LabVIEW subscription model for 2022

[rolfk]

---

@Jay14159265 wrote:

Deal or no deal?

It's listed at ~$14K for that P/N from one of their distributors:

https://www.newark.com/ni/776678-35/labview-software-professional/dp/15AJ5392

-AK2DM

---

It's a little more than 3 times the price before subscription started. That's quite a price increase even if you consider the high inflation of some 10% per year in these two years. Of course what really happened is that when subscriptions were introduced they were priced just a little below the previous perpetual license price and since that, this price has gone up substantially like most other NI products.

Re-introducing perpetual licensing poses a problem. You can't put it at the same price as a one year subscriptions, as there would be no incentive to purchase a subscription in that case. So you have two options, substantially lower the price of the subscription to have a reasonable perpetual license price, or put a moon price for the perpetual license in the price list that makes other professional software packages out there look like cheap pancakes.

It's obvious that NI hasn't been able to decide for the first yet and doesn't feel like publishing moon prices either. Seems to me that they got themselves stuck between a rock and a hard place here somehow. Not sure what it takes to remove that heavy rock. 

- LabVIEW is not innovative, it was 20 years ago but it is no longer. Ever since subscriptions started the year to year new features have predictably turned to a trickle. 

---

You should get the facts right here when you say something like this: The list of new features have been relatively low since about 10 years before the subscription started. At that time the main reasoning was that everything will be better in NXG and they put all their effort in there. After subscription started that didn't improve despite the "improved ongoing communication with the user base" that subscription was supposed to allow. It's only now that NI kind of seems to wake up out of a deep sleep in their cave and realize that LabVIEW could be actually a flagship product that many other companies would love to have.

[ooth]

Could someone explain to me why there is the aversion to have a model that is a perpetual license with a SSP? I mean the subscription option is $2750/yr. Why can't NI just offer a perpetual license with a SSP model at that price instead of making it SAS. My company would be on board to pay that every year and I could view/edit my code forever if sometime in the future my company decides to stop paying.

Or is that the rub, that NI wants the guarantee that I will never stop paying because I'll lose access to my code?

[crossrulz]

---

@rolfk wrote:
It's only now that NI kind of seems to wake up out of a deep sleep in their cave and realize that LabVIEW could be actually a flagship product that many other companies would love to have.

---

In my opinion, you can thank Emerson for that.  We all thought NI being bought out was a horrible thing, but I am currently of the opinion that Emerson is turning NI around.

[LLindenbauer]

---

@Jay14159265 wrote:

Further reading: 

- I learned a good/hard software architecture lesson with LabVIEW; never align your team with a tool that requires a license (even a free one) to see your own source code. 

---

This is by far the biggest point for me.

I do not understand why this was seldom talked about, even before the recent rough times.

Also, this is not about money. Even with the perpetual or SaaS licenses, users are still banking on the licensing services staying available forever (and/or their setup never failing) and future versions staying compatible.

NXG would have gone with the general industry trend of moving away from proprietary, undocumented formats.

Its cancellation came just at the moment when I was looking into migrating all important code over.

I assume that releasing documentation on the binary format is also impossible, at the very least due to legal reasons

[rolfk]

---

@LLindenbauer wrote:

---

NXG would have gone with the general industry trend of moving away from proprietary, undocumented formats.

Its cancellation came just at the moment when I was looking into migrating all important code over.

---

I heard that often, but it is still not true. Just because you put something into XML format doesn't mean it's not proprietary anymore. Sure it is text and it is a little easier to guess the meaning of elements in an XML file format than in a binary file format, but that doesn't make it suddenly non-proprietary.

The same about a text format like XML making version control software easier. Any reasonably complex XML format is not simply diffable with standard text diffs. The diff algorithm needs to understand the schema of the XML file in order to make consistent diffs. Otherwise it is more likely to corrupt the actual data in the XML file than anything else.

Most diff algorithms out there only work for sequential text programming languages where it is safe to assume that the structure of the programming language is line by line organized.

[Kevin_Price]

---

---

It's a little more than 3 times the price before subscription started. That's quite a price increase even if you consider the high inflation of some 10% per year in these two years...

Re-introducing perpetual licensing poses a problem. You can't put it at the same price as a one year subscriptions, as there would be no incentive to purchase a subscription in that case. So you have two options, substantially lower the price of the subscription to have a reasonable perpetual license price, or put a moon price for the perpetual license in the price list that makes other professional software packages out there look like cheap pancakes.

It's obvious that NI hasn't been able to decide for the first yet and doesn't feel like publishing moon prices either. Seems to me that they got themselves stuck between a rock and a hard place here somehow. Not sure what it takes to remove that heavy rock. 

I don't have a lot of widespread exposure to the industry at large, so I'm mostly influenced by remarks here in the forums and my own reasoning.

1. The SAAS model and its pricing was a really big mistake.  A lot of dept managers saw it as a greedy cash grab and a lot of developers saw it as something akin to ransomware.  Plenty stopped considering LabVIEW to be a viable option and have now moved on, unlikely to return.

2. I remain perplexed at the disconnect between the luxury pricing and subscription lock-in strategy on the one hand and the Community Edition zero cost of entry for hobbyists (but not degree-pursuing students) on the other.

3. The pre-subscription pricing level was *already* high enough to inhibit growth of the user base.  Schools seem to have migrated heavily toward Python, judging by the skillsets of recent graduates.  They're ready to be at least moderately productive immediately for free.   It's increasingly hard to "sell" the value proposition of LabVIEW to the not-yet-initiated due the high upfront cost of both the dev environment and the typically long learning curve to achieve payback-level productivity.

4. When the subscription model was first getting reactions here, I kinda figured it to mark the beginning of a death spiral for LabVIEW as a generally valuable skillset in engineering, turning it into a specialists-only niche instead.

    I don't see how the kinds of pricing we're hearing about now can hope to avoid or reverse that death spiral.  Or even slow it down appreciably.  I think at minimum there would need to be a far more drastic reduction in pricing, especially for students and part-time users to keep repopulating the on-ramp.   Make it make sense for the "tinkerers and dabblers" that many of us were when we were first getting started!  

-Kevin P

[Gregory]

---

@AhmedEisawy wrote:

If anyone would like to renew their expired SSP for perpetual development license, for the next quarter (up until end of September 2024) we can waive the late fees and just pay the 25% of the perpetual list price.

---

Hi Ahmed,

I like the idea of waiving the late fees, but given the pricing structure, it seems like the perpetual license is purposely made to be an unattractive option. I assumed the perpetual vs. subscription models would breakdown as follows:

Subscription: lower upfront cost, higher average cost after the break-even point

Perpetual: higher upfront cost, lower average cost after the break-even point.

If the $13,800 price for the LV pro perpetual license is accurate, then SSP renewal is $3450. Given the $2750 subscription price, there is no break-even point.

In my opinion the the break-even point has to be at 3-5 years to make the perpetual license a serious offering. Even at 5 years, this would put the LV pro license at $6,875 with SSP renewal of $1,718.75 per year.

[rolfk]

---

@Kevin_Price wrote:

---

---

2. I remain perplexed at the disconnect between the luxury pricing and subscription lock-in strategy on the one hand and the Community Edition zero cost of entry for hobbyists (but not degree-pursuing students) on the other.😁

---

You missed one important point here. Since NI Week (I mean NI Connect) this year, NI has officially stated that the Community Edition is free for all students, independent of their degree level. So that is one point less to complain about. 

[LLindenbauer]

---

@rolfk wrote:

---

@LLindenbauer wrote:

---

NXG would have gone with the general industry trend of moving away from proprietary, undocumented formats.

Its cancellation came just at the moment when I was looking into migrating all important code over.

---

I heard that often, but it is still not true. Just because you put something into XML format doesn't mean it's not proprietary anymore. Sure it is text and it is a little easier to guess the meaning of elements in an XML file format than in a binary file format, but that doesn't make it suddenly non-proprietary.

---

This is technically correct. The particular point depends on what value one assigns to the "it is a little easier to guess" part. From what I have seen on the relevant efforts to do either, I would estimate the difference as about the same as getting to the roof of the Empire State Building using the stairwell vs. trying to climb up the outside.

A similar argument can be made regarding git and text compare engines. It is a secondary issue, though: Visual programming produces tree-graphs, not pieces of linear text. The graphs might be flattened in a way that accommodates linear text comparisons, but I will grant that this might require undue effort. It is still a vast improvement to versioning binary data blobs. It becomes a non-issue when the format is open so that graph-comparison tools become feasible.

Source code of text-based programming languages is by definition in an open format, with the format given by the grammar of that language. NXG would also conform to a grammar, even if the particulars were not published. NXG was a huge step towards an open format and got cancelled. Thus the particular size of the step is no longer a concern to me.

It is beside my point, though. The source remains in vendor lock, regardless of the price tag on the license, or the length of its term. That there was once a perceivable effort to remove the lock is just a sharp reminder of its continued presence.

[rolfk]

---

@LLindenbauer wrote:

---

A similar argument can be made regarding git and text compare engines. It is a secondary issue, though: Visual programming produces tree-graphs, not pieces of linear text. The graphs might be flattened in a way that accommodates linear text comparisons, but I will grant that this might require undue effort. It is still a vast improvement to versioning binary data blobs. It becomes a non-issue when the format is open so that graph-comparison tools become feasible.

---

SubVersion has a reasonably efficient binary diffing algorithm. It can be done, but GIT decided not to do it.

As to reverse engineering the LabVIEW binary format, it has been done and there are tools out there available for download for anyone wanting to look at that. LavaG has some threads about that, and the effort has pretty much stopped since even hackers need to live from something else than an occasional reader saying: Great effort, but can you add this and that too? But they still work for pretty much everything since the actual LabVIEW resource format hasn't really changed at all. There might be occasional new resource types that the current tools don't know about but all the rest is fully decodable.