On Sat, 19 May 2012 12:02:03 +0200, Pascal Bleser wrote:
As an intermediate solution, would you be able to use a management tool of your choice to maintain your own overview, importing/exporting information/lists/calendars to Google Docs?
I believe that there would be real added value in using a "project management tool", especially to - track things that need to be done - dependencies between tasks - whom is taking care of what
Yes. The more I look at it, the more I think maybe the MediaWiki "Todo Tasks" plugin provides sufficient functionality to do this, with the added functionality of e-mailing the task owner. It doesn't track status, but it tracks what needs to be done, and in that regard it's a step in the right direction.
We definitely don't need gantt charts and probably not even estimations of effort (although it might come handy for a few things if it is used by everyone on openSUSE, e.g. thinking of OBS development).
I think we need to track dependencies, though (something a gantt chart would do) - and time estimates are useful to ensure we move forward on stuff. If we have to adjust the time estimates, that's fine - project management isn't about the PM plan being a "roadmap" but more of a "directional guidance" - and if we see something bogging down, we can solicit additional help on a task that's holding things up. But that's more effective if you have an estimate as to the effort to complete a task to begin with - otherwise you don't know (falling back into PMP- speak) what the "earned value" is on the task and how it's tracking to the "planned value" - which is how you know if you're ahead of or behind schedule.
- topics/categories/groups/projects
- status (todo, in progress, done, frozen, canceled)
- dependencies/links to other tasks
- description
- comments
- integration with opensuse.org accounts and SSO
- obviously web based
- maybe subtasks
- maybe effort estimation and time tracking
These are the reasons why I initially was thinking Vibe would be a good choice - its task management module has nearly all of these features. The integration with opensuse.org accounts/SSO does mean some level of integration with Access Manager, which will slow things down (in terms of implementation), though OpenID might be an option (which I think means we can just implement and point to the OpenID provider) - I can look and see if that's something that can be added and what would be involved. But that is a bit of "cart before the horse", too, if we organize a project to select a tool.
If we go down that route, and I believe we should, let's make it properly for once and take our time to pick the right tool:
- let's gather requirements * let's talk to a few people and teams to see whether they would use it (if there is very low adoption and we just end up with 5 people using it, it's just a burden)
- let's make a list of candidates (tools) and see whether people have experience with them and feedback to give (ChiliProject, RedMine, Trac, Trello, Retro, ...)
Essentially, we need a project to decide on a project management tool. :) So the first step would be to identify who the stakeholders are so we're sure we're assessing the needs from the proper audience.
Anyone interested in driving this? I'm definitely interested in taking part as I have some experience with such tools but I'm too swamped to drive it.
I'm interested, but at the same time, while my time is presently somewhat free, that's subject to change depending on availability of contract work or taking on a full-time position. Perhaps you and I could work together (would you have time if we split the load?) to drive this.
I must say that personally, I'm a big fan of the Features/"RFC" kind of pages the Fedora folks are doing in their wiki for specific tasks and topics (e.g. introducing systemd, moving everything to /usr, etc...), e.g. https://fedoraproject.org/wiki/Features/systemd https://fedoraproject.org/wiki/Features/UsrMove
IMHO we should set it up like that and make an RFC page on our wiki to collect all findings, give context, list stakeholders, and ultimately the decision. Document for posterity in a more visible and clear manner, if you will, rather than having it buried in lengthy email threads.
Agreed. E-mail is good for bashing ideas out, but it gets difficult to follow after a while (even following via NNTP on gmane.org, which is what I'm doing) and we need to make sure we don't start repeating ourselves. :)
No need to slavishly follow the structure of Fedora's pages, they have a slightly different purpose, but most of the document structure could be mimicked (Summary, Owner, Current Status, Detailed Description, Benefits, Roadmap, add "Potential Issues" etc...).
Agreed, but a consistent structure is something I'd highly encourage, as having a repeatable process means that we can more accurately predict the outcomes - and predictability is something that I think is highly desirable. Jim -- Jim Henderson Please keep on-topic replies on the list so everyone benefits -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org To contact the owner, email: opensuse-project+owner@opensuse.org