Hmm, seems like I scared everyone away? No replies on this one? Crickets? I am still worried about this...
This is my fault. As I got sick and have fallen behind.
The short answer is the PMC waited to discuss this at the PMC meeting. The PMC meeting was dominated by the discussion of this see the minutes: Project Management Committee (PMC) meeting notes - #11 by yanokwa
I focused on getting the Mission and Values out as fast as possible and then started working on the official response that has to get all PMC member approval. We were planning on doing this in a two week time frame as shown in the minutes, but I pushed it to 4 weeks because of my limitations.
We are examining different points of view and the PMC plans to have an official response, but the official response takes longer to come to consensus and since I am driving it, it's my fault on the delay. I was working on fine tuning the official response this morning.
My focus while being sick was to get the mission and values going as that proceeds any naming discussions.
UPDATE: Because of the differing opinions on the PMC we were unable to reach consensus on a joint response statement in a timely fashion. Therefore, each member of the PMC @Richard_Anderson, @W_Brunette, @Carl_Hartung, and @yanokwa will be posting an individual response in the next couple of days.
@tomsmyth, thanks so much for the work you are doing on the @TAB and the offer to help.
I've spent a couple of sleepless nights thinking about the messages in this topic and the ongoing transition. None of this is easy, but I figured I'd put my perspective on where we are and propose a way forward.
And to be clear, while I'm a member of the PMC, I'm not speaking for the PMC in this message. I'm speaking for myself.
Quick PMC background
From my perspective, the PMC has been wrestling with two problems: the website and naming.
The current website is out of date and it's confusing to users. The site text is not relevant, most of the binaries are old, certain resources now live elsewhere (e.g., forum, docs).
The ODK 2.0 name is confusing. "ODK" means a certain thing to people and "ODK 2" is not a perfect map to that thing. Further, the "2" implies a replacement which it isn't.
The PMC, after some difficult conversations, decided to start with the website with the hope that it would get to completeness quickly. It's been slow going because community processes take time. And it's also been slow because PMC members all have a lot of other work to do.
The feedback on this topic so far suggests that the PMC should put the website on hold and complete a renaming. This is a tough pill to swallow given how hard it's been to get to this stage.
My current thinking
I strongly believe that open source works best when the people who are doing the day to day work make the decisions. And so when long-time and active contributors who've not complained for years speak up, I think the community should bias to their suggestions.
I was hopeful that the website could help the community find positioning that would work well for ODK and ODK 2, but it has not turned out that way. I've come to realize that the push back from long-time contributors is a manifestation of a deeper problem that should have been taken on earlier.
And it's a problem that I can relate to. At the convening, I committed to help ODK transition out of the UW and I committed to focusing on the most widely deployed tools. We've made great progress but it's been extremely difficult to provide support, write docs, recruit contributors, build marketing, raise funding to push ODK forward when there is a seemingly competing set of tools operating under the same brand. I've spent a lot of time talking to other open source maintainers and working on the positioning over the last year. It's not solvable with disclaimer text. It is confusing and that burden falls on the folks doing the day to day work.
From what I can see, sharing a brand isn't great for ODK 2 either because of their relative size. Everything from pinned repos to forum posts to grant proposals from ODK tends to drown out ODK 2. It's hard to grow a community and ecosystem that way. And as the lines between ODK and ODK 2 continue to blur, it will get harder to live under the same brand. Case in point, ODK has started work on improving repeat groups and longitudinal data collection and ODK 2 has started reducing the technical skills required to deploy it. Building a consensus around positioning is not going to get easier.
I understand that the PMC laid out a tentative transition plan, but I think plans should be readjusted as the situation changes.
A way forward?
I think a website launch or a minor renaming does not help ODK or ODK 2 maximize their potential and while I understand that the following might be a painful thing to propose, I believe it might be a good way forward. I propose the following:
ODK 2 spins out as a standalone sister project with shared goals and ideals.
ODK 2's active contributors and most trusted users come up with a name without the ODK prefix and positioning that they agree on.
The various TSC's launch websites that work well for their respective tools with prominent links to each other as sister projects. The various pieces of infrastructure that ODK 2 needs are migrated.
The current PMC disolves and assigns project leadership to the various TSCs and calls the transition finished.
I think the beginning of this idea was proposed at one of the breakout sessions at the convening and I'm raising it again because I think it's a reasonable solution to a complicated problem.
I understand that there was some opposition at the convening, but I strongly believe that before we take an option off the table, contributors who have had a long history with the project and have actively contributed during the transition should have a chance to weigh in publicly.
I also like this idea because it addresses the remaining tasks that the PMC proposed to take on. The reality is that we are eight months beyond the initial transition end date and I argue these remaining tasks are no longer necessary because the transition is functionally finished.
And by functionally, I mean members of the community who are interested in the various tools have already stepped up to take ownership and those folks should be given the control needed to move things forward.
To support this proposal, I will commit my time to helping make sure ODK and ODK 2 has whatever resources it needs to ensure a smooth spin out.
I think it is better to have an improved website that removes out of date information and increases people’s understanding of the ODK project. We have been through multiple iterations of different websites using different platforms, designs, etc. I am worried we have ended up in a situation where “Perfect is the enemy of good” as there is always a list of content improvements that can be made to the website. We have been working on a revised website for over a year and the need for a restructured website has only gotten worse since I first raised this in my transition announcement in August 2016.
The confusion between ODK 1 and ODK 2 is greatly exaggerated. The differences between the two can be described easily on project pages. Both of these systems are targeting moderately sophisticated users – and the bottom line is that people should understand the software that they install and deploy. As ODK 1 and ODK 2 are deployed across multiple devices there needs to a system manager who knows what is going on. Also, this is not a scenario where there is an “automatic upgrade” across versions of the system.
Historically the Open Data Kit project has had multiple efforts to explore different aspects of mobile data collection and management. This goal of creating multiple tools to handle various needs is a very important component of the work to understand how mobile devices could help people around the world. One of the early projects – ODK 1 – has turned out to be very successful – and no one denies its impact. Gaetano and the many members of the ODK team made the decision to develop a new project – named ODK 2 (perhaps an unfortunate name) – to explore a different set of scenarios from ODK 1. ODK 2 has reached the point where it is fulfilling its mission to cover a different set of use cases, and I am impressed with the technical work of the team to get it to this stage.
I completely disagree with the suggestion of jettisoning ODK 2 from the Open Data Kit brand – this completely undercuts Gaetano’s legacy and the efforts of the many people who made ODK what it is over the years. My goal is simply to see the project transition outside of the University of Washington as an Open Source, Global Good Software project. We have been planning the transition of ODK for what seems to me like a very long time – and there have been a number of frustrating (and to me unnecessary delays). We have an overall structure with the PMC managing the cross cutting issues – and separate technical groups managing the individual projects. I would like to see the website finally released as I believe renaming is a separate issue.
an improved website that removes out of date information and increases people’s understanding of the ODK project
A better website is sorely needed. I don't think anyone would disagree with that.
The confusion between ODK 1 and ODK 2 is greatly exaggerated.
I respectfully disagree but I also acknowledge that changing to a new brand would not be trivial and could add more confusion in the short term at least.
this completely undercuts Gaetano’s legacy and the efforts of the many people who made ODK what it is over the years
This is a fair point. Removing the ODK brand from ODK 2 seems like it's a non-starter for some and I don't think it's necessary.
If the site is to go live with the ODK1/2 thing, I think there are some things that could be done with the information architecture and styling to minimize the confusion and give more prominence to the brands of the actual tools than to the suite brands. I'd be happy to consult a bit on this if desired. It would be nice to at least do that before flipping the switch.
Here's my general take on things as a PMC member giving my own opinion.
I think that...
- The website desperately needs an update.
- ODK 2 needs to be renamed.
- Renaming ODK 2 is a red herring as it relates to updating the website.
The name "ODK 2" has been around since 2013, and while it has caused some confusion, I've not seen evidence that it has hindered the growth of the community in any way over the last 5 years, so I wonder if we'd still be having this conversation if the project had been originally named "ODK Sync" or anything other than "2". The ODK 2 tools were designed as a set of complimentary tools to ODK 1 tools to handle a set of problems organizations brought up that ODK 1 didn't address. To that end, those tools were designed to be part of the ODK ecosystem, and removing them from the project seems like an unnecessary step towards updating the website. If people have reasons they don't want those tools to use the ODK name in general then that's a separate issue that should be taken up with the community as a whole, but that seems like an entirely different, larger discussion and still unrelated to moving the website from one platform to another.
Given that the goal of the new website is to revise out-of-date information and provide a way for community members to update and navigate the information more easily. It seems the quicker the new website is brought to light, the more accessible the project will be in general. Releasing a new website is not going to create more/new confusion because of ODK 2. It's an opportunity to present the information people currently find confusing in a more coherent manner, which I think will be even more necessary when ODK 2 is actually renamed to avoid new confusion in half the community in the form of: "What happened to ODK 2?".
Perhaps said another way... if you were to switch code repositories from SVN to Git, would it be necessary to fix all the bugs in your code first? No, because you'd never move repos, and that's what it feels like we're trying to do with this website transition.
I totally agree about confusion between the two ODK 1 and ODK 2. Many of the collaborators I'm working with ask me how it is possible to upgrade to ODK2, and it's not easy to understand for some that the two are not related.
Having them on the same website on my personal opinion will increase the confusion and I agree that having two separated site or different names would help. Some more evident distinction.
And I also agree that same name and website and forum only drowns out ODK 2.
A better website is for sure needed, at the moment is as well a cause of confusion, but I think that the naming issue should be first sorted out.
Honestly, I also tried to support for some new projects ODK 2, but people get confused. The brand ODK in it sort of bias the people. I think that , if the name ODK should be there for both at least a logo, a color...some restyling is needed for ODK2.
@tomsmyth thank you for volunteering to help participate in community facilitation. We are always looking for ideas and skills from community members and value your feedback. The PMC has spent the majority of our last two meetings discussing your post and looking for ways to decrease confusion.
The reality of the transition is that there are strong opinions about priorities and sequencing, and it's not always easy to find consensus. While a year has passed since the transition process started, the PMC has continued to focus on the list of tasks that were formed by the concerns raised in the ODK transition management surveys, at the ODK convening, and on the ODK forum. My primary concern as a PMC member is that keeping the status quo has caused confusion; therefore, we need to make progress on project improvements in a timely fashion to reduce the confusion. The question that I have been pondering is whether the naming or the deprecated/out-of-date website is causing the most confusion and which to address first. It should be noted that the name ODK 2 already exists on public ODK websites so in my opinion it’s a pre-existing concern not something new. I appreciate suggestions on how to balance multiple concerns about confusion and make expeditious progress on the transition.
My impression is that the PMC and community in general recognize that the current ODK 1 and 2 naming scheme could cause confusion so most people are supportive of a change. The ODK 2 developers suggested that the rename process occur sooner rather than later and asked that the rename be completed last fall (2017). However, there was concern from the ODK 1 side that the renaming discussion would distract from the TSC 1 formation discussion. The PMC considered the formation of TSC 1 a high priority so decisions could be made about ODK 1’s roadmap. Therefore, the PMC decided to delay renaming to not interfere with TSC 1 formation and to keep the momentum around TSC 1.
From our discussions the PMC is worried about 1) the ODK 1 and 2 naming schemes and 2) the confusion that has occurred with having inconsistent information distributed over multiple public websites. Currently there is not a clear updated authoritative website to represent ODK that has all of the current information on it. The documentation is inconsistent and out-of-date across the multiple public websites as well as the binaries. Overall, this is causing confusion among users since there is not a single landing page to get correct information and correct binary versions. An example of user confusion is demonstrated via the emails we receive at email@example.com such as: “I have been trying to use the ODK platform, however I am not able to open the build, collect and aggregate link, the page opens with This page has been deprecated.” Additionally, 13+ users/organizations that may be new community members have had their requests to participate in the “help-for-hire” page denied because we have been in a perpetual website transition (transitioning to using the “marketplace” in the forum instead).
Historically the ODK project has valued the creation of multiple tools to explore different aspects of mobile data collection and management. This has led to the creation of different tools some with overlapping capabilities (e.g., Build and XLSForm, Aggregate/Briefcase/Google Sheets). As stated by others, ODK 2 was designed as a set of complementary tools to ODK 1 to handle a set of problems organizations brought up that ODK 1 did not address. Gaetano (ODK visionary and founder) thought it would be cool to follow Python naming convention (has the 2.x series and 3.x naming) and have a 1.x series and a 2.x series of tools. As stated above, I think most people agree we should change the naming scheme as it is a source of confusion, this seems sensible to me. However, removing a set of complementary tools from the ODK community seems to be overkill (almost tangential) and does not reflect the ODK value “flexibility to select suitable components from an ecosystem of complementary tools.” There has always been confusion around the ODK ecosystem when it came to KoboToolbox, CommCare, SurveyCTO, Enketo, GeoODK, etc. The ODK project has historically valued multiple tools solving different use cases, thereby, helping us “serve users with a wide range of technical ability and organizations with a wide range of technical capacity”.
One approach the PMC has been discussing to speed up the rename is to separate the naming discussion into two stages. The first stage would be an expeditious rename of ODK 2 (based on suggested names from the community) with a second stage focused on broader community discussions regarding naming and branding. However, there was also the concern that most transition tasks have taken longer than originally planned (the website is now almost a year behind schedule, TSC 1 took longer, etc.) and we would just continue to kick the can down the road. We have multiple public websites that do not contain the most up-to-date software or documentation. We also need to rename ODK 2. My fear is that we continue to find reasons to delay the launch of the website as it was originally going to come out around the ODK convening, then September 2017, then the end of 2017, then by end of Jan 2018, etc. Waiting until the ODK project is in a perfect state I do not think is helpful because the multiple out-of-date websites continue to generate confusion. I would also like to see ODK 2 renamed as soon as possible so we can move on to other ODK project improvements.
I do echo and support @martinjr sentiment on having a representation of the other players in the ODK ecosystem.
On the issue of ODK 2 renaming, I think it is something that has to happen soon. ODK 2 as a name does allude to an upgrade which is pretty confusing to most users that ask about this. I do think ODK 2 should still be part of ODK, but a name change will make all the difference. Let's get the website out as we look into how to move forward with the naming.
Thank you very much for bringing this important discussion to a public forum. I really think open source works best when most conversations are had in public with opportunities for different opinions to be shared with civility and respect.
I strongly agree with this point. The challenge has been that four things have been conflated:
- a website platform change to make contributions easier
- a website theme update to modernize the look
- a website content update to make the latest binaries available, remove stale content, etc
- a website update to project positioning
It's the last one that has made the website transition so difficult. To follow @Carl_Hartung's analogy, it's like adding underspecified features to a git repository as part of migrating from svn.
My preference has been and continues to be to take on a subset or all of the first three and only addressed the 4th after challenging issues of branding and positioning have been addressed. That is, transferring and updating content from the existing website largely as-is, making it current and then considering what questions need to be answered to evolve that content.
I agree with many of the posters here that the ODK 2 rename is a separate discussion that shouldn't block the website launch. As @Carl_Hartung said:
Rather, I think the current state of the old website is causing more problems than a continuation of the status quo with the naming. It is so out of date and full of deprecations that I don't even use it as a resource anymore; I just link the forum and the docs. Yet it is still the first thing anyone sees when they search Open Data Kit. The new website, even an imperfect one, would be a valuable resource to everyone in the full ODK community. And, as a living document, it can be updated when the rename occurs.
I think updating the website is the most pressing issue currently on the PMC's plate. I think a rename is important but it should not block the website. In my opinion, these tasks could be completed in parallel. In the meantime, we currently have this explanation:
The ODK community currently produces two suites of tools, currently known as “ODK 1” and “ODK 2”. Some users will look at the version number and assume the latest is the greatest, but this is not always the case. For this reason, ODK 2 will be renamed in the near future. The ODK 2 Suite was designed to co-exist with ODK 1 tools, and does not replace any ODK 1 software."
Additionally, every summary of the ODK 2 tools on the website or in the docs includes some version of the phrase: a new set of tools that will co-exist with the ODK 1 tool suite. I think these work fine until the rename is complete.
Finally, I think ODK 2 should remain part of the ODK community. My view mirrors some other posters that Open Data Kit is larger than any of its tools. There was a great discussion of the mission and values recently on this thread: Feedback on Draft of ODK's Mission and Values - #28 by LN. I think the collection of all the current ODK tools fulfills these values better than either suite on their own. Particularly
We serve users with a wide range of technical ability and organizations with a wide range of technical capacity
We believe users should have ownership of their data and the flexibility to select suitable components from an ecosystem of complementary tools
are more fully realized by supporting both tool suites. My experience with the typical ODK 2 user is that they are very familiar with ODK 1, and often continue to use it, but have certain scenarios where the tools are not meeting their needs. For users with a stronger technical ability or organizations with a higher technical capacity, the ODK 2 tools provide an option built under Gaetano's vision and that meets those same core values.
I agree with other posters that ODK 2 should not compete with ODK 1. It is my opinion that it currently does not, and with more appropriate naming under the same ODK umbrella this would be clearer. However, I think separating the two projects would do the opposite: it would cause more confusion and imply that they do compete and are separate choices to the same problem.
To wrap up, I think that current state of opendatakit.org is the most confusing and detrimental problem for the PMC to fix, and none of these other issues should distract from it. The website should be updated as soon as possible, and a renaming discussion should happen in parallel.
Happy Monday, everyone. It's been really good to hear the voices of so many people on this latest discussion. (Note to mods: I wonder, maybe we should start a separate topic/title so more people recognize what's happening inside?) "Product strategy" discussions are one of the most important types of things the ODK community can work on, even if they don't happen often. I want to specifically thank the PMC for being so open to hearing feedback from everyone -- and I hope more people add their voices to be heard.
I wanted to add my personal thoughts on this topic. TLDR: Don't wait to launch the web site. Now, of course, it's more complicated than that, but one of the great temptations in any open source software project is to get caught up in what engineers often call "analysis paralysis". It's an easy thing to do, and I'm worry that is what's starting to happen here.
A few points in more detail:
The ODK 1/2 naming issue described here is an important one. In fact, I was in a discussion just last week where there was some confusion about the two products and their visions, and how they relate to each other. IMHO, the community shouldn't keep "kicking the can down the road" any longer -- working through this issue will be good for both sets of products and the overall ODK & ICT4D/aid communities in whole. It could be a really great opportunity for the new-ish ODK 1 TSC to work alongside the ODK 2 team (new TSC?) to really clarify their collective visions.
That said, the website need not be a blocker for this. In fact, it can help make that communication effort easier! As many people have seen over the past few weeks, it's pretty easy to make edits with the new platform, and as whatever group of people work through the communication strategy for the ODK 1/2 questions, the website can fairly easily be updated to communicate this. The newly-staged site may not yet be "perfect" in all of our minds, but as many of us know, there's a real temptation in
"letting perfect be the enemy of good". As Steve Jobs once said before the launch of the original Macintosh computer, "real artists ship". The web site can be rapidly fixed/improved, so why wouldn't the community ship it ASAP? That's the open source way.
Doing "both at the same time" helps the community grow. More volunteers can get more involved either doing web site improvement activities (have you seen the list of issues on GitHub?) or helping with the product vision work and communication work related to the two mainline products. These may be different people (although there will also undoubtedly be overlapping volunteers!) but it also opens a different door that may attract more minds to becoming more active in the community.
I'd love to hear what others think about these thoughts, and to others who have been posting over the past several days. And most importantly, if you have ideas on these topics, or especially if you feel strongly, please take this chance to add your voice to the conversation -- the ability to do so is one of the great things about this community!
You know with all this talk over whether or when or not to rename, we probably could have chosen a new name in the same amount of time. Why does it have to be so hard? How about a one week nomination period and then a one week round of online instant runoff voting?
Anyway I am persuaded that this need not block the website launch. My offer to help with information architecture still stands if there is interest.
I'm going to putting on my PMC chair hat to remind about process and try to get to something actionable. Our governance requires lazy consensus to make decisions. If a proposal doesn't have strong opposition, that proposal moves forward.
So with that in mind, here's what I'm seeing across the discussion:
- On splitting, there is strong opposition and weak at best support.
- On website launching, there is consensus on the website launching soon.
- On naming confusion, there is consensus that the names are very confusing.
- On sequencing, there is consensus that naming shouldn't block the website.
Here's what I propose:
The folks who are concerned about confusion commit to helping clean up the information architecture and language and positioning to minimize the confusion. @tomsmyth are you up for facilitating this effort? And I'm guessing @martijnr, @aurdipas, @Ukang_a_Dickson, @ln, @adammichaelwood will help.
We lock in a launch date of whatever is on staging. Close of business on June 15th seems as good of a date as any, but I think the folks from the previous point should have a say on if this works for them. I think the more aggressive the timeline, the better.
We start a new public topic to discuss how we go about naming, what process we use to decide on a final name, and a timeline. I think this should start now and that the PMC be public with the deliberations so we aren't forced into a last minute discussion like this one.
How does this proposal sound?
The PMC met an hour ago (see Project Management Committee (PMC) meeting notes - #13 by yanokwa) and has decided to adopt the most recent proposal, but moving the launch date up to next Tuesday, May 29th.
The process for making changes to the website will be through our normal file issue, send in a PR, merge by a committer process at https://github.com/opendatakit/website. The caveat is that @yanokwa and @W_Brunette must both approve review and approve changes before they can go live. Changes that are contentious will require a tie-breaking vote from another member of the PMC.
If you'd like to help with this effort, a number of us will be meeting in #website on the Slack starting tomorrow at 2 pm EDT to help clear up any confusing parts of the site. We'll work asynchronously in that channel until the site is live.
So to summarize, http://staging.opendatakit.org will launch as the ODK website on Tuesday, May 29th at 00:00 UTC (8 pm EDT). If you'd like to help, go to http://slack.opendatakit.org and join the #website channel.
I'm late to post but have followed the conversation. Thanks to everyone for their thoughtful posts and discussion here in the open; it's been an interesting read. I'm glad the PMC decided to adopt the most recent proposal.
- On splitting, I would add my voice to the opposition. A new brand would probably result in some short term confusion and have some long term benefits; but I agree with the sentiments around ODK project valuing multiple tools to solve different use cases. I would like to see both remain under the ODK brand. We (Red Cross) plan on using both in the future depending on the situation and needs. Keeping the two linked by the ODK brand is helpful for us.
- On website launching (and sequencing), I think updating the website can come before figuring out the branding/naming. I think it would be better to launch it soon with some information missing rather than wait and continue to have a mix of outdated and deprecated information on the old website. In my opinion, would be great just making it easier for people to end up looking at the new documentation and end up on the forums.
Will try and comment sooner and contribute in the #website Slack channel and other places where this conversation continues!
Lots of discussion on the renaming. Want to add a response to the below (which others have raised in some version as well).
In the Red Cross Movement, technical people responsible for implementing (install and deploy) are going to be moderately sophisticated and able to understand the software. But these people have to explain software recommendations to sometimes very non-technical audiences (management, senior leadership, media, etc.). And in such cases, the names do create a major hurdle that would hopefully be lower if not for the implied sequencing etc by "1" and "2".
This is such an important point.
Just an example of confused users: Difference between Suitcase and Briefcase - #2 by abdoul