:24

Comments (0)
Trackbacks (0)

User Input

We are about three months into the beta testing, and its going well. About 50 labs and institutions, and a fair number of individuals are participating, from 14 countries. So far so good!

We really appreciate it when we get feedback -- its what we need to keep improving e-CAT and make it responsive to the community’s needs. So thanks to all those who have made forum posts and sent emails. Professor Larry Gonzalez, Director of the University of Oklahoma Health Sciences Center, Department of Psychiatry & Behavioral Sciences, has kindly agreed to let us reproduce an email he sent recently commenting on his experience with e-CAT and including his suggestions for improvements. This serves two purposes. It will let others see what one active and creative tester has been doing with e-CAT, and, through our responses to Larry’s comments, it will give you an update on some of the changes we are planning over the next several months.

Here is Larry’s email, with our responses in italics.

I am generally favorably impressed with e-CAT. I have been working to develop e-CAT classes that are suited for my data formats.

1. I find class editing to be fairly easy. I would like to be able to copy a class with a new name so that I can create new classes by editing copies of old classes instead of having to recreate the entire class.

This will be possible in the mini release we will be bringing out in January.

2. Also, I haven't figured out how to get a drop-down box as an item in a class.

We will add this capability, but do not have a definite date for release.

3. Another thing that I would like to see is the ability to add new items to a selection list on-the-fly. That is, if I have a Choice item with a list of items, I would like the user to be able to add new items to the list and have this modified class be saved with the new item added. For example, if I have a choice list that includes different mouse strains that we are using I would like the user to be able to add a new strain to the list when they are entering information for that class, and then have that strain included in the list the next time that the class is used.

We may add this capability in the future, but it requires careful design. It carries some potential downsides in that when new items are added on the fly, they can easily become messy, e.g., with typos or multiple spellings for new items. To be useful it probably would require a new field type, predictive typeahead, and suggestions for fixing mistakes– quite a bit of sophistication.

4. Another thing that I haven't been able to use is the new CSV importer that was added in the last release. When I try this with a 1-dimensional array nothing is imported, but I seem to get a new class created with the number of string items that I had in my CSV file. A 2- dimensional array didn't work either.

We will be improving the CSV importer, which as you point out still has problems associated with it. These improvements will be in the commercial release which is scheduled for March/April.

5. As I noted in my forum posting, the text editor doesn't scroll properly with long documents. The cursor will scroll off the screen, but the text isn't shifted, so you can't see where the cursor is located and you can't get to the text that is off the screen.

This is a bug, and it will be fixed for the commercial release.

6. I am looking forward to being able to do a real delete, because the trash bin gets pretty busy. I've also noticed that links to items that were moved to the trash bin still link to the item in the trash bin.

The delete capability will be in the commercial release.

7. One more item on my wish list would be the ability to paste an image rather than having to create a file and then import the file.

This is one where it would be useful to get a bit of elaboration as to exactly what you would like to be able to do. As background, it is possible to drag an image from a web page in another browser (firefox or safari but not IE) and drop it into an e-CAT record. It also is possible to right click and copy an image on another web page and then paste it into an e-CAT record. In addition to this, what would you like to be able to do?

You asked for more info about my comments about pasting an image: I haven’t tried this with an image from a browser, but my data acquisition software allows a right-click copy of plots to the clipboard that I can paste into programs like ‘paint’. I haven’t been able to paste this into the editor in e-CAT. A control-V will paste the plot into Paint, but nothing happens when I do this in the e-Cat editor. The advantage of this is that I could quickly copy and paste plots of data into e-CAT during an experiment without interrupting the experiment.

A direct paste of an image probably will not be possible. The editor we use in e-CAT is an HTML editor, so the output it produces does not really cater for a direct image paste - it can only be a URL pointing to an image somewhere on the web. This is why an image needs to be uploaded to e-CAT first before it can be displayed in a record.

8. I would also like to be able to copy an entire project or experiment with all of the child records included. This would allow me to set up a project template that could be reused with all of the standard operating procedures and other items that would be recreated each time a new experiment is performed.

Or perhaps there is a better way to set up a project than the way I am trying to do this. Here is how I am setting up a project:

• The project describes the overall study, research goals, general approach, etc.
• To this I attach experiments that use a specific set of standard operating procedures which are linked or copied as children to the experiment.
• Then each time we run a subject, a subject record is created as a child of the experiment.
• Each subject record will then have child records that include information from individual tests that are done on each subject and the test record will have child records for each phase of the test.
• The next time a new subject is run, the same record structure is used. -- here it would be helpful to be able to copy a 'subject template' that would copy the structure of this set of records.
• If there is to be a change in the standard procedures, a new experiment is created, but all the rest of the parent/child structure is the same, so I would want to copy the entire 'experiment template'.

The way in which you are setting up projects sounds very sensible to us! We will aim to have the 'copy/template' capability you describe in the commercial release.

Hope this helps. Thanks for your efforts in developing this.

Thanks again to Larry, whose comments were very useful to us. We hope they are useful to other testers, too. We welcome comments, suggestions and questions from everyone, in the suggestions forum or if you prefer by email to beta@axiope.com.

Friday, October 31. 2008

Take Control of your Data Posted by Jonathan Ritchie at 14:51

Comments (0)
Trackbacks (0)

Take Control of your Data

We have released a new version of e-CAT and this blog post discusses the changes in the release and why we decided to make those changes.

No important data warning removed!

First of all, existing users will notice there is no orange box on the login screen warning them not to store important data in the beta. We feel that, although the software is still definitely in beta form, we are confident enough in our system and its ease of backup to allow users to start using it properly to store data that is important to them. To offer increased peace of mind to hosted service users we will take a daily backup of the database that runs e-CAT on axiope.com so that in the unlikely event of something going wrong there will be very little, if any, data loss. For customer managed users we will be speaking to your administrators and they will inform you of the backup procedures they have in place.

We have changed more than just the login screen however!

Permissions system improved

The previous release did not make it as easy as it should be to take control of your own data. By this we mean that if you are able to create new Projects in Projects and Experiments then you should also be able to make those Projects private and decide who is allowed to view and edit them. In fact this is a general principle that should apply whenever you are creating anything anywhere on the system. If you create it you should control it.

To make this possible when you "Add" or "Import" any records (Projects, Experiments or other data types) you are now given two choices:

The first, "Add/Import with full permissions" will add whatever record you are creating and make it private (only you can view it) but also gives you full control over it, so you can then browse to the created record and change its "Sharing" settings to enable the users and groups that you want to see your record to do just that. If you want to collaborate with other users, give them "Edit" permission as well.

The second option, "Add/Import with parent permissions" will add a record as expected but does not set any permissions for that record. Because of the way e-CAT's permissions work, this means that whatever permissions existed for the parent record will also apply to the newly created record. What does this mean? Well imagine you had used the option "Add/Import with full permissions" to create a Project in Projects and Experiments. You then have full control over the Project, so you decide to add some Experiments to it. If you were to create those Experiments using "Add/Import with parent permissions" then you would also have full control over those Experiments because they inherit their permissions from the parent Project. The real utility of this option is that you can then change the permissions for ALL the Experiments AND the Project simply by setting the permissions on the parent Project record. This is a very powerful feature that saves setting permissions for individual experiments.

To make this a little more obvious, when you are creating a Project "Add with full permissions" is recommended to you. When you are creating an Experiment the "Add with parent permissions" option is recommended to you. When you are importing data it’s impossible for us to make a recommendation - you choose the option which best applies to what you are doing. If you are ever in doubt, then "Add with full permissions" is the best option because you can then go back and change all permissions later. The boxes next to the buttons - "What does this do?" - are help guides. You can mouse over them to see what the permissions would be were you to choose either option.

To make working with the actual permissions themselves easier (there can be an awful lot of checkboxes if you have many users or groups) we have added four convenience buttons to the top of the list of both user and group checkboxes. These can be used to quickly set or unset all the parent or other checkboxes in the table of checkboxes so you no longer need to go through them all one by one. Note here though that if the parent checkbox is set for a user or group then the values of the other checkboxes are in effect meaningless - the permission will still be taken from the parent record.

CSV importer for spreadsheet data

We have also added a simple CSV importer to this release which should enable you to get at some of your Excel data. You can import data from spreadsheets into sets of records with each row in the spreadsheet being converted into a record in e-CAT. You can then view the data in tabular form on the web. This functionality will be improved further in future releases.

Better attachment stores

Attachment stores can now be on UNC network drives when e-CAT is running on Windows servers. You just supply the server and share name and e-CAT can display the files. The interface for creating all kinds of attachment stores has also been improved.

Coming soon...

There is a lot more in the pipeline - the next release after this one will enable users to build their own classes. This major feature, currently only available to administrators, will give much of the power currently available with records to the objects that define them - classes. For example, if you don't like having all these fields in Experiments you can remove some, add some or change the type of some to build your own custom Experiments that record exactly the data you want with the structure you require.

We also plan to add a proper delete function (no more moving to Trash!) with the ability to restore deleted records and add more importers, dealing with a range of custom image types, to allow for the automatic creation of metadata rich record structures from your files.

We hope you like the latest e-CAT and are as excited as we are about the features to come. As always we welcome all feedback, both positive and negative. It’s the only way we can make the software better.

Monday, September 15. 2008

Welcome to e-CAT! Posted by Rory Macneil at 10:23

Comments (0)
Trackbacks (0)

Welcome to e-CAT!

A warm welcome to the first beta testers of e-CAT. e-CAT is the first fully fledged online lab notebook specifically designed to enable lab scientists to conveniently document experiments and manage data online. I thought it would be useful to give you a quick overview of how e-CAT came about and what we are trying to achieve with it, so here goes.

Back in 2003 Fred Howell and Robert Cannon (who have since gone on to establish Textsensor) and Nigel Goddard at the University of Edinburgh’s School of Informatics founded Axiope to commercialize a user-friendly XML database designed to help scientists manage their data. They called it Catalyzer. For the next several years Axiope had some limited success with Catalyzer, and some loyal early adopters like Martin Spitaler at Imperial College London, Alex Swarbrick at the Garvan Institute in New South Wales, and Jeff Forbes at NIAMS have made Catalyzer work well for them.

But as 2005 and 2006 (when I was getting more actively involved) rolled around, Catalyzer’s constraints—restricted scalability, an overly complicated user interface, client-server configuration, and high price ($1,000 a seat non profit)—limited uptake. And these problems were compounded by the emergence of Web 2.0 apps. It was a good start that Catalyzer had been ‘designed by scientists for scientists’, but the client-server model was no longer what people wanted.

In early 2007 we decided to bite the bullet, go back to the drawing board and develop a product we were pretty sure the scientific market was crying out for: a browser based lab notebook with a familiar front end like a paper lab notebook that links in with a user friendly, configurable and scalable database backend. We decided to call it e-CAT, and that’s what we have been working on over the past 18 months.

We got a major boost in October 2007 when a group at the FDA led by Ron Rabin and Robert Fisher, who were looking to adopt an ELN, said they liked what they saw of our plans for e-CAT. Since then Ron, Robert, and their group have patiently stuck with us and provided invaluable input into how to design e-CAT to meet the real needs of working scientists.

In early 2007 we also became aware of developments in scientific collaboration and communication that were happening in conjunction with the advent of Web 2.0 – for example OpenWetWare and Nature Network. We began to follow the dialogs that were springing up and were pleased when in Corrie Lok’s February 2007 forum, “Tell us what you want next on Nature Network”, Egon Willighagen, Pedro Beltrao and Deepak Singh started talking about how useful it would be to have a hosted ELN which sounded a lot like what we intended e-CAT to become.

These discussions have of course moved on, and we have followed with interest the more recent discussions involving Cameron Neylon, Jean-Claude Bradley, Martin Fenner, Neil Saunders, Deepak, Bench Press and others about blogs, wikis, OLNs and how they might relate to the development of the web as a platform for scientific research. I am particularly intrigued by one of Cameron’s latest thoughts, that what is needed is a “lab book system, or a virtual research environment” whose “core is an authoring tool (or tools)”. We believe that e-CAT is well suited to be a core tool and a key part of the next generation research environment.

I had the pleasure of meeting Martin and Cameron at the recent Sciblog 2008, hosted in London by Nature Network. We are keen to establish an active dialog to keep in touch with the latest thinking of these Life Scientists on Friendfeed, and ensure that e-CAT evolves in ways which are relevant to their needs. I will be speaking in more detail in future posts about the issues the Life Scientists are coming across as they continue to push the envelope of using the web as a platform for conducting scientific research.

Our immediate objective is more limited than the leading edge innovation of the Life Scientists: to make online documenting of experiments and data sharing as routine as paper lab notebooks are today by providing lab scientists with a convenient and affordable online lab notebook.

In that regard it was also great to meet Heather Etchevers at Sciblog. As part of a panel with Jean-Claude Bradley and Bob O’Hara, Heather gave a presentation called, “Laboratory Notebooks Online: Perspectives from the Bench”. Heather’s presentation is a thoughtful overview of the bench scientist’s perspective on online lab notebooks: what scientists want, what they want to avoid, what is needed in an OLN, etc.

One way we are trying to make e-CAT useful and accessible is to provide meaningful online support - this was the number one request that came back from a survey we conducted – ahead of any single feature request! There are two introductory videos here and here as well as a Help section which has been written in plain English with a view to enabling the ordinary (i.e. non technical) user to quickly and naturally become familiar with e-CAT and fit e-CAT into their normal working practices.

So what happens next? After the private beta we will open things out through a public beta. The two beta periods in total will last for around 6 months, and we intend to launch the commercial version of e-CAT in Q1 2009.

Oh, one final thing, as we know from our earlier experience with Catalyzer, scientists are definitely not crying out for an OLN which costs $1,000 or more. We want everyone in the community to have access to e-CAT , so our pricing model is radically different from the earlier one we used with Catalyzer, and the models of ELN vendors. When the commercial version of e-CAT is released it will be priced at $50 non profit/$100 commercial per user per year, with scaled volume discounts.

If you would like to join the beta testing, email us at beta@axiope.com. Thanks to everyone who gives e-CAT a try! Please give us your feedback, preferably on the forums so other users can also benefit from your experiences, but if not by email to beta@axiope.com. Its only through your feedback that we will be able to make e-CAT a more useful tool.

Rory Macneil

« previous page (Page 1 of 1, totaling 3 entries) next page »