Circuitmaker vs Eagle vs KiCad vs ??

[tooki]

I'll just add 3 cents about KiCAD.
I've used it for years, during uni and after.
Then I tried Altium at work which had a license, and man, never looked at KiCAD again

Looking back, from a perspective: KiCAD is great for begginers and overall entry level stuff. But it's Altium (or some other proffesional studio) that will get you places and will count on your resume. Altium's license cost is like nothing from company's POV.

Typical open source GUI app, IMHO. Tons of fans with modest needs calling it the [commercial app] killer, but lacking the features true professionals need. No, Writer is not equivalent to MS Word. No, the Gimp is not equivalent to Photoshop. And no, KiCAD is not equivalent to Altium.

The open source products don't need to be equivalent to the top end closed source product to be of value. True professionals who need and would benefit from Altium are pretty common around here, but not so common in the average workplace.

Much more frustrating are the people who swear black and blue they need the likes of Word or Photoshop, but really have basic needs and could use nearly any similar product. I see this very often at work with MS Office products. Brand whoring at its finest.

It’s not brand whoring. It’s that there’s often one deal-breaker feature (or annoyance) that is missing in LibreOffice, and just because you don’t know about it, or understand why it’s important, doesn’t mean it’s not a valid complaint.

Wanna talk about brand whoring? Look at the open source sycophants who insist emphatically that Gimp and LibreOffice are more than anyone would need. No amount of showing them that there’s a critical missing feature will ever convince them that you’re anything but an Adobe or MS “sheep” blindly following.

I’m the first person to say that the average home user, and even many business users, needs nothing more than Writer or Google Docs. But when people insist that those tools are sufficient for all uses, it makes me angry, because it’s basically telling me that I’m wrong or stupid for relying on useful or necessary features in Word. But I’m not wrong, I use Word because a) I know it well and like it, and b) for certain documents, rely on features not present in the others. It’s all about using the right tool for the job. Sometimes it’s one thing, sometimes it’s another.

Yes work places like standards and despite MS office offering to use open document formats it still warns you you could loose data if you save a new file with them and encourages you to use MS formats

Which is absolutely correct, because the open formats do not actually fully represent the internal object model, never mind all the features, in Office (especially Word). It’s not that Microsoft doesn’t try, it’s that the open formats simply aren’t designed for Word.

[Mr. Scram]

I can't stand the incessant evangelism of quite a few open source zealots. If you want to privately make a hobby out of being as open souce as possible, that's fine and probably adds to it being and becoming a worthwhile alternative. There does regularly seem to be a complete lack of realism when it comes to other people than themselves, though. Most people don't want to live with all kinds of concessions to avoid a few binary blobs or anything else not completley in line with the open source mindset. They just want their software to work and work reasonably well. Until the zealots get this, it's just holding the open source community back as a whole. It's never going to go prime time when people insist on being strict. People want solutions, not more problems.

I really like the concept of doing things open source, but understand the practicalities of the real world and how it's not always feasible to go that route. That makes an "everything open or bust" attitude rather counterproductive. It's not just the application itself either, it's how it meshes with its surroundings and infrastructure and practicalities of being able to hire people who know how to use it too.

[poorchava]

I'm wondering how will things go with Eagle. It seems that Autodesk has been pumping quite a lot of money into it's development since they've bought it. IIRC recently they've added a push&shove routing and 3d interface to Fusion 360. Diptrace has been better than Eagle for a few years, but now unless Novarm steps up their development game, they are gonna loose customers to Eagle.

KiCAD is the same as any other open-source engineerign related software - kinda sorta resembles the real thing, and some people are almost able to show you that you can create a professional-looking product with it, but in the end it's lacking some basic features. Reason still the same: it's created by software developers who are at best hobbyists, not professionals and don;t know how this should work and what is required of such software.

[hendorog]

I'll just add 3 cents about KiCAD.
I've used it for years, during uni and after.
Then I tried Altium at work which had a license, and man, never looked at KiCAD again

Looking back, from a perspective: KiCAD is great for begginers and overall entry level stuff. But it's Altium (or some other proffesional studio) that will get you places and will count on your resume. Altium's license cost is like nothing from company's POV.

Typical open source GUI app, IMHO. Tons of fans with modest needs calling it the [commercial app] killer, but lacking the features true professionals need. No, Writer is not equivalent to MS Word. No, the Gimp is not equivalent to Photoshop. And no, KiCAD is not equivalent to Altium.

The open source products don't need to be equivalent to the top end closed source product to be of value. True professionals who need and would benefit from Altium are pretty common around here, but not so common in the average workplace.

Much more frustrating are the people who swear black and blue they need the likes of Word or Photoshop, but really have basic needs and could use nearly any similar product. I see this very often at work with MS Office products. Brand whoring at its finest.

It’s not brand whoring. It’s that there’s often one deal-breaker feature (or annoyance) that is missing in LibreOffice, and just because you don’t know about it, or understand why it’s important, doesn’t mean it’s not a valid complaint.

Wanna talk about brand whoring? Look at the open source sycophants who insist emphatically that Gimp and LibreOffice are more than anyone would need. No amount of showing them that there’s a critical missing feature will ever convince them that you’re anything but an Adobe or MS “sheep” blindly following.

I’m the first person to say that the average home user, and even many business users, needs nothing more than Writer or Google Docs. But when people insist that those tools are sufficient for all uses, it makes me angry, because it’s basically telling me that I’m wrong or stupid for relying on useful or necessary features in Word. But I’m not wrong, I use Word because a) I know it well and like it, and b) for certain documents, rely on features not present in the others. It’s all about using the right tool for the job. Sometimes it’s one thing, sometimes it’s another.

We are talking about different things. My problem is with _non-professional_ users who insist they need Office/Photoshop et-al because it is Office or Photoshop. They are the definition of brand whores. These people who actually can't do anything in Office, but will bitch if you give them anything else.

If you are an advanced or expert and can properly drive Word or Excel or Photoshop or Illustrator or whatever, then no problem, get what you need. However in your average workplace those people are not common. Instead they should get Google Docs.  LibreOffice isn't really a good alternative despite the saving in licences. The support costs are the same or greater for that. The challenge though is when one person needs something specific and expensive but also needs to collaborate with someone else who doesn't.

Really, the open source zealots are no worse than the Apple zealots and the Microsoft zealots and the x zealots. I have sat in so many meetings and copped abuse from numpties who think Outlook is the only email client in the world, Exchange is essential for life, and think everyone should have a Domain, OTOH they have never heard of LDAP and think unix is some sort of weird sex cult. Conversely, I don't ever recall sitting in a meeting where someone tried to stuff open source software down my throat.
Hopefully we all grow up at some point and realise that one size doesn't fit all.

Funnily enough LibreOffice Writer and it's ancestors back to Star Writer have always been (IMHO) more intuitive than Word in its use of Styles, and your average business user would do well to learn it as they will then not produce such complete shit Word documents. However my impression is that now it is starting to look more and more like Word so that advantage is going if not gone.

Yes work places like standards and despite MS office offering to use open document formats it still warns you you could loose data if you save a new file with them and encourages you to use MS formats

Which is absolutely correct, because the open formats do not actually fully represent the internal object model, never mind all the features, in Office (especially Word). It’s not that Microsoft doesn’t try, it’s that the open formats simply aren’t designed for Word.

Hmm, Microsoft did actually specify the xml format themselves... They just didn't bother doing everything in XML which is the issue. I have spent some time in the depths of MS Word files, and the XML stuff is manageable enough, but the binary blobs are not. I also found that Word was reasonable  to work with at the XML level, but Powerpoint was not. It was quite buggy and would crash and generally misbehave badly loading files which were valid but which were modified outside of Powerpoint.

[paulca]

I think the Open Source haters need to look around and remind themselves how much of it they actually use.  Also how much their proprietary software uses underneath and the infrastructure that allows them to use it uses.  Almost everything on the internet is built on Open Source tech stacks.  Open Source pretty much dominates general computing and network infrastructure because it encourages interoperability.  Even Windows is full of it if you look close enough.

I can understand specific professional domains favouring some commercial software because they provide certain defacto features, but at the same time the disadvantage of this paradigm is it stifles innovation and agility.  The software become slave to the big spenders who have the clout to demand features or in some cases demand the lack of them, or even more often the lack of change.

The biggest thing that attracts me to Open Source software is freedom.  Freedom from corporate boxing in, proprietary formats which don't work outside of that companies golham licensing.  Freedom from being told how I should use someone's software.  The fact that, in a large part, things work together. This allows entire stacks to come from different vendors, which allows component-ization, reuse and component interchange which is simply not available in commercial proprietary software. On Linux you can swap whole layers and expect the layers above and below will still play nice.  Linux entire desktop environments, logon managers or window managers.

Analogously - Microsoft will provide you with a fully working PCB for your design, but it's potted and has only proprietary connectors for other Microsoft extension PCBs.  Linux on the other hand is not potted and provides open standard connectors to work with any other PCBs.   You can swap ICs or components because not only are the component standarised (though custom components exist), the full schematics are publicly published, they are at least fairly well documented and there is nothing stopping you modifying the design should you need.

In commercial ventures however this is often seen as a disadvantage.  It is often much more preferencial to have a single vendor, single support contract, single point to complain to and that point should be large, persistent and hopefully at the right price responsive.

That said most proprietary vendors are embracing support for open source software.  You can program Python in Visual Studio, although in a limited way.   Still they peddle their locked in formats and openly discourage inter-operability with other software to keep you locked in and pumping revenue for them.  At the same time, they steal innovations from Open Source software just as often as open source tries to emulate or be compatible with closed source software.  KDE and Windows have been exchanging features for decades.  It's an odd relationship.

[Mr. Scram]

I think the Open Source haters need to look around and remind themselves how much of it they actually use.  Also how much their proprietary software uses underneath and the infrastructure that allows them to use it uses.  Almost everything on the internet is built on Open Source tech stacks.  Open Source pretty much dominates general computing and network infrastructure because it encourages interoperability.  Even Windows is full of it if you look close enough.

I can understand specific professional domains favouring some commercial software because they provide certain defacto features, but at the same time the disadvantage of this paradigm is it stifles innovation and agility.  The software become slave to the big spenders who have the clout to demand features or in some cases demand the lack of them, or even more often the lack of change.

The biggest thing that attracts me to Open Source software is freedom.  Freedom from corporate boxing in, proprietary formats which don't work outside of that companies golham licensing.  Freedom from being told how I should use someone's software.  The fact that, in a large part, things work together. This allows entire stacks to come from different vendors, which allows component-ization, reuse and component interchange which is simply not available in commercial proprietary software. On Linux you can swap whole layers and expect the layers above and below will still play nice.  Linux entire desktop environments, logon managers or window managers.

Analogously - Microsoft will provide you with a fully working PCB for your design, but it's potted and has only proprietary connectors for other Microsoft extension PCBs.  Linux on the other hand is not potted and provides open standard connectors to work with any other PCBs.   You can swap ICs or components because not only are the component standarised (though custom components exist), the full schematics are publicly published, they are at least fairly well documented and there is nothing stopping you modifying the design should you need.

In commercial ventures however this is often seen as a disadvantage.  It is often much more preferencial to have a single vendor, single support contract, single point to complain to and that point should be large, persistent and hopefully at the right price responsive.

That said most proprietary vendors are embracing support for open source software.  You can program Python in Visual Studio, although in a limited way.   Still they peddle their locked in formats and openly discourage inter-operability with other software to keep you locked in and pumping revenue for them.  At the same time, they steal innovations from Open Source software just as often as open source tries to emulate or be compatible with closed source software.  KDE and Windows have been exchanging features for decades.  It's an odd relationship.

I think you're touching upon another often seen issue with open source, which is the massive fragmentation. Every disagreement seems to be settled by a fork, so there's a huge amount of mostly similar alternatives out there. That's both a blessing and a curse, but tends to be the latter if you want to standardize something in an organisation and expect people to actually work with it. The variants that seem to do well generally have large orgnisations with good support behind them, which for many intents and purposes is the same as any proprietary supplier.

Like you say, the world wouldn't be the same without open source and I think the practical sweet spot is a healthy mix of both.

[VEGETA]

Every disagreement seems to be settled by a fork

You just summarized the biggest problem in one line. You take my "Sayian prince award" for sure!!

One big example of this is mplayer. This one is very good alternative to Windows-only packages which relies on codecs based on Microsoft's directshow. However, they perhaps disagreed and someone said "hey let's fork it and make it the best ever player!".

They went on with "mplayer2". The problem is now we have both mplayer and mplayer2 being developed at the same time! which one to choose? However, VLC gave a comprehensive solution despite being weaker than mplayer-based players simply because it is one package from one source which gets developed by a true team.

You guess it... later on some people made "MPV" which is the latest one forked or inspired by both mplayer and mplayer2... this one is good and maintained by true experts in the field.

Same thing happened to media player classic as they made "media player classic - home cinema" and other made even more forks...!

They could have just stayed with one package, one name, and one team!

[agehall]

Holy crap! This thread got me to watch the guided tour of Diptrace and that just knocked me out of my chair!

I started with Eagle a few years ago and quite liked it. I was about to buy a license for it instead of just using the free license when they were bought and switched to the subscription model which I feel isn't very good for me. When I looked at the different packages available, I ended up going with KiCad because it was free had "no" limitations etc etc but boy, this choice made my life painful!

I don't think there is anything you can't design in KiCad if you put your mind to it, but there are so many small things that are just annoying that I don't even know where to begin. Just turning layers (i.e. more than one at the time) on and off is a pain. I still have not figured out how the heck traces are routed when I move the wire tool around 100%. Moving a component after wiring it up (either in schematic or PCB view) seems to make the worst choices available etc etc.

Basically, nothing that makes it impossible to use, but the usability of the package is horrible.

I'm going to download Diptrace ASAP and give it a spin. If it does what I expect it to, I'll be buying a license from them pretty soon. I'm a pure hobbyist and when I design things, I want the best tools I can afford that helps me accomplish what I want and that are easy to work with. As I can afford more than $0 for my tools, I'm quite happy to buy myself a better experience when designing PCBs.

[poorchava]

Holy crap! This thread got me to watch the guided tour of Diptrace and that just knocked me out of my chair!

I started with Eagle a few years ago and quite liked it. I was about to buy a license for it instead of just using the free license when they were bought and switched to the subscription model which I feel isn't very good for me. When I looked at the different packages available, I ended up going with KiCad because it was free had "no" limitations etc etc but boy, this choice made my life painful!

I don't think there is anything you can't design in KiCad if you put your mind to it, but there are so many small things that are just annoying that I don't even know where to begin. Just turning layers (i.e. more than one at the time) on and off is a pain. I still have not figured out how the heck traces are routed when I move the wire tool around 100%. Moving a component after wiring it up (either in schematic or PCB view) seems to make the worst choices available etc etc.

Basically, nothing that makes it impossible to use, but the usability of the package is horrible.

I'm going to download Diptrace ASAP and give it a spin. If it does what I expect it to, I'll be buying a license from them pretty soon. I'm a pure hobbyist and when I design things, I want the best tools I can afford that helps me accomplish what I want and that are easy to work with. As I can afford more than $0 for my tools, I'm quite happy to buy myself a better experience when designing PCBs.

Don't get it wrong. While Diptrace has been better than Eagle for a few years because it was developing, while Eagle was pretty much stuck, now things may change since Autodesk has bought Eagle and is apparently putting some serious resources into development. Diptrace development was never very fast, although support seemes to respond to issues and questions pretty quickly. Still, Autodesk can throw orders of magnitude more resources at making Eagle a useful software package.

As for open source being open and free to use it how u see fit and so on, I think in case of PCB development, your job is to develop PCBs, not fiddle with software and make imprevements to it. It does not carry any value added with it, so no employer will devote resources into it. In reality neither my employer nor myself as a business owner after hours give a flying shit if something is open and I can improve it so that it finally works. It either works and produces something that can be sold or it doesn't. If it doesn't - it doesn't make money and is not worthwhile to use.

[agehall]

Don't get it wrong. While Diptrace has been better than Eagle for a few years because it was developing, while Eagle was pretty much stuck, now things may change since Autodesk has bought Eagle and is apparently putting some serious resources into development. Diptrace development was never very fast, although support seemes to respond to issues and questions pretty quickly. Still, Autodesk can throw orders of magnitude more resources at making Eagle a useful software package.

Well, until Autodesk changes it back to a product I can buy, it will probably not be my first choice ever. My preference is to own software.

[VEGETA]

now things may change since Autodesk has bought Eagle and is apparently putting some serious resources into development

We all said that until we knew about subscription based approach. No one will ever be convinced that this is the best way to guarantee product's development. It is mainly for gaining tons of money from users. I don't believe anyone should get subscription software, especially that there are lots of better and cheaper packages elsewhere.

Autodesk can throw orders of magnitude more resources at making Eagle a useful software package.

Yes, Autodesk can release the kraken on it and it will be better, but still subscription based.

Autodesk has the ability to make eagle one of the best packages, but this approach is just never gonna work. will you pay 500$ per year for a software that you can buy its alternative for the same price and keep it for you? NO! because next year you should pay another 500$ and so on! while if you got CS or diptrace, it will be yours forever... you just pay for maintenance and vault.

[^_^]

I don't think there is anything you can't design in KiCad if you put your mind to it, but there are so many small things that are just annoying that I don't even know where to begin. Just turning layers (i.e. more than one at the time) on and off is a pain. I still have not figured out how the heck traces are routed when I move the wire tool around 100%. Moving a component after wiring it up (either in schematic or PCB view) seems to make the worst choices available etc etc.

Greatest pain KiCAD gave me was when trying to multiply single circuit on a PCB. On schematic you can do dumb Ctrl+C Ctrl+V, but on PCB...
Also after you see how one can make footprints in Altium (using footprint wizard for standard packages) it's all just "what have I been doing with my life "
Haha

[paulca]

Greatest pain KiCAD gave me was when trying to multiply single circuit on a PCB. On schematic you can do dumb Ctrl+C Ctrl+V, but on PCB...
Also after you see how one can make footprints in Altium (using footprint wizard for standard packages) it's all just "what have I been doing with my life "
Haha

Of course the plus point of KiCad is that all the files are free text format.  So you can write scripts to copy blocks with a lateral offset and place them back into the schematic.  There are a few canned solutions for this out there already.

Someone mentioned dragging/moving wired components on a PCB with the traces still attached in KiCad... is that even possible?

[DrGeoff]

Holy crap! This thread got me to watch the guided tour of Diptrace and that just knocked me out of my chair!

I recently switched out of Protel99SE after using it for 17 years and migrated all the custom libraries, many circuit and PCB designs within the evaluation period to see how it performs as a replacement (as P99SE is long unsupported). I found it is be a satisfactory replacement and bought a license, and have since migrated another 100 or so designs across. It appears stable and very usable.

[bson]

I can't stand the incessant evangelism of quite a few open source zealots.

That view is a fallacy; I know, because I've stood in the same trap myself in the past.  I worked on the kernel network components and related libraries (remember libsocket) of Sun Solaris at Sun Microsystems for five years in the 90s, and looked at Linux: nah, not going anywhere.  Fragile.  Not usable for production systems.  Lacks critical production system features.  Sun CC is so much better than gcc.  Teamware beats plain sccs/rcs.  Where is Sun Microsystems today?  Doesn't exist.  Sun CC?  Doesn't exist.  Teamware?  Doesn't exist and git dominates.  Solaris?  Just kicked to the curb by Oracle last year, all but replaced by Linux.

If you imagine this can't happen to Altium or anyone else you're delusional.  Why did Linux become so prominent?  Because it was worked on by competent, skilled engineers hired and paid to do so.  Gcc?  Same.  If you think successful open source projects are the result of "hobbyists" you've missed the boat completely.  KiCAD development is led by equally competent paid engineers - in fact it wasn't until CERN started funding it that it actually started to become more than a toy.  And if you go look at the roadmaps, planning materials, presentations, schedules, and commit logs closing out items on the roadmap you can see this is not someone's hobby project.  And THIS is the critical aspect of it, and exactly why Altium and every other vendor in this space should be concerned.

[poorchava]

With Linux it's the case of software developers making something they'd use on their own and they know how they'd like to use it. I'd say one can be either a competent software developer or a competent pcb designer, but both is probably very, very rare and such people have enough stuff to do. For the said Linux case, one has only got to be a software developer.

What open source engineering-related softwares need is constant feedback from the industry that is using those tools on what to improve and/or change.

[Electro Fan]

With Linux it's the case of software developers making something they'd use on their own and they know how they'd like to use it. I'd say one can be either a competent software developer or a competent pcb designer, but both is probably very, very rare and such people have enough stuff to do. For the said Linux case, one has only got to be a software developer.

What open source engineering-related softwares need is constant feedback from the industry that is using those tools on what to improve and/or change.

+1
These are subtle, very insightful, and important distinctions

[hendorog]

With Linux it's the case of software developers making something they'd use on their own and they know how they'd like to use it. I'd say one can be either a competent software developer or a competent pcb designer, but both is probably very, very rare and such people have enough stuff to do. For the said Linux case, one has only got to be a software developer.

What open source engineering-related softwares need is constant feedback from the industry that is using those tools on what to improve and/or change.

+1
These are subtle, very insightful, and important distinctions

Yeah, nah.

Did you guys miss the comment above? High profile open source software like KiCad and certainly Linux is not developed by hobbyists. Read that again.

That means it is mostly developed by paid professionals - just as commercial software is. Therefore what you are saying applies just as much to  commercial products as it does to the open source ones.

_All_ software development needs constant feedback from subject matter experts.

[Mr. Scram]

That view is a fallacy; I know, because I've stood in the same trap myself in the past.  I worked on the kernel network components and related libraries (remember libsocket) of Sun Solaris at Sun Microsystems for five years in the 90s, and looked at Linux: nah, not going anywhere.  Fragile.  Not usable for production systems.  Lacks critical production system features.  Sun CC is so much better than gcc.  Teamware beats plain sccs/rcs.  Where is Sun Microsystems today?  Doesn't exist.  Sun CC?  Doesn't exist.  Teamware?  Doesn't exist and git dominates.  Solaris?  Just kicked to the curb by Oracle last year, all but replaced by Linux.

If you imagine this can't happen to Altium or anyone else you're delusional.  Why did Linux become so prominent?  Because it was worked on by competent, skilled engineers hired and paid to do so.  Gcc?  Same.  If you think successful open source projects are the result of "hobbyists" you've missed the boat completely.  KiCAD development is led by equally competent paid engineers - in fact it wasn't until CERN started funding it that it actually started to become more than a toy.  And if you go look at the roadmaps, planning materials, presentations, schedules, and commit logs closing out items on the roadmap you can see this is not someone's hobby project.  And THIS is the critical aspect of it, and exactly why Altium and every other vendor in this space should be concerned.

What exactly is the fallacy? I don't see an argument that's different from anything I've said. I addressed the larger Linux distros and applications being close to or the same as closed source, as they're developed by large paid teams with all the benefits and drawbacks that brings. I also addressed similar things happening to closed sources alternatives.

[Electro Fan]

With Linux it's the case of software developers making something they'd use on their own and they know how they'd like to use it. I'd say one can be either a competent software developer or a competent pcb designer, but both is probably very, very rare and such people have enough stuff to do. For the said Linux case, one has only got to be a software developer.

What open source engineering-related softwares need is constant feedback from the industry that is using those tools on what to improve and/or change.

+1
These are subtle, very insightful, and important distinctions

Yeah, nah.

Did you guys miss the comment above? High profile open source software like KiCad and certainly Linux is not developed by hobbyists. Read that again.

That means it is mostly developed by paid professionals - just as commercial software is. Therefore what you are saying applies just as much to  commercial products as it does to the open source ones.

_All_ software development needs constant feedback from subject matter experts.

Hi hendorog,

I think we might be saying the same thing, or something similar.

The intention wasn't in any way to take a shot at KiCad, or Open Source.

The point was just that in the case of Linux a software developer can often be both an end user and a software developer.  As an end user a software developer has a pretty good idea of what he or she would like in software in general and software OS's and tools in particular.

In the case of other non-core IT applications (hardware circuit design software for hardware engineers, medical imaging software for radiologists, etc, etc, etc.) the end result (software that is enjoyable and productive to use, and hopefully reliable and affordable too) often requires input from the end users to the software developers.  This collaboration between end users and software developers happens very naturally in the case of Linux (or other similar open source projects) because the end users and the software developers are often the same people – just wearing two hats.

Whether an application development project is open source or a conventional commercial project the benefit of having end user input to the software developers is generally very important to the success of the project, of course.  In both cases (open or commercial) the software developers can have a range of skills (from beginners to pros, and pro’s pro’s); the better the skill level the more likely the success.  In both cases, the key is often knowing which end users to listen to.  Often this requires some decision as to who the application is going be built for (beginning end users, intermediate end users, advanced end users, power users, etc – or some combination.  Or maybe the market is segmented by enthusiast, intermediate/advanced amateur, freelance pro/small company pro, big company pro, etc.).  In any event - to your point - collaboration between software developers and end users is generally pretty important for both commercial and open source projects.

The point I think poorchava was trying to make (and that I agree with) is simply that Linux is an interesting (and somewhat unique) use case (relative to the broad spectrum of uses cases) for software development because many of the end users are software developers and vice versa.  This isn’t generally the case with most application software.  As poorchava pointed out, perhaps some hardware circuit design engineers (end users) are also software developers, but if so, they are probably pretty busy and they have to choose how much time they want to spend wearing one hat or the other, or both of their two hats (or maybe yet other hats if they are that talented.)

My guess is we have some forum members who would like EDA tools that will help automate design and build processes - and there are others here who would like to help build EDA tools as well as use them.  Most of us would at least like to find a tool that helps us become comfortably productive, and to the extent we don't find such tools we sometimes share our input as end users regarding our "requirements" for ease of use, functionality, price, etc.  The chances of our specific end user requirements/preferences making it into some list of product requirements is not real high so we get to choose between the products available, unless we happen to be software developers who can make modifications - to the extent that is practical and permissible.  In the case of Linux it's reasonably practical and permissible, in the case of EDA it's maybe somewhat less common but still feasible.
 
EF

[jmelson]

Holy crap! This thread got me to watch the guided tour of Diptrace and that just knocked me out of my chair!

I recently switched out of Protel99SE after using it for 17 years and migrated all the custom libraries, many circuit and PCB designs within the evaluation period to see how it performs as a replacement (as P99SE is long unsupported). I found it is be a satisfactory replacement and bought a license, and have since migrated another 100 or so designs across. It appears stable and very usable.

Switched to Diptrace, I'm assuming?  Is it really that much better?  I've been using Protel 99SE for ages, and know where all the bodies are buried.  It certainly does just about everything I need.  Not only plain schematic and PCB, but I also know how to build schematic and PCB library parts, how to use the pick & place output, etc.
And, I trust the PCB design rule checks completely.  Just because there's no updates to P99 doesn't mean it is dead.  It still works fine.  I've also made a number of customizations to P99 so I can use keyboard shortcuts to do a lot of stuff, that is much better for me than clicking on menu items.

I got a design in KiCAD, and it does work, but I don't completely trust the design rule checks, and the rest just seems cumbersome.  I think I can make MUCH better schematics in P99, but the KiCAD PCB stuff seems to be pretty good.  But, I'm nowhere close to an expert in KiCAD.  I've been hoping KiCAD development continues, and maybe someday I'll make the switch.

Jon

[Simon]

Just downloaded KiCAD for a quick looksy at what progress has been made. Thing that i would have a problem with is the need to manually link footprints to the schematic. I'm at the point now where I pick the symbol, footprint, 3D model and MPN the moment I put a symbol on the schematic. that way design is much faster.

[jgalak]

Just downloaded KiCAD for a quick looksy at what progress has been made. Thing that i would have a problem with is the need to manually link footprints to the schematic. I'm at the point now where I pick the symbol, footprint, 3D model and MPN the moment I put a symbol on the schematic. that way design is much faster.

My understanding is that it's a deliberate design decision, letting you choose components in the schematic before worrying about footprints. 

Honestly, I don't love it, especially since it's inconsistent - it works fine for generic components - passives, diodes, etc., but not for more complex ICs which are only available in a few packages and those packages may have different numbers of pins (necessitating different schematic symbols).

I also often start by thinking about footprints, especially if trying to fit a certain size budget.

But I can live with the annoyance.

[poorchava]

Just downloaded KiCAD for a quick looksy at what progress has been made. Thing that i would have a problem with is the need to manually link footprints to the schematic. I'm at the point now where I pick the symbol, footprint, 3D model and MPN the moment I put a symbol on the schematic. that way design is much faster.

My understanding is that it's a deliberate design decision, letting you choose components in the schematic before worrying about footprints. 

Honestly, I don't love it, especially since it's inconsistent - it works fine for generic components - passives, diodes, etc., but not for more complex ICs which are only available in a few packages and those packages may have different numbers of pins (necessitating different schematic symbols).

I also often start by thinking about footprints, especially if trying to fit a certain size budget.

But I can live with the annoyance.

It's a good example of a difference between "nice feature" (thought the developer) and "WTF" (said almost anybody, who's ever worked in a bigger engineering company). In my book the one and only correct way to do libraries is when you have a single library entry for a single order code that is automatically put onto the BOM. And that generally has one footprint, perhaps some 2-3 variants in case of THT parts where leads can be bent in different ways or SMD L-N-M options.

[DrGeoff]

Holy crap! This thread got me to watch the guided tour of Diptrace and that just knocked me out of my chair!

I recently switched out of Protel99SE after using it for 17 years and migrated all the custom libraries, many circuit and PCB designs within the evaluation period to see how it performs as a replacement (as P99SE is long unsupported). I found it is be a satisfactory replacement and bought a license, and have since migrated another 100 or so designs across. It appears stable and very usable.

Switched to Diptrace, I'm assuming?  Is it really that much better?  I've been using Protel 99SE for ages, and know where all the bodies are buried.  It certainly does just about everything I need.  Not only plain schematic and PCB, but I also know how to build schematic and PCB library parts, how to use the pick & place output, etc.
And, I trust the PCB design rule checks completely.  Just because there's no updates to P99 doesn't mean it is dead.  It still works fine.  I've also made a number of customizations to P99 so I can use keyboard shortcuts to do a lot of stuff, that is much better for me than clicking on menu items.

I got a design in KiCAD, and it does work, but I don't completely trust the design rule checks, and the rest just seems cumbersome.  I think I can make MUCH better schematics in P99, but the KiCAD PCB stuff seems to be pretty good.  But, I'm nowhere close to an expert in KiCAD.  I've been hoping KiCAD development continues, and maybe someday I'll make the switch.

Jon

Yes, Diptrace.
It's better in that it is supported for future OS changes, whereas P99SE is likely to stop working one day after an OS update. I needed to be able to transfer significant amounts of IP to a new system before this happens. Diptrace appears to be very close in functionality to P99SE, so the time to get started was very short. The evaluation exercise convinced me that it was an appropriate replacement as I managed to easily reproduce my schematic and PCB libs from P99SE to Diptrace and then reproduce a few designs (as an exercise in navigating the software). There are a few differences with schematic/PCB interaction which are relatively minor however overall the results have been satisfactory and PCB's have been manufactured from it now. I rarely use autorouter functionality so I cannot compare this adequately.