Verdy_p's Comments
| Changeset | When | Comment |
|---|---|---|
| 90644017 | about 5 years ago | En plus ctout ce que tu as fait quelques cjours avant c'était le changeset/90539071 où:
|
| 90644017 | about 5 years ago | Non je n'ai pas dit ça. Et le seul ton désagréable a été le tien justement, dès la première intervention, avec un ton très impératif et professoral pour me dire de ne pas faire une chose ou faire autrement sans donner de solution applicable, comme si j'avais fait une prétendue erreur (pire volontairement) pour quelque chose dont je ne suis absolument pas responsable puisque c'est ici toi qui n'a pas fait le travail approprié pour mentionner ta "source" (pas une seule note). Il faut bien voir aussi que les commentaires de changeset ne sont jamais visibles à l'édition: les objects sont trop nombreux, et même en les interrogeant un par un on a une série de changeset: ces commetnaires ne sont pas faits pour qualifier un objet individuel.
|
| 90644017 | about 5 years ago | C'est bien de le mentionner maintenant mais RIEN ne l'indiquait (et même si c'est le cas, ne pas le dater ne permet pas de savoir quoi comparer).
|
| 90644017 | about 5 years ago | Moi je peux comparer en plein d'endroits je trouve l'inverse, mais il n'y a strictement aucun moyen de savoir quand la source a été mise à jour, car la date affichée n'est que la date globale et pas la date de chaque cliché: les orthophoto proposées sont presque toutes des assemblages de photos réalisées sur des périodes différentes, puisqu'aucun plan de vol ne peut couvrir le monde entier. Au mieux ça peut porter sur quelques kilomètres carrés, et la BDOrtho proposée ici est en fait pas directement la BDOrtho mais comprend aussi des mises à jour depuis les plans de vol de Rennes Métropole ici: il y en a bien plus souvent que ceux qu plan national BDOrtho, et Maxar Premium en a encore moins souvent; il peut donc arriver (et il arrive très souvent) que l'un soit donc plus en avance que l'autre localement, ou l'inverse; ce n'est pas disponible sinon on verrait les bordures et la géométrie (orthomodifiée) de chaque cliché, et on aurait une série de dates et pas une couche unique.
|
| 90644017 | about 5 years ago | c'est le problème: il n'y a AUCUN moyen fiable de dater les clichés sur les sources d'image: si une route ou un chemin est supprimé, le supprimer d'OSM ne sert à rien, tu imposes un travail inutile aux autres. Il vaut mieux le garder avec un attribut de cycle de vie, et mettre une note datée décrivant la source incorrecte et celle correcte.
|
| 89137456 | over 5 years ago | Deux "contributions" et deux abus ! Merci de ne pas mettre n'importe quoi dans OSM, c en'est pas un terrain de jeu ou de promo d'idées personnelles (surtout quand elles sont insultantes ou illégales) ! |
| 89218078 | over 5 years ago | Deux "contributions" et deux abus ! Merci de ne pas mettre n'importe quoi dans OSM, c en'est pas un terrain de jeu ou de promo d'idées personnelles (surtout quand elles sont insultantes ou illégales) ! |
| 88211516 | over 5 years ago | Don't be abused by the name "Rio", it does not necessarily mean it is a "river".
|
| 88176811 | over 5 years ago | Anyway postal adresses are not necessarily geographic (see also the recuring debates with "addr:*=*" fields, also not all geographic; and then non-geographic postal addresses better fit in "contact:*=*" fields) |
| 88176811 | over 5 years ago | @cquest, you reiterate what I said and summarized "In summary, what applies to printed postal addresses does not apply in OSM" |
| 88176811 | over 5 years ago | Ys for Wikipedia, they were wrong, but the debates there were highly biased a few influencers with too much power, insisting only because this convention was used in some specific databases or applications that would otherwise not treat addresses correctly (even if printed postal addreses never contain any hyphen, even after "Saint(e)" where they should be replaced by spaces, and capitalized without accent, a preconisation in fact abandonned for printed addresses using clear fonts with decent sizes and no decorative features, but still recommanded for handwritten addresses or when manually filling paper forms with a pen with irregular letter shapes that could be incorrectly read) |
| 88176811 | over 5 years ago | Oh thanks, I effectively meant "ther'es never been any hyphen (not "space") between the leading article and the rest, or between the generic and the specific name".
|
| 88176811 | over 5 years ago | In summary, what applies to printed postal addresses does not apply in OSM. We don't need these transforms for default "name=*"; we may jsut need "short-name=*" if possible abbreviations are not common and basic truncation of long words (without predefined abbreviations) could break the interpretation and could become ambiguous. |
| 88176811 | over 5 years ago | in a language read easily by the French poste; this last line with the country name has no use when delivered later in that country so it does not change the total limit of 5 lines for the rest of the printed address. |
| 88176811 | over 5 years ago | Note also that you are not required to write the country name on a fina line if this name is a "city-state" whose name is usually alreayd written at end of their local address: "MONACO", "SINGAPORE", "VATICANO"; provided it is written in Latin and capitalized when posting from France (if you ant to write Singapore in Chinese, then repeat it in Latin on an additional line).
|
| 88176811 | over 5 years ago | @trial: I meant this in the street name.
Printed postal addresses are special, and FANTOIR contains some additional fields capitalized that exhibit the suggested non-ambiguous and capitalized abbreviations of street names (also without accents or cedilla). We don't want these altered names in OSM (except possibly in "short_name", but this is not needed if all what is abbreviated is the generic, like "RTE" instead of "Route" or "BD" instead of "Boulevard", as these abbreviations are predefined in a wellknown dataset, fully documented.
But the last line with the postal code must remain below the limit so the municimality name may be abbreviated in addition to being capitzlized and written without accents/punctuations (it may be followed by an additional line for the country name when posting from another country, capitalized "FRANCE", as the convention of prefixing the French postal code by "FR-" is not universal in other countries, even if it could avoid an additional line; for example in it not understood when posting from outside Europe, notably countries where the Latin script is not well understood: using "FRANCE" in capital only makes this clear for the local postal services)
|
| 88176811 | over 5 years ago | Well, the hyphens in French toponyms is not a strict rule. It used to be a rule for communes, but no longer for "communes nouvelles" (the hyphens are kept in their "communes déléguées").
|
| 83083550 | over 5 years ago | Note that the boundaries for "Kalrsruhe" are very simple, even these two:
|
| 83083550 | over 5 years ago | And it's definitely NOT more complex thant any typical international boundary which also requires care and lot of work to keep them connected! These bays were not so complicate (in fact their resilution is much coarser than any typical international boundary or any existing coastlines also used for these boundaries and other administrative subdivisions. They did not "pollute", they represented very low amount of data compared to all the rest, and they did not cause major issues in renderers (given that seas have very low density of data compared to everything else on the borders and lands) |
| 83083550 | over 5 years ago | This is a very massive deletion, even if all was sourced from international documents. The reason given does not even hold on OSM: things are always "difficult" to edit everywhere, and here this is jsut destruction of work made by many contributors, jsut because ofg a single user abusing his rights to do what he wants, without even consulting anyone and without notice... |