X___

4 hours ago, hooovahh said:

Yeah one thing I wanted to do was only show it when the number of signals were greater than the number of rows shown.  If I have room for 4 signals but only 3 were selected to be graphed, I figured I could hide the scrollbar.  But that isn't true if the user scrolled down previously.  A pain for sure.  Always showing it at least never leaves the user being stuck.  Oh turns out I did actually post my code that attempted to do graph and selection of signals differently here, with a youtube demonstration.

 

I am certainly not going to rewrite a Plot Legend, as there are so many functionalities I am relying on (however buggy they are), that this would be way too much work.

Another interesting tidbit about the Index Display is that if you edit its Description and Tip, the Tip does show up as expected when hovering over the index (and its description in the Help window), but if you have not set a Description for the array itself, that index's description will show up for the whole array (but not the index's tip).

Define both Description and Tip for the array AND its index, and things work as expected.

It's tough to write code that does behave properly in all cases...

At this point I am just showing the index display at all times and hiding it with a disabled button whenever I don't need it.

I am not expecting this to be fixed before 2030 and since I am not planning to upgrade to 2022 and later, that will be it.

As the title says, you can show and hide the plot legend index display (it is an array after all) in Edit mode, but there is no property to do that programmatically.

Since the NXG style plot legend scrollbar is buggy (see this report and the linked thread), providing this index display as a backup for users to navigate the plot list in the absence of a mouse wheel, is not elegant has its visibility cannot be controlled (AFAIK).

The bug has been described a while ago here but might never have been submitted...

It is still present in LV 2021 SP1.

23 minutes ago, ShaunR said:

I don't have a use case for this but I love that the API uses plymorphic VI's-a vastly underused feature IMO.

If you want to organise the functions instead of having a huge linear list, you can group them by separating the menu item with a colon ":".

As described here, I just discovered: https://zone.ni.com/reference/en-XX/help/371361R-01/lvhowto/editing_the_shortcut_menu/

This looks very interesting.

I am curious about your mention of HDF5 support. I am using the h5labview wrapper, which is not really updated anymore. Are you developing a complete support (read/write) of the HDF5 format as well?

Which version are you going to release it in? I am not planning to upgrade past 2021 due to the new NI licensing scheme (can't pay for the hostage fee with our meager grants if our Department decides to drop LabVIEW from their supported toolkit).

Kudos for the project and developers!

2 hours ago, A Scottish moose said:

To his credit it takes a brave presenter to take live comments, especially when they are posted directly behind you. 

Right... Feels like the 100 flowers campaign to me. They will learn their lesson quickly.

1 hour ago, Stagg54 said:

Well NI just solved the distribution and licensing problem didn't they? No one's complaining about that any more.

Just to make it clear to those who did not watch the video (or attend the meeting): these were LIVE inputs from the audience, which were fed directly into the presenter's screen.

But, yes, I don't think the "distribution and licensing" comment was addressed by NI's new licensing policy. But again, he just laughed all these comments off and went on to the next question.

6 hours ago, ShaunR said:

It depends. How do you do it on Linux and Mac?

I am not using Linux or Mac.

4 hours ago, ShaunR said:

The RTF bug is in the save function. You can either save it as UnicodePlainText (lose all the snazzy rendering stuff like images) or get the RTF string and save it with LV file primitives. But your problems are only just starting with the RichTextBox. Like I said. It will get you 80% of the way there.

Mixing text and graphics, as well as allowing style and color, undo and redo, etc. is precisely why I am using the RTB .NET object. Care to enlighten me how you can do that otherwise in LV?

4 hours ago, Rolf Kalbermatter said:

There is a reason that it is still marked Beta (and likely will remain so for the foreseeable future). It is a telltale sign that even the RichTextEdit control, which is a Microsoft technology has problems with that setting.

Basically enabling UTF8 as a codepage feature would be a nice idea, IF all Windows applications were properly prepared to work with codepages that can be more than 1 byte per character. But since this simple assumption of 1 byte == 1 character works for all English speaking countries, there have been many sins committed in this respect and nobody ever noticed. Enabling this feature tries to solve something that can not really be solved since there is simply to much cruft out there that will fail with it (and yes LabVIEW has also areas where it will stumble over this).

Linux is in that respect a bit better of. The Linux developers never were shy about simply abandoning something and put people up with facts and tell them, this is how we will do it from now on. Take it or leave it but don't complain if it doesn't work for you in the future if you do not want to follow the new standard. Most desktop distributions nowadays simply use UTF8 as standard locale throughout, pretty much what this setting would do under Windows. And distributions simply removed applications that could not deal with it properly.

The fix, BTW, was quite simple: delete the offending "u" character in the header (using Notepad ++ or any neutral editor). This fixed the problem with Word or the .NET rtf loading method.

In the meantime, I have switched off the Windows beta function.

On 3/11/2022 at 10:55 AM, ShaunR said:

Just to side-track the whinging completely (most of it mine  ) and to show that it's not as hard as NI make it out to be; this will get you about 80% of the way with UTF8 (UTF8 FTW) on English Windows systems.

Note that the Windows ANSI functions - the underlying calls for the LabVIEW primitives - do notionally support UTF8. Rolf can give a much better explanation of why this works (Codepages) and the pitfalls of using it.

There is no change to the VI, by the way. It is just changing the setting and restarting the OS.

I found clever to try out that setting, only to realize the hard way that it was messing with the RTF .NET support. Namely, saving a .NET rich text box content to rtf generated a file with header starting with {\urtf1\ansi... instead of {\rtf1\ansi..., which prevented reloading the file in the rich text box.

(similar to this error report: https://supportcenter.devexpress.com/ticket/details/t993371/import-from-rtf-richeditcontrol-does-not-correctly-display-specific-characters-when-the)

Not only that but Word would not open that file as rtf, but as plain text (which made it fairly easy for me to spot the culprit).

Buyer beware.

On 3/9/2022 at 7:24 PM, Stagg54 said:

We only had 1 track. Maybe in the future we'll graduate to 2 or more. 

As far as Eric's presentation, a. we didn't have a whole lot to chose from (being our first year and generally a smaller conference due to COVID) and b. as you mentioned a prominent NI employee talking on the future of LabVIEW seemed an interesting proposition. We knew the general topic of his presentation but didn't know exactly how it was going to go. 

Having been there, I haven't watched the video, so I don't know how it comes off in the video, but I thought it was an alright session in person. Looking back on it with all the things that have happened since then, I can see why you might look on it with a negative light. For all the talk of listening to their customers more, they don't seem to be heeding that advice.
 

The problem with that talk is that it says nothing about what the stated goals of NXG were, why the progresses were so slow, what their roadmap looked like at the time they canned it, and what their plans are for the future. Instead, there was a mention of their audit by a bunch of Agile zealots, a free advertisement for online post-it boards and other kind of nonsense about why the lessons they learned (essentially regurgitated slogans) should be taken very seriously by the audience (no kidding). And then there was the now infamous series of botched polls and Q & As which of course they have diligently ignored ever since.

Not that it departs much from what the scripted presentations at NI week used to be (at least the few I watched online) so it shouldn't have been a surprise that it would end up that way.

I am ahoping that those really interested in actual facts managed to ask him those questions later on (if so, please tell us). In my world (academia), the most important bits are gathered that way, by post=presentation discussions. If your format doesn't constrain presenters to be around during the whole series and be available for discussions, this calls for changes.

This being said, thanks for posting those presentations!

On 3/13/2021 at 3:10 PM, X___ said:

It appears that the Green Lady of NI has moved to even greener pastures: IBM - Biographies

Can't abbreviate IBM (I?) but the website could use some of that pastel green and washed out pink palette NI has ambitiously engineered his website with.

Hmm... I guess not anymore:

https://www.forbes.com/sites/martyswant/2022/01/18/ibm-chief-marketing-officer-carla-pieyro-sublett-departs/?sh=18db8f8b4daa

I have no idea why the name Boeing seems to be colliding with that of NI in my mind lately...

You will love this: https://forums.ni.com/t5/LabVIEW/NI-s-move-to-subscription-software/td-p/4215663

In essence: we failed you, but this is because you were not paying us enough, so we will change this, which will be good value for you and us.

What was he saying about denying reality?

7 hours ago, thols said:

Wiebe has some answers here.

Thanks. So it is a bit as I feared. NI dumped a bunch of VIs, briefly mentioned them in Upgrades Notes and that's it.

I have jumped from 2019 SP1 to 2021 SP1 and incidentally discovered the existence of new Multiple Errors VIs introduced in LV 2020 after listening to AQ's presentation at the previous GDevCon (which otherwise went right above my head).

The upgrade notes only say:

Multiple Errors VIs
The Dialog & User Interface palette includes the new Multiple Errors subpalette. Use the Multiple Errors VIs to convert an error cluster into different formats or to manipulate the
attributes of an error cluster.

Digging cursorily into the VIs, I see the use of JSON instead of the ad hoc LV error message syntax, but I am unclear what changes (or not) as far as handling "standard" errors. I am assuming not everything has been rewritten has this would break former code (the Simple Error Handler.vi seems not to have changed and only deals with <append> and <err> syntax.

I couldn't find anything in the examples or much in the Help.

Do I want to add "Attributes" to an error cluster? How can I take advantage of the new features for better error reporting/formatting?

On 12/7/2021 at 12:36 PM, X___ said:

I provided all the logs I could, had a live demo session with no less than 2 support engineers, and they finally decided to give up after supposedly escalating the issue. Unsurprisingly, their last resort suggestion was to upgrade to the newest version (I cautiously stuck to 2019 SP1 to avoid the ripple effects of the pandemic and the NXG fiasco). So if this was due to a bug, the hope was that this undiagnosed bug had been fixed in the latest version...magically I suppose.

Update after upgrading to 2021 SP1: no changes. Back to NI support...

I thought you meant unicorn...

I guess they don't really need unicode support at SpaceX (not that I do myself, but it's more like I never enjoyed the benefits of it, to know that I am missing it).

For those unsuspecting folks who like me are dropping files in a front panel-wide hidden path control to speed up file loading, pay attention to this nasty bug: #1811630 described at

https://www.ni.com/en-us/support/documentation/bugs/22/labview-2021-sp1-known-issues.html

What's kind of ironical (related to the first video above), is that the LabVIEW Chief (or whatever his official title now is), after a contrite mea culpa in which we indeed don't learn anything beyond the fact that they have decided to go from a semi-organized long term chaos to semi-chaotic short term organization model based on the lessons they have not understood... yet? learned from external consultants, announces that from now own they will listen to their customers.

As you noted, in the second video, he perfectly demonstrates what listening means at NI, by having flash polls which he interrupts arbitrarily as people are still trying to understand what the multiple to the power 2 choice questions mean. It is probable that nothing is going to change in the way LabVIEW is developed, except for the TPS report cover sheets...

There we are: Subscription model and related questions. I fail to see any mention of a subscription model though.