I like Mingle, but it's not perfect. I've posted a few times about the benefits of using Mingle, so now I'll mention the problems.
We had read of performance issues, but I experienced where the cliff is. We have Mingle running on a virtual server with 4 GB of memory. We created a custom project using one tree and approximately 15 fields in the card. As we approached 1200-1500 cards in the project, performance dropped significantly. Users could count to 10 - 30 before items loaded, especially in grid view. As we neared 2000, grid view failed completely. Users new to Mingle started asking for another tool. Stakeholders in planning meetings would lose their patience waiting on it. My advice is find a way to keep bugs and features separate. Delete old, fixed bugs, and be disciplined to keep the features list (product backlog) short. Why keep low priority items when there are months of highs in the list?
Exporting data is a pain. Cut and paste text into Excel gets you by, but you can't use any filters (WHERE clause) nor choose what fields to include or exclude. You get it all. Massaging it into something useful for management can become hours per week if you have large backlogs.
As far as I can tell, you can't enforce data integrity. I tried without success. And garbage in\garbage out feeds the two items above that can become a monster.
The project templates look wonderful (XP, Scrum, Agile Hybrid), but there's only one thin page of documentation on their site for each one on what comprises each template. Trying to figure out how to use them (what the workflow is, the objects, details) is essentially up to you. Don't get me wrong - I really appreciate these views into how Thoughtworks does agile project management, but as I kept poking around, feeling like I was trying to put together the lifestyle of some ancient race based on artifacts, I kept thinking, "The creator of this template could have taken one day to document it and the whole world would be ready to go." As is, the whole world will play Sherlock Holmes and my guess is many will give up and go to what they know and lose out on a great agile mind-share opportunity.
We had read of performance issues, but I experienced where the cliff is. We have Mingle running on a virtual server with 4 GB of memory. We created a custom project using one tree and approximately 15 fields in the card. As we approached 1200-1500 cards in the project, performance dropped significantly. Users could count to 10 - 30 before items loaded, especially in grid view. As we neared 2000, grid view failed completely. Users new to Mingle started asking for another tool. Stakeholders in planning meetings would lose their patience waiting on it. My advice is find a way to keep bugs and features separate. Delete old, fixed bugs, and be disciplined to keep the features list (product backlog) short. Why keep low priority items when there are months of highs in the list?
Exporting data is a pain. Cut and paste text into Excel gets you by, but you can't use any filters (WHERE clause) nor choose what fields to include or exclude. You get it all. Massaging it into something useful for management can become hours per week if you have large backlogs.
As far as I can tell, you can't enforce data integrity. I tried without success. And garbage in\garbage out feeds the two items above that can become a monster.
The project templates look wonderful (XP, Scrum, Agile Hybrid), but there's only one thin page of documentation on their site for each one on what comprises each template. Trying to figure out how to use them (what the workflow is, the objects, details) is essentially up to you. Don't get me wrong - I really appreciate these views into how Thoughtworks does agile project management, but as I kept poking around, feeling like I was trying to put together the lifestyle of some ancient race based on artifacts, I kept thinking, "The creator of this template could have taken one day to document it and the whole world would be ready to go." As is, the whole world will play Sherlock Holmes and my guess is many will give up and go to what they know and lose out on a great agile mind-share opportunity.
No comments:
Post a Comment