Closed squeak9 closed 2 weeks ago
I agree. In SCEE I change it very often by myself in the Tag edit mode.
TL;DR: it seems to be subcategory of stands
, and is not hugely popular (0.41%
). Probably better fit for SCEE, as it might be confusing in SC, where bicycle_parking=stands
is currently defined as "Stand (supports bicycle frame)"
Some extra information:
safe_loops
has 2228 uses (0.41% of bicycle_parking=*
values)but I thought they would just be classified as
bicycle_parking=wall_loops
because it's technically correct
Is it technically correct, though? Main characteristic of bicycle_parking=wall_loops
is that the whole bicycle is kept upright just by wheel being attached to parking (resulting in popular name "wheelbenders" for those parkings, as that is the damage they often cause due to ). It also has issues with securing a bike (as an U-lock will be able only to lock a wheel, allowing for the rest of a bike to be stolen).
So I would say that bicycle_parking=safe_loops
is instead a sub-category of bicycle_parking=stands
, as the bike is secured from falling / balanced by its frame (instead of only by its wheel), and both a wheel and a frame can be securely locked preventing partial thefts. There are many other sub-categories of stands
, see some examples here. Some have separate (documented or not) OSM values, some don't - what is common for all of them for all is the parts I've marked in bold.
That being said, the bicycle_parking=safe_loops
seems like acceptable category if one prefers to distinguish it (although I personally would prefer bicycle_parking=stands
+ stands=safe_loops
method of specifying every-increasing level of detail, as it would avoid huge burden on data consumers[^1] and actually help increase the acceptance of the new tag!).
It is just that if it was to be included in StreetComplete, other answers should be redesigned to include subtypes of parking along the types in order to be usable -- perhaps just change descriptions of more generic types, or better yet change it to quest type like "building types" which supports generic-types (like residential
) as well as sub-types (like house
or apartments
).
[^1]: for example, if some data consumer (e.g. OsmAnd or cyclemap or whatever) does support that level on detail, with bicycle_parking=safe_loops
it won't be able to show appropriately different icon or any type of information except "there is some bike parking here". It it were tagged as bicycle_parking=stands
+ stands=safe_loops
instead, even if it did not know about stands=safe_loops
sub-tag, it would still be able to show the use that there are Stands there where user can safely lock their bicycle without damaging it, even if it could not say additional details (e.g. is it "safe loops" or "wide_stands" or "artistic" etc. subtype of "stands")
it seems to be subcategory of stands
I would disagree with that. Short: wide_stands are a subcategory of stands.
stands (German: Bügel) only attach the frame wall_loops (German: Felgenkiller) attach only the rim safe_loops attach rim and frame
From my point of view, this is a huge difference. Not only the design but also the safety aspect.
The use started quite late (end of 2022), I am not yet able to find out what triggered this: https://taghistory.raifer.tech/?#***/bicycle_parking/safe_loops Maybe? https://github.com/simonpoole/beautified-JOSM-preset/pull/380
There are some additional variants documented here: https://wiki.openstreetmap.org/wiki/User:JeroenHoek/bicycle_parking_additional
stands (German: Bügel) only attach the frame
I don't know about German wiki, but English wiki for bicycle_parking=stands explicitly states:
A bent piece of metal against which you can lean your entire bicycle. Makes it possible to lock the frame and a wheel to it. Moderate security. Rectangular stands are sometimes called "staple racks" in North America due to their shape, and "Sheffield stands" in the UK. Use this tag for non-rectangular stands as well (e.g. round ones, fancy artistic ones, long ones allowing to attach more than two vehicles).
(emphasis mine)
safe_loops attach rim and frame From my point of view, this is a huge difference
I agree it would be big difference, if it were correct. However it seems to me that "only attach the frame" for bicycle_parking=stands
is incorrect. :man_shrugging:
I'd say it is German translation problem then, as both pictures on the wiki for bicycle_parking=stands
(even on German wiki) clearly show structures that allow both the wheel and a frame to be locked to it. It also seem supported by numbers (it seems quite unlikely that the most popular type of bicycle_parking would allow locking only frame but not a wheel -- that would seem to me like extremely rare structure). It also seems unlikely to me that bicycle parking which allows "locking both frame and a wheel" (as probably most popular type of bicycle parking) did not exist in OSM at all until late 2022, while many other (much less popular) types did.
stands (German: Bügel) only attach the frame
why you think so? You definitely can attach also wheels on typical inverted U stand (though in areas with lesser amount of bicycle theft attaching only frame is popular)
https://www.stolenride.co.uk/resources/how-to-lock-your-bike/
https://www.rota.mt/bicycle-racks/
You're right, that was a problem with the translation. I didn't mean that you can't lock the rim to it, but that the rim is not held in place. You can only lean the bike against it. This is not ideal without a stand. With safe_loops or wall_loops, you don't need a stand because the wheel is held between two metal bars.
The difference I want to point out is the possibility of supported standing vs. unsupported standing of the bike without a kickstand. As the owner of a carbon frame, this is important.
You lean bicycle against stand, why it is not "supported standing"?
It is "supported standing of the bike without a kickstand" but only with the (carbon) frame not with the rim.
A small addition with more context from the perspective of a carbon frame bike owner without a kickstand (roadbike):
stands
:I can lock my bike at almost any point with a bike lock, this includes the frame and both wheels (rim) if I have so many locks with me. However, I have to lean the carbon frame against the "metal stand", which is usually made of stainless steel. This harbours great risks of scratches or hairline cracks in the carbon.
wall_loops
:The bike stands on its own, secured by it's rim/wheel, there is no need for a kickstand on the bike. However, I can only lock one tyre/rim there. As racing bikes are often fitted with a quick release on the thru axle, it is child's play to remove the thru axle and steal the bike together with the frame but without the one locked wheel.
safe_loops
:An attractive combination of the two previous systems (at least for carbon frame road bike owners). The bike stands on its own, attached to a wheel in a "metal clamp". I don't need a folding / kick stand on the bike. In contrast to the wall_loops
, with this variant I can secure the frame itself without it having any contact with the metal.
Does that help and does it make my view of the differences a little clearer for you?
P.S.: I created the wiki sub page for safe_loops
to make the differences visible: https://wiki.openstreetmap.org/wiki/Tag:bicycle_parking%3Dsafe_loops
Any feedback more than welcome!
@mcliquid thanks for the explanation above, it does make it more clear!
I agree that safe_loops
has advantage over wall_loops
in that in allows (in addition to wheel) the frame to be locked, which makes it significantly more theft-proof (while still not as good as stands
; as judging by the pictures it only allows securing one wheel, not both)
However I am confused about parts on the wiki where you seem to imply that safe_loops
would avoid "wheel bending"? If I understand you correctly above, the main advantage for carbon frames over stands
is precisely that safe_loops
balances the bike only on a wheel (just as wall_loops
do) and bike frame does not lean / is not supported by parking structure at all, right?
So any lateral force would still lean the whole of a bike to a side, putting all the pressure solely on the wheel rim? At least until the frame hit the structure, which would only happen in 50% of the cases commonly depending on the direction (For example in picture if force was pushing bike from left to right, the damage to the wheel would be limited as frame would hit the upper part, but if the force was pushing from right to left, the damage to the wheel could be as high as with regular wall_loops
, right?)
P.S. (I'm not trying to argue here which bike parking is "better", but just trying to get a clear understanding of all the practical differences, so they could be better documented)
However I am confused about parts on the wiki where you seem to imply that
safe_loops
would avoid "wheel bending"? If I understand you correctly above, the main advantage for carbon frames overstands
is precisely thatsafe_loops
balances the bike only on a wheel (just aswall_loops
do) and bike frame does not lean / is not supported by parking structure at all, right?
That is correct and in my view this is also clear from the description? But my English is not ideal for the wording. By ‘wheel bending’ I understand the deformation of the wheel like ‘wheel distortion‘ or ‘wheel warping‘. And this is not the case here. Is the term used wrong?
I think you can see that very well in this picture. The bikes stand with the front wheel in the metal ‘thing’ at the front and do not touch the high metal bar / bow but are only locked there. Yes, the black one on the right is leaning against it, but that could have been avoided. You can also see very clearly that no pressure/force is exerted on the rim itself on the centre bike.
That is correct and in my view this is also clear from the description?
Thanks, I'll try to improve wiki description later when I have access to regular computer again in about 2weeks..
You can also see very clearly that no pressure/force is exerted on the rim itself on the centre bike.
it is not at the moment (when picture was taken), but it would be the same with regular wall_loops, no?
It is about what would happen if such force were to be applied (e.g. to take that same picture; if someone parking their bike in empty spot close to the leftmost bike were to accidentally push that leftmost bike hard to the left, its front wheel might get damaged, right?)
but it would be the same with regular wall_loops, no?
It all depends on the design, I don't think you can generalise. If the wheel is placed from top to bottom, the load is less than if the bracket is in front of the wheel and the wheel is pushed in horizontally. With a bicycle parking of this type and a bike without a kickstand, the expensive carbon rims are not at all happy.
With an expensive carbon bike (frame and rims made of carbon) it's like riding a raw egg. You shouldn't actually park it in public, but sometimes you can't avoid it. And if I have the opportunity to find a bicycle parking in advance where I can lock the bike to the frame and I don't have to lean the frame directly against metal, but ideally only the rubber of my tyre is exposed, I prefer this. But that's already getting very scientific. My point is just that there is a meaningful distinction between wall_loops, safe_loops and stands.
TL;DR: Can the new value be added without introducing more confusion than the value it brings? I'm not sure :shrug:, but it does not look easy to me. Suggestions solicited.
My point is just that there is a meaningful distinction between wall_loops, safe_loops and stands.
I absolutely agree with you on that, as well as that there is meaningful distinction between all other documented & used bicycle_parking
values. If they weren't important to someone, people would not invent more detailed tag and use them... Yet SC only shows a subset of them (and there are reasons for that).
One question is thus "is importance big enough to warrant separate SC answer". Now it obviously is important to you :smiley_cat:, and I personally would like it. It is however a matter whether such answer is better match for SCEE (where it would fit well, as requiring subject expertise is not a big issue), or "is it easy enough answer to include for general StreetComplete population?" - which is IMHO more important second question?
Because, if it is included in SC, current answers for both Stand
and Wheelbender
(and not only the new Safe loops
!) will likely need to change, and change in such a way that it is very clear to anyone (even persons not owning a bicycle!) which answer is correct one in each case.
I.e. currently user can visually exclude all "special" rack types (lower two rows) and only needs to distinguish between
So basically, most confusion to users is just between stands
and wall_loops
, and current explanation texts for those are (hopefully) clear enough to distinguish them.
Adding a third contender to the mix muddies up a situation a lot, and now requires three separate explanations how to clearly delineate all three them with using just 2-3 words (the picture will become invisible with more words), which seems like a hard problem (given most of target SC audience is not familiar with names of such bicycle parking types. Even some bicycle users might not be familiar with those names; and I don't think e.g. Croatian even has a separate name for those).
Or if more than 3 words are needed the quest would probably need to move to Surface quality Quest
format, allowing for more text, but then again that would complicate the quest (although hopefully not to the point of making it disabled by default).
Given above, do you @mcliquid have suggestions how that problem should be approached? What would be your suggested texts for stands
, wall_loops
& safe_loops
(if the last one were to be added)?
Given above, do you @mcliquid have suggestions how that problem should be approached? What would be your suggested texts for
stands
,wall_loops
&safe_loops
(if the last one were to be added)?
Thank you for the mention. I don't feel I have the expertise to answer this as English is not my first language. In German, I see that there is the term "Vorderrad-Rahmenhalter" in official publications:
According to the German Wikipedia, these models are listed under number 7 in Normnahe Modelle.
Die niederländische Variante eines "wheelwell secured racks" wird auch "Tulpenparker" genannt, da der Ausleger mit seiner großen Anschließöse einer gebogenen Tulpe ähnelt. Das Modell hat die Prüfung nach der niederländischen FietsParKeur-Norm bestanden.
A German manufacturer calls this model series "Lean-to parker" or in German "Anlehnparker".
Here is another document from the province of Salzburg. It also recommends using a bicycle stand that secures the bicycle frame and wheel at the same time, i.e. a ‘safe_loops’ in our wording. On the right you can see a model:
On the last pages of this document they also call it "Vorderrad-Rahmenhalter". A complicated term, but obviously the right one.
You are warmly invited to find a suitable translation. The word components directly translated are "front wheel" + "frame" + "holder"
For completeness:
stands
is "Anlehnbügel"
wall_loops
is "Vorderradhalter"
So if you remove the wheelbenders from the safe loops, you end up with stands? (Why have the wheelbenders in that construction at all, then?)
No, the construction is different as already written above.
So if you remove the wheelbenders from the safe loops, you end up with stands?
Judging by the pictures (included that last one in https://github.com/streetcomplete/StreetComplete/issues/5774#issuecomment-2288858645) I would say that if you were to saw-off the part holding the wheel from safe_loops
, you would end up with something looking and functioning like "regular" stands, yes.
I don't know why @mcliquid seems to disagree here (if we've both understood the question in same way?)
(Why have the wheelbenders in that construction at all, then?)
One (only?) of the reasons mentioned in discussion was that carbon frames are extremely sensitive to bumps with the metal (which might case invisible microcracks making frame prone to catastrophic failures when later subjected to stress when riding technical track sections, IIUC). And I guess even owners of many fancy aluminum/steel frames owners might dislike paint being scraped off from their new bicycle (which is likely when parking a bicycle on inverted-U type stands).
Having the bicycle balanced by the wheel only is in such cases lesser of the two evils (truing the wheel or even buying new one is faster and cheaper than a new paint job if one is sensible to scratches, and certainly better then buying a new carbon frame or being involved in an accident). And some models of safe loops might even reduce the extensive damage to the wheel (by transferring damage from the wheel to the frame instead in there is too much lateral motion - see 4th paragraph of https://github.com/streetcomplete/StreetComplete/issues/5774#issuecomment-2260335448)
I didn't want to add stuff that requires new translations in v59.x while it was already near-release. I think the implementation as contributed to SCEE by @mcliquid makes sense.
But anyway, sorry still that it took so long.
No problem, thanks for adding it! 👍
Use case I feel like I see so many of these types of types of bicycle parking in Sweden but I thought they would just be classified as
bicycle_parking=wall_loops
because it's technically correct. I just recently looked into it and found out this more specific tag that's more specific:bicycle_parking=safe_loops
which makes it more accurate.It's like a wall loop but there is a thing on the side where you can lock the bike frame to the stand instead of the wheel which makes it safer.
Here is a nice image that's public domain: https://wiki.openstreetmap.org/wiki/File:Safe_loops_bicycle_parkint.png , more here: https://wiki.openstreetmap.org/wiki/Key:bicycle_parking EDIT: more images:
Proposed Solution Add it as an option to the bicycle parking type quest,
You might say that the tag is rare but that might be because people didn't bother looking it up and just tagged it as
wall_loops
.