TM1Tools is there one for PAX

[yyi]

Been using this excel add-in for ages, the one made by Alan, Steve & Martin - namely the hierarchy documenter and formula trace. The forum search gave me a 0 result, maybe noone uses it anymore or maybe I spelt it wrong.

Just wondering if a PAX/PAFE version was ever put together.

[Alan Kirk]

Been using this excel add-in for ages, the one made by Alan, Steve & Martin - namely the hierarchy documenter and formula trace. The forum search gave me a 0 result, maybe noone uses it anymore or maybe I spelt it wrong.
Just wondering if a PAX/PAFE version was ever put together.

Storm.jpg (220.95 KiB) Viewed 8150 times

I thought you were on a break?

The short answer is "Not at this time".

The longer answer is this.

First (and as I've discussed previously), being in a Windows shop I have thus far been unable to get PAW working and therefore to get PAfE working, since for some ridiculous reason IBM chose to make the latter dependent on the former.

I was on a web presentation with IBM last week where I raised this issue. On the webinar you submitted questions but had no way of responding to their answers. I shall therefore give you my thought-responses below.
Them: "Oh, you DO need to make sure that the Windows server is set up the right way first."
Me: "Yes, I know. I read your FOUR HUNDRED AND FOURTEEEN frapping page installation guide, the SQL Server equivalent of which is a guided dialog."
Them: "Shrug, I've never had a problem with it."
Me: "I can think of more people than me who have, but it's good to know that when you build a Windows setup that's configured juuuuuust for your own software, it works."
Them: "Everyone's experience will be different."
Me: "Just as everyone's experience of catching typhoid will be different. That doesn't mean that it's ever good."
I did compare and contrast how friction-free installing and using Docker Desktop to run RDBMSs is compared to PAW, to which they replied that it wasn't a fair comparison since they use a NUMBER of different Docker instances for PAW, not just one.
Me: "Ah. So you haven't merely overcomplicated it, you've overcomplicated the HELL out of it. You have provided us with not just ONE potential point of failure, but multiple potential points of failure. And because they're all running inside little black boxes, not only will we be blind to WHAT has failed, but WHERE it has failed. And because you've made PAfE dependent on PAW, if one of PAW's little Docker boxes collapses, you lose the entire platform outside of using Python to do direct queries. Oh, I expect that Web will still work, but Web doesn't support hierarchies. I feel so reassured now."

I am currently seeing whether I can get a Linux server just for PAW, because you really want your network infrastructure to be disrupted just for one user interface, don't you? Sure you do. It's an IBM interface, it's more important than mere customers. If I can, then I may, MAY be able to run PAW, and therefore PAfe, and therefore start developing software for it.

EXCEPT that brings us to the second point, the absurd concept of "Short Cadence" (SC) releases where there are updates for PAW and PAfe Every. Fracking. Month.

No, no, don't sit down and design the functionality of a piece of software, build it, test it, release it and fix any bugs, that's OLD STYLE thinking. All the COOL kids are following the "Move fast and break things" paradigm. (This is also known as the "make crap up as you go along and hope it works in the wild" software design methodology.)

Except…a business tool is not a social media site. If Suckerberg and Jack The Weed Eater were to move fast enough to break Faceplant and TWIT-ter into a thousand irreparable pieces then first, it would be a net positive for society and second, it would be a net positive for most people once they had allowed themselves a couple of weeks of digital detox.

A business tool requires functionality and stability. Much as IBM seems to think otherwise, NOBODY'S goal in life is to re-learn their software every time IBM comes up with a New! User! Experience! Our work lives do not revolve around IBM, they revolve around the tasks that we use IBM software to perform. To my mind IBM is completely incapable of understanding that concept.

To illustrate, they conducted a survey on the webinar participants asking whether we would be interested in IBM Cloud Paks (sic) for Data. I see a lot of you staring blankly. As did I, since IBM had not bothered to explain in plain English (a language that IBM seems incapable of mastering, incidentally) what these things are. The assumption seemed to be "Well, it's IBM software, you MUST know about it! Don't you spend your life scouring the Web and reading every one of our technical releases to see what we're doing?" Thus, the only answers were "Yes" or "No". There was no option for "I have not heard of this", for such a thing is apparently inconceivable. (Apparently IBM Cloud Pak for Data are a way of managing Docker containers like the one (sorry, oneS) that PAW runs in, but I've yet to wade through the reams of technojargon bulldust on their pages which "explain" the product. Edit, 20 March: I think they may have a broader role than that. I still haven't managed to drill to the bottom of the Jargonese that the relevant documents are written in.)

EVERY company that has developed a "constant release" cycle breaks things. I've got yet another problem with Windows Update at the moment because the poor little petal has an issue with the Sennheiser software for my headset. A couple of years back an update to Office 365 broke the writeback capability of DBRW formulas for a few weeks. (But that's OK, since Office 365 is not supported. IBM is the only one who is allowed to move fast, notwithstanding that at least Office updates aren't monthly.) Adobe tanked a critical function in its decrepit Photoshop scripting language called Actions (which makes even TI look cutting edge in comparison) which caused masses of photographers to need to roll back to a previous version. I seriously doubt that IBM has a golden touch that will prevent it from ever stuffing up one of its trademark short cadence releases.

And that sentence there? It's the lead contender for "Understatement of the Millennium".

So IF I get PAW running and IF I can get PAfE running and IF I can then find time to build a PAfE equivalent of TM1 Tools THEN I would need to check it EVERY SINGLE MONTH to make sure that the latest SC release does not stuff the living crud out of it.

Since TM1 Tools is a freebie, I cannot help but think that I would not be paid enough to do that on an ongoing basis.

Then there is the question of whether to continue to do the add-in in VBA, shift it to VSTO, or the All! New! Office add-in format which is hosted on line because Microsoft's Office team seems to have lost sight of the fact that people do actually work offline, so no, it won't be that.

In summary then, I cannot rule out the possibility that there will be another version of TM1Tools aimed at PAfE. Getting around the second of the points above is likely to be the biggest issue, because anybody who believes that they can continue to support a freebie which is aimed at a constantly moving target off into eternity is delusional.

[Harvey]

Well said, Alan!

You said a lot of things out loud that I've been thinking about Planning Analytics for a long time now. I remember hearing them talk about Docker and Kubernetes and all these technical aspects of the software and thinking "how about you just build a really good front end?" But no, IBM just want to build over-complicated technical marvels to 80% quality and then chase their tails forever on feature requests and bug reports.

Not to mention, not even a company with a modest sized IT department can keep up with destructive updates every 4-6 weeks that may or may not bring them any benefit -- what hope do consultants have to keep it running on a laptop?

It's almost as if IBM want everyone on cloud, so deliberately make it too damn difficult to get PAW working in it's local form, isn't it?

[Alan Kirk]

It's almost as if IBM want everyone on cloud, so deliberately make it too damn difficult to get PAW working in it's local form, isn't it?

The! Cloud!, where they would then have you over a barrel with pricing because of the inertia that would be involved in trying to move your data from their platform if, two or three years down the track, you find that the bang for buck isn't there any more?

The Cloud that would thus lock you in forever, or face massive disruption in moving to an alternative platform?

That cloud?

SHHHH!!! You aren't supposed to say that out loud!

Not to mention, not even a company with a modest sized IT department can keep up with destructive updates every 4-6 weeks that may or may not bring them any benefit -- what hope do consultants have to keep it running on a laptop?

Indeed. They did mention something about a process which will handle this automatically which, IF I remember correctly, is Red Hat OpenShift Container Platform. (I may have misunderstood that; it was, after all, 3 am my time given that EVERY Ask Me Anything session is at 11 am Ottawa time, regardless of the time in other parts of the world. They NEVER shift that time to provide everyone on the globe SOME coverage over the course of a year. This is why they put the "I" in "IBM".)

Why wasn't I up to date on this?

Because. I. Work. In. A. WINDOWS. Environment.

I have neither the time nor the inclination to become an RHEL specialist as well. It provides me with no benefit. It provides my company no benefit for me to need to split my attention in that way. My focus is SUPPOSED to be on USING the software, not installing it and keeping it running.

If anyone at IBM is reading this I'm sure that they'll be saying "Why doesn't this guy shut up about this, we've heard it before?"

To which my response is "Give me software that has a FRICTIONLESS, EASY, RELIABLE install, software that runs well and reliably, software that does NOT require the user base to be frequently re-learning it, and then hey, not only will I shut up but I may even praise it."

For all their faults, Microsoft, Adobe, Oracle via MySQL, Docker Desktop, every one of them is improving EVERY YEAR in providing an easy deployment experience. The only company that has elected to walk in the diametrically opposite direction is IBM.

Some might say that me saying things like this is giving ammunition to IBM's competitors.

To which I would reply "I'm not doing that, IBM is doing that." Even if people are conflict averse and won't say this sort of thing to IBM's face or put it in public, do they really think that they aren't saying it? That they may not be saying it privately to potential customers?

Am I not trashing my own career in doing that? Trust me, I'm looking at alternatives as well because if IBM continues down this path rather than stopping, taking stock and making the user experience better, easier, faster, less "frictiony" and less high maintenance... then the Max impersonator at the top of this page isn't the only one who can sense a storm coming.

Oh, and STILL they haven't provided hierarchy aware DBRW formulas. They might, maybe get around to it by the end of the year. That tells you a lot.

[bgregs]

No, no, don't sit down and design the functionality of a piece of software, build it, test it, release it and fix any bugs, that's OLD STYLE thinking. All the COOL kids are following the "Move fast and break things" paradigm. (This is also known as the "make crap up as you go along and hope it works in the wild" software design methodology.)

If I could give this a "thumbs up" or "+1" 1 million times I would. Hit. The. Nail. On. The. Head.

Now I'm not saying the Waterfall approach to software development isn't without its own intrinsic flaws, but it sure beats the hell out of throwing half finished crap at customers just so it can be said that "look at us, we're Agile now". Well, it may be Agile, but my work flow sure as hell isn't anymore.

Anyway, thanks for that post Alan. As always, well said and thoroughly entertaining.

[Harvey]

Oh, and STILL they haven't provided hierarchy aware DBRW formulas. They might, maybe get around to it by the end of the year. That tells you a lot.

Don't get me started

After becoming an early adopter of virtual hierarchies and discovering this omission live with the client, I asked IBM at every opportunity when that fundamental and absolutely necessary piece of functionality would be delivered. Each time I was promised the next cycle until eventually I gave up asking.

It's now gotta be over 3 years later(?), and still using virtual hierarchies means ruling out custom PAX reports.

[Harvey]

Now I'm not saying the Waterfall approach to software development isn't without its own intrinsic flaws, but it sure beats the hell out of throwing half finished crap at customers just so it can be said that "look at us, we're Agile now".

I recently had some training on Agile Project Management and the guy was pushing multi-coloured sticky notes, Blu-Tack and duct tape. It was like he was describing how to fix the landing module on Apollo 11.

The only explanation I could think of was that he was sponsored by 3M, because the very idea of preferring "low-tech" tools to track something as complex and nuanced as software development is a terrible idea, especially in an age where we have mobile PCs, tablets and Kanban software. It might have been fun in the 90s, but we can no longer party like it's 1999.

Agile software development has some virtues, but man it's got some silly **[Expletive Removed]** in it too. IBM's handling of PA is an example of exactly how NOT to do it, IMO.

[yyi]

Thanks Capt'n for the really detailed reply! yeah I'm still doing a bit of tm1/paw stuff WFH.

I can relate to that annoying paw install routine - the day I was introduced to powershell, after the nth time scrapping previous failed installations, by some great fluke the ./start.ps1 actually launched admin tool in chrome.

Well I hope you're doing well and perhaps be able to kick off the tool again and someone can prolly maintain it from there on