Hi! Is there a way to get a query string that can tally up contribution data across all Wikimedia deployed extensions, or a larger subset of them? I can imagine that if we write them all out individually, that would probably hit up against some query string character limit. Part of that query string, would be to set a date range, and that would further add to the overall number of characters in the string. Thanks!
Talk:Community metrics/Archive 2
If it helps, the canonical listing of Wikimedia-deployed extensions is make-release/settings.yaml.
Thanks, James! It does confirm for me that there are *a lot* of extensions! Is there a known query string character limit for Biterg? So, that if I add in all 200+ extensions and a date range, what is the character count I need to stay under?
For now I'm not aware of a better way than creating a huge manual query for those 190 extensions. Potential future config ideas to avoid could be phab:T146135 and phab:T187661. Not sure if there is a character limit, I'm afraid it's "try and see". I'm sorry that I don't have a better idea for now.
Passing 198 repos (exts and skins) as the query seems to work for me: https://wikimedia.biterg.io/goto/0267e89edb357fda6d3c8f8b74eda062 (last 90d)
Hi, I read this page along with https://chaoss.github.io/grimoirelab-sigils/panels/community-structure-by-organization/ and https://gitlab.com/Bitergia/c/FAQ/-/blob/master/how-to-identities/README.md#how-to-add-a-new-organization but I still don't understand where the "organization" data in our Developer Advocacy bitergia dashboard comes from. For example, the Community > Organizational Diversity dashboard lists organizations like WikiWorks, WMDE, Wikimedia, Independent, Google, Wikia, Inc...where is this data coming from?
@TBurmeister (WMF) The database contains an "organizations" table (id, name) which includes the orgs and which can be expanded. There's also an "enrollments" table (id, start, end, uuid, organization_id) that stores which user profiles (uuid) were affiliated (organization_id) at which timeframe (start, end). It's manual work to keep this updated, mostly based on SUL username prefixes, email address host names, and other creative non-perfect magic.
I am looking at your ),sort:(columnIndex:!n,direction:!n)))),gridData:(h:24,i:'12',w:24,x:0,y:8),id:all_onion_developers_active_quarters_by_project,panelIndex:'12',title:'Developers%20by%20Project',type:visualization,version:'6.8.6'),(embeddableConfig:(title:Projects,vis:(params:(config:(searchKeyword:),sort:(columnIndex:!n,direction:!n)))),gridData:(h:20,i:'13',w:24,x:24,y:24),id:all_onion_projects,panelIndex:'13',title:Projects,type:visualization,version:'6.8.6'),(embeddableConfig:(title:'Community%20Structure'),gridData:(h:12,i:'15',w:24,x:24,y:0),id:all_onion_projects_community_bar_chart,panelIndex:'15',title:'Community%20Structure',type:visualization,version:'6.8.6'),(embeddableConfig:(title:Contributions,vis:(colors:(Contributions:%23447EBC),legendOpen:!t)),gridData:(h:12,i:'16',w:24,x:24,y:12),id:all_onion_projects_contributions,panelIndex:'16',title:Contributions,type:visualization,version:'6.8.6'),(gridData:(h:8,i:'17',w:12,x:0,y:0),id:'77316e20-c23a-11e8-8aac-ef7fd4d8cbad',panelIndex:'17',title:'Data%20Source',type:visualization,version:'6.8.6'),(gridData:(h:12,i:'18',w:24,x:0,y:32),id:bf1fc450-3393-11e9-b68b-31a1aa44b23a,panelIndex:'18',title:'Community%20Structure%20by%20Project',type:visualization,version:'6.8.6')),query:(language:lucene,query:(query_string:(analyze_wildcard:!t,default_field:'*',query:'*'))),timeRestore:!t,title:'Community%20Structure%20by%20Project',viewMode:view) contributions visualisation and it gets me curious about how to use similar data for my team, am I assuming correctly that this is using the phabricator data? (still struggling a bit to understand and how to analyse which datasets a saved visualisation uses) I would like to look at contributions per project, where I can differentiate projects by looking at phabricator tags as in the case of my team one project is not necessarily one repository. So I would have projects on the x-axis while looking at contributions on the y-axis.
@KSiebert (WMF) Hi, this is not using any Phabricator data (the backend to index Phabricator activity is very rudimentary). See "Data Source" in the upper left corner: by default it is set to "git" and I highly recommend to set "gerrit" instead, because Git also includes imported code repositories which were imported from non-Wikimedia development (e.g. upstream projects that we rely on).
Depending on your use case(s), looking at data in certain code repositories and creating a custom dashboard (if supported) might be the closest.
Sorry, for being quiet here a while, I was experimenting quite some time and realised that I am somehow limited in what I want to do for my team, as we work on features all over different codebases and repositories, so looking at contributions within a specific repository doesn't show me what I would like to see, which is how engaged technical contributors are with our features.
Hmm, any specific criteria in mind that you would like to filter on?
Yes, let me tell you an example, I build a dashboard for my team: https://wikimedia.biterg.io/goto/6b7734f5f3fc2d360c0d8cb94cd2497a
In the top left there's the visualisation "Open Ticket Count for Maintained Project" which compared ticket counts per ticket with certain tags. That's super interesting and I hope it's a meaningful use of the phabricator data as you also mentioned some limitations with that data, please feel free to warn me if any of my visualisations don't make sense from your point of view. But I would also be interested in tracking contributions but filter them by projects tags.
@KSiebert (WMF): Indeed this is super interesting and great. Thanks for having created that and sharing it. (And I admit that I did not know that some of these widgets are possible, from the top of my head, e.g. by using the "Include" parameter for the X-Axis bucket.) Wondering if this dashboard should be documented on Community metrics in a new dedicated section of, maybe, "custom dashboards" or such?
Sure, share it, maybe it's interesting for others that want to make a custom board for their teams as well. @NHillard-WMF made one for this team as well, maybe he wants to share. Talking about the "Include" parameter, I was just thinking today to build a separate dashboard to differentiate between volunteer and employee contributions, to do that I could use the "exclude" parameter and exclude specific users, in this case team mates. I wonder if the author data has a boolean to show that as well.
Yeah! Thanks@KSiebert (WMF) for looping me in here. The dashboard I built is ),sort:(columnIndex:1,direction:desc)))),gridData:(h:15,i:'9',w:22,x:24,y:5),id:maniphest_issues_submitters,panelIndex:'9',title:'Phabricator%20Submitters',type:visualization,version:'6.8.6'),(embeddableConfig:(),gridData:(h:5,i:'10',w:46,x:0,y:0),id:maniphest_main_numbers_timing,panelIndex:'10',title:'Phabricator%20Summary',type:visualization,version:'6.8.6'),(embeddableConfig:(),gridData:(h:5,i:'11',w:46,x:0,y:35),id:gerrit_main_numbers_new_changesets_statistics,panelIndex:'11',type:visualization,version:'6.8.6'),(embeddableConfig:(),gridData:(h:15,i:'12',w:24,x:0,y:50),id:gerrit_authors_changesets,panelIndex:'12',title:'Gerrit%20Authors',type:visualization,version:'6.8.6'),(embeddableConfig:(),gridData:(h:16,i:'13',w:14,x:0,y:80),id:fe1fe260-cae7-11ea-8358-4d35848e335d,panelIndex:'13',type:visualization,version:'6.8.6'),(embeddableConfig:(),gridData:(h:16,i:'14',w:12,x:34,y:80),id:e8c305e0-cae8-11ea-8358-4d35848e335d,panelIndex:'14',type:visualization,version:'6.8.6'),(embeddableConfig:(),gridData:(h:16,i:'15',w:20,x:14,y:80),id:'77a31740-cae5-11ea-8358-4d35848e335d',panelIndex:'15',type:visualization,version:'6.8.6')),query:(language:kuery,query:'repository:%22design%2Fcodex%22%20OR%20tags:%22Design-Systems-Team%20(Design-Systems-Sprint)%22'),timeRestore:!f,title:c_dst_phabricator_plus_gerrit_highlevel,viewMode:view) here for reference, it basically just has basic phab and gerrit stats, composed for now against already-existing visualizations. My team only has a single project, so this is not as relevant to the current discussion, but perhaps some of the constituent visualizations could be repurposed to report on the status of individual projects?
Thanks. I've created a section Community metrics#Custom team dashboards
Hello, my team, Community Tech, is looking into different ways for measuring maintenance burden. Does anyone have ideas on how to measure maintenance burden with wikimedia.biterg.io? Would you track open tasks, patches, etc? Thank you!
@HMonroy (WMF) Hi, how is "maintenance burden" defined? Are there clear criteria shared across engineering teams in Wikimedia, aligned with the work of the Quality and Test Engineering team in this area to come up with metrics? If not I'd recommend to contact them first to avoid duplicated work.
@HMonroy (WMF) Hi, did this happen?
Hey,
so the truth is, maintenance burden is something we made made up and we were looking for ways to measure it, it would be some combination of: time spend on a ticket + ticket count + number of patches created for one project. That was a useful comment to check in on what's already there, thanks a lot @AKlapper (WMF). Will report here what we end up going with.
@AKlapper (WMF) The dashboard mentioned here doesn't seem to include anything about retention rates, is that the right place to link to? Also, is it intentional that the two sections are identical?
@Tgr Uh, thanks, that was indeed the wrong link. Fixed. Identical is intentional, yes. :)
There are no older topics