[OSM-dev] OSMF Engineering Working Group inauguration

Matt Amos zerebubuth at gmail.com
Tue Aug 23 13:06:25 BST 2011

On Tue, Aug 23, 2011 at 5:21 AM, Kai Krueger <kakrueger at gmail.com> wrote:
> Hi,
> as I in the end wasn't able to attend the irc meeting (got busy at work), I
> will write down my thoughts to the topics discussed today.

thanks :-)

> * Barriers for entry for getting started:
> I agree that if setting up a development environment is too complicated it
> will stop contributers (e.g. it has stopped me from trying to contribute
> some minor i18n patches to P2), but I don't think the situation is that
> bleak for the rails_port.
> At lest on a standard Ubuntu installation, I thought the step by step
> installation instructions were actually quite good and accurate. And so even
> though it might take a while to go through all the steps, it wasn't too
> difficult.
> So the question is what part do people find too difficult/tedious? Is it
> that people use other operating systems / distributions, for which there are
> no instructions? Are there simply too many steps in the process for people
> to bother? Do the steps not work under some conditions? Do people not find
> the existing documentation?

indeed. we have some people working to see what the hurdles are to
installation. it may well be that these can be documented better, or
possibly even scripted into some sort of automatic installer.

> The second aspect that didn't get mentioned during the irc meeting is that
> there is the osm dev server environment. Once you have a dev server account
> with db access, setting up a rails_port instance is actually pretty easy.
> You basically only need to download the source, extract it in a directory
> and update a few config files and you have a working rails_port instance up
> and running in your home directory with which you can experiment. All the
> dependencies on rails and gems have already been taken care of so you don't
> need to worry.
> Are there things that can be done to make this resource even easier? Easier
> account creation? Better advertisement that this resource exists?

my first thought was that many people, myself included, prefer to
develop on a local machine. but you are right - it is a simpler way to
get up & running. we should discuss this at the next meeting.

> * Motivation to contribute
> Given that currently pretty much everyone is working on coding in their
> spare time for free, imho having some form of (non monetary) reward to
> motivate people to put in the effort to contribute is important. Usually
> this is in form of seeing your code deployed and getting positive feedback
> from the users that they very much enjoy the feature.
> For the first part (and partially second part) having a list of "top ten
> features / bugs that developers want help with" would possibly be a good
> start. If a feature is on that list, and the patch is reasonable, then
> hopefully chances of getting it deployed would be high. (and chances of
> getting told it is a stupid idea and will never get deployed anyway, lower).
> It also provides ideas for people who might want to help development, but
> might currently not have great ideas them selves.
> With respect to the criticism voiced in the irc meeting, that such a list
> would likely be outdated and useless, well, recruiting new developers does
> require a bit of effort and if we can't even put the effort in to keep a
> list of top 10 wanted features up-to-date, perhaps it is not surprising that
> not many people start contributing to the rails_port.

again, we've had volunteers for looking through trac to see what can
be cleaned up, but my feeling is that trac isn't a particularly good
way to organise this anyway. part of the reason for setting up EWG is
to connect people willing to put in a bit of effort (i.e: working
group) with these sorts of tasks. or, as steve would put it, get a
tempo going.

> The second part of the motivation "getting positive feedback from the users
> that the new features are appreciated" is perhaps much harder to solve. One
> possible way though would be to have a user oriented feature wish list a la
> uservoice (or better to say an open source equivalent). If a feature has
> e.g. a thousand up-votes and is the top _feasible_ feature on that list, I'd
> be a lot more motivated to spend my weekends coding that feature /
> enhancement / bug fix than if I don't know if anyone would even care about
> the patch or if the community would even think it is rubbish. But perhaps
> there are better ways of obtaining that feedback.

getting good, constructive bi-directional communication between users
and developers is important (and possibly hard). it's clearly
frustrating and pointless for both groups if they maintain separate
feature lists with different priorities. i think the concern with
uservoice and similar systems was that it's not a good forum for
bi-directional discussion about things like feasibility and
compromise. but the difficulty of the task should not dissuade us ;-)

> But overall, it seems like it was a reasonable  productive first meet for
> the EWG.

thanks to everyone who came. hopefully we'll have another productive
meeting next week.



> Kai
> On 01/-10/-28163 12:59 PM, Kate Chapman wrote:
>> Hi Matt,
>> Just wanted to mention I'm interested in helping.  Sadly I am
>> unavailable today, though normally I would be at the time selected.
>> Couple thoughts I have about the rough agenda:
>> * Barriers for entry for getting started:
>> People seem to have issues getting the rails port working.  Perhaps
>> creation of VM with development environment?  That would allow those
>> who maybe can just do JavaScript and CSS still to contribute
>> something.
>> Documentation (this of course depends on the specific project).  Are
>> there holes in the documentation that could be improved?
>> * Small Projects that are mentored to get developers started:
>> I don't have any great ideas about this.  I don't actively contribute
>> to any of the main OSM projects by actually coding, though I'm willing
>> to help mentor if there is a way that makes sense.  Are there
>> individuals who come in and want to contribute code and don't map that
>> need basic OSM mentoring?
>> Thanks!
>> -Kate
>> On Mon, Aug 15, 2011 at 8:09 PM, Matt Amos<zerebubuth at gmail.com>  wrote:
>>> the "Engineering Working Group"[1] is a new group set up with the
>>> purpose of helping stimulate the development community surrounding
>>> OSM. you don't need to be a developer to join - we're interested in
>>> hearing from everyone and ideas are welcome. software enables us to do
>>> the mapping we love, and to make use of the data we collect, so making
>>> our software better is part of making OSM better.
>>> although IRC isn't a perfect communication medium it's easy to set up
>>> and it's interactive, so i've set one up over at #osm-ewg on OFTC.
>>> just to kick off, i suggest a meeting next week on:
>>>  Monday 22nd at 5pm GMT
>>> (5pm CEST, 6pm BST, 1pm east-US, 10am pacific-US)[2] to start the
>>> conversation.
>>> a very rough agenda for the meeting, including but certainly not limited
>>> to:
>>>  * what are the barriers-to-entry for new developers?
>>>  * are there small projects (perhaps mentored?) which new developers
>>> can take on?
>>>  * what can be done to improve the usefulness of communication between
>>> developers and users?
>>> hope to see you there,
>>> matt
>>> [1] it's not necessarily the best or most descriptive name, but it'll
>>> do for now. other suggestions include "Development Support Group" and
>>> "Software Working Group".
>>> [2] yeah, i know, but no matter what time is chosen it'll be bad for
>>> someone. i suggest we start with this and i'm open to suggestions
>>> about rotating the time or setting up other communication channels
>>> which are less time-zone sensitive.
>>> _______________________________________________
>>> dev mailing list
>>> dev at openstreetmap.org
>>> http://lists.openstreetmap.org/listinfo/dev

More information about the dev mailing list