Best Answers

Hi Folks, I'm having a brain-freeze - how do I convert a line to its constituent vertices resulting in an output point dataset? Logically the GeometryCoercer should do it, but it doesn't.  CoordinateExtractor with a loop could probably do it, but that seems massively overkill for what should be a simple task. Is there a transformer I'm missing?  Thanks, Jonathan

Hi , Tried running few workbenches in built in desktop 2012 built on 2013 had macro issues hence translation failure.Had few question regarding interoperability and compatibility of desktops and servers.  1.Is there a tool/way to migrate a 2012 workbench to 2013 ?  2.If we use a 2013 fme server would that seamlessly support 2012 workbenches?  3.Eager to implement transaction support for oracle spatial ,is there a suite for the same in desktop 2012 ?

Hi all,  is there a way to make sure that data from a shape file transfered to a gdb file using FME have the exact same coordinate values?  I happen to have to rounding happening in some features. This means that when I compare features from the shp and the gdb feature class, there will be some geometry discrepencies flagged up even though there should be.  I don't know when the rounding occurs (i.e. when it's written the first time to the gdb or when it's read the following time from the gdb)  Any idea? Anything I can set on the GDB writer or FeatureReader to make sure this doesn't happen?  Thanks in advance,  Maryse

Why does Fme workbench  is still running after Translation was successful message is displayed in the translation log window ? 

This should be simple... I'm using FME 2013 SP1.  I'm reading values from Excel.  When I encounter any alphabet character, I want to replace the full value with a null.  I'm fiddling with a StringReplacer and using a regular expression [a-zA-Z], but not sure what I can put in the Replacement Text.  My source data in column looks like this:  2013-05-31 00:00:00 Today Day 0 Day 1 Day 2 Day 3 Day 4 Day 5 Day 6  I want the date string to remain in the column.  I want the "Today" and "Day x" values to be converted to null values...  Sarah

This should be simple but I’m ready to throw up my hands.  I have a workspace that fetches an OpenStreetMap chunk of XML for a requested area.  I can save the data and then read the data using the FME OSM reader, a two-step process.    What I want to do is pass the OSM data it to the OSM reader in a single workspace.  I guess I could call the OSM fetcher in a startup Python script but that seems more convoluted than it should be.  Am I missing something obvious?  Thanks, Alan 

We recently upgrade to FME Desktop 2013 sp1 Build 13450 and we are noticing significant decrease in time of two workspaces that are doing a simple point on area overlayer process of our county address point with several polygon features.  The one process took 12-15 min in 2012 and is now running in 55 min.  The second one was running in 10-12 and is now running in 35 min.  I tested utilizing the area first option in the new transformer in 2013 but that did not help.  I put a ticket into Safe Support but no  response yet so I thought I would see if anyone else is noticing this type of performance downgrade.    It is simply taking several SDE FC doing the operation and spitting out to SQL Table in one and shapefile in the other.  Any help or insight of what might be going on and what I can do to help this get back to the performance we were seeing would be much appreciated.  Thanks Travis 

Hi,  I want to use a Dynamic ORACLE SPATIAL writer in UPDATE MODE. But when i try to update a row through this, two extra attributes 'ORIENTATION' and 'TEXTSTRING' appends in the query so that when i run the workbench oracle throws an 'invalid identifier' exception and the process stops.  Can anyone tell me how to use a writer dynamically to update a row in oracle spatial  database. Thanks in advance.

Hi guys,   I am trying to migrate some .dwg file to ESRI raster/symbols.  1) when i use inspector on the .dwg reader, the colors for polygon or hatches are missing from the drawing.  May I know how resolve this?  2) How should i migrate .dwg file to geodatabase raster dataset or symbols?

I am using the trial version of FME desktop to see it's suitability to my uses. One of my tasks is working with Mentum Planet grds which are Mapinfo raster grds. It appears FME cannot input these grds natively? Is there an easy way to load a grd into FME to work with it then output it?  I'm completely new to FME.

Hello, is there any way how to write compound polygon (lines + arcs) to Oracle with FME? When i try to write this kind of geometry i always get "ORA-13347 Description:the coordinates defining an arc are not distinct". But arcs seems fine to me (no redundant points, distance between them is larger than oracle tolerance).   FME 2013 (Build 13261 - WIN64) WIN 7 64bit Oracle 11gR2   Thanks for any advice, David  

Using the CSV reader. Encountering an extremely annoying problem ... the columns by default get named col0, col1 .. etc. up to col20 and are output in the same order which is fine EXCEPT that col9 is added at the end and not where it should be ... between col8 and col10.  Can't see how I can re-order.  Any ideas? Am I doing summat wrong.

I would like to transform a 3D multipatch SHP into a 2.5D SHP whereby the attributes for the 2.5D file are the calculated elevation and slope characteristics from the multipatch.    I have tried using the 2DForcer transformer and the BoundsExtractor calculator, which allowed me to hook the maxz up to the BLDG_HEIGHT attribute.  I had some limited success with this configuration.  However, it would seem things come across at strange scales, there are of course no slopes, and in general the it looks like the resulting buildings are just smashed together faces.    Does anyone have any guidance or workflows that you can share to transform 3D multipatch SHP into 2.5D SHP?    Thanks!

Hi there, I am trying to migrate Oracle dump datas into gdb. The reader reads every column except raw format type and giving a warning "skipping column". Has anybody have any idea?

Hello,  I started using FME 2013 a couple of weeks ago and I am having problems in importing arcpy module in a startup python script. Every time I try to run the process I get this message:  "Python Exception <ImportError>: No module named arcpy"  I checked the following article in FMEpedia and set the Path Environment Varible but the problem continues. On the other hand, I don't belive that I am having the second problem described in the article since I don't get the error dialog box.  http://fmepedia.safe.com/articles/Samples_and_Demos/Using-a-Python-Startup-Shutdown-Script-to-Perform-Geoprocessing-with-Arcpy  Anyone have an idea of what could be happening?  Thanks, Gabriel

Does anyone have any examples of reading an SDE feature class with fmeobjects in Python?  I was able to do it with a shapefile without any issue but am having trouble figuring out how and where to specify the parameters for the GEODATABASE_SDE format.  I am using an SDE connection file if that makes it any easier. However, whenever I try to open() the reader it pops up the SDE connection properties dialog. I get the feeling I'm just not specifying the reader parameters/directives correctly, but without finding any examples of the proper format I can only guess.  Thanks!

Hi All,  I've been working all day on the following issue (fme2009). I have a small custom transformer to log some information to a seperate log, this logging is done using a pythoncaller. I want to use this custom transformer on several places in my model, and retreive custom messages using published parameters.  How do I access these published parameters, I know that within custom transformers I should use FME_MacroValues['transformer name_published parameter name'], however the transformer name is the name of the transformer in the main model, and as such always unique.  Does anyone have suggestions to solve this issue?  Paul

Hello,  I am currently using a feature type reader with the Merge Feature Type filter * - wilcard to load a reader from command line, ex:    --folder all --data fullcoverage (the path formed with publish params: C:\\all\\fullcoverage\\area.shp What I would like to do, is to allow a user to choose between a default reader (let's say, the YES choice, which would be the above example) or to allow him to insert a completely new path (the NO option). So, if the choice is YES, it will take him on a branch, if the choice is NO, on a different branch. It should work somehow like this: "--choice yes" -> would load the default reader C:\\all\\fullcoverage\\area.shp "--choice no" (should require the "--path" param where the user could indicate the path for a completely different shapefile) which should create a reader like C:\\cities\\test\\dummy.shp if the command is: "--choice no --path C:\\cities\\test\\dummy.shp" I hope I stated t

Hello,  I am quite new to FME and I am curious to know if there is a way to change the reader(s)/writer(s) using an interactive prompt box. To be more specific, in my projects I have 3 readers: ca/sacramento/address ca/sacramento/roads ca/sacramento/county Each time I go to another county, I delete the readers and bring the new ones in: tx/wise/address tx/wise/roads tx/wise/county I would like to make a variable %path% which should get a value from the prompt box (or command line) so in the end I have a general fme template which I can use by only changing the %path% once. In this way, maybe I cand run the program from the comand line, specifying the two parameters (state, county) and just let it run  

Hi  I have a shp file with a numeric attribute.  I can sort the attribute table on this attribute using a Sorter, but how can I then filter out only the top, say, 1000 features?  Thanks  Dag

Hi,  May I know how can I find all the feature class that has not been committed after a run?  Yes I know how to save those feature that have error/ignored by setting the writter but there are those that are miss out because they are not commited anywhere.  I do not wish to manually route them to a database or shape files because it can be quite a tedious task.  May I know if there is anywhere to say all feature classes not saved to be commited into a file or database table?     

Hi, I got a translation going on with few sql executors and a workspacerunner .A sql executor is placed after Workspace runner as well.Now a strange thing is observed when I set the "wait for job to complete" as yes the oracle procedures are called from executors and are reflected in the table. But when set to no ,none of the procedures seem to work though the log shows them being called . Is this a bug ?

so i have a point cloud (lots of them actually) in a LAS file and all of the data is unclassified.   I am wanting to extract the ground heights from the LAS file and then use those height values to create a DTM.  my thinking is to create a 1metre grid over the top of the area covered by the LAS file (or get FME to do this for me?) and then find all points within that 1m grid cell and assign the lowest Z value to the grid cell.  then once each cell has the lowest Z values i can use that grid to create a DTM. I know there will be cells that have tree tops etc in them and therefore the lowest Z value will be much higher than the ground, how to fix this i am not sure as i do not believe FME has convultion filter style transformers availabe?  so is my approach a effcient way? or is there other ways i can go about doing this? any other ideas on how to solve the inevitable issue of cells with no ground points in them?  any help would be apprecaited.

Has anybody had issues with the Reprojector? I'm converting some shapefiles from EPSG:4617 to CANA83-10TM115 into a GeoMedia SQL Server Warehouse.  I'm getting the error   "Reprojection from 'LLCSRS' to 'CANA83-10TM115' requires grid shift file(s) for datum transformation" Reprojector: Could not create reprojection engine    Any thoughts? 

Hi All, i'm quite new to FME and i have to work with FME Work Bench 2010 to migrate the Landbase data(Map Info TAB files) to Smallworld PNI. I have to migrate the smae set of data to different databases, for this every time i'm performing the same tasks like add the readers and writers, mapping etc. My question is, to avoid this is there any way to create a standard template file so that i have to change only the input file(reader) and output dataset based on the data(writer) or we can automate this process? Please help in this regard. I want to avoid performing same task again and again so that i can save the time. Thanks in advance.