All Activity

Yes! It is so much better now. Excellent work!

I was relieved to login this morning and not see pages worth of spam when I clicked on the "Unread Content" link. Awesome job to all!

@hooovahh Is still weeding out the spam. I think he's in the eastern US time zone so he's 3 hrs. ahead of me ☺️. Much thanks to him. But I'm also improving the filters. Unfortunately, I think there are some sleeper accounts that were created before the changes that are starting to post. But, yes, I think it's getting much better. BTW, I just discovered that if you ctrl+right click a posted image you can set its' size! neat.

It seems to start to work! Unless of course you have hired some elves that keep checking the forum and deleting any account created with nefarious motives. 😀 But it is quite some time that I came in the office in the morning and didn't get greeted by a stream of products of humanities lowest when opening LavaG. In fact there seems none, and the latest new member is already 21 hours ago and still has 0 accounted posts for, so he may be actually genuine. Just hope that the moderators new message queue isn't overflowing with held back posts. Although you can always hope that it will die down once they realize that it has exactly 0.0% reach factor.

If you have a controller in the PXI chassis it can either run LabVIEW RT (used to be a PharLaps derivate but they are now switching over to something built on Linux) or Windows. If you want the PXI system to run as an embedded system, and if you need any real-time capabilities, then LabVIEW RT is the way to go. If you don't need that I do not suggest running Windows on the controller. We have a system where we do that and unless you really don't have space for a rack computer or some other external PC I don't see any advantages. What you get is basically a more expensive computer with worse performance. Just connecting to the PXI system with an MXI link is much better (which we do in all our other systems) and if I understand you correctly that's already your idea.

Says the account with "AI" right in the name. Hiding in plain sight! eta: In fact, you can't even pronounce it without saying "AI" - "A I va lee oh tis". Well, I can't, anyway...

I noticed that this morning. However, I'm adjusting some knobs behind the scenes. There will still be some that get through and I will be monitoring the forums for the next few weeks to optimize the settings.

Unfortunately, it seems that the site upgrade did not fix the spam issue. Are there any new options at your disposal?

I've had to disable all external services used to login to LAVA such as Google, Facebook etc. If you were using these services and now cannot login. Please send an email to s u p p o r t (at) l a v a g (dot) o r g with your login email address and I will reset your password so you can use the built-in login method. This is a permanent change moving forward. Sorry for the inconvenience.

Hello I am working with USRP and GNSS Test Toolkit 3.0 to generate GNSS signals. As I transmit the simulated signal through the USRP, I have difficulty getting anything on my GNSS device(no satellites visible/no location fix). What might be wrong, especially about the generation parameters/configuration? What should be the TX power level? Are there any other aspects that I need to consider? Description of hardware setup: I am using NI USRP-2974 with the default streaming project, where I read the GNSS binary file created using GNSS toolkit 3.0 and transmit the IQs at a rate of 10MS/s (read from the binary file) at a center frequency of 1602 MHz using an omni antenna which is appropriate for the L band. All other USRP TX settings are at default values. I have tried attenuators ranging from -50 dB to -90 dB. My GNSS receiver/test device is a surveying GNSS tablet with a GNSS antenna. Configuration of the GNSS test toolkit: I have used niGLONASS Write Waveform to File (Simple, Automatic Mode).vi from the GNSS toolkit 3.0 with the default values (including the almanac and ephemeris files provided with the toolkit) I will be looking forward to hearing from you. Best regards,

Hello. I am not a bot... I'm planning on taking the site offline this weekend to perform long overdue upgrades and to investigate ways to curb the spam attacks. Thanks to everyone for all the help cleaning up the forums. Hopefully I can find a solution and we can get back to the usual next week.

China's LabVIEW users are indeed more, most of them are concentrated in coastal areas, I think it may have something to do with China's huge manufacturing industry, factories inside a lot of test software are developed using LabVIEW.

And yet, for this year, Google shows peak monthly search interest for LabVIEW occurring in China. Weird. The following does not represent total search interest, but peak interest listed by Google in any given month this year: Here's the data from the tree map above: 1 China 100 2 Taiwan 96 3 Tunisia 74 4 Switzerland 74 5 South Korea 67 6 Ecuador 58 7 Armenia 57 8 Mexico 57 9 Singapore 56 10 Germany 54 11 Slovenia 52 12 Lebanon 47 13 Austria 42 14 Hungary 35 15 Denmark 33 16 Norway 33 17 Czechia 32 18 Israel 32 19 Finland 31 20 Costa Rica 31 21 Ireland 30 22 Colombia 29 23 Sweden 28 24 Belgium 27 25 Romania 27 26 Malaysia 26 27 United States 25 28 Netherlands 23 29 France 23 30 Italy 23 31 Hong Kong 23 32 India 22 33 Poland 22 34 Peru 22 35 Canada 21 36 Portugal 21 37 Japan 20 38 Pakistan 20 39 New Zealand 19 40 Morocco 19 41 Spain 17 42 United Kingdom 17 43 Slovakia 16 44 Thailand 16 45 Russia 15 46 Algeria 13 47 Serbia 13 48 Australia 12 49 Iran 12 50 Greece 11 51 Egypt 11 52 Venezuela 11 53 South Africa 10 54 Vietnam 10 55 Philippines 9 56 Chile 9 57 United Arab Emirates 7 58 Brazil 6 59 Ukraine 6 60 Türkiye 5 61 Argentina 5 62 Indonesia 5 63 Saudi Arabia 4

Google search engine not available in China😅

Because of a thread over on the darkside, I got the motivation to improve this code, and include the Google Material icons in it. I posted the package over on VIPM.IO. This uses the native 2D picture control for displaying icons like I wanted. It still requires Windows due to how icons are resized, but maybe that could be worked around if there is interest. https://www.vipm.io/package/hooovahh_boolean_vector_controls/ Install the package and its dependencies and you'll have a Tools >> Hooovahh >> Boolean Control Creation... Once ran it will start trying to display all the icons the toolkit installed. In the background it will be converting the vector images to 56x56 PNGs to be able to display them in the window. I tried being smart and having it prioritize icons that you scrolled to, but I honestly don't know how well it works. It basically takes about a minute after first launching it to have all of its icons displayed properly. You can use the tool during that minute but not all the icons will be available yet. From that point on you can scroll around and resize the window and it should work as expected, just a little bit slow at times. There is a single constant on the block diagram where you can change the icon side. At one point I had icon size be a control on the front panel but since it took about a minute to process all the images for every change I just left it. Some of the icons have multiple versions. If you left click on an icon and a window pops up you can pick from what version of that icon you'd like to use. Then create a control using that icon. You can theoretically put your own EMF files in the folder with the rest but at the moment it doesn't scan for new files since it is relatively slow to find all icons on every launch. What I'm saying is compromises had to be made. Maybe I could have a separate program that gets ran in the Post Install VI that starts processing the icons right away in parallel. That way the tool might be done processing icons by the time the user launches it for the first time. I did use the Post Install and Post Uninstall to do extra work since there are so many individual files. Normally you'd have VIPM handle the files but it took a long time. So the package just installs a Zip, and the Post Install will unzip them. This also means Post Uninstall needs to delete the extracted files. Not ideal but the install time was much longer otherwise.

Completely agree. I wonder how the number of people with internet access affect the data. In the past the percentage of technology-agnostic people was probably higher on average than nowadays. Especially for niche topics like LabVIEW. It also makes a difference whether you look at global data or just a specific area. For example, interest for LabVIEW in China and USA are somewhat distinct: China USA There are also distinct changes to the slope when Google decides to improve their categorization systems. We may also see a sudden decline for languages with the adoption of AI assistants as most developers don't need to search the internet anymore and instead rely on the answers produced by the Copilot integrated in their IDEs. Maybe this is what's happening to the Python chart?

I was interested in downloading it to have a look. However, the download page will not accept an email address from providers such as gmail, hotmail, etc.

Hi, LogMAN – Thanks for sharing. This is an important point. One thing I've noticed with Google Trends is to be careful about selecting their suggested search term categories. For example, "Python" vs. "Python Programming language" graphs look very similar; however, "C++" vs. "C++ High-level programming language" look very different. Maybe it's the addition of "High-level" in the case of C++. Either way, I like to look at the raw data so we're not prematurely filtering. We can intuit some things from the shape of the graphs. For example, are many more people searching for the python snake vs. the programming language today vs. 20 years ago? Are people searching for "C++" for some other reason than the programming language? Keeping these things in mind and looking at the raw plots, while risky, is at least a "known" risk. What is "unknown" is how Google is deciding when a "C++" raw search gets categorized as a "High-level programming language". And, here is the raw graph for LabVIEW, C++, Python, and C. As you can see, C++ and C follow a similar trajectory, which hints, to me at least, that we're not talking about "the letter C". Again, it's raw data with no categorization by Google, so who knows: Looking at how LabVIEW disappears from view compared to C++, C, and Python, umm yes? This is why I wanted to develop TestScript. It's ~3,000 lines of Python code that lets us abstract sequencing logic away from LabVIEW. Effectively making LabVIEW a manual-control-screen builder. Again, I'm not there anymore but the tool is solid, in my biased opinion, and free: https://www.genuen.com/products/testscript For context, C++ was released in 1985, LabVIEW 1986, and Python 1991.

These statistics can be very misleading as they show interest over time and cannot easily be compared. For example, this is a comparison between LabVIEW (blue), Simulink (yellow), and Node-RED (red): Here is another comparison between PHP (blue), C++ (red), C (yellow), JavaScript (green), and C# (purple): Relatively, each of these languages shows considerable less interest over time and converge towards a similar value. Unfortunately, we can't get the full picture as we do not have earlier data. Otherwise, it would probably show pretty much the same curve as Python does. We are just further along the curve (the hype is over 🙁). For the sake of completeness, this is the graph for Python: While everyone is clearly interested in Python, this doesn't mean that any of the other languages is suddenly going to be obsolete or non-functional. People (and businesses) will always follow the hype train, of course. Though, we should certainly not fall victim to ignorance as clearly the LabVIEW community is not quite as enthusiastic as it used to be. I wonder if we should pitch a Python port of LabVIEW to NI... 😈

Oh that's hilarious 🤣

There should be plenty of work for LabVIEW pros for a long time to come maintaining all the code that's out there, especially as newer developers pass over LabVIEW for Python. AI just makes coding in Python soooo fast. I'm no longer at Genuen, but we did make a LabVIEW-Python connector for those who would like to cross the streams (e.g. run Python scripts that control your LabVIEW manual control screen). It's solid. I'm attaching a screenshot from Keywords Everywhere so you can see some numbers on search interest. Yes, Google has adjusted the way it reports this a few times, but fwiw, they have Python heading the other direction.

One other odd side effect is I've seen several reports that are clearly bots, that think it is a reply system. The report will be a normal spam message but reporting some content. So at least those ones only I need to deal with and you won't see. Sorry again about all this. I have talked to the site admin again emphasizing the severity, and frequency.

Using an old SFX installer is likely the case. I don't recall deleting anything, but my C:\National Instruments Downloads directory is empty.

Well, yes. There is only so much we can do on the client side. It also appears that some of the bots have recently figured out how to reply to their own posts. It is probably just a matter of time before our notification areas get bombarded...

Yeah, but when the picture are suddenly in the "Unread Content" stream, it makes it a lot harder to ignore.