Share your ideas ,thanks!
1.What kind of report do you use in your application? I think HTML and TXT don't have a good view, XML needs a stylesheet which is annoying(I know there is a XML package tool, but it still doesn't work when users only want to pick up sveral pcs of report, and they don't to copy the stylesheet at the same time). I think generating the report myself by parsing the ResultList at the end of the sequence, but obviously it will take a lot of time for every single project.
2.OK,let’s say that there is no ideal solution for question 1. I save the report by “NewFileNameForEachUUT”, how to MERGE all report files to one file for data analysis(other than database)?Does NI have an APP to merge XML/HTML .. reports?
1- You hit it right on the head with the style sheets. We had that issue and so we created our own Plugin to generate an HTML report that then gets rendered to PDF (you wouldn't need to generate your own HTML first, you could use the existing plugins and then render to PDF). I suggest you read up on Plugins. If designed correctly you won't need to do it for every automation you create. You will only need to do it once and have it be part of your framework. http://zone.ni.com/reference/en-XX/help/370052N-01/tsfundamentals/infotopics/pmpcreatingplugins/
2- Why are you against databases? In addition to our HTML/PDF plugin we also write all of our data to a database. It is easier to analyze this way. Ultimately you could just use Access which generates one file. However, you are limited on how big it will get. You have the same problem with your option as well (unless you are using Windows 64 bit) where you merge the files into one really large file. I'm pretty sure NI doesn't have a tool for this. At least I've never seen one.
Hope this helps,
Thanks for your attention!
1.Process model plug-ins is new to me.I need to read it.By the way,is it easy to deploy plug-ins to a target PC which has only TestStand-RunTime installed?
2.I'm not against database.But there're several things I'm not very clear about.When I save data to database(.mdb),I have to put a database template along with the sequence release,TestStand seems to be unable to create a MDB file itself, right?Second ,on the deployment PC,sometimes there is debugging time that all test data is useless in that time.With a database,I have to manually disable the database logging so not to pollute the formal test data.I don't know whether there is a smart way to do it,but if I use XML/HTML.. report,I will save it by DATA\PASSFAIL..,so it is easy to sort/remove useless data and keep the whole test data clean.And also in this way ,I don't have a very big MDB file if I only want to analyze some DATA's data.Some deployment PC don't have mdb application either.
1- There is a little more effort when deploying plugins. But it is trivial. If you develop them in the TestStand public folder then you just need to check the From TestStand Public Directories checkbox on the System Source tab of the TestStand Deployment Utility. Everything else is done for you.
2- I understand all of your concerns here. .mdb is already installed for you by TestStand. All you have to do is simply turn on Database logging and you are good to go. No configuration required. No extra deployment required. As for debugging, in your case you would just need to remember to turn Database logging off and on before and after you debug. In our process model we created a different entry point called Debug UUT. In this entry point we turn off all of the reporting and logging. So when our technicians and engineers are debugging they run with that entry point. Something to think about. It was actually very simple to implement but we did need to deploy some extra things.
Hope this helps,