Hi,

This was a long time ago and it never matured into something worth proposing. The tag serves no purpose now.

Absolutely—whether deliberate or accidental, I don't know. This entire airport has seen some detrimental edits in the past few months, not just from this user.

Where it was laid out carefully before with deliberate alignments, etc. there's now buildings inside buildings that appear to be placed by users who know of new information to add but not much about etiquette or techniques in doing so (also the that the entire military boundary in the airport has been removed for some reason?)

I was the user who originally drew out this area carefully but I am no longer contributing much to OSM (mostly due to the inability to prevent work being changed like this) so I'll leave the repairs to another user. See my changesets from a year or two ago if you want to see the initial work put into drawing Ivato carefully.

I understand your point Geofreund and did mention similar as reasoning I'm familiar with. I also understand the drive to fulfil and correct as best a routing system as possible. But one of my points was that it is a flawed goal that came from an archaic principle of OSM that was once pertinent but it has outgrown (that all routes must connect. See an example above for just one reason that is not realistic as a map becomes more mature). Routing is incredibly important but it is not the only functionality of a map. Forgive me if I'm repeating myself but the ability to visually read a map has significant worth in itself. I'm not questioning whether one is more important than the other but suggesting that bettering one aspect should not harm, disintegrate or molest another. Creating highways that do not exist is harming the visual reading of a map, no matter what size they are.

What is the reason for the series of non-existant highways leading onto pedestrian areas around the town?

While I understand that some see OSM as primarily functional for routing, etc. it should also serve as a map that can aesthetically guide people, as traditional maps did. Isolated pedestrian areas are highly relevant to this and can be responsible for confirming a location is that which they're now in, due to distinctive shapes, etc.

I think there was a rise of third-party utilities that provided collective OSM goals, such as highlighting highways that don't connect to the the wider highway network and incentivising users to make them connect. I understand the logic but I think the incentive has overlooked the damage this causes to certain map uses and may misguide those reading a map. For example, I've seen highways drawn to pedestrian areas at junctions that were dangerous to cross and may well have frustrated anybody relying on that way in the belief that the map showed actual features of the junction.

I have also seen it stated the OSM is ultimately a collection of connected highways and any isolated highways fails in this goal. Personally I believe this is nonsense. How would this equate with wanting to map old mountain tracks that are highly useful to hikers but have long worn out their connection to the existing road network? I believe they should be mapped but any ways drawn to suggest a link to main roads or paths would be misleading and may endanger hikers. Etc. It made sense in the beginning of OSM but has matured to the point that it is far more nuanced now.

Apologies if this is a bit long-winded but I've seen so many pedestrian areas linked to roads by non-existant paths or highways and usually it's done to prioritise routing (without actually helping it much) and causing damage to traditional map reading techniques. I'm happy to be corrected.

Privatemajory - apologies for my late response. Maybe I should rephrase my statement about appropriateness to say I believe it to be somewhat inappropriate to use dots in a name of an organisation if the acronym, abbreviated name, is the dominant name used by the organisation.

This is not to say it is wrong to do so, just not commonplace throughout the world as it's just cleaner to drop the dots in terms of branding, marketing and signage and suggests an official title. As for the WWF, look at any of their official communications. Go to their main website. They dropped the dots from their name a long time ago (if they ever had it?) Any use of dots in their name is usually the mistake of a local entity or third-party. Anyway, I agree with a uniform title and think VOIMMA is more appropriate to use than V.O.I.M.M.A. with the long-term in mind.

Thanks for the response Majory and good point regarding the common use of Malagasy and French acronyms without dots. The title should really be what is officially used by the entity. I guess this rules out using the full 'Vondron'olona Ifotony Mitia sy Miaro ny Ala' as - although this is correct - it is not used on any signage or information by the organisation that I came across. I wonder if this is an unusual case of the organisation inappropriately using dots in their name in a manner that would be better for us to omit? See WWF for another relevant example.

Hi Majory - where did you source that this is a typo? As well as the name being an acronym, which would typically be spelt with dots, official signage shows the dots. See: http://aupaysdelando.over-blog.com/2016/02/le-parc-villageois-v-o-i-m-m-a.html

Thanks Andre, I'll sort that out today

Hi Andre, I think you've made a mistake here. This area was tagged as a forest area inside a park, which I believe is correct. Some may argue how many trees constitute a forest, and some argue should a group of trees be a forest or a wood, but a significant group of trees close together is typically tagged as a wood or forest in OSM.

As you've tagged it now, it's tagged as a park inside a park, which doesn't make sense.

I realised I've just commented on the two contributions you've made, so I'm happy to help explain any confusion if this isn't making any sense. Cheers.

I believe this is wrong and will confuse users of the map as they will find a large concrete plaza rather than a park. It is a pedestrianised area that has facilities like a playground (which is already marked as such). The name change to 'Smithfield Square Park' is also inaccurate as I cannot find any records of this as a title.

If you don't mind, I'm going to revert the changeset back in a little while but I'll leave it unchanged for a short while incase there's objections to what I just mentioned.

There's some good advice there, thanks SK53. You get familiar with common visual deceptions with practice. One thing I'd add is that it helps to recognise that a building will likely have predictable shape patterns, such as 90 degree corners rather than 86 degree corners, etc. and matching parallel edges, whether from the same or a neighbouring building (if they already look pretty close). Take a look at the buildings up in the neighbouring Grand Canal Square to see how lots of lines are parallel.

Appreciated, cheers.

The updated nodes of this building, while using better satellite imagery, are far less accurate than the previous changeset. They fail to take into account the perspective warping and, instead, trace just the flattened outline of the combined roofs as presented. See how far they are from the building's footprint, for example.
The previous nodes were carefully considered and calculated before drawing, taking into account any perspective warping. I would highly recommend a revert to the previous changeset.

You're probably very familiar with the Burren, so you'd have an idea of the geology of the Tsingy. They're both large areas of limestone that have eroded into unusual shapes over time. The Tsingy eroded far more drastically than the Burren but vegetation has found its way in small numbers through some of the cracks in both.

I should note that there are some succulents and underlying vegetation that appear in small parts of the satellite photos of the rock areas. When traversing the area on foot they appear in some large crevices under the surface level, growing upwards to find light, having wound their way out of a crack even deeper below. The limestone in these areas is very much bare but the occasional protrusion of these gives a misleading impression from overhead satellite pictures. Anywhere that the vegetation has grown significantly dense to cover the bare rock, the satellite images show as vegetation.

Hi Rory. What brings another Irish user of OSM to this part of Mada? I can see the bare rock formations quite clearly on both the Mapbox and Bing satellite images. The reason I used Mapbox's version was for the higher contrast between bare rock and vegetation, which aided tracing.
Have you been to this area? The reason I ask is that it's famous for its rock formations and there's plenty of them on the surface. The tell-tale limestone strata lines are a giveaway in the satellite images. Perhaps I should have added a second source as hiking the area and familiarity with its rock and vegetation structure.

Used last changeset comments by accident in original upload. I added a new amenity to the area, pedestrianised areas for the large footspaces, changed parking access and added a new parking space

The Mapbox Satellite image seems to be enhancing limestone visible near the surface (I assume this isn't deliberate) as opposed to all sparse mud areas so this felt like a good source to translate to map data while it's still available freely for OSM use. Other satellite imagery is harder to differenciate between what's limestone and what's mud on the surface.