Author Topic: Eagle will be part of Fusion360  (Read 2517 times)

0 Members and 1 Guest are viewing this topic.

Offline ehughes

  • Regular Contributor
  • *
  • Posts: 240
  • Country: us
Eagle will be part of Fusion360
« on: February 14, 2017, 01:28:49 AM »
Pure speculation at this point but I think EAGLE as a standalone product will be a thing of the past.    It doesn't take a lot of brainpower to read these tea leaves.  Not saying that this is a bad thing for those who do integrated design but my *guess* is that  you will soon be purchasing Fusion360, not EAGLE.

http://www.autodesk.com/products/fusion-360/blog/fusion-forecast-dont-resist-current/
http://schnitgercorp.com/2016/06/29/autodesk-acquires-pcb-design-fusion-360/


I am willing to best a basket of chicken wings that the UI will be dumped(which is good news),   the guts will be thrown out(which is good news) and EAGLE will be simply a name for the ECAD component of Fusion360.    The internal architecture will *never*  work with an integrated tool as the current code base implements a system that is stove-piped *by design*.     The current method of move into the 3d space in EAGLE is an ugly hack and I can't see a large company like Autodesk having a tool that looks this primitive.   

 When it is all over,   I am guessing that you will have the the name EAGLE and the XML file format.     There is absolutely no way to sell EAGLE in its current form as an Autodesk product in the long term.   If I was a pointy haired boss,  I would probably be doing the same.   The "old customers" will go to KiCAD or Dip Trace regardless of what Autodesk does.    They are after competing with Dassault and fish with much bigger pocket books.   

I am not saying any of this is bad per say,  just that I can't see the old business strategy working in the long term.    Autodesk is about integrated product design.   EAGLE in its current form will never able to be part of that mission.

To them,  I say onward and upward.  We need better tools!   I am a Solidworks and Altium User but some healthy competition from Autodesk will be a very good thing.





 

Offline Karel

  • Frequent Contributor
  • **
  • Posts: 938
  • Country: 00
Re: Eagle will be part of Fusion360
« Reply #1 on: February 14, 2017, 02:54:29 AM »
I am willing to best a basket of chicken wings that the UI will be dumped(which is good news),   the guts will be thrown out(which is good news) and EAGLE will be simply a name for the ECAD component of Fusion360.

You think they paid £20 million for Cadsoft just for the "name"?

The difference between theory and practice is less in theory than
the difference between theory and practice in practice.
Expensive tools cannot compensate for lack of experience.
 

Offline janoc

  • Super Contributor
  • ***
  • Posts: 1706
  • Country: fr
Re: Eagle will be part of Fusion360
« Reply #2 on: February 14, 2017, 03:11:01 AM »
Pure speculation at this point but I think EAGLE as a standalone product will be a thing of the past.    It doesn't take a lot of brainpower to read these tea leaves.  Not saying that this is a bad thing for those who do integrated design but my *guess* is that  you will soon be purchasing Fusion360, not EAGLE.

I am Fusion360 user for my tinkering and I don't believe such move would make any sense whatsoever. Perhaps Fusion will get some PCB import capability and Eagle gets better support for mechanical CAD, but otherwise those are completely different worlds and user bases. Few mechanical engineers need PCB design support (and even know how to design boards) and few EEs need a full blown mechanical CAD package - most need only import and export. Merging those two products wouldn't really help anyone. Ever heard about jack of all trades master of none?

Also Fusion is mostly an experimental "sandbox" for Autodesk for trying out new ideas and to have a foothold in the low end market but the real money is in AutoCAD and Inventor, not in Fusion360.
 

Offline ehughes

  • Regular Contributor
  • *
  • Posts: 240
  • Country: us
Re: Eagle will be part of Fusion360
« Reply #3 on: February 14, 2017, 06:06:11 AM »
Quote
You think they paid £20 million for Cadsoft just for the "name"?

http://amigobulls.com/stocks/ADSK/income-statement/annual


For a name,  a starting point and an existing customer base  £20 million is a noise bit to them.   








 

Offline janekm

  • Supporter
  • ****
  • Posts: 391
  • Country: gb
Re: Eagle will be part of Fusion360
« Reply #4 on: February 14, 2017, 07:58:35 PM »
Have a look at this page: http://www.autodesk.com/products

Notice anything? That's right, Autodesk aren't in the habit of retiring products if they have active users. They have Fusion 360 but Autocad isn't going anywhere.

I think they're hoping they can fix Eagle bit by bit (seems crazy to me too), but if not I think they would consider building a new product from the ground up using the experience from the Eagle team but as a new separate product that would try to gain new customers through new approaches and new licensing models. Seems like a bit of a challenge though even though compared to the mechanical CAD industry all PCB design software is embarrassingly bad.

I was actually quite disappointed that Autodesk bought Eagle, because I think if they had instead made an attempt to build a completely new tool from the ground up, in the vein of Fusion 360, it could have been something really interesting, taking some ideas from parametric CAD, design history, more "smarts" throughout.

Even autorouting would be interesting to have another look at... So far almost all autorouting algorithms are terrible because they are purely rules-based, but applying some of the recent advances in convolutional neural networks and reinforcement learning, effectively teaching the algorithm what good routing looks like (so it should learn what good decoupling looks like, or BGA fanout, good groundplane, etc) would be really interesting.

 

Offline janoc

  • Super Contributor
  • ***
  • Posts: 1706
  • Country: fr
Re: Eagle will be part of Fusion360
« Reply #5 on: February 14, 2017, 08:23:28 PM »
Even autorouting would be interesting to have another look at... So far almost all autorouting algorithms are terrible because they are purely rules-based, but applying some of the recent advances in convolutional neural networks and reinforcement learning, effectively teaching the algorithm what good routing looks like (so it should learn what good decoupling looks like, or BGA fanout, good groundplane, etc) would be really interesting.

I think that is a bit overly optimistic to expect from AutoDesk. That is not a company too known for innovation - most new things that came from them were thanks to acquisitions when they have bought out competitors. All their major products came to being like that, with maybe the exception of AutoCAD.

They also don't have any previous experience in EDA/PCB design. All their tools are about 3D modelling and mechanical CAD, that is what they are known for. So all such know-how would have to come from the Eagle team - and if those people had such abilities I think Eagle wouldn't be the PCB software known for its unusable crappy autorouter ...
« Last Edit: February 14, 2017, 08:27:08 PM by janoc »
 

Offline Josephine85

  • Contributor
  • Posts: 32
Re: Eagle will be part of Fusion360
« Reply #6 on: February 14, 2017, 10:03:10 PM »
Hey guys,

you can still purchase EAGLE licences on autodesk: http://www.autodesk.com/products/eagle/subscribe.

Do you know of any changes on EAGLE already? i find the new presence is very neat, but i hope they keep the product.
 

Offline janoc

  • Super Contributor
  • ***
  • Posts: 1706
  • Country: fr
Re: Eagle will be part of Fusion360
« Reply #7 on: February 15, 2017, 01:38:33 AM »
Hey guys,

you can still purchase EAGLE licences on autodesk: http://www.autodesk.com/products/eagle/subscribe.

Do you know of any changes on EAGLE already? i find the new presence is very neat, but i hope they keep the product.

Did you miss the huge recent thread with the uproar when AutoDesk made Eagle subscription only, jacked up the price and now requires periodic online check-in? All after steadfastly denying they are going to do that only few months ago. That's why Eagle is a dead horse for many, even though it is still technically sold (err, rented).

The fact that people are abandoning it in droves because of this is not a sufficient change for you? Or maybe you are the exception to the rule that likes the new business model  :-//

« Last Edit: February 15, 2017, 01:41:19 AM by janoc »
 

Offline bitwelder

  • Frequent Contributor
  • **
  • Posts: 591
  • Country: fi
Re: Eagle will be part of Fusion360
« Reply #8 on: February 15, 2017, 05:54:36 PM »
 

Offline helius

  • Super Contributor
  • ***
  • Posts: 1784
  • Country: us
Re: Eagle will be part of Fusion360
« Reply #9 on: February 15, 2017, 06:30:01 PM »
Autodesk CEO Carl Bass: "The factory of the future will have only two employees, a man and a dog. The man will be there to feed the dog. The dog will be there to keep the man from touching the equipment."
 

Offline janekm

  • Supporter
  • ****
  • Posts: 391
  • Country: gb
Re: Eagle will be part of Fusion360
« Reply #10 on: February 15, 2017, 08:52:32 PM »
Even autorouting would be interesting to have another look at... So far almost all autorouting algorithms are terrible because they are purely rules-based, but applying some of the recent advances in convolutional neural networks and reinforcement learning, effectively teaching the algorithm what good routing looks like (so it should learn what good decoupling looks like, or BGA fanout, good groundplane, etc) would be really interesting.

I think that is a bit overly optimistic to expect from AutoDesk. That is not a company too known for innovation - most new things that came from them were thanks to acquisitions when they have bought out competitors. All their major products came to being like that, with maybe the exception of AutoCAD.

They also don't have any previous experience in EDA/PCB design. All their tools are about 3D modelling and mechanical CAD, that is what they are known for. So all such know-how would have to come from the Eagle team - and if those people had such abilities I think Eagle wouldn't be the PCB software known for its unusable crappy autorouter ...

Yes, I'm sure you're right. It's a bit of a pipe-dream... I look at the state of PCB design tools and all the major ones have source-base going back to DOS days. Must be horrific to work on. Like Altium which apparently is still in large part based on Delphi. I didn't even know Delphi still existed... Amusingly the example UI from the recent version of Delphi is basically the Altium UI: https://www.embarcadero.com/products/delphi (they must be the largest remaining customer...)

So I think there is an opportunity for Autodesk to take the Eagle team and move half of them to a team building a completely new EDA tool, and the other half onto maintenance of Eagle. I'm guessing that's more or less what they did for Fusion 360. But I have no reason to believe that's the plan (and it's looking like they're trying to add lipstick to the existing Eagle codebase instead...), and if they did we wouldn't hear about it for a few years until they have something to show...
 

Offline technolomaniac

  • Contributor
  • Posts: 32
  • Country: us
Re: Eagle will be part of Fusion360
« Reply #11 on: February 20, 2017, 09:25:06 AM »
Eli, if I never told you you were insightful, let this be the first time.  And if we aimed to be coy in our strategy, we should have picked a name other than "Fusion" when we named the product.  ;). (They say the ship leaks from the top...so I guess I'm the guilty party).

Truth is, this is a direction we are extremely interested in and this was even announced (albeit a bit anecdotally) at Autodesk University this past year.  We (Autodesk) don't view these as separate processes any longer and my charter is in ensuring that we treat ECAD, MCAD & Manufacturing as all sides of the same coin. 

Curiously, I think Altium figured this out when they suddenly put Eagle and Altium Designer on the same footing by providing a massive discount to EAGLE users.  The fact is, it's only time & resources that creates the lead. 

Our goal at Autodesk goes beyond feature parity.  Sure, we know we need comprehensive solutions for routing, design reuse, hierarchy, libs, data management, mcad integration, etc. ...but there's SO much more we can do if we integrated these capabilities at some level.  And truth is, we have a bit of a blank slate insofar as we have a lot to build from, but a lot of "net-new" stuff to build.   It's a shame really, just trying to be patient long enough for this all to take place.

The question is, what happens to the $12K tool with $1250/yr maintenance when someone does the same thing for less than $1000?  And what if 80% of that is under $500?  Then what?  :). ...Game on.

Best regards,

Matt - Autodesk.
 

Offline timb

  • Super Contributor
  • ***
  • Posts: 2362
  • Country: us
  • Pretentiously Posting Polysyllabic Prose
    • timb.us
Eagle will be part of Fusion360
« Reply #12 on: February 20, 2017, 12:31:34 PM »
Eli, if I never told you you were insightful, let this be the first time.  And if we aimed to be coy in our strategy, we should have picked a name other than "Fusion" when we named the product.  ;). (They say the ship leaks from the top...so I guess I'm the guilty party).

Truth is, this is a direction we are extremely interested in and this was even announced (albeit a bit anecdotally) at Autodesk University this past year.  We (Autodesk) don't view these as separate processes any longer and my charter is in ensuring that we treat ECAD, MCAD & Manufacturing as all sides of the same coin. 

Curiously, I think Altium figured this out when they suddenly put Eagle and Altium Designer on the same footing by providing a massive discount to EAGLE users.  The fact is, it's only time & resources that creates the lead. 

Our goal at Autodesk goes beyond feature parity.  Sure, we know we need comprehensive solutions for routing, design reuse, hierarchy, libs, data management, mcad integration, etc. ...but there's SO much more we can do if we integrated these capabilities at some level.  And truth is, we have a bit of a blank slate insofar as we have a lot to build from, but a lot of "net-new" stuff to build.   It's a shame really, just trying to be patient long enough for this all to take place.

The question is, what happens to the $12K tool with $1250/yr maintenance when someone does the same thing for less than $1000?  And what if 80% of that is under $500?  Then what?  :). ...Game on.

Best regards,

Matt - Autodesk.

If have to rent it? Nothing, that's what happens. You guys are losing a huge chunk of the market by not having perpetual license. ;)

Seriously though, I'd say the vast majority of people that use EDA software want it to do two things well: Let us capture a schematic and layout a PCB. That's it. We don't need integrated FPGA tools, MCAD, MCU programming or the ability to toast bread.

We just want to layout circuit boards. If I want to do MCAD I can export my board as a STEP file and bring it into a 3D package. (That said, easy export with a leading CAD package wouldn't be a bad thing per say, but the majority of users don't need the actual MCAD functionality integrated into their EDA tools. I think Altium has proven that people don't want tons of disparate things bolted onto their EDA software.)

When you try to integrate everything under the sun into one package, it *always* ends up a huge, bloated mess. Functionality is likewise *rarely* better (or even close to) standalone packages.

Sure, there is a small percentage of users (large companies mostly) that could benefit from a streamlined ECAD-MCAD-Manufacturing package, but they're already using $100,000 packages that do this and you're not likely to break into that market for quite some time.

Why not stick to making Eagle a modern EDA tool first? And keep it simple, please. Engineers *hate it* when companies take otherwise good software and ruin it by bolting on everything plus the kitchen sink.
« Last Edit: February 20, 2017, 12:34:50 PM by timb »
Any sufficiently advanced technology is indistinguishable from magic; e.g., Cheez Whiz, Hot Dogs and RF.
 
The following users thanked this post: ar__systems

Offline ehughes

  • Regular Contributor
  • *
  • Posts: 240
  • Country: us
Re: Eagle will be part of Fusion360
« Reply #13 on: February 21, 2017, 01:19:55 AM »
Matt:

he he he.

I can see where things are going and it looks promising.   

I know many here don't care for an integrated approach but I have been doing the export to STOP, import into Solidworks since Altium 1st allowed for STEP integration.   While it was a big step forward,     it is  static process.

Some other thoughts:

1.)    PCBs are a mechanical device with electrical design rules.     Any good PCB design starts at the solid modeling phase.   Having a tool that can go both directions seamlessly is a big deal.     There is quite a bit of back and forth when doing anything non-trivial.

2.)  STEP is a very bloated format.    I have assemblies that are several hundred megabytes without working very hard.    A unified tool working in the native 3D model format is where things need to go.   Working with non-trivial boards is difficult.    I generally use very accurate models which always pushes envelope.   (Pretty pictures sell new projects).

3.)   Exporting step and importing into another tool means you lose all of the parametric constraints/mates.     This is one aspect of PCB design that needs improved.   Parametric relationships between parts is where things need to go.

Altium/Dassault have a PCB tools that does some level of b-directional communication (it was branded PCBWorks) but I don't believe  you can get the plugin for the full version of Altium.

I will be curious (and patient) to see what you guys come with.
 

Online ebastler

  • Super Contributor
  • ***
  • Posts: 1111
  • Country: de
Re: Eagle will be part of Fusion360
« Reply #14 on: February 22, 2017, 03:52:22 AM »
Our goal at Autodesk goes beyond feature parity.  Sure, we know we need comprehensive solutions for routing, design reuse, hierarchy, libs, data management, mcad integration, etc. ...but there's SO much more we can do if we integrated these capabilities at some level.  And truth is, we have a bit of a blank slate insofar as we have a lot to build from, but a lot of "net-new" stuff to build.   It's a shame really, just trying to be patient long enough for this all to take place.

Matt,
While you and us Eagle users are being patient, and waiting for all that magic to happen -- why can't Autodesk be patient too, and postpone the fleecing of Eagle users until you have something really strong to offer?
 
The following users thanked this post: ar__systems


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf

 

http://opalkelly.com/