Recent diary entries
Maproulette challenges have become fairly popular recently, especially due to Jochen's Area fixing project. But it seems this has gotten out of hand now and creates serious damage to the OpenStreetMap project.
In general this kind of tool is prone to inviting mechanical work. But with the recent Island and Shoreline Alignment challenge this really gets over the top. I first saw this when various edits turned up in remote areas of the world by various mappers in very high frequency editing islands in changing locations far apart within minutes, often without factual basis and often factually incorrect.
This challenge does everything wrong that can be done wrong with a fixing effort:
- there are no useful instructions to the mapper what to do and what problems to consider. It only says: ''Align the highlighted island to match imagery''.
- there is no documentation who created this task and how the allegedly misaligned islands are detected.
- and most importantly: the task covers areas where the global images routinely available offer no basis for improving the existing data.
Task: Align the highlighted island to match imagery (link)
Especially the last point is a big practical issue now since the edits made through this challenge misalign and worsen a lot of data in OSM. I commented on two most obvious cases where Bing offers no image at all but mappers none the less blindly followed the task to Align the highlighted island to match none-existing imagery with obvious results. But even in areas where Bing offers low resolution images these hardly ever allow improving existing data. These images in Bing are mostly from >15 years old L1G Landsat 7 images which have positional errors of sometimes more than 100m and rarely allow substantially improving existing mapping. In most cases attempts to do so made within the challenge worsen data which has often been mapped from either better data or with better alignment of the images.
With edits like this the island in question is not necessarily less accurate than before - the mapping before was done based on - if at all - only slightly better aligned images. But it is no improvement since it is likely not more accurate than before in absolute terms and is definitely less accurate relative to the surrounding features.
Yes, these images show a lot of potential to improve mapping here
The least that needs to be done here is
- stop the challenge as it is now
- limit it to areas where high resolution images are available in common sources
- create clear instructions for the mapper advising them to properly check image alignment and find the best quality sources in the area and check if existing mapping might already be of better quality based on other sources. This is not possible to do within a few minutes if you pick random locations all around the world of course.
Of course not all edits made within these challenges even in areas where Bing is poor are bad, there are also some experienced mappers participating here who know how to properly assess and select images.
If whoever created this challenge seriously wants to improve mapping in remote areas the most basic, most productive and most obvious way would be to provide better quality images.
And in general i think we need to put a review regime on organized mapping efforts like maproulette challenges requiring at least basic documentation of the process used to generate the task and ensuring there are proper instructions for the mappers and no nonsense tasks on a larger scale. A lot of thoughtful tasks have been offered in maproulette in the past but apparently this is not something that can be relied upon to be ensured without a QA process for the QA process...
OSMF board elections are done with Single transferable vote (STV). Since this has caused confusion and misunderstanding in the past occasionally here a quick and politically neutral helper how to vote with STV:
The name says it already in fact: Single transferable vote. You only have a single vote and you cannot split it. But you can specify an order of priorities who to give this vote to. In almost all cases your vote will go to the candidate you put on top of your list. Only in the rare cases where your vote for this candidate would be wasted it is transferred to the next highest candidate on your list (and subsequently possibly even further down the list). This happens in either of two cases:
- your top candidate does not stand a chance because he/she has too few votes overall.
- your top candidate got so many votes he/she does not need your vote to win.
To help your decisions you can find the list of candidates and their manifestos on the wiki as well as questions by the community and answers of the candidates - where you also can still ask questions if you have any.
Recently the OpenStreetMap Foundation issued the OpenStreetMap Awards.
The whole thing was primarily organized by Ilya Zverev who deserves thanks for doing this and for the courage to try something new.
When this was first suggested it seemed like a good idea to me but during the process i already had some critical thoughts on the way it turned out. I did not want to speak up while the votes were still running not to influence the procedure but now i think it is time to bring this up.
First of all the whole process was quite biased towards English language activities. There were non-native English speakers among nominees and winners but almost everyone on the list was nominated for activities in English language. Since the whole process was done in English only it was not possible for someone who does not understand English to competently participate in nomination and voting and assessing someone nominated for activities in a language you don't understand is not really possible either - the few suggestions in the first nomination round that were formulated in languages other than English never stood a chance. This is a hard problem. But still i think this can be done better with not too much additional effort.
The three stage process - open nomination, preselection by committee and final open vote again - does not really work in reality. It gives an impression of manipulation since it appears the preselection is used to eliminate undesirable nominees and the final vote therefore appears staged. In the future i would probably either skip the committee selection (making it a fully open process) or eliminate the final open vote making the final choice by the committee - which would of course require this committee to be selected in an open process somehow.
Somewhat related to this the award categories do not really work either. The initial nomination round showed that people often simply wanted to nominate someone and put them into a category that seemed to fit best. As a result in many categories nominees were not really comparable because they were nominated for very different things which kind of defeats the purpose of having categories. The categories should either be more strictly defined or nomination should be across categories and votes decide on which category they are awarded for.
All of this of course does not mean the winners do not deserve their awards - all winners and nominees should be commended for their work. I have slight misgivings only about Frederik - who specifically said before he did not want the award and about Manuel Roth and Lukas Martinelli who certainly deserve an award although IMO not in the category 'Innovation'. The technology they thankfully make more accessible to a broad range of users is for the largest part the innovative work of others. Now i don't say that Mapbox employees should have been awarded here instead because awards like this should primarily be given to those who volunteer their free time and not to professionals who get paid for their work. However if you strictly evaluate the innovative merit of the nominees' work this seems a somewhat odd choice to me. But of course voters will usually consider who of the candidates they think deserves an award most and don't care what particular award this is.
I hope these comments will help improving future award processes and maybe start some further discussion on how the OSM community wants to reward and acknowledge contributions.
Last week BushmanK wrote about the use of up-to-date open data satellite imagery for mapping in OSM and noted what i also frequently experience - that awareness and interest within the OSM community regarding the large bandwidth of up-to-date near real time open imagery that is available today is astonishingly very low. Mappers do complain that imagery in Bing and elsewhere is frequently outdated and poor quality but few are aware that newer imagery exists and is available and in contrast to Bing etc. is often truly open data.
The real problem here is that as a result of this mappers keep wasting energy and time on tracing things from images that are hopelessly outdated and at the same time often also poorly aligned. At the moment approximately 15-25 percent of the Earth land surfaces are shown in Bing and Mapbox with 15 year old imagery that is poor quality in a lot of aspects.
With this blog entry i hope to somewhat further increase awareness of this subject among mappers. I have been for quite some time making available recent imagery from open data sources for mapping in OSM. This is only a small contribution for select areas but shows that a huge body of primary data is available today and is largely unused for OSM-mapping.
Northern Greenland July 2016
Images from 2016-07-17, the most recent ones of this remote area, better detail and more up-to-date than current mapping in OSM.
Northern Ellesmere Island July 2016
Images from 2016-07-08 to 2016-07-15, recent images, partly overlapping the previous, poor and largely faulty data there in OSM based on imports.
From 2016-06-23, showing most recent building activities.
From 2016-06-07 - the new locks.
Darwin and Wolf islands, Galapagos
From 2016-03-11 - two small islands with poor coverage in other sources.
From 2014 to 2016, quite a few islands missing or poorly mapped in OSM just a short distance from Singapore.
All of these are prepared from open satellite data, of course the main advantage of this is you do not depend on my or others' services to make use of it. Processing raw satellite data is something you need to learn to do it but it is not that difficult in principle. You just need time and an open mind to get the necessary experience and some background in photography or color physics definitely helps. There are quite a few mappers who routinely map from custom processed Landsat images for example.
And since the remark will inevitably arise - yes, these are all lower resolution than what is necessary for tracing smaller buildings or other small scale features. That is the downside of having up-to-date open data for everywhere in the world. But as said the main target here is abolishing the 15 year old even lower resolution and much poorer quality imagery. A nice secondary use is supplementing older high resolution data with information on recent changes like in case of the Vostochny Cosmodrome and the Panama Canal.
Yesterday evening there was a public OSMF board meeting. I was one of the few non-boardmembers attending so i thought i'd give a report of my impressions here.
This was not the first public board meeting, there was one previously last July but this was a singular occurence so it was possibly more of a mock up meeting demonstrating publicly how board meetings go. The one yesterday was held under the premise that this is how board meetings are going to be conducted in the forseeable future which is a very different sitation. A big thanks to the board for taking this step and i hope the OSMF members and the OSM community as a whole acknowledge this by coming to the meetings. With the short announcement and the Friday evening date in Europe the small participation this time was understandable though. This is really public by the way, everyone can listen in, you don't need to present your OSMF membership number or something like that before you are allowed to enter.
I was about ten minutes late so i did not get the start, i came in during some discussion on SotM regarding finances between the board and Rob Nickerson from the SotM working group as invited guest. There were very few non-board members present overall - i think apart from Rob and me there were two others overall.
A few general words on procedure: The meetings are conducted with Mumble which i was already familiar with from the German OSM podcast which was usually recorded with audience via Mumble. You connect your Mumble client to the HOT mumble server (talk.hotosm.org), move into the OSMF board meeting room and can immediately listen to the conversation. Since i was not present at the start i missed any initial statements on procedure. There were no constraints in place this time so i probably could have said something at any time but in general it is likely expected from guests to not speak up freely but only talk when being given the word by the board. You can also mute yourself (which i did) to indicate you are not actively participating.
In general Mumble is not quite like a face to face meeting, you have only acoustic and no visual communication and there is always a small but inevitable time lag in communication. It is more like radio communication. You usually configure your client to only transmit when you press a button to eliminate any background noise when you are not talking. There is also a text message/chat system connected to it which can be used for communication without interrrupting the audio conversation.
On the meeting itself - my general impression was that it was easy to follow, everyone was understandable and none of the current board members has a really problematic accent - Paul a bit of Canadian which you need to get used to, Peda quite strong German tone (which i of course have no problem with) and Ilya a bit of Russian tint (which i find enjoyable). There was occasionally somewhat strong background noise while people were speaking but not everyone can move to a tone studio for the meeting of course.
What happens during a board meeting is the members talk about various topics make decisions on some of them via vote and so on. My general feeling of the whole thing is - i hope this does not sound too harsh - that it is kind of unproductive. I am probably somewhat biased here, being self employed i am not really that used to regular organizational meetings any more - when i am at a meeting these days i tend to get paid by the hour which usually tends to expedite things. But i know from past experience that meetings are often fairly unproductive at least by outward appearence and this board meeting was not an extreme case in that regard at all.
This impression is probably partly because of the setup in Mumble - although you are talking to each other you are not really stitting together. Quite a lot of time is spent essentially on waiting if someone has something more to say on the matter because you cannot indicate this using body language. There is also the occasional conflict when two people try to speak up at the same time and then both back off to let the other have the word. Another factor probably was that because the meeting was public everyone was very guarded and careful with voicing a strong opinion. To get progress on a subject it tends to help if you try to work out topics of disagreement by expressing your standpoint in a very pointed way and possibly even insituating a disagreeing standpoint from someone else. This did not really happen. So this is probably something that will improve in the future when everyone gets more used to the public setting.
I will give two examples of subjects that were discussed:
One topic was the collective database guideline which was approved by the board during the meeting. Procedure for votes is apparenly very formal by the way, Kate (who was chairing the meeting) called every member individually to approve or disapprove. There was some discussion about the examples to be included with the guideline - i did not really understand that, maybe because it was about a third example which was not part of the guideline draft on the wiki. What astonished me about the procedure a bit is that although this was a decision with quite some impact - after all this is now an official statement on the interpretation of the license by the organization holding the rights on the OSM data - there was no recap of the process leading to the guideline, the reasoning behind making the guideline the way it is and how the board thinks this fits into the OSMF mission (which it probably does - but still). Also i would have expected a kind of outlook in lines of where to go from here in terms of developing additional community guidelines or modifying existing ones.
Part of this could have been due to the fact that no one from the license working group was present and the board probably considers the guidelines to be managed mostly independently by the LWG and their role being purely oversight in terms of preventing possible gross blunder in these.
Another topic which was still in a much earlier state of discussion was a possible donation drive for the OSMF to be conducted later this year. Here my understanding was somewhat hampered by the fact that apparently this idea has been already extensively discussed on the face-to-face meeting of which there is not yet a comprehensive record. The discussion was mainly about how to proceed about this regarding the purpose of the donations (what the money is needed for), possible legal implications (if the donations can only be used for the purpose they were announced to be needed for) and timing (what is the best moment to start such a drive). The impression i got from this is that the board considers a fairly general donation drive to support their efforts to put OSMF finances on a less volatile basis (meaning less from hand to mouth and more of a cushion to compensate fluctuation in either income or expenses). One topic touched in that regard was the matter of trust in the OSMF board (specifically by the operations working group which is considered instrumental for a donation drive but also in a broader sense). My own impression is that this in general are important matters and it is good to see these are discussed although i kind of see the risk of starting to build the house from the top. Trust in the board regarding finances is a prerequisite for a successful donation drive and having an overall concept and realistic plans for both income and expenses is necessary to build such trust. With plans for corporate membership and widening general OSMF membership still somewhat vague and little long term (i.e. beyond yearly budgets) plans and directives on expenses (what the money will be spent on and what it will not be spent on) there is little basis to form an informed opinion on an individual matter like a donation drive. I can see the possibility of all of this developing into a solid an trustworthy concept but there is still a lot of work to get it there.
Overall i see the concept of public board meetings on a good way. Mikel made an interesting suggestion to have alternating formal board meetings and more informal talks in between which could make the whole thing more participative and more interesting for the community (although care needs to be taken for this not to degrade into a general chat).
A final suggestion to the board: It would probably be good if in the future you were all easily identifiable on Mumble by your Mumble name for anyone entering at any point in the meeting. IIRC Mikel was kind of cryptic. Maybe just agree on a common form (first name, first + last or OSM username).