Comments on: The Editable Grid Antipattern http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/ Chris Stevenson's blog Wed, 07 Jan 2009 08:08:31 +0000 http://wordpress.org/?v=2.6 By: Chris Barrow http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-691 Chris Barrow Wed, 31 Oct 2007 23:38:18 +0000 http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-691 I too have experienced these same problems, and have concluded that a fully featured, client-heavy grid should never be used in web apps. The alternatives are far easier to implement, maintain, and extend. Long live the editable grid antipattern! I too have experienced these same problems, and have concluded that a fully featured, client-heavy grid should never be used in web apps. The alternatives are far easier to implement, maintain, and extend.

Long live the editable grid antipattern!

]]> By: Michael Goldobin http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-243 Michael Goldobin Fri, 18 May 2007 02:14:25 +0000 http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-243 Hmm... Very-very familiar problems :) I've re-read the "Lean development" one more time and yes, I really missed the "5 times". Great notion! Also, I finally retried your idea of reflecting on ASP.NET page to test something like validation. Your TM is mentioned :) - http://lazyloading.blogspot.com/2007/05/testing-aspnet-pages-outside-http.html Hmm… Very-very familiar problems :)
I’ve re-read the “Lean development” one more time and yes, I really missed the “5 times”. Great notion!

Also, I finally retried your idea of reflecting on ASP.NET page to test something like validation. Your TM is mentioned :) - http://lazyloading.blogspot.com/2007/05/testing-aspnet-pages-outside-http.html

]]> By: Claus http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-216 Claus Sun, 01 Apr 2007 17:19:29 +0000 http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-216 I think the caveat at the end is better than the warning: If there's no spec, other than (Like the stuff we're used to) - don't write the software. Just use the stuff they're used to. I think the caveat at the end is better than the warning: If there’s no spec, other than (Like the stuff we’re used to) - don’t write the software. Just use the stuff they’re used to.

]]> By: Kim http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-138 Kim Mon, 05 Mar 2007 10:46:53 +0000 http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-138 I mostly agree with you, but James' point of usability (massive dataentry) is also a factor that must be taken into consideration. One issue that I don't understand from your post, is that you aximatically assume that any databinding must be done to the underlaying table. There's no reason why a grid couldn't be bound to a business object which could trap and enforce all the issues you mentioned in your post. I mostly agree with you, but James’ point of usability (massive dataentry) is also a factor that must be taken into consideration. One issue that I don’t understand from your post, is that you aximatically assume that any databinding must be done to the underlaying table. There’s no reason why a grid couldn’t be bound to a business object which could trap and enforce all the issues you mentioned in your post.

]]> By: Lance Walton http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-137 Lance Walton Mon, 05 Mar 2007 08:37:45 +0000 http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-137 Agreed, Chris. It can be particularly bad when the requirement is actually expressed in terms of the underlying database tables >:-{ Gerald Weinberg has an analytical technique similar to “Ask why? five times” which consists of alternating "Why?" with "What?". e.g. "Why do you need that?" "What will that do for you?" "Why do you need that done?" etc. If I recall correctly (and I read the book 10 years ago so my memory migh be faulty), he says that you often have to ask about 6 questions to get to what he terms "The Data Question" which is where you get the information you really need. Regards, Lance Agreed, Chris. It can be particularly bad when the requirement is actually expressed in terms of the underlying database tables >:-{

Gerald Weinberg has an analytical technique similar to “Ask why? five times” which consists of alternating “Why?” with “What?”. e.g.

“Why do you need that?”
“What will that do for you?”
“Why do you need that done?”
etc.

If I recall correctly (and I read the book 10 years ago so my memory migh be faulty), he says that you often have to ask about 6 questions to get to what he terms “The Data Question” which is where you get the information you really need.

Regards,

Lance

]]> By: Asking why five times. « Rui Silva http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-122 Asking why five times. « Rui Silva Sat, 03 Mar 2007 18:32:38 +0000 http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-122 [...] March 3rd, 2007 in Other Today, I have found an interesting post named “The Editable Grid Antipattern”. The illustrated scenario it’s a perfect [...] [...] March 3rd, 2007 in Other Today, I have found an interesting post named “The Editable Grid Antipattern”. The illustrated scenario it’s a perfect [...]

]]> By: James Crisp http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-113 James Crisp Wed, 28 Feb 2007 00:44:50 +0000 http://skizz.biz/blog/2007/02/23/the-editable-grid-antipattern/#comment-113 Hi Chris, You make a convincing argument and I certainly agree that editable tables are often not appropriate plus require more development effort. However, I would like to suggest that an editable grid has a valuable place in the world - data entry. In the brokerage industry for example, it is necessary to enter huge amounts of information by hand. A shipment for example can have hundreds or thousands of individual products included and each of those products has quantity, weight, volume, classification, goods description, etc. An efficient way of entering this data is do do it in an editable grid. Alternatives such as clicking an add button which brings up an add form, with a save button which then updates the DB and refreshes the grid, would be a lot slower for the user with hundreds or thousands of these product lines to enter on a daily basis. On the front of addressing validation in an editable grid, I would suggest that rather than working almost directly with a database table, you introduce an intermediate domain object layer which contains validation (perhaps of declarative style) which is run before data is saved to the database. James Hi Chris,

You make a convincing argument and I certainly agree that editable tables are often not appropriate plus require more development effort.

However, I would like to suggest that an editable grid has a valuable place in the world - data entry. In the brokerage industry for example, it is necessary to enter huge amounts of information by hand. A shipment for example can have hundreds or thousands of individual products included and each of those products has quantity, weight, volume, classification, goods description, etc. An efficient way of entering this data is do do it in an editable grid. Alternatives such as clicking an add button which brings up an add form, with a save button which then updates the DB and refreshes the grid, would be a lot slower for the user with hundreds or thousands of these product lines to enter on a daily basis. On the front of addressing validation in an editable grid, I would suggest that rather than working almost directly with a database table, you introduce an intermediate domain object layer which contains validation (perhaps of declarative style) which is run before data is saved to the database.

James

]]>