Comments on: The Big Table Issue http://www.8164.org/the-big-table-issue/ Jin Yang's blog on web design, philosophy of Taoism and life in general. Tue, 09 Mar 2010 23:08:54 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: Marion http://www.8164.org/the-big-table-issue/comment-page-1/#comment-8483 Marion Tue, 11 Aug 2009 05:48:49 +0000 http://www.8164.org/?p=1158#comment-8483 I had found a great option for this, but now I can't find it! Basically, it enabled the viewer to collapse a series of columns- which was perfect for what I have in mind.... While trying to find that solution again, I came across this one: http://www.massless.org/%5Ftests/grid1/?s=1&r=40&c=40 I had found a great option for this, but now I can’t find it! Basically, it enabled the viewer to collapse a series of columns- which was perfect for what I have in mind….

While trying to find that solution again, I came across this one:
http://www.massless.org/%5Ftests/grid1/?s=1&r=40&c=40

]]> By: Olivier Vanbiervliet http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1640 Olivier Vanbiervliet Tue, 03 Feb 2009 14:56:16 +0000 http://www.8164.org/?p=1158#comment-1640 The frozen column demo is here: http://demos.telerik.com/aspnet-ajax/grid/examples/generalfeatures/frozencolumns/defaultcs.aspx The frozen column demo is here: http://demos.telerik.com/aspnet-ajax/grid/examples/generalfeatures/frozencolumns/defaultcs.aspx

]]> By: Olivier Vanbiervliet http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1639 Olivier Vanbiervliet Tue, 03 Feb 2009 14:55:21 +0000 http://www.8164.org/?p=1158#comment-1639 Telerik has a great grid control with column & row freeze that works on most major browsers. Here's a live demo: http://demos.telerik.com/aspnet-ajax/grid/examples/client/scrolling/defaultcs.aspx Telerik has a great grid control with column & row freeze that works on most major browsers. Here’s a live demo: http://demos.telerik.com/aspnet-ajax/grid/examples/client/scrolling/defaultcs.aspx

]]> By: Best of the web - January 2009 - Morning Break http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1637 Best of the web - January 2009 - Morning Break Tue, 03 Feb 2009 11:19:26 +0000 http://www.8164.org/?p=1158#comment-1637 [...] The Big Table Issue [...] [...] The Big Table Issue [...]

]]> By: Dmitry http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1287 Dmitry Tue, 13 Jan 2009 22:45:08 +0000 http://www.8164.org/?p=1158#comment-1287 Hmm.. I've never had to create an interface for such large data sets, so I've honestly no idea how to best approach this, especially the column locking issue. However, I think Steven and steve are spot on. The data here is overwhelming and I don't see how a human can process it if it can't even fit on one screen. To design an interface, you need to know exactly what people want to do with it. You don't need to focus on the content as much as the process that the user will go through. What are they going to look at on this table? What are they going to compare? etc. Once you know what people really need, the resulting interface may not even be a table. Maybe all they actually want is a set of charts. I quite like the way Google Analytics presents their data. The data itself is overwhelming, but Google crunches it down into focused graphs and charts, showing you how different areas of your website perform. You can customize the graphics on the fly, selecting the fields you want to see, as well as their presentation (bar, pie, line etc). I think the best UI for a huge table isn't a table at all, but a presentation interface and a search interface. The faster you can get meaningful data out, rather than all of the data, the more useful it would be. Just my opinion though ;-) Hmm.. I’ve never had to create an interface for such large data sets, so I’ve honestly no idea how to best approach this, especially the column locking issue. However, I think Steven and steve are spot on. The data here is overwhelming and I don’t see how a human can process it if it can’t even fit on one screen.

To design an interface, you need to know exactly what people want to do with it. You don’t need to focus on the content as much as the process that the user will go through. What are they going to look at on this table? What are they going to compare? etc. Once you know what people really need, the resulting interface may not even be a table. Maybe all they actually want is a set of charts.

I quite like the way Google Analytics presents their data. The data itself is overwhelming, but Google crunches it down into focused graphs and charts, showing you how different areas of your website perform. You can customize the graphics on the fly, selecting the fields you want to see, as well as their presentation (bar, pie, line etc).

I think the best UI for a huge table isn’t a table at all, but a presentation interface and a search interface. The faster you can get meaningful data out, rather than all of the data, the more useful it would be.

Just my opinion though ;-)

]]> By: Jin http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1286 Jin Tue, 13 Jan 2009 15:26:20 +0000 http://www.8164.org/?p=1158#comment-1286 Hi Martinos. I've experimented with YUI before. I didn't see any solutions they have for freezing columns. Hi Martinos. I’ve experimented with YUI before. I didn’t see any solutions they have for freezing columns.

]]> By: Martinos http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1283 Martinos Tue, 13 Jan 2009 04:40:09 +0000 http://www.8164.org/?p=1158#comment-1283 Hi Jin, Did you get a look at this? : http://developer.yahoo.com/yui/examples/datatable/dt_fixedscroll.html Hi Jin,
Did you get a look at this? :
http://developer.yahoo.com/yui/examples/datatable/dt_fixedscroll.html

]]> By: Steven Clark http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1276 Steven Clark Mon, 12 Jan 2009 07:13:46 +0000 http://www.8164.org/?p=1158#comment-1276 Although, I confess it was a scan read because it was a long post. I didn't spend that much time Steve :) Although, I confess it was a scan read because it was a long post. I didn’t spend that much time Steve :)

]]> By: Steven Clark http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1275 Steven Clark Mon, 12 Jan 2009 07:12:28 +0000 http://www.8164.org/?p=1158#comment-1275 Yep I read the post, and I agree with what you're saying... I just said it slightly differently. I like your term "actionable intelligence", it fits well and I seem to have heard it before somewhere... My point really was that large data sets are unusable, and even more often they are there for stakeholder appeasement rather than desire to meet a user need. Apologies if I rambled that into a confused muddle... Yep I read the post, and I agree with what you’re saying… I just said it slightly differently. I like your term “actionable intelligence”, it fits well and I seem to have heard it before somewhere… My point really was that large data sets are unusable, and even more often they are there for stakeholder appeasement rather than desire to meet a user need. Apologies if I rambled that into a confused muddle…

]]> By: steve http://www.8164.org/the-big-table-issue/comment-page-1/#comment-1266 steve Sun, 11 Jan 2009 18:09:18 +0000 http://www.8164.org/?p=1158#comment-1266 you obviously didnt read his post, but I dont blame you. Not only can't browsers handle thousands of rows at a time, humans cant either! It all boils down to what I like to call, "ACTIONABLE INTELLIGENCE" and a table with 20,000 results is just not actionable, i.e. you cant DO anything with it. you can barely read it, and you cant visually verify the data 'cause there is just too much (Using intelligent statistical sampling methods you can take a few hundred records and verify them independently). It needs to be presented as: 1) A data dump download, readable and parseable by client-side tools. 2) A summary that describes the data : We have 50 columns, Name,sex,age,height.... and 10,000 rows : we can drill down or filter/sort by Name,sex,height and reigion and any combination of those .... in essence... a THUMBNAIL! :) you obviously didnt read his post, but I dont blame you.

Not only can’t browsers handle thousands of rows at a time, humans cant either!

It all boils down to what I like to call, “ACTIONABLE INTELLIGENCE” and a table with 20,000 results is just not actionable, i.e. you cant DO anything with it. you can barely read it, and you cant visually verify the data ’cause there is just too much (Using intelligent statistical sampling methods you can take a few hundred records and verify them independently).

It needs to be presented as:
1) A data dump download, readable and parseable by client-side tools.
2) A summary that describes the data
: We have 50 columns, Name,sex,age,height…. and 10,000 rows
: we can drill down or filter/sort by Name,sex,height and reigion and any combination of those
…. in essence… a THUMBNAIL! :)

]]>