02-08-2006 10:30 AM
tst wrote;
"
I have several guesses:
02-08-2006 11:10 AM
02-08-2006 11:10 AM
02-08-2006 01:01 PM
Perhaps my issues are mostly a result of my background, primarily with text-based, embedded and PC environments and inexperience with LabView in general.
I probably represent the most dangerous combination, someone with a lot of experience in a number of langague and environments trying to force my interactions with LabView to fit my text-based paradigm.
But a few points that I think are germane:
The ability to link in pre-built libraries
Since LabView (prior to 8 anyway) ran only from memory,
linking in libraries is probably not an issue. However, I notice that when I load my main VI, it sometimes spends tens of seconds looking for files that it loaded the last time it was launched. Why is that?The ability to run post-processing programs on built systems.
This is generally important when you need to modify a created executable (or hex file in embedded land) by adding serial number, checksums, watermarks, etc.
The ability to run simulations and inject signals.
There certainly is a lot of help available, and yes there is context-sensitive (although the response to the mouse motion can be annoying).
My comments are more along the lines of trying to find things before you know what VI to use.
It seems that I need to spend a lot of time setting up palettes the way that I want, and just do a lot more reading.
LabView has a tremendous amount of flexibility for the developer, but it seems that the flexibility is a compensation, not a bonus.
Still, it is much easier to use (and a lot less buggy) than the LabWindows version 2.2 that I was using ten years ago.
02-08-2006 01:38 PM
I probably represent the most dangerous combination, someone with a lot of experience in a number of language and environments trying to force my interactions with LabView to fit my text-based paradigm.
Yes, I can understand the difficulty in that. Overall, LV has an incredibly low learning curve and is very rewarding to the experienced LV programmer, but those coming into it, especially those needing to make a paradigm shift into the data flow model do tend to find it hard and often produce "graphical text-based code" which looks bad and makes LV look bad.
The ability to link in pre-built libraries
Since LabView (prior to 8 anyway) ran only from memory, linking in libraries is probably not an issue. However, I notice that when I load my main VI, it sometimes spends tens of seconds looking for files that it loaded the last time it was launched. Why is that?I don't think there is any difference in LV 8. Everything still has to be loaded into memory. The only difference is that you can have more than one application instance (something which was available with previous versions, BTW, as I said in the other thread, just not in a project environment). LV code is constantly complied at design time. This allows for immense ease of development. For example, it's what allows LV to immediately break your VI if anything is wrong, even if the problem is 6 levels down.
But to do that, all the code needs to be in memory, so it needs to be loaded. Now, this should take a while for VIs with many hundreds of subVIs, but even my fairly slow and old laptop doesn't take tens of seconds to do that. At most, it will take roughly 20 seconds to load a hierarchy of about 500-600 VIs, and that's just after I started LV and it's still not up to speed.
If LV is really looking for files (something which would take a lot of time) it means that you didn't save the calling VI the last time or that you moved some of the files. When LV can't find a file where it expects it, it will start looking through various places (which can be set through the options menu).
The ability to run post-processing programs on built systems.
This is generally important when you need to modify a created executable (or hex file in embedded land) by adding serial number, checksums, watermarks, etc.
Well, I still don't understand this one, but if you're talking about modifying an LV EXE, I believe this can be done with a resource hacker, although most changes would probably break the file. There are ways to protect and validate LV code (although not built into LV). If you tell us exactly what you want, we can probably help.
The ability to run simulations and inject signals.
That may be one of its purported strengths, but I... cannot get my development system to "pretend" that it has a 7344 card installed.
I don't think that you can do that (although I haven't worked with any of NI's motion cards yet). When needing to simulate hardware, you can usually temporarily replace the hardware VIs with simulator VIs, but that would require designing such a simulator. How would you do such a simulation in a text-based language?
...the response to the mouse motion can be annoying
Can't argue on that one.
My comments are more along the lines of trying to find things before you know what VI to use.
Yes, knowing that things exist and where to find them is tricky because LV comes with many hundreds of built-in functions (not to mention add-ons) and that probably takes time and practice. You might consider taking one of NI's courses (haven't done them myself).
It seems that I need to spend a lot of time setting up palettes the way that I want, and just do a lot more reading.
I never really found any need to set up the palettes. I just changed the express view to the dynamic one and that was enough.
One extremely good way of learning is through the on line communities. The three largest ones are probably this one, the LAVA forums and Info-LabVIEW, all with intelligent, funny people who can help you learn about LV. You can try helping yourself and you will find out that the more you read and participate, the more you will know. Check out the BreakPoint for some fun.
In general, most people who get to use LV regularly get spoiled by it and don't want to go back to other languages later. The ease of development is quite addictive. It just requires understanding how it works.
02-08-2006 07:01 PM
02-09-2006 03:40 AM
02-09-2006 08:52 AM
wiebe@CARYA wrote:
Here is the same VI, but it steps into arrays. I also added "show all" and "hide all" buttons.
Wiebe.
02-09-2006 11:11 AM
I thought that might be the problem... I use Outlook Express, and the attachments show up just fine.
Perhaps now it shows up (I use forums.ni.com now).
02-09-2006 06:13 PM
Hello all:
There are three known issues with attachments:
1 - Cannot attach files larger than 5MB.
2 - If you are posting in the forums, and you preview your post, you will need to readd your attachment before submitting your post.
3 - Files attached using the NNTP newsgroups are not sent to the forums.
These are all still on the list of things that we would like fixed, but I don't have timelines on any of them. The first is the highest priority of the three
Thanks!