Context Navigation

Upload

Timestamp:: 2010-07-04T12:56:10+02:00 (16 years ago)
Author:: pinkduck
Comment:: Improved English text

Legend:

: Unmodified
: Added
: Removed
: Modified

Help/Action/Upload

-              v36
+              v37
 '''Upload new, modified, and deleted objects in the current data layer to the server.'''
 Geo data edited in JOSM is only available locally and only usable by the user who entered the data unless the user decides to make it available to the community too. In order to publish data the user has to '''upload''' it. Uploading means that the geo data is transferred to a central server run by the OSM community where it is integraded with geo data from other mappers in a central database. It's by uploading data that locally edited geo data both becomes part of the [http://www.openstreetmap.org public maps] and can be used by other members of the OSM community.
+Geo data edited in JOSM is only available locally and only usable by the user who entered the data unless the user decides to make it available to the community too. In order to publish data the user has to '''upload''' it. Uploading means that the geo data is transferred to a central server run by the OSM community where it is integrated with geo data from other mappers in a central database. It's by uploading data that locally edited geo data both becomes part of the [http://www.openstreetmap.org public maps] and can be used by other members of the OSM community.
 {{{
 #!html
 <p style="background-color:rgb(253,255,221);padding: 10pt; border-color:rgb(128,128,128);border-style: solid; border-width: 1px;">
 Please be careful with editing and uploading data. When in doubt that everything is ok, rather refrain from uploading and save the area locally on disk to check back to it later.
+Please be careful with editing and uploading data. When in doubt that everything is okay, refrain from uploading and save the area locally to disk, check what you need and try again later.
 </p>
 }}}
 …
 == Changesets, upload requests, and upload strategies ==
 Geo data contributed to the OSM server consists of [wiki:/Help/Concepts/Object nodes, ways, and relations]. Because there are hundreds of mappers publishing their geo data on the OSM server it is important to track who published what data in which context. It would be tedious for mappers to describe for every single node and every single way why it is published. A more convenient approach is to package a group of related objects and assign the package a comment, not the individual objects.
 The OSM project calls such a package a '''[wiki:/Help/Concepts/Changeset changeset]'''. A '''changeset''' is a collection of related changes (new objects, object modifications, or object deletings) applied to OSM data.
 Changesets are different from '''upload requests'''. A changeset is a '''logical''' grouping of [wiki:/Help/Concepts/Object objects] whereas an upload request is a '''technical grouping''' for the purpose of transferring geo data to the central OSM server only. In JOSM we say, that objects are uploaded '''using''' upload requests '''to''' a changeset. The JOSM Upload Dialog allows one to configure aspects of the upload requests to be used and aspects of the changeset objects are uploaded to. Some of the settings are compulsory (i.e. the user has to configure them before he or she can upload geo data) but most of them are optional and JOSM works with reasonable default values.
 Geo data can be retrieved later by the changeset it was uploaded to, it can't be retrieved by the upload request used. Changesets have a unique id, upload requests have not. For instance, [http://www.openstreetmap.org/browse/changeset/3274448 this changeset] includes geo data from tracing over orthofotos of the city of Bern, Switzerland. It isn't possible, though, to see which and how many upload requests have be used to fill it in.
+Geo data contributed to the OSM server consists of [wiki:/Help/Concepts/Object nodes, ways, and relations]. Because there are hundreds of mappers publishing their geo data on the OSM server it is important to track who published what data in which context. It would be tedious for mappers to describe why every single node and every single way is published. A more convenient approach is to package a group of related objects and assign the package itself a comment, instead of the individual objects.
+The OSM project calls such a package a '''[wiki:/Help/Concepts/Changeset changeset]'''. A '''changeset''' is a collection of related changes (new objects, object modifications, or object deletions) applied to OSM data.
+Changesets are different from '''upload requests'''. A changeset is a '''logical''' grouping of [wiki:/Help/Concepts/Object objects], whereas an upload request is a '''technical grouping''' for transferring geo data to the central OSM server only. In JOSM we say, that objects are uploaded '''using''' upload requests '''to''' a changeset. The JOSM Upload Dialog allows aspects of the upload requests and changeset objects to be configured. Some of the settings are compulsory (i.e. the user has to configure them before geo data can be uploaded) but most of them are optional and JOSM works with reasonable default values.
+Geo data can be retrieved later via the changeset it was uploaded to, though it can't be retrieved by the upload request used. Changesets have a unique identifier, upload requests have not. For instance, [http://www.openstreetmap.org/browse/changeset/3274448 this changeset] includes geo data from tracing over orthofotos of the city of Bern, Switzerland. It isn't possible, though, to see the upload requests that were used to fill it.
 === One upload request into one changeset ===
 …
 [[Image(one-changeset-individual-object-upload-request.png)]]
 Please note that this is a legacy feature. Until recently, this was the only upload option in OSM.  Technically, it uses a slightly different approach to communicate with the OSM server. It is still available in JOSM although the OSM server now supports upload requests with up to 50'000 objects. For the casual and normal user it is almost obsolete. In rare cases, it could be useful for power users.
+Please note that this is a legacy feature. Until recently, this was the only upload option in OSM.  Technically, it uses a slightly different approach to communicate with the OSM server. It is still available in JOSM although the OSM server now supports upload requests with up to 50,000 objects. For the casual and normal user it is almost obsolete. In rare cases, it could be useful for power users.
 === Uploading large datasets into multiple changeset ===
 JOSM also supports uploading large datasets which don't fit into one changeset only.
+JOSM also supports uploading large datasets that don't fit into a single changeset.
 [[Image(multiple-changesets.png)]]
 This is an advanced option which only is useful for power users. They can select this configuration if they have to upload a dataset with more than 50'000 new, modified, or deleted objects.
 == Choosing your upload strategy - when and how often  to upload ==
+This is an advanced option which is only useful for power users. They can select this configuration if they have to upload a dataset with more than 50,000 new, modified, or deleted objects.
+== Choosing your upload strategy - when and how often to upload ==
 There is no simple answer to the question when and how often to upload. You neither should edit the whole day and then upload everything at once nor upload every minute after adding a way.
 Here are some rules of thumb:
  * '''Time required to upload''': the smaller the upload requests you choose the longer it takes to upload. It takes more time to upload 100 objects with 100 individual upload requests than to upload them in one request including 100 objects.
  * '''Collisions with other mappers''': if you upload 30'000 objects in one request and if the server encounters a problem on the 29'999th object the whole upload is rolled back. One has to fix the problem (i.e. by resolving a [wiki:/Help/Concepts/Conflict conflict]). The whole 30'000 objects will have to be uploaded again. On the other hand, if you upload 30'000 object with 10 request à 1000 objects and if the server encounters a problem on the 29'999th object then you only have to repeat the last upload request including the 29000th to the 30000th objects. Object 1-28999 are already successfully uploaded. If you're mapping in areas where other mappers are active too you should therefore prefer smaller sizes for upload requests.
+ * '''Time required to upload''': the smaller the upload requests you choose the longer it takes to upload. It takes more time to upload 100 objects using 100 individual upload requests than to upload them in one request containing 100 objects.
+ * '''Collisions with other mappers''': if you upload 30,000 objects in one request and if the server encounters a problem on the 29,999th object the whole upload is rolled back. The problem has to be fixed first (i.e. by resolving a [wiki:/Help/Concepts/Conflict conflict]). The whole 30,000 objects will then have to be uploaded again. On the other hand, if you upload 30,000 objects with 10 requests containing 1,000 objects each and the server encounters a problem on the 29,999th object then you only have to repeat the last upload request for the 29,000th to 30,000th objects. Objects 1-28,999 are already successfully uploaded. If you're mapping in areas where other mappers are active too you should therefore prefer smaller sizes for upload requests.
 == Uploading data  ==
 …
 If the [http://wiki.openstreetmap.org/index.php/JOSM/Plugins/Validator Validator plugin] is installed, JOSM performs a quick check on the data. If it finds warnings or errors you'll be presented with a list of them and get the chance to correct the problems and try again or upload anyway. It's not necessary to fix all of that, but please don't go over it easily, especially not the errors.
 JOSM also checks for limititations imposed by the OSM server:
+JOSM also checks for limitations imposed by the OSM server:
  * tag names and tag values must be shorter than 255 characters
  * ways can't consist of too many nodes
 …
 If the objects you're uploading have cyclic dependencies (relation 1 refers to relation 2, relation 2 refers to relation 3, relation 3 refers to relation 1) JOSM can't upload them. JOSM will ask you to break up these dependencies first.
+'''Note''': On large dataset these checks can take some time and for the time being they don't provide user feedback. Please be patient if it takes a couple of seconds to launch the Upload Dialog on large datasets. This will be fixed soon.
+'''Note''': On large datasets these checks can take some time and for the time being they don't provide user feedback. Please be patient if it takes a couple of seconds to launch the Upload Dialog on large datasets. This will be fixed soon.
 === The Upload Dialog ===
 …
 If you want to upload to an existing changeset you have to select one of the available changesets from the drop down list. This list is only enabled if there are open changesets which you can upload data to because you own them. Click on [[Image(source:/trunk/images/dialogs/refresh.png)]] ''Refresh'' to load the list of available open changesets from the server. If there is at least one open changeset available you can select the radio button '''Upload to an existing changeset''' and select a changeset.
+Click on [[Image(source:/trunk/images/closechangeset.png)]]''Close Changeset'' to close the currently selected open changeset. This is for convenience only. You don't have to close a changeset here in order to run a succesful upload.
+Click on [[Image(source:/trunk/images/closechangeset.png)]]''Close Changeset'' to close the currently selected open changeset. This is for convenience only. You don't have to close a changeset here in order to run a successful upload.
 After a successful upload JOSM can either close the changeset used or leave it open for another upload. The default setting is to close it. You can configure whether JOSM should leave it open by unselecting the checkbox '''Close changeset after upload'''.
 …
 Select '''Upload data in one request''' to upload all object in one request. If you're uploading a medium to large number of objects this might take some time and JOSM will not be able to inform you about the progress. There are only two outcomes of such an upload: the upload either succeeded or it didn't. In the former case everything is fine, in the later no objects at all have been uploaded. This kind of upload will never lead to a partial upload which is both its strength and its weakness. It can be its weakness if the entire upload fails because there is a problem in the very last object in the upload request.
+Select '''Upload data in chunks of objects''' to upload the objects in a sequence of requests with a predefined size. You can enter a preferred size for an upload request (called the "chunk size"). Depending on its value JOSM will submit a number of upload request (the number is displayed on the right of the input field for the "chunk size"). In contrast to the former option, every upload of a chunk can either succeed or fail. You will get some coarse grained progress feedback because JOSM will inform you when an individual chunk has been uploaded successfully or when its upload has failed. Uploading in chunks is in most cases slower than uploading in one requests, though.
+Select '''Upload data in chunks of objects''' to upload the objects in a sequence of requests with a predefined size. You can enter a preferred size for an upload request (called the "chunk size"). Depending on its value JOSM will submit a number of upload requests (the number is displayed on the right of the input field for the "chunk size"). In contrast to the former option, every upload of a chunk can either succeed or fail. You will get some coarse grained progress feedback because JOSM will inform you when an individual chunk has been uploaded successfully or when its upload has failed. Uploading in chunks is in most cases slower than uploading in one requests, though.
 Select '''Upload each object individually''' to upload each object with an individual upload request. You will get very fine grained progress feedback because JOSM will inform you when an individual object was uploaded successfully or when its upload has failed. Note that this option is in most cases the slowest option.
 '''Recommendations'''
+ * For small (<1000 objects) to medium upload sizes (<5000 objects), choose '''Upload data in one request'''
+ * For medium to large upload sizes (> 5000 objects) choose '''Upload data in chunks of objects'''. A chunk size of 1000 is a good value to start with.
+ * For small (<1,000 objects) to medium upload sizes (<5,000 objects), choose '''Upload data in one request'''
+ * For medium to large upload sizes (> 5,000 objects) choose '''Upload data in chunks of objects'''. A chunk size of 1,000 is a good value to start with.
  * Don't use '''Upload each object individually''' unless you have a specific reason to do so. This is basically a legacy strategy from former versions of the OSM server.
 === Running a very large upload ===
 An upload is considered to be very large if its size exceeds the maximum size of a changeset on the OSM server. Currently, the upper limit for the size of a changeset is 50'000 objects and if JOSM detectes that your upload is bigger it displays the following information in the Upload Dialog:
+An upload is considered to be very large if its size exceeds the maximum size of a changeset on the OSM server. Currently, the upper limit for the size of a changeset is 50,000 objects and if JOSM detects that your upload is bigger it displays the following information in the Upload Dialog:
 [[Image(large-upload-basic-settings.png)]]
 …
 [[Image(large-upload-advanced-settings.png)]]
 For very large uploads, uploading in one request isn't possible and the respective option is therefore disabled. Please select a chunk size to be used in the upload. It has to be smaller than 50'000, too, because the upper limit for the size of a changeset also applies to the chunk size. You can upload a very large dataset with an individual request per object but you're not recommended to do so.
+For very large uploads, uploading in one request isn't possible and the respective option is therefore disabled. Please select a chunk size to be used in the upload. It has to be smaller than 50,000, too, because the upper limit for the size of a changeset also applies to the chunk size. You can upload a very large dataset with an individual request per object but you're not recommended to do so.
 A very large upload doesn't fit within a single changeset. Please select
 …
 === Uploading to a closed changeset ===#ChangesetClosed
+If you upload to an open changeset, the upload may fail because the changeset has been closed in the meantime. It could have been closed by the server because of a timeout (changesets don't stay open for more than 24 hours) or because you closed it explicitly in another JOSM instance.
+If you upload to an open changeset, the upload may fail because the changeset has been closed in the meantime. It could have been closed by the server because of a timeout (changesets don't stay open for more than 24 hours and are also closed after an hour of inactivity) or because you closed it explicitly in another JOSM instance.
 JOSM displays the following error message:
 …
 === Changeset becomes full during upload ===#ChangesetFull
 There is an upper limit for the size of changesets. On the OSM server it's currently set to 50'000 objects per changeset. An upload may exceed this limit while uploading, mainly if you are uploading a sequence of upload requests to the same changeset.
+There is an upper limit for the size of changesets. On the OSM server it's currently set to 50,000 objects per changeset. An upload may exceed this limit while uploading, mainly if you are uploading a sequence of upload requests to the same changeset.
 If JOSM detects that a changeset became full it displays the following warning message:
 …
 === Deleting node which is still in use ===
+=== Deleting nodes that are still in use ===
 {{{
 #!html