Eagle Premium no longer included with Fusion for hobbyists?

[ataradov]

Is there a way to buy out old Eagle sources (latest CadSoft version, something around v7-v8) from Autodesk?

I would assume not in  a reasonable way. For the amount of money you would have to pay, it is much cheaper to just write it from scratch.

[jpanhalt]

Search on it!  First hit from Google: http://eagle.autodesk.com/eagle/software-versions/7

Of course, that is contaminated with Autodesk, but I am sure there are other sites.

Edit:  I have versions 7.2 to 7.7 on my desktop.  I still use 7.2 almost all of the time.  Labeling tracts (as in later versions) is redundant.  I set up hotkeys for the important ones, like VDD's and GND.

[ataradov]

I assumed by "sources" was mean the actual source code.

You can find downloads for all versions, but it is no longer possible to buy a Pro license. So, you are stuck with the lite eval version.

[Unixon]

I assumed by "sources" was mean the actual source code.

Yes, this is exactly what I meant -- actual source code. This would save a lot of research time.
Not that these sources are exactly necessary, but it would be nice to see how some of things were implemented like DRC or CAM.

[KE5FX]

I assumed by "sources" was mean the actual source code.

Yes, this is exactly what I meant -- actual source code. This would save a lot of research time.
Not that these sources are exactly necessary, but it would be nice to see how some of things were implemented like DRC or CAM.

One of the (few) valid points Matt Berggren makes on the Autodesk forum threads is that the computational geometry library is the key to the whole ball game, and Autodesk is uniquely experienced in that area.  For its part, EAGLE's geometry library is suited to its original purpose, which is implementing a 2D PCB CAD program that does what EAGLE does.  If you had access to that foundation, or an equivalent one, things like DRC and CAM would be fairly straightforward to implement. 

The potential was always there to build a better-EAGLE-than-EAGLE on top of Autodesk's superior 2D/3D geometry toolset, but nothing has emerged except a lot of hot air and promises.  There was a lot of talk about how subscription revenue was needed for continued improvement and development, but we've seen how that worked out.  All of the R&D effort was directed toward Autodesk's goals, and not their customers'. 

[Unixon]

If you had access to that foundation, or an equivalent one, things like DRC and CAM would be fairly straightforward to implement. 

Well, not that straightforward but... Computational geometry is science, a well defined set of methods for doing particular things, if you know that you need to solve a particular geometry problem you will. Methods could be complex, but they are strictly formal, so at the end of the day this is just a huge pile of math and algorithms that has to be worked out. However, application of these methods and efficient application is more of an art because one problem could be solved multiple ways and it is not immediately evident that one is better than another and what's even worse -- the very process of coming up with a list of things to formalize and solve is not formal on itself, one can always skip and forget something important.

[james_s]

The potential was always there to build a better-EAGLE-than-EAGLE on top of Autodesk's superior 2D/3D geometry toolset, but nothing has emerged except a lot of hot air and promises.  There was a lot of talk about how subscription revenue was needed for continued improvement and development, but we've seen how that worked out.  All of the R&D effort was directed toward Autodesk's goals, and not their customers'.

I would be curious to know how many subscribers there actually are/were. I suspect the numbers were dramatically lower than predicted since virtually the entire hobbyist market evaporated after Autodesk lied and then took it subscription anyway.

[jpanhalt]

@james_s

I agree.

Been with Eagle since 3.x.  Just a hobbyist, but I liked it so much I bought a low level license and updated as needed.  Still use it.  But any subscription while affordable is not in my future.  Some years, I may design half a dozen PCB's; others none.  FWIW, I still use bought versions of all my CAD software, Photoshop, and office software.  Had to switch from WordPerfect and Quattro Pro (which were perfect for my needs) to Word Office several years years.  I grimace every time I have to use them.  Jorge at Eagle was/is a real prince.  Always helpful.  I hope he is able to stay on at Autodesk.

[langwadt]

The potential was always there to build a better-EAGLE-than-EAGLE on top of Autodesk's superior 2D/3D geometry toolset, but nothing has emerged except a lot of hot air and promises.  There was a lot of talk about how subscription revenue was needed for continued improvement and development, but we've seen how that worked out.  All of the R&D effort was directed toward Autodesk's goals, and not their customers'.

I would be curious to know how many subscribers there actually are/were. I suspect the numbers were dramatically lower than predicted since virtually the entire hobbyist market evaporated after Autodesk lied and then took it subscription anyway.

I'd think the Kicad maturing and becoming a more common format for opensource projects also did it's bit

[Keith Ward]

This sort of thing happens all the time with big corporations.  If they can't be bothered to compete with competitors, buy them and resume their profits, or absorb them and make the product go away.  It's rarely a good thing but that's how they roll in the big money league.

[james_s]

I'd think the Kicad maturing and becoming a more common format for opensource projects also did it's bit

I think KiCad did a fair mount of maturing in response to Eagle becoming non-viable for a lot of people.

[TomS_]

I thought it might be worth giving a little update. Ive been playing with KiCad (6.0.5) on and off over the past month or so, and long story short, it hasnt been a painless experience, but not as painful as I might have thought it would be to switch to another PCB CAD package. There have been some frustrating times, and a bug that Ive had to work with in the symbol editor (hopefully that is fixed in 6.0.6), but generally I would say the experience is positive.

I think Ive pretty much mastered making my own footprints and symbols, and have managed to put together one simple board and sent it off for manufacturing - it came back exactly as I would have expected, so thats a good sign that I might have figured things out.

I have a more complex project in the works, but theres still a lot of work to do on that one before I go ordering PCBs including building a prototype first. And I have a second project that I am currently prototyping/breadboarding that I'll get to some time in the next couple of months.

In summary, while I think EAGLE does some things better (or in a way that I prefer), there are also aspects of KiCad that I like, and I think it would be fair to say that I wont be looking back. If youre stuck on what to do post EAGLE, my advice is jump into KiCad sooner rather than later.

[jpanhalt]

@TomS_

Since you have experience with Eagle, you may have found that plated through slots are not easily supported in the Gerbers  made by the Eagle CAM.  Maybe that was fixed in versions 8.0 and higher.

Have you tried plated through slots in KiCad?  Are they a regular function?

[TomS_]

@TomS_

Since you have experience with Eagle, you may have found that plated through slots are not easily supported in the Gerbers  made by the Eagle CAM.  Maybe that was fixed in versions 8.0 and higher.

Have you tried plated through slots in KiCad?  Are they a regular function?

This is not something Ive had to work with to date so I dont think I can be much help with it.

But I just had a quick look in KiCad and it seems you can make oval shaped pads, so this should make it quite trivial to do plated slots I would imagine.

[ebastler]

you may have found that plated through slots are not easily supported in the Gerbers  made by the Eagle CAM.  Maybe that was fixed in versions 8.0 and higher.

I have certainy used plated slots with Eagle (version 7.x and earlier). The PCB houses I have used will interpret slots which are surrounded by copper, e.g. are located within pads, as plated slots. How else would slots be designated as "plated" in the Gerbers?

[jpanhalt]

Admittedly, I don't quite understand the issue with early Eagle.  Making oval pads is easy and has been available for a long time (even teardrop pads are possible). It's the plated slot per se.  I had communications with JLCPCB about it, and its requirement for a particular Gerber layer wasn't met.  Oshpark developed its own/modified CAM processor to create them, and as usual, the boards from Oshpark were perfect.  It's just an additional hoop to go through. 

Thanks for the prompt reply.

EDIT: @ebastler
Yes, I have been told that here.  So, I submitted Gerbers made that way to JLCPCB and never heard back from it.  Rather than take a chance, I went the sure route to Oshpark.

[ebastler]

Strange; I have definitely had JLCPCB manufacture several of these Eagle-generated boards. The slots, whether plated or not, were all defined in the milling/outline layer, I believe. I can check my archived files if that's of interest!

[jpanhalt]

Yes, I would be interested in those details. 

I tried a lot of things to view the Eagle gerbers and nothing (except the ones made by Oshpark) convinced me the slots would be plated.  The most important ones were signal tabs from an Omega TC socket, but mounting tabs were also wanted.

[ebastler]

I have attached two screenshots from a Gerber viewer. The GML milling layer includes all slots and large circular holes -- some plated, some non-plated. The second screenshot shows it overlaid with the top copper layer (GTL).

The slots which are surrounded by copper (the longish ones near the right board edge, as well as the near-circular ones to the left) were produced as plated slots. The large round holes were produced non-plated, since they are surrounded by a copper-free area.

No idea why the slots to the right are shown as fully milled-out, while the others are defined by contours only. There was no particular reason for that, as I recall, and it does not make a difference in what JLCPCB manufafctured.

[jpanhalt]

That is interesting.  Thank you for finding and posting those screenshots.  I was using EasyLogix.de as my Gerber viewer. 

I will save your post and use it as a guide with my next version of that board.

John

Edit:  Just FYI.  Here's a link to the Oshpark comments about slots.  It has been considerably updated since is did that board that needed slots.
https://docs.oshpark.com/design-tools/eagle/cutouts-and-slots/

[Northy]

They explicitly stated that they want to kill Eagle and re-implement the whole thing from scratch using fusion graphic engine. So, Eagle is not long for this world anyway. It will be dropped any time now.

Hi Alex,

When and where did they state this please?

Thanks,

G

[ataradov]

There was some blog post that described a graphical engine rewrite, which would only happen in Fusion and will not be backported to Eagle.

It was linked here, but I can't find the thread or a link right now. The only thing I could find is this forum post that talks about the engine rewrite https://forums.autodesk.com/t5/eagle-forum/future-of-eagle-and-fusion/td-p/10124680 . The blog post must be from around the same time.

[jpanhalt]

Here you go (11.27.21):

https://forums.autodesk.com/t5/eagle-forum/is-eagle-retired/m-p/10785448#M36081

Our decision to focus on the Fusion capabilities right now are not based on the move to cloud-storage nor is “cloud” the reason feature development has been slowed or stopped on EAGLE in recent months while our efforts focused on Fusion Electronics.

[Northy]

There was some blog post that described a graphical engine rewrite, which would only happen in Fusion and will not be backported to Eagle.

It was linked here, but I can't find the thread or a link right now. The only thing I could find is this forum post that talks about the engine rewrite https://forums.autodesk.com/t5/eagle-forum/future-of-eagle-and-fusion/td-p/10124680 . The blog post must be from around the same time.

Here you go (11.27.21):

https://forums.autodesk.com/t5/eagle-forum/is-eagle-retired/m-p/10785448#M36081

Our decision to focus on the Fusion capabilities right now are not based on the move to cloud-storage nor is “cloud” the reason feature development has been slowed or stopped on EAGLE in recent months while our efforts focused on Fusion Electronics.

Thank you both