seadynamic8 (10) [Avatar] Offline
#1
The subtitle of this book is "With Node, Electron, and NW.js", but the summary doesn't mention it at all.

So my question is how is Electron covered since even though it's end goal is same as NW.js, it approaches it differently.

Do you also show all the examples in the book using Electron?
184533 (3) [Avatar] Offline
#2
My thought exactly - I bought it for it's stated Electron coverage and there's nothing there. I wish I'd looked at the intro chapter first and would have avoided this one - terrible diagrams, full of typos and woolly hand-waving content.
paulbjensen (39) [Avatar] Offline
#3
Hi,

I'm the book author, I understand where you're coming from.

The book was originally going to cover only NW.js, but given the similarity of Electron, it's scope has been expanded to cover that framework too, and it was announced recently that the book's title was changed from "NW.js in Action" to "Cross-Platform Desktop Applications" to reflect this change. If you've bought the book after the title change, then I can see how confusing it is for readers expecting to see Electron in the book.

At this point in time I'm adding new content for Electron into the book, but this is no small task and at the moment the book's content is still covering NW.js only, but that will change shortly.

I apologize for the confusion. I'm working as hard as I can to make sure that the content on Electron gets into the book soon.

Regards,

Paul Jensen
184533 (3) [Avatar] Offline
#4
Hi Paul

Thanks for getting back - I was indeed expecting plenty of Electron coverage! I was considering looking for a refund but will just hang onto it now. No offence, but I think the diagrams need a bit of work - they are so fuzzy they're sometimes impossible to read. I also felt a few of them didn't really impart any meaningful info and that the first chapter in general was a bit hand-waving in terms of explaining the architecture, but I do realise it's a MEAP book.

Kind regards, Bob
paulbjensen (39) [Avatar] Offline
#5
Hi Bob,

Thanks for the feedback, I'll adjust the graphics to not be fuzzy, and I pushed up a new draft of Chapter 1 in the past week, so once it goes into the MEAP please let me know where the weak points in the chapter are, hopefully they may have been ironed out in the new draft, but if not let me know.

Regards,

Paul Jensen
seadynamic8 (10) [Avatar] Offline
#6
Hi Paul,

That's great to hear, but yes it was confusing because I couldn't find any Electron info in the book. I'm eagerly awaiting the Electron coverage as well.

Thanks.
49465 (9) [Avatar] Offline
#7
Recommend abandoning nw.js, focus exclusively on Electron
You can't be blamed for starting this project 9 months ago NW.js as an exclusive focus, as Electron hadn't by then superceded NW.js.

However, at this stage the decision to pay any more attention to NW.js effectively dooms this book. There isn't an informed developer that doesn't believe Electron has taken over.

I share your pain as I've personally had to convert a project from nw.js to Electron, but it was obviously the right decision. There are hundreds of other OSS projects that have done the same. I don't know of a single switch in the other direction.

The right course for the MEAP's buyers, for Manning as a business, and you as the author:
1) focus all your resources on Electron, and don't let NW.js distract you. Don't fool yours of into thinking you can do both, especially as Electron rapidly takes on exciting new functionality that nw.js will lack. This is exactly the kind of situation that leads to MEAPs being terminated.
2) put "Electron" in the title of the book (the new title won't capture the distinct and large market for Electron)
3) for the sake of the less informed buyer, stop promoting NW.js as a viable alternative to Electron. You're doing them a big disservice.
49465 (9) [Avatar] Offline
#8
Recommend abandoning NW.js, focus exclusively on Electron
49465 wrote:You can't be blamed for starting this project 9 months ago with NW.js as an exclusive focus, as Electron hadn't by then superseded by NW.js.

However, at this stage the decision to give NW.js equal treatment to Electron effectively dooms this book. There isn't an informed developer that doesn't believe Electron has taken over, and that NW.js is in rapid decline.

I share your pain as I've personally had to convert a project from nw.js to Electron, but it has proven the right decision. There are scores of other OSS projects that have made the same switch. I don't know of a single switch of late in the other direction (from Electron to NW.js).

I've paid my money and invested time in reading this MEAP, so I'm comfortable offering what I believe is right path forward for all parties -- the MEAP's buyers, Manning as a business (which ultimately will have to justify its investment in this book), and you as the author. Here it is:
1) Focus all your resources on Electron, and don't let NW.js distract you. Don't fool yours of into thinking you can cover both, especially as Electron rapidly takes on exciting new functionality that NW.js will lack. This is exactly the kind of situation that leads to MEAPs being terminated by Manning.
2) Put "Electron" in the title of the book. The new title doesn't capture the distinct and large market for Electron, and will lose you sales.
3) For the sake of the less informed buyer, stop promoting NW.js as a viable alternative to Electron.

paulbjensen (39) [Avatar] Offline
#9
NW.js
[ 327 KB ]
Stars graph
[ 249 KB ]
Hi,

Thanks for your comment. I'm open-minded (you have to be to adapt a book just towards reaching the end as we did when we decided to cover Electron as well at Xmas). There's a few things I wanted to query.

In terms of the popularity of both projects, I think that Electron is definitely hot right now, if the stars graph below is anything to go by (see attached doc).

Is NW.js in rapid decline? I'm not sure, if anything the new releases last month indicated that there is a flurry of activity at the moment, which is good on one hand but annoying on the other when you have to incorporate changes to your code examples just as you're trying to wrap up the book.

I agree that I've been evidence of people moving apps over to Electron from NW.js (Light table for example), if you know of other examples it would be good to find out. In terms of the other way I don't see anyone porting from Electron to NW.js. Where NW.js does come in handy though is the ability to run its applications on Windows XP. To you and me that is an ancient operating system, but in other parts of the world (China being a significant factor) it is still in use. Electron works on only Windows 7 and above.

I will pass on the suggestions to the development editor and we will discuss it.

49465 (9) [Avatar] Offline
#10
I just saw the newly announced Electron MEAP, expecting to find that it was in fact this book, rebranded and, appropriately, leaving NW.js behind. But, it looks like Manning now has 2 MEAPs covering the same topic.

Based on Manning's history in these situations, I'd expect this MEAP to be cancelled soon. I hope Manning will inform buyers of its plans ASAP so as to save us time.
paulbjensen (39) [Avatar] Offline
#11
Hi,

Thanks for your reply.

We did discuss your suggestion a few weeks back, and we agreed that Electron does have a lot of momentum right now so we'd tailor the content as appropriate, but the book would remain covering both frameworks. After all, we are talking about 2 projects that have over 28,000 and 30,000 stars on Github each, making them numbers #23 and #25 on Github's most starred projects of all time. Also, although the frameworks have distinct differences in architecture, they share a lot of common functionality, no doubt in part to the fact that Cheng Zhao has worked on both of them.

With regards to this whole "cover Electron only and drop NW.js entirely" premise that you've suggested, I did a bit of homework to try and ascertain the data behind what you suggested. I agree that Electron has definitely got greater momentum (as evidenced by the reaction to the 1.0 release as well as the fact that they've got more stars than NW.js now, and are close to surpassing Ruby on Rails' count of 31k stars). But does this mean that NW.js should be dropped entirely?

If you do a quick search on Google, you'll find quite a few links comparing both frameworks, their feature sets and which one to go for:

- https://www.xplatform.rocks/2016/02/09/nw-js-vs-electron/
- http://tangiblejs.com/posts/nw-js-and-electron-compared-2016-edition
- https://www.linkedin.com/pulse/electron-nwjs-zac-anger
- http://www.tivix.com/blog/nwjs-and-electronjs-web-technology-desktop/
- http://www.akawebdesign.com/2015/05/06/electron-vs-nwjs/

Some of those articles are within the last couple of months. They compare feature sets, what can and can't be done with both frameworks, and which option to go for (depending on your needs). This is clearly a question that desktop app developers have, and one that this book is uniquely positioned to answer. In fact, dare I say it, if someone found themselves having to port a legacy NW.js app to Electron (which you did), they may want something like this book to help them see how the frameworks differ architecturally, and what the equivalent APIs are across both frameworks. That said too, for those wanting to create Chrome apps for installing on Chromebooks (which outsold MacBooks in the US recently), they too will want to know that NW.js supports that whereas Electron currently doesn't.

I also looked at the number of issues on Github for both projects in order to compare them. Both look like active projects, but yes Electron is definitely on top of their Pull Requests and manages to close issues pretty quickly. When you said that NW.js was in rapid decline, what was the basis for that? Was it the number of issues (they seem to get created just as quickly as Electron), or was it the number of commits over time? Is it activity in the Gitter channel? It would be nice to have some data to relate to that statement.

The book has 11 out of 12 chapters produced, with content being added for remaining chapters to cover Electron. It's close to being finished, and to kill the book because it covers building cross-platform desktop apps across 2 frameworks instead of 1? I'm stumped.
49465 (9) [Avatar] Offline
#12
If you are close to competing the book, I wouldn't presume to tell you to stop.

In the meantime Manning has just launched a brand new MEAP dedicated exclusively to Electron, which would make it the obvious choice for Electron developers.

Electron has not only a large dedicated development team at GitHub, but is now strategic to Microsoft and Facebook. It's future is guaranteed.

You might enquire in the NW.js community whether Intel, for whom the project is not strategic, intent.ds to continue paying the 1 full-time developer. Perhaps Intel might be openly suggesting that the emergence of Electron as a better-funded and more modern solution makes it the sensible choice for NEW projects.