Jump to content

NateJames

Members
  • Content Count

    20
  • Joined

  • Last visited

Community Reputation

0 Neutral

About NateJames

  • Rank
    member
  1. Hello 360Works, I am looking to refresh my current database into a new cleaner version. During this process I am looking to modify the primary keys so that I am not using user friendly serial numbers (example invoice ID's). To do this requires some work on my end, and I have consulted the advanced topic section regarding Mirror Sync Managed vs Developer managed. I imagine that developer managed keys are faster than MirrorSync managed. Is there any data to support my assumptions? It would also seem that there maybe other reasons that you would want to use developer managed over MirrorSync managed, for instance I am thinking that using multiple sync configurations might benefit on this setup. Can you elaborate more on this as I have not had much success with multiple sync configurations and have not adapted to using this in my solution. OR Is there a possibility that there is going to be a new version of MirrorSync that will use internal ID's (field and record's) and I shouldn't invest to much time into doing this (I am transitioning a file from FMSP 4.5 to FMSP 7)
  2. Caveat to that method that is not talked about very well... if you are using SQL filtering using say the MIRRORSYNC_USERTOKEN. The server file does not know what the MIRRORSYNC_USERTOKEN is example is I am filtering timesheets to only give the user there timesheets and no other. than my sql statement is where (id_staff = '{MIRRORSYNC_USERTOKEN}' ) but the initial sync on server will never know the token and well it doesn't work than. Fishtech, I have struggled half the morning with coming back to trying and setup multiple configurations for one file. I tried 3 configurations in my file and did not have the same tables in any of the configurations my 3rd sync was a one way with 2 tables (products and product images). I was able to do one initial 2 way sync, and than tried the one way product sync (initial). It failed miserably. I would suggest staying away from multiple configurations until MS tells us otherwise.
  3. I can answer the second part. During setup you provide a name for the second sync configuration, so mirror sync will append that sync name to the new sync scripts. So you will have to paste the additional scripts into your file as normal. In my testing it doesn't seem to matter which one you distribute for your download link, because your triggering 2 different sets of scripts. For your first question I have struggled and have had some good discussions and have multiple trial and errors to setup exactly what your looking to do. MirrorSync imho really needs to clarify having multiple sync configurations on one application. One technician says you can't have the same tables in each sync configuration, but in the advanced setup it says you can. I too want to set up exactly what your saying and my lack of understanding and having to infer what is going on has been painful. I can tell you that I tried to be clever and have my one way sync run at startup. And this is BAD (unless you create a skip on first run step to bypass this first initial one way sync). My one way sync was filtering the user data, and once completed the 2 way sync was no longer functional and would fail. (If you run the 2 way first which also has the same filters because its a "full" sync, than I have had no issues with running a one way sync (spoke --> host) later on). 😀One thing to share with you also, that I recently caught and makes sense. If you have large attachments and containers and those records get modified. Your users end up having to redownload the container. That might help your field guys if your solution uses a lot of containers, and for example you have a photo gallery full of cat photos and user manuals...Then you may be downloading them all the time
  4. Joshua, At one time I thought there was a caveat to the full sync configuration and partial sync configurations setup that's referenced below. They way I understood it was that this would work, but its false to call it a 'full sync'. That your full sync could not contain the tables that your partial sync was using. Or is the trick that the full sync has to be both directions and the partial sync has to be one way only. I have wanted to implement this feature but during initial setups it didn't work well for me so just left it alone and stayed with a full sync. Thanks Nate
  5. I have considered it, and tested that option. It takes 10 minutes do to the products and image files associated with the database. But because MirrorSync is so awesome, i have been testing the setup with a pre - populated and different mobile file. By different file I mean a clone of the original, and than i deleted all the tables i do not want the user to have (i.e. invoices). I left everything else within the database, and than setup a synch relationship using the separate mobile file option (Please see the other topic, and problems I had setting that up). But i can report that it does work nicely and I do not get any gaps in the serial numbers on the office version anymore quite simply because we dont sync it. The other tables can elevate the serial numbers as it is not critical for an accounting auditing purpose. In the future I will definitely follow Tim and your advice and solution
  6. again, I haven't pasted anything yet, What tables are we talking about, is this the crazy ones that i mentioned earlier on ?
  7. so to finish the sync setup you need to not select "sync with a separate mobile file" but instead go through like a normal setup So than I click edit on the config engine and proceed to walk through to make the change to a separate mobile file when you hit next your are presented with this screen once you get past that and setup your relationships etc ( i left that alone but added my sql filter i was presented with this option to past the script steps into a file. but what file as i am given misleading information -- i am going to assume the iPad copy and i haven't pasted anything yet so why is it checked Green? when you hit next again this is the new screen
  8. ok so this is what i was saying happens on a clean configuration, same file duplicated, invoice and invoice line items removed including layouts all tables, scripts, and layouts of mirror sync have been removed before uploading to the server. After selecting "sync with different mobile file" this is the window i get. Interesting enough is that none of those items exist on my iPad configuration. I purposely deleted them earlier on and if you hit next to skip past that
  9. http://docs.360works.com/index.php/MirrorSync_advanced_topics#Can_I_pre-populate_my_offline_database_to_speed_up_initial_sync.3F so i tried to do this part today ok so one thing would be nice. when you are making reference to the different files and you need to be working in one vs the other, can you bold the file name and change the colour. That would be nice to distinguish between the file you need to be working in When i deleted all the records in the new duplicate file, i also reset all the serial numbers back to default on every table but the one table i was keeping populated (products has 12000 records and images so this is what i wish to ship out with the data already intact). I had some major issues doing this with a new configuration, so i ended up doing a new configuration synching to the same file in order to be able to finish the configuration. I than went back and hit edit the config so that I could select 'sync with a separate mobile file' (remember it wouldn't let me go any further by selecting this as default). I managed to get all the way to the end, jumping between files first the desktop version, and than to the duplicate file. Last step it was asking me to past tables into my duplicate file and to copy the script step into the duplicate. I could get a green check mark for the script steps, but i could not get the paste tables to achieve a green check mark (it remained with a triangle). The tables that where being pasted in made no sense (please see attached photo) and because i could not get it to give me a green check mark i hit finish anyways. On sync everything looks okay (except some conditional formatting that used primary keys that are now mirror sync managed and not the same value), and these new pasted tables remain at a zero records and never used. Not sure what to think about that. Tables are sorted by creation date, so you can see that after pasting in the mirror sync table, that I ended up pasting the other tables above it
  10. Thank you for the reply, And in the future when I build something I will definitely go that route you suggested, its great advice Unfortunately for this solution that we are using that would mean going through all the relationships and changing out the primary keys, including all the scripts etc etc to change the dependancy on that user number. (we have a heavily modified version of FMSP). We have used the write back solution on the estimate module and it works incredibly well for our offline users (for the very reason you suggested it being a user visible serial number), keeping everyone on the same page when discussing the estimates in progress and keeping track of the offline versus online file. What are your thought on actually keeping data off the offline file. If you give the users a full copy than they get everything at once (we do this as mention and than do a first sync to remove the invoice we don't want on there device, causing the elevated invoice numbers). (full offline file is around 400mb ) if we give them an empty file it takes 10 minutes to sync all the products and estimates and staff etc that we do want them to have (empty offline file is 40mb) But the user never gets an invoice record, there is no sync_invoice layouts, and the invoice ID will than never elevate Have you had experience with pre-populating the data, (personally I am a little uncomfortable to always have to make that additional file all the time any major changes happen, and than having to remember to delete all the records and such. Seems like a pain I guess what I am getting at is that I understand why mirror sync creates these gaps, what I don't understand is why mirror sync doesn't take corrective action to fix what there tool does to the file. Do you know what happens if change the numbers back myself ? Does that affect the Apache tomcat database keeping track of the sync, because the last thing I want to do I cause sync issues to a live file that we are using to run our business on. Thanks webko for your insight Nate
  11. http://docs.360works.com/index.php/MirrorSync_advanced_topics#Configuring_MirrorSync_creates_gaps_in_serial_numbers Hello Mirrorsync So I understand now why mirror sync creates these gaps in the serial numbers, ( I originally asked for the clarification on why this does this). However what happens to the internal database if I go back to these serial numbers and adjust them to be the next sequential number. Or better yet, why doesn't the sync tool on setup capture what the next sequential number should be, do its testing and than write back to the database the captured value so our files do not end up with these sequential gaps? I need you to consider what happen if a company gets a visit from the taxman during an audit and they question comes up why the gaps in invoice numbers. Where invoices deleted, is the company hiding something etc,etc. The other fact is let's face it we are not pro mirrorsync users when we set this tool up, we make mistakes, we don't understand the relationships correctly, we add something to tables we want to include. All these things we do as we are learning greatly accelerate these serial numbers within our file every time we run the sync configuration. I personally have an estimate table that only has 700 records, but serial number wise because of mirror sync is showing my next serial number as est1200. In my learning of the tool I recently realized that when giving my remote users a full copy of the database, even though I was not syncing my invoice table with them, the remote user did In fact have all the invoices from date of them downloading there working copy. I have now learned to make the script run on startup, checking if the device has an active connection or not, and remove all invoices from there device, but, from now on anytime I run the configuration I am also going to elevate those invoices numbers. And yes I realize that I can distribute a different file, and not include those tables in the iPad version, but now I have to maintain 2 different files if we go that route. As I said I am very new to the FileMaker world, and maybe there is something I don't understand still about why this happen, and maybe it's critical to the internal database integrity. Can I change the numbers back? OR, if I can ask for anything this upcoming Christmas from 360works..... Thanks again Nate
  12. Hello Jesse, From a previous thread Started by OceansWest So I wanted to test MS by setting up a duplicate configuration one using Java, the other xml. So this new testing database was setup --(which I used as a duplicate from an existing in use configuration...i deleted all mirror sync tables, layouts, scripts and external references before setting this up). Step one was to modify an existing jdbc setup, reinstall new scripts, layouts external reference etc etc. step 2 was to duplicate the test configuration and modify to using xml When using the configuration tool, I selected "duplicate" a configuration that I just setup for my JDBC. I renamed the configuration so MS would create a new set of scripts, and proceeded with next. At the point of recognizing that i needed to paste the new script steps into my file, the program would not paste the script steps in. So it would seem that perhaps duplicate is not the best feature to use because i could not paste the new script steps in, or because maybe duplicate doesn't quite work the way I think it does. Can you elaborate when we can use the duplicate function? I was hoping to test the speed on 3.1 with synching an empty file, as there are items that i don't want the field users to have, and the only way is to give them an empty file or create a synch and filter Update --- so after closing the duplicate file, and than reopening the duplicated configuration, it seems that was necessary for the config engine to "re evaluate" my synch configuration. (Much like when you want to test the SQL statement on the tool, you have to finish the config, and than open it with edit, navigate to the sql setup, and than MS will recognize what your testing and provided results). Anyways after doing that I was able to past the new xml based synch config scripts into my test file (Note to anyone else doing this, you also need to change to xml for your synch and finish all the way through, mirror sync will not create the relationship needed for the primary keys, and will not find the relationship until you complete the setup and than re edit the sync, than the tool will be able to setup the proper relationships as need......FYI)
  13. I am glad I was able to help. So thinking about your solution, you might not need to always create a new duplicate tables and layouts for your solution. I used to do the same thing all the time as well, but realized very quickly that Filemaker gives us all sorts of creative ways to filter the data we see. (Disclamer...Thats not to say for other reason that you want to have the data separated, in which case its a great idea to do that). But you could create a "flag" field that is set if the contact is a donor =1. Than you could do a script step that would filter all contacts if flag_donor = 1. This would return only the donor contacts. You can than have fields that are specific to being a donor hidden if that flag does not = 1. Conditional formatting on the banner for colour could also be changed on that flag =1 value to give these a visual cue that they are working with donors vs non donors. Nice thing is what ever choice you decide it comes down to the work you wish to invest into it and the time you need to have it up an running. If you ever go do the path of offline synching i mention, than you may be very happy to have made the decision to create a new table and new layouts, (and from memory, the video does a great job of walking you through doing that) Just remember that all the internal links will also have to be fixed on the tabs to reference the new tables you are making as well, so it becomes a bit more work to keep everything functioning as a nice neat package.... Cheers and happy filemaking Nate
  14. Hello bioplasma, What you are going to need to do is create a new table to contain the data that will be related to your donor. its under file/manage/database on the tables table you will need to make a new table "Units" now you need to add fields within this table you will need id_contact (or what ever you called your new donor contact id) as one of the fields. its also a great idea to have the fields, created and modified date and timestamps and possibly the id_staff (if you are using that module for the staff entering the data). This will allow you to relate the data to other modules within FMSP. in the relationships tab, you will create a link from the former contact table "ID_contact" to the new table you created above "Units", linking to id_contact (this is a field within the new table). Notice the naming convention that RCC uses in the relationship table (its called anchor bouy). This creates the relationship of one donor to the new table. (you will have to drag the new units table over to the contacts module. expand both the anchor table and the new "unit" bouy table" and than connect the 2 id's together). Give this new relationship a name (see the notes above in the relationship module to see how its named). go to your donor layout and enter layout mode. pick a portal tool and when you draw it onto the layout you will have to point the data source to the new relationship you created above. This is explained way better in the video training from RCC in video #1104 What is a portal... and a one to many Relationship. at this point you have a little more work to do to get the data to work together............ or (and this is only my opinion if you would have used the accounts module and modified that to be your "donor module" instead, the products module is already related to the accounts. Repurposing the products module will allow you to see all the products--"donations" that a particular account--"donor" has in the system already. The products module will assign a unique serial number to each donation. And the created timestamp in the "products-donations" module could be used to reference when the last donation was made. (just a thought but you could repurpose contacts to be a family/next of kin contact information). Welcome to the wonderful world of filemaker, you did yourself a big favour on purchasing FMadvanced to start with. I am a newbie myself going on 1 year of using FM (so if I didn't explain it very well above I apologize), and its been a great experience (so much I went to Devcon this year). Biggest thing I did to help myself was purchase the video training series from RCC after watching a few free ones. The best thing is that Richard takes apart this very file and shows you how to modify it, explaining the process as he goes along, its well worth the dollars. And disclaimer alert, I do not work for Richard Carlton, I have my own business that is using FMstarting point and I am thrilled on how it works, including the offline mobile synching that we do with it. (think taking your app and having iPad users fill out the forms, and than pushing all the data into one server for all users to have available with them when you have an internet connection again...its very cool) hopefully I didn't confuse you more Nate
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.