Atlassian JIRA (project management suite)

I installed and configured Atlassian JIRA on our OSEG server and acquired an evaluation license. If it suits us, we may apply for a free Open Source license after the trial time.

Everybody can sign up and login here:
http://opensourceecology.de:8080

Feel free to add projects, bugs and so on.

Shure

Great! Thanks!

Till when is the evaluation license valid?

30 days from yesterday on.

Hi, how did it turn out? Or shall we go with a open source development? Not
https://opendesignengine.net/projects/far-horizons-project

I don’t know if I really need a project management tool. Who shall put in all those infos? It’s a lot of work … time that we could invest into development. :mrgreen:

Otherwise, this is what I found:
This one from this list (see *) looks like a (useful? but probably overfeatured?) tool:
http://www.libreplan.com/features/
Demo: http://demo.libreplan.org/libreplan/planner/index.zul (extremely lagging, or only for me?)



Blog - Overviews

There’s no active collaborative development without the ongoing documentation and collaboration in a project management system :neutral_face:
For now there’s github, mediawiki, etherpad, trello, this forum, mailing lists and so on. there’s still no easy go-to solution for collaborative development at OSEG. I’d prefer a ticket system for any support and e-mail stuff, plus a project management with versioning. for now, the latter is github.

Jira was not used much and is no longer active on the server. At least for now I would even prefer to use git with issue tracking rather than a grown up project management suite.

Thanks for the input on alternatives!

You are right, I also prefer a ticket system like trac. What impressed me most is the really nice integration of their SVN and Git repository in 0 A.D. Oh wait, I don’t want to loose our developers to 0 A.D., please don’t leave us, … I just want to use their trac system as an example.


I like their trac tool setup (we have also used it at university, but it is does not look that convincing there). Its forum integration is still not perfect, but it’s at least a bit integrated into the forum (e.g. ticket status is shown in the forum too, if a ticket is cited there, all automatically without user input). The loss of information in the masses of data in the forum is still a problem though.

Generally, in a forum I would like to have a summary page to every thread, but using the first thread for this reason - though possible - is no good idea, as the first post should be permitted to allow HTML and Wiki-tags and I don’t know if that’s possible (tbd) … If it’s not then either only administrators and moderators could create new threads/topics because only they have the rights for HTML and Wikitags (this hierarchy I dislike anyway) or the moderators have to activate html and Wikicodes for each new topic - a quite heavy workload. :confused:

I know there is a possibility to put a small textarea that is editable for all users that allows Wiki-tags … but still I did not get around getting it done (via php templates). … yes … oh well, why? I don’t really know, perhaps I just am not good enough. :wink:

If this works, this would be the solution for the forum mess that is seen around the web everywhere (even at microcontroller.net).

The trac/ticket system, which itself integrates versioning of the dev-files with easy option to display all changes of each and everything by clicking on this icon

, when you reach this example changes overview page Note that of course this is not a question of Git vs SVN vs Mercurial vs Bazar. It’s about the ticket integration. For example the tickets that are solved are automatically scored through [durchgestrichen]. Every member can freely change the ticket state and post answers. I think it’s quite cool.)

I also have written a monolog about wiki-development-site-Integration and hope it will turn into a heated discussion: ( :
https://discourse.test.opensourceecology.de/t/sollte-das-wiki-entwicklerseiten-spiegeln/504/1

Another benefit of trac vs. bazaar/launchpad is that the search function is more obvious. In Bazaar/launchpad, my impression is, that everything looks hidden at first (see for example the missing or hard to find search function) and they often use a vertical approach instead of an horizontal one. This of course might well be a benefit for Android devices et al. but I like the compact design of trac in file browse view. Bazaar of course is online … whereas Trac has to be installed on a server or webspace (if I’m getting it right) and I’m not sure that we want to do that. I also have no idea about the functionality of the in trac integrated wiki. To move our Wiki style engine to the trac is one important feature we need as Tony and colleagues already had a lot of work with our websites.

Also I don’t know currently how to combine Forum and Trac system and how to make it possible to use one Account for all our sites. Currently we have Wiki, Forum, blog accounts separated …

One thing I currently question myself is how we could and should ideally deal with Excel/Spreadsheets related with Single-Source-principle. In buildtheenterprise development we use Google hosted spreadsheets but I don’t think it’s perfect because if we want to build a model out of the parameters in our spreadsheets then we have to parse the google sites. And that is no good idea. Especially it’s very maintenance and traffic heavy as things break easily. If anyone knows a good solution to that please tell us.

We could of course try to manipulate ODF (Open Document Format, read: OpenLibreOffice) spreadsheets directly. This is our best option currently and I could do it using JAVA Server pages without a dependency of running Open/LibreOffice or having it installed. For automatic conversion to PDF of course it would be good to have LibreOffice installed. I only know of PDF creation of DOCX/XLS files that does not use any program instance. For ODF/ODS I have no alternative idea currently but is it a problem to have LibreOffice installed on the server - I think it’s not.

Once we have a PDF we could even extract plots from the spreadsheet and directly mirror them onto the Wiki page to show the results of the new parameters in almost realtime. Alternatively we could just use HTML5 /javascript d3.js directly for creating the preview but there would be a lot of code adaption required once for each plot.


You see, I’m a bit floating in what approach to take. Also the overall strategy I prefer currently is:

  1. Derive basic requirements/max-min values from Physical considerations (only those r e a l l y needed).
  2. Generate 3D model from these requirements. Best by using a generator addon in blender.
  3. Design details in the single-source 3D model to spec and in real size.
  4. Generate documentation, BOM and and physical sizes and constraints from this model. Also automatically calculate final mass and inertia.
  5. Render FEM analysis result.
  6. Generate manufacturing files.

Is the Atlassian JIRA open source? Is it a good choice? I share your opiion that we should bundle our efforts.