Interresting new TM1 site

[Olivier]

Hi All,

Just received a notification regarding the launch of "Bedrock" by Cubewise consulting company.

http://www.BedrockTM1.org

I have not read all the content yet but it sounds like a good source of ressource for most of us...

[Craig Trevor]

This seems to be a promising addition to the community.

Now to find some time to be able to go through it all .

[Ben H]

With respect to the mention of Cubewise by Olivier, I want to make it clear that the intention of Bedrock is completely non-commercial, in the spirit of real sharing and trying to forge better TM1 standards collaboratively with the global TM1 community.

Bedrock is not a Cubewise product - it is a point for collaboration for the benefit of all existing and future TM1 models.

The materials and information on the Bedrock website are free and you are not required to provide any information in order access all of its contents.

We hope many of you will take a look and contribute to it. This is the point - the more you participate, the more valuable it will be for everyone.

Also, to be clear, Bedrock is not a forum - it is a collaborative project.

[dnicol]

Interesting, but too little too late for me. I have spent the last six months wading through thousands of lines of pretty terrible TI code (superfluous, duplicated, untestable). Something like this will be useful for those starting out with new models, but the fundamental issue with TM1 is you can't change anything without having to regression test the whole model - manually. Throw in some complex calcs/rules/feeders between each cube and I've found it takes a while just to figure trace what the calculations are meant to do. Not to mention the IDE sucks massively, the scripting language is proprietary and very limited and there's no such thing as a design pattern. A collaborative project may help, but for me, the shortcomings of the product are just too many.

[Martin Ryan]

I guess that is a drawback that goes with having a tool that is a blank piece of paper. It can be very badly built if the designer doesn't know what they're doing. Then unpicking someone else's code can be tedious and difficult.

I think you'll find a lot of agreement on those points from other people who have inherited poorly built models, but the flexibility is a real boon when you're building the model from scratch (provided you know what you're doing of course).

Martin

[Alan Kirk]

I guess that is a drawback that goes with having a tool that is a blank piece of paper. It can be very badly built if the designer doesn't know what they're doing. Then unpicking someone else's code can be tedious and difficult.

I think you'll find a lot of agreement on those points from other people who have inherited poorly built models, but the flexibility is a real boon when you're building the model from scratch (provided you know what you're doing of course).

Agreed; the problem is less the limitations of the tool, more the limitation of the skills and/or imagination of some developers particularly those who don't know the tool but are prepared to charge for developing in it anyway. That can leave the kind of model that was referred to earlier; a model which then needs to be unpicked by someone who does know what they're doing.

I don't think that the fact that the scripting language is proprietary is material; it's similar enough to other languages for anyone who's familiar with pretty much any other language to get the hang of it fairly easily. (And for people who don't know any programming language... well, it doesn't make a difference because they don't know how to program anyway.)

This is not to say that IBM doesn't need to put considerable effort into the development environment; debugging, code snippet libraries, formatting, find and replace... all things that they have thus far shown little interest in providing.

However it's far better to have the flexibility of a product which can be developed as you need it to be (skills permitting) than one which forces you try to shoehorn what you need into a "standard" structure.

[lotsaram]

Interesting, but too little too late for me. I have spent the last six months wading through thousands of lines of pretty terrible TI code (superfluous, duplicated, untestable). Something like this will be useful for those starting out with new models, but the fundamental issue with TM1 is you can't change anything without having to regression test the whole model - manually. Throw in some complex calcs/rules/feeders between each cube and I've found it takes a while just to figure trace what the calculations are meant to do. Not to mention the IDE sucks massively, the scripting language is proprietary and very limited and there's no such thing as a design pattern. A collaborative project may help, but for me, the shortcomings of the product are just too many.

As someone who has inherited just such a model and was part of the beta testing for the bedrock project I can verrify that there is a lot of benefit to adopting this as a platform for an existing model. Clearly there are greater benefits for any greenfield development but I wouldn't discount the significance and potential benefit for aleady established models (especially those that may not be shining examples of best practice or good design.)

[Kyro]

As someone who has inherited just such a model and was part of the beta testing for the bedrock project I can verrify that there is a lot of benefit to adopting this as a platform for an existing model. Clearly there are greater benefits for any greenfield development but I wouldn't discount the significance and potential benefit for aleady established models (especially those that may not be shining examples of best practice or good design.)

I think this is definitely a start in the right direction for the wider TM1 Community. The more people adopt Bedrock's modular methodology (or any single methodology in general in relation to TM1 Development), the easier it will be for consultants who have to pickup a new model every few months cleaning up messes left by the noobs. It just shows that CubeWise isn't only looking out for themselves and instead is looking to drive the innovation in this space. Well done guys, All aboard!

[stex2727]

Yes great site and a doff of the cap to those behind the project.

I'd also like to add a quick comment on the quickness of TM1 developers to blame fellow developers for train wrecks (noob etc). I'm sure we've all had our share of inherited train wrecks and also work that we're not particularly proud of. As a rule, I never pour scorn a fellow developer as it reflects badly on us as a community and also on TM1 as a product. More importantly though is that we are generally not there to witness the cause of the train wreck so how do we know it was the developer’s skill rather than client demands, existing data standards or the way the project was managed.

So as a courtesy, can I ask fellow developers hold back on judgement when inheriting a train wreck. Just imagine how short the TV program air crash investigations would be if they blamed the pilot for everything without understanding the circumstances.

Steve

[Malcolm MacDonnell]

Actually, If they took out the repeated bits (eg. the first two minutes after an add are the same as the two minutes before) and the fillers, air crash investigation would only go for six or seven minutes anyway.

[Alan Kirk]

Actually, If they took out the repeated bits (eg. the first two minutes after an add are the same as the two minutes before) and the fillers, air crash investigation would only go for six or seven minutes anyway.

Amen to that; not to mention the repeated (and repeated and repeated) misty scenes of the passengers enjoying themselves in the pre-disaster stages of the flight. At the risk of skewing the thread wildly off topic (who are we kidding, that's already happened) in this case reading beats audio-visual. If your interests lean that way, Macarthur Job's books are as good as they get:
https://shop.australianaviation.com.au/shop/browse.37

{Obligatory on-topic redirection} Now if we could find a way to rebuild a bad model in 6 or 7 minutes...

[stex2727]

Actually, If they took out the repeated bits (eg. the first two minutes after an add are the same as the two minutes before) and the fillers, air crash investigation would only go for six or seven minutes anyway.

Amen to that; not to mention the repeated (and repeated and repeated) misty scenes of the passengers enjoying themselves in the pre-disaster stages of the flight. At the risk of skewing the thread wildly off topic (who are we kidding, that's already happened) in this case reading beats audio-visual. If your interests lean that way, Macarthur Job's books are as good as they get:
https://shop.australianaviation.com.au/shop/browse.37

{Obligatory on-topic redirection} Now if we could find a way to rebuild a bad model in 6 or 7 minutes...

..... you'll have 10 minutes to bring peace to the middle east.

Having just enjoyed several long haul flights I don't get the misty eyed bits either. Probably not good TV showing grumpy smelly old men crunching into economy class I guess. Speaking of design and knowing many of you travel can I put in a bad recommendation for Lufthunza. Some bright spark in their design section thought it was a good idea to put a big metal box right where your left leg should stretch out. Not just the odd unlucky seat but every second seat in economy class. I wonder if this was ever brought up in the design sessions before signoff and implementation.

Steve

[Alan Kirk]

Actually, If they took out the repeated bits (eg. the first two minutes after an add are the same as the two minutes before) and the fillers, air crash investigation would only go for six or seven minutes anyway.

Amen to that; not to mention the repeated (and repeated and repeated) misty scenes of the passengers enjoying themselves in the pre-disaster stages of the flight. At the risk of skewing the thread wildly off topic (who are we kidding, that's already happened) in this case reading beats audio-visual. If your interests lean that way, Macarthur Job's books are as good as they get:
https://shop.australianaviation.com.au/shop/browse.37

{Obligatory on-topic redirection} Now if we could find a way to rebuild a bad model in 6 or 7 minutes...

..... you'll have 10 minutes to bring peace to the middle east.

Having just enjoyed several long haul flights I don't get the misty eyed bits either. Probably not good TV showing grumpy smelly old men crunching into economy class I guess. Speaking of design and knowing many of you travel can I put in a bad recommendation for Lufthunza. Some bright spark in their design section thought it was a good idea to put a big metal box right where your left leg should stretch out. Not just the odd unlucky seat but every second seat in economy class. I wonder if this was ever brought up in the design sessions before signoff and implementation.

Sounds... restful. Oh well, the 787 (if it ever arrives in significant numbers) should fix that!

Well, except for...

The most noticeable change for passengers stepping into the Dreamliner will be windows that are 65 per cent larger than those in competing aircraft, giving passengers a view of the horizon from any seat on the plane and reducing the feeling of being trapped in a metal tube.

and

Rather than the typical cabin altitude of 1980 metres to 2130 metres, the cabin altitude will be below1830 metres, a level Boeing says was revealed in testing as optimum for passengers.

, he says, remembering what brought down the first generation of Comets. (Yes, I know that metallurgy has moved on since then but I'm betting that Boeing remember as well which is what precipitated the emphasis that:

The larger windows have been made possible by the use of composite materials in the fuselage; materials that can handle the loads of larger cutouts.

)

Also:

The 787 will have a wider cabin, allowing for wider seats and aisles. Boeing says the Dreamliner cabin could offer the widest economy seats in the industry, depending on the individual airline's chosen configuration.

Though history suggests that the standard configuration is "set the seat pitch to jam 'em in, then when you're done, jam some more in. Then put in a steel box."

{Increasingly desperate and failing attempt to retain on-topicness} Hmm, I wonder what feeders we'd need for the rules in a cube which models the material fatigue in fuselages which are highly pressurised and with large cut-outs...