Make iD default (clean PR)

[lxbarth]

Let's make iD default. If a user misses the drop down, and just clicks on "Edit" they should go to iD, not Potlatch 2.

• We had a great testing period
• iD's doing much more to onboard new users and we want to particularly reach new users who might miss that there's a drop down.
• iD is the first one in drop down list but isn't default.

This PR just modifies the example config.

iD in edit drop down on osm.org right now

This pull request is a cleaned version of #261

[tomhughes]

This is not an available option until we implement support for falling back to an alternative (presumably Potlatch 2) for at least those browsers which don't support iD at all.

There is a secondary question about browsers which technically support it but where performance may be poor enough to render it non-optimal.

[woodpeck]

I think it would not be well received by the OSM community if we made iD the default on the first day it is even available on osm.org. I think there's a large group of people who are interested enough to try it out, but not so interested that they would have sought out iD while it was not yet available on osm.org; let's give these people a chance to get familiar with iD (so that, when questions from new users appear, they are not met with "sorry but I'm not familiar with that editor you're using, why don't you switch to [other editor] and then do ...").

I think that having a great testing period within the developer community is not sufficient for rolling out iD as the default editor to the unsuspecting masses.

Let's set a date ~ 2 months in advance and say: We'll make iD the default on that date unless major road blocks appear. That would also be a good incentive for existing power mappers to learn iD at least well enough to be able to answer questions about it - and answer them usefully, rather than "why don't you switch to ...".

[lxbarth]

@woodpeck - don't follow as iD is particularly geared towards new users and very much intended as a successor to Potlatch 2.

[openstreetmap-website]

I agree with Frederik,

although it is great to see that iD is now an option on osm.org, I'd
also prefer to see it mature a little and get more wide spread feedback
and acquaintance from the community before we make it the default option.

After leaving it a little while as it is now, an option, to iron out any
potential major feedback from the existing community it might even be a
candidate for some A/B testing. I.e. a proportion of new users get iD
set as a default option and a proportion of new mappers P2. Then see if
there is any statistical difference in editing behaviour between the two
groups.

Kai

On 05/07/2013 08:51 AM, Frederik Ramm wrote:

I think it would not be well received by the OSM community if we made
iD the default on the first day it is even /available/ on osm.org. I
think there's a large group of people who are interested enough to try
it out, but not /so/ interested that they would have sought out iD
while it was not yet available on osm.org; let's give these people a
chance to get familiar with iD (so that, when questions from new users
appear, they are not met with "sorry but I'm not familiar with that
editor you're using, why don't you switch to [other editor] and then
do ...").

I think that having a great testing period within the developer
community is not sufficient for rolling out iD as the default editor
to the unsuspecting masses.

Let's set a date ~ 2 months in advance and say: We'll make iD the
default on that date unless major road blocks appear. That would also
be a good incentive for existing power mappers to learn iD at least
well enough to be able to answer questions about it - and answer them
usefully, rather than "why don't you switch to ...".

—
Reply to this email directly or view it on GitHub
#262 (comment).

---

rails-dev mailing list
rails-dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/rails-dev

[tmcw]

A fallback is added for browsers that are known to not support iD at all: IE. Opera is workable with a few bugs. iOS obviously doesn't work, but neither do any other editors.

[tomhughes]

Shouldn't ie_editor be defined somewhere?

[tmcw]

It's defined in the chunk above. @jfirebaugh is working on a more robust version of this.

[systemed]

@woodpeck: that's all very well for you to say but you're not the poor unfortunate who has to maintain P2. :) I am and frankly, as far as newbie editors are concerned, I want shot of it (refashioning it into something nice to map cycle routes with is another matter!). Please, please, let's move to something actively maintained and get iD as the default.

[Firefishy]

@woodpeck I support @systemed here.

Newbies will use whatever they are given, it already makes sense to give them the easier iD editor instead of PL2. Those already more comfortable with OSM, can switch (back) to PL2 via dropdown or profile preference.

[apmon]

I don't think anyone objects to plan to make iD the default editor.

But so far I don't think the wider community has had a real chance to weigh in on it with their opinion of iD, as I don't think people realized that iD was so close to being ready to deploy on osm, let alone become the default.

P2, if I am not mistaken, also was available as an option for a while before it became the default.

[woodpeck]

@Firefishy my point was not that iD is more difficult than Potlatch (I sure hope it is not). My point was that I would like to have a healthy and experienced community of iD users before we send every newcomer to iD on the first click so that when they ask for help, which they eventually will, there are enough people to provide that help.

Making an editor the default editor for OSM on the same day it shows up on the radar of 99% of the community is asking for trouble, and I don't see the benefit of rushing it.

[AugustusKling]

I second Frederick/woodpeck in saying that things should not be rushed. iD looks promising but has not yet reached the maturity to justify releasing it as default editor.

I could not get it to run in Firefox 20 in a way that is smooth enough to use iD. Performance in Chromium 25 is okay but tends to slow down in areas with lots of features or complex geometries. Given that this are results from an oversized development machine, I doubt the majority of users can enjoy editing with iD.
Even though the comparison is unfair due to the technology in use JOSM runs notably smoother for much larger data sets.

The main thing that speaks against promoting iD is the missing support for relations. At times one can't even see the presence of relations and thus risks to mess up data without noticing. An obvious example are route relations.

The UI is nicely done and probably makes editing possible without knowing much about OSM's data model. It would be nice though if less screen space would be wasted by whitespace. Consider moving the tags of the main editing area closer together and merge the title (for example “Editing Forest”) with the icon below. The micro-size scrollbars are a no-go because they can hardly be used – especially in the drop-down boxes.

In my opinion the known bugs that risk messing up or destroying data unintentionally should be fixed before having iD on the OSM page. Afterwards I'd love to see a more mature version of iD on OSM's page because the concept does suit beginners better than any editor I've seen before.

Instead of having a default editor OSM could present the most popular editors with a wee description when a user clicks the “Edit” tab without using the drop-down menu. Such a presentation could recommend iD for beginners and JSOM for power-users.

Can OSM shed some light into the current visitor demographics, please? Things like browser window size, browser versions, available plugins and languages should be considered before advancing forward with changing the default editor.

[lxbarth]

This needs a fresh PR. Coming.