When runnning my Coded UI Tests through a VM (deployed via OctoDeploy incase that matters) 2 of my tests will run properly. The other 44 however do not. I am presented with the exception:
Microsoft.VisualStudio.TestTools.UITest.Extension.UITestControlNotAvailableException: The Internet Explorer window is not available.
When looking at the output, it shows that it passes 2 steps, which the second step involves clicking a link. It states that it passes this step, and when I watch it physically, it doesn't look like it does. (I even attempted to take control real quick and click it, but it did nothing).
Has anyone else had these problems?
So looks like I found the answer to this through different means entirely. It was due to the security of the VM and I had to enable Java applets. Sorry for this everyone.
Related
I am using Application Insights in Azure very intense but very often I experience loading issues.
So many times when I want to inspect the logs, I find the following loading animation taking forever:
The interesting thing ist, that nothing is really "loading" in the background. You can see my dev tools here:
So waiting forever does not help - its not "slowly", it just does not load at all.
None of my http requests seems to be pending... is there anyone experiencing the same behaviour? Or could I have misconfigured anything?
I get this all the time.
Press F12 (in Chrome) to open Dev Tools.
Hold left click on the refresh icon in the tool bar.
Click 'Empty cache and hard reload'.
The transactions will now load.
I am getting "Not Connected" issue while executing a blue prism object. Currently, I'm working for browser application and I have given browser URL and window title in application modeler.
If I launch the application by navigation stage it's working fine but I'm expecting like already running browser we need to attach and has to do some automation operations on it.
We tried with navigation stage with actions attach/Active application, getting issue "Not connected".
Can anyone help on this issue?
Blue Prism best practice is to call “Attach” page at the beginning of each action to make sure that BP is connected to the application. This page should look like this:
Start
Read stage – to check connection status (it will return Flag with connection status)
Decision stage – connection flag
If connected – go to End
If not connected – Navigation stage – Attach
You also need to provide proper input parameters for Attach action, for Browser automation (Internet Explorer) you need to provide at least process name: “iexplore” & child index: 0 (it depends on your TabProcGrowth registry settings).
Blue Prism also recommends to start browser apps using Utility – Environment VBO & Start Process action. There are 2 inputs: Application – Path to internet explorer exe file & Arguments – webpage address.
If you want to learn more about Browser Automations in Blue Prism I recommend you to familiarize with Blue Prism Guides available on Blue Prism Portal - https://portal.blueprism.com/.
Do you have Attach bookmark created? When you are working with object you are creating bookmarks resposible for different things in your obcject(login, start etc) you need to create Attach bookmark.
I never had any problems with attaching to app, except mainframe app.
Copy your Navigate step from Launch bookmark, then create new one(call it attach) and paste that Navigate step. Then Edit it, Chane from launch to attach and in Properties you need to fill "Window Title" and "Process Title". Window title, you have it at the top window bar. Process title, youll find it in Windows Task Menager - Processes.
Hope it helps!
Well, I am not sure exactly sure if this is your problem but let me try as my issue is close to your issue.
I had encountered this: There is no current stage
So what I did was to trigger the reset button in the upper left corner of the Process Studio. Screenshot below.
Then it works like a charm!
This is because the launched instance of the application is different from the one BP is searching for. To solve this issue close the current instance of the app and Launch it from the Application Modeller.
I've 4 monitors connected to a pc running CentOS 7. Each physical monitor/display should show a website in full screen and auto refresh it every x seconds.
I've tried:
opening from bash and moving with xdotool -> how can I distinguish between the 4 chrome or firefox windows? If I don't do it and apply to the last active this might become unreliable with other programs open.
researched devilspie, seems it has the same problem (can't distinguish between multiple chrome/firefox windows)
using DISPLAY variable -> does only find one display 0.0
https://addons.mozilla.org/de/firefox/addon/monitor-master/ -> does not work with extended workspace, contacted developer
I've only a xrandr bash script to rearrange the displays when they become connected (three of four may be switched off to save power).
I would use a firefox/chrome add on for the auto refresh functionality.
Maybe someone can give me a hint how to reliably do this?
In addition to the "easy auto refresh" chrome plugin. I have another idea for you, why not use a chrome extension for handling the positioning.
It is of course, only for chrome(maybe thats fine for you).
Background: We had related difficulties. Internal webapp that opens multiple documents in windows, and need to be placed in other monitors.
The javascript does not support this, for security reasons and only a native extension can properly work with the tabs/windows objects.
Therefore, we have created an open source chrome extension for doing exactly that: flexible windows position across multi-monitor setups.
In your case you can define for each monitor a website-rule that the window would appear there.
The chrome extension is called "MultiWindow Positioner" and its complete free. You can get it at the chrome store here
The actual source code you find in github in the project chrome-multiwindow-positioner
Disclaimer: I am the maintainer of the open source (MIT) github project. If there any interesting idea, or comments feel free to share them here.
Our products coded in Designer are usually structured as:
database1.nsf,
database2-production.nsf,
database3-devel.nsf,
then, all of them use the same template:
database-template-verX.X.ntf
Agents I code, are coded in NTF. After ntf is ready to be moved to NSF files, I usually just do Application > Refresh design. Everything works well. Now here is the thing...
When refreshing design and checking domino.log, usually newly created agents work great on:
database1.nsf,
database3-devel.nsf
However,
- database2-production.nsf doesn't seem to get agents properly. Even If I open NSF file directly in Designer, everything looks good. The agents just don't run. All agents are ENABLED, have proper permissions and are signed. That -production.nsf database just doesnt want to cooperate.
Usually we solve this with restarting domino and even then, sometimes we arent sure if agent will start running as scheduled or not.
Did someone else had similar problems or maybe knows what is wrong? It looks like there's some kind of cache or something..
We have no idea...
Any help would be great!
Thank you in advance!
You can use tell amgr debug * to have the agent manager output detailed information to the console. Unfortunately, if it seems that restarting the server solves the problem temporarily, you may have to collect debug information over a long period of time to see a before and after picture of what's going on -- and even then it might not point you toward the reason.
Check all the design elements property, Prohibit Design refresh or replace to modify should not be checked. Also check your acl for this database.
I know this might be a no-brainer, but please read on.
I also know it's generally not considered a good idea, maybe the worst, to let a browser run and interact with local apps, even in an intranet context.
We use Citrix for home-office, and people really like it. Now, they would like the same kind of environment at work, a nice page where every important application/document/folder is nicely arranged and classified in an orderly fashion. These folks are not particularly tech savvy; I don't even consider thinking that they could understand the difference between remote delivered applications and local ones.
So, I've been asked if it's possible. Of course, it is, with IE's good ol' ActiveX controls. And I even made a working prototype (that's where it hurts).
But now, I doubt. Isn't it madness to allow such 'dangerous' ActiveX controls, even in the 'local intranet' zone? People will use the same browser to surf the web, can I fully trust IE? Isn't there a risk that Microsoft would just disable those controls in future updates/versions? What if a website, or any kind of malware, just put another site on the trust list? With that extent of control, you could as well uninstall every protection and just run amok 'till you got hanged by the IT dept.
I'm about to confront my superiors with the fact that, even if they saw it is doable, it would be a very bad thing. So I'm desperately in need of good and strong arguments, because "let's don't" won't do it.
Of course, if there is nothing to be scared of, that'll be nice too. But I strongly doubt that.
We use Citrix for home-office, and people really like it. Now, they would like the same kind of environment at work, a nice page where every important application/document/folder is nicely arranged and classified in an orderly fashion
I haven't used Citrix very many times, but what's it got to do with executing local applications? I don't see how "People like Citrix" and "browser executing local applications" relate at all?
If the people are accessing your Citrix server from home, and want the same experience in the office, then buy a cheap PC, and run the exact same Citrix software they run on their home computers. Put this computer in the corner and tell them to go use it. They'll be overjoyed.
Isn't it madness to allow such 'dangerous' ActiveX controls, even in the 'local intranet' zone ? People will use the same browser to surf the web, can I fully trust IE ?
Put it this way. IE has built-in support for AX controls. It uses it's security mechanisms to prevent them from running unless in a trusted site. By default, no sites are trusted at all.
If you use IE at all then you're putting yourself at the mercy of these security mechanisms. Whether or not you tell it to trust the local intranet is beside the point, and isn't going to affect the operation of any other zones.
The good old security holes that require you to reboot your computer every few weeks when MS issues a patch will continue to exist and cause problems, regardless of whether you allow ActiveX in your local intranet.
Isn't there a risk that Microsoft would just disable those controls in future updates / versions ?
Since XP-SP2, Microsoft has been making it increasingly difficult to use ActiveX controls. I don't know how many scary looking warning messages and "This might destroy your computer" dialogs you have to click through these days to get them to run, but it's quite a few. This will only get worse over time.
Microsoft is walking a fine line. On one hand, they regularly send ActiveX killbits with Windows Update to remove/disable applications that have been misbehaving. On the other hand, the latest version of Sharepoint 2007 (can't speak for earlier versions) allows for Office documents to be opened by clicking a link in the browser, and edited in the local application. When the edit is finished, the changes are transmitted back to the server and the webpage (generally) is refreshed. This is only an IE thing, as Firefox will throw up an error message.
I can see the logic behind it, though. Until Microsoft gets all of their apps 'in the cloud', there are cases that need to bridge the gap between the old client-side apps and a more web-centric business environment. While there is likely a non-web workaround, more and more information workers have come to expect that a large portion of their work will be done in a browser. Anything that makes the integration with the desktop easier is not going to be opposed by anyone except the sysadmins.
The standard citrix homepage (or how we use it) is a simple web page with program icons. Click on it, and the application get's delivered to you. People want the same thing, at work, with their applications/folders/documents. And because I'm a web developer, and they asked me, I do it with a web page... Perhaps I should pass the whole thing over to the VB guy..
Ahh... I know of 2 ways to accomplish this:
You can embed internet explorer into an application, and hook into it and intercept certain kinds of URL's and so on
I saw this done a few years ago - a telephony application embedded internet explorer in itself, and loaded some specially formatted webpages.
In the webpage there was this:
Call John Smith
Normally this would be a broken URL, but when the user clicked on this link, the application containing the embedded IE got notified, and proceeded to execute it's own custom code to dial the number from the URL.
You could get your VB guy to write an application which basically just wraps IE, and has handlers for executing applications. You could then code normal webpages with links to just open applications, and the VB app would launch them. This allows you to write your own security stuff (like, only launch applications in a preset list, or so on) into the VB app, and because VB is launching them, not IE, none of the IE security issues will be involved.
The second way is with browser plug-ins.
For example, skype comes with a Firefox plug-in, which looks for phone-numbers in web-pages, and attaches special links to them. When you click on these links it invokes skype - you could conceivably do something similar for launching your citrix apps.
You'd then be tied to firefox though. Writing plugins for IE is much harder than for FF, I wouldn't go down that path unless forced to.