Project status before tonight's Code for Seattle Civic Hack Night [was: SEANetMap project status and documentation]
6 views
Skip to first unread message
John Tigue
May 7, 2015, 7:06:44 PM5/7/15
to sean...@googlegroups.com, Michael Mattmiller, Bruce Blood, Saunders, Will (OCIO), Brett Miller, Chris Ritzo, Will Scott, Nathan Kinkade, Ginger Armbruster, Georgia Bullen, Ryan Biava, Kendee Yamaguchi, Tony Perez
Three point summary
1. The demo site is now live and hosted in Amazon's AWS cloud:
http:/broadbandtogether.com/
2. My document set has been updated and is being migrated to the wiki for tonight's Civic Hack Night.
3. The minimal basis infrastructure foundation of an open source software development project is now in place. This includes a new fully public mailing list, to which this message is addressed (so be mindful of doing a Reply All).
The demo site
The code is nothing at all to get excited about, as you will quickly surmise by checking out the site. The goal is simply to have a way for folks to get an idea of the state of the code. Honestly, there is nothing there except the two main components from M-Lab (net pert tester and test results mapper) poorly mashed together with a Code for Seattle logo slapped it and a welcome dialog.
The domain name, broadbandtogether.com, is just a placeholder until such time as Seattle's real site is live; it is just somewhere for the code to be running. The site is hosted on one of my Amazon Web Services instances.
nginx is deployed on the site acting as a reverse proxy, fronting for the SEANetMap NodeJS server. Currently getting nginx going really did nothing except waste my time. Nonetheless, laying out a real-world deploy system now ensures that as code actually develops we can ensure through testing that it is written in such a fashion that we can scale to multiple servers when it goes live to the public. Also, with nginx set up I will be able to keep multiple versions of SEANetMap running so progress between versions can also be tracked (more on in a later message).
Note: The deploy process is still being filled out so www in the url (http://www.broadbandtogether.com/) will not work. At this time only the following version of the URL works http://broadbandtogether.com/. There is going to be another overhaul this weekend, after which both versions will work. Expect the site to come up and down over the next few days until the deploy process really is a one-click affair.
Project documentation
My document set has been updated on http://tigue.com/seanetmap/.
New: whiteboard-captured diagrams now are starting to get included:
Once I get feedback and the ideas settle I will generated equivalent diagrams via software tools, rather than by hand on a whiteboard.
Enough progress has been made that the documents have now been migrated to the project's wiki so that anyone can contribute from here on in. There will probably be no further updates of those particular documents on tigue.com, although other document by me will continue to be published at http://tigue.com/seanetmap/. That page is becoming my project dashboard.
At this time the tigue.com hosted HTML version of the docs are more pleasant to read that the GitHub wiki hosted version. But the tigue.com version should get stale quickly as further updates will happen in the wiki. So, now is the best time to read them.
Project infrastructure
I first got involved with this project at the meeting of April 19th at the Montlake SPL branch. I have no idea what occurred before that but since then I have seen the following progress:
1. Code repository was started on 2015-04-19
2. Wiki was started on 2015-04-23
3. First round of documentation enabling others to know how to contribute: live on 2015-05-06
4. First live deploy of the (very early) codebase 2015-05-07
The above milestones are not really that exciting except that they represent a minimally acceptable level of software development process.
Next steps
The above marks a milestone where I can actually get down to playing with the code, rather than all the other less fun things I have been doing like: writing documentation, laying out process, and admin'ing servers.
My next goal is to get all codebases running on the same server with minimal but clean integration between the components. These efforts include:
1. M-Lab's bq2geojson tool generating new maps (with test results aggregated by District, by Census track, by Census block, and by hex grid) on some schedule basis (that is scheduled NodeJS code, not realtime web serving).
2. Code for Seattle's NodeJS-based work in the codeforseattle/seanetmap repo far enough along to where it can show the maps generated in #1.
3. The NTD-JavaScript client running in-browsers but somehow integrated with SEANetMap (this will involve hooking the tester's onprogress and other events).
I will speak with Brett Miller tonight to get an idea of how we can incorporate his work.
The above will represent a major milestone for software dev process and project cohesion. (And I can actually start having some fun hacking code in this project).
I will update the live site with the latest code by end of business tomorrow, Friday, but do not expect much to have changed.
I would rather not be broadcasting messages to a large group like this. I prefer to send announcements and status reports to a mailing list so that interested parties can pull, rather than me pushing to their email inbox. Also, a fully public mailing list with a searchable archive leaves an info trail for others to follow later. So, please join the list if you want to continue to receive messages and contribute yourself to the fully public list.
-John
John Tigue
May 12, 2015, 6:44:40 PM5/12/15
to Nathan Kinkade, sean...@googlegroups.com, Mattmiller, Michael, Blood, Bruce, Saunders, Will (OCIO), Brett Miller, Chris Ritzo, Will, Armbruster, Ginger, Georgia Bullen, Biava, Ryan, Yamaguchi, Kendee, Perez, Tony
(This message is addressed to the project's fully public mailing list.)
On May 12, 2015, at 12:19 PM, Nathan Kinkade <kin...@opentechinstitute.org> wrote:
Two updates on bq2geojson:
Excellent. You are making this very easy for us.
1. Being able to feed the map-maker (bq2geojson) any arbitrary map is very useful. With this new functionality we can generate test result maps aggregate by: City Council District, US Census block, US Census track, and regular grids as was done in the original hex grid.
2. TopoJSON will definitely help with performance.
How are things going, generally?
Things have gone quite after the big push of the last few weeks to get the most basic stuff set up. Personally I will not be able to give the project my full attention for the next two to three weeks. Most generally, judging by GitHub commit activity, things are stalling out:
Getting more specific, to address the biggest things I can think of:
1. None of the City's requirements have been met.
2. bq2geojson has not been incorporated into the code at all, yet.
3. There is essentially zero implemented in terms of collecting data, besides what M-Lab already does.
The deployed code simply has one (just one and not generated by our codebase) of your generated maps, and many features of your demo have not been faithfully reproduced in the SEANetMap codebase (e.g. map animation has been lost):
The code is still basically a frankenstein mashup of the various tools M-Lab provided, with lots of things having gotten lost in the migrate-and-merge process, and very little novelty has been added by Code for Seattle, yet.
Tigue's status and near term plan
Even my commit graph is trending down after having made the big push over the previous weeks:
Note: that graph does not capture even one tenth of the work I put into the project, as much of my time was spent on all the other aspects of the project. Only maybe one full day of actual coding has been done by me.
I have been and will continue working on the project daily (to keep it fresh in my mind) but it is on a back burner currently. I was idle over the last few weeks and so was able to put in a lot of time on the SEANetMap project. I think I got to where there is enough structure so that other folks can productively contribute in a coordinated manner. (That was my complete plan with the window I had, so not bad.)
My current freelance project started this week and it is going to eat most of my time this month as the customer wants to get a site live by the end of the month. In June I might be able to make another big push on SEANetMap.
During May, though, I will continue to try to raise general awareness of the project as well as attempt to recruit more folks to the project:
1. Tonight I am going to the Seattle Citizens Technology Advisory Board meeting:
Certainly they should be kept aware of developments, and who knows maybe I will find a recruit.
2. Wednesday I am going to CUGOS 2015 Spring Fling:
CUGOS is the core local community of GIS enthusiasts and the Spring Fling is their big event so I am going to spread the good word and hope to find some volunteers there.
3. Thursday I will be at Code for Seattle's Civic Hack night:
I will continue to make folks aware of what exists in the wiki and issue tracker, as well as explain what the plan is currently. As with last week I will spend time describing how things work and on-boarding folks. I get very, very little actual coding done at those meetings, but c'est la vie.
If I do manage to get anyone active, note how I made Issues #33 through #43 each a single requirement from the City. This way a contributor can take on a single ticket in order to address a specific, defined need:
If anything worth making public does come out of Thursday's meetup, then I will update http://broadbandtogether.com/ over the weekend.
You M-Lab folks have been extremely helpful with the tools you have provided us. As well, you have been providing great feedback in the project's issue tracker. The functionality recently added to bq2geojson is really just you serving up soft balls for us to hit out of the park.Please don't hesitate to let us know if there is some information or assistance we can provide.
Thank you.
-John
Reply all
Reply to author
Forward
0 new messages