Dojo

How is Dojo 2+ trending?

Is it still too early to tell? Does 2+ have buy-ins from companies that are too big to let 2+ fail? Are there any metrics available?

It seems to be a Site Pen project only. Dojo Toolkit 1 was strongly backed by quite a big development team from IBM which also attracted many other programmers. So to speak, Dojo Toolkit 1 was (inoffically?) IBM’s JavaScript framework - like YUI was for Yahoo. But as far as I can see, IBM has completely dropped out of the project.

Thanks for the info. It sounds right.

Can’t say that I am surprised. Dojo has always been ahead of its time. I wrote an app in 2008 that is still running today. I didn’t appreciate the value of Dojo then and the owners of the app still don’t. Since I have the freedom to do so, I am going to invest further in 2+ just out of trust in the folks at SitePen. I think it is a brilliant move that they have built a company around support of all forms of JS frameworks. This allows them to see in depth what is out there and provide a model to keep 2+ alive until the rest of the world catches up.

Dojo 2+ has been difficult to assimilate, but it doesn’t make sense that for decades we have remained basically the same distance away from the machine. Coding obviously must continue to abstract the target. Seems like it has been overdue.

One question that I have is whether or not 2+ is actually a React clone or is there much more to the picture than reactivity?

This answers some of my questions about whether or not Dojo 2+ is a React clone:

The given article by Site Pen’s Kitson Kelly (who is no longer with Site Pen) does not mention a third strategy: The strategy of a “frameless” framework for everybody.

If you have a look at the original Dojo 2 release candidate by the IBM team, you can clearly see this strategy and its benefits: http:// ibm-js. github. io/ deliteful/ .

It is based on well maintained third party libraries which are too popular to fail: DCL.js, Handlebars.JS, Require.JS, jQuery and so on.

It adds in the missing pieces to make it a full blown UI library based on Dojo 1, which has both, a profund i18n and a11y support. It carefully adds new features like the possibility to invalidate a widget to improve DOM rendering and to prevent inefficient frequent updates. Furhermore, it is flexible, since you can easily instantiate widgets either programmatically (the React way) or by HTML markup as web components (which I prefer). It also has a CLI now. Someone used to Dojo 1 should therefore be able to programm easily with Delightful.JS. Site Pen’s Dojo 2 is a completely different product. It just uses the established brand name “Dojo” in my eyes - and I’m quite annoyed about that. I’m really angry especially about Dylan not mentioning the unreleased real Dojo 2 candidate by IBM and the overall project situation und changes behind the scenes.

The Site Pen team did an excellent job in the past (see e. g. dgrid and dstore). And no doubt: The team is highly competent. But you have to bear in mind, that behind the scenes, the Dojo team has changed and the fast majority of developpers has gone, developpers maybe, that made Dojo ahead of its time, resisting trends and main stream buzz-wording like VDOM, state containers, injectors.

Dylan thinks, Dojo has to look up to React and thus he wants to make Dojo React-like https://www.youtube.com/watch?v=lHRuIA7cyU0 . I don’t think so. Bill Keese has done an excellent job on his rewrite of Dojo 1 called Delightful.JS. We don’t really need a virtual DOM but programmers who understand what they do with the DOM and how they do it. By reinventing the wheel, Dojo 2+ is a complete rewrite of a completely new UI widget system which (yet) misses widgets and features: Dialogs e.g. are not draggable by mouse and so on. In Dojo 1, this is not a problem. Also in the rewrite of Dojo 1, Delightful.JS .

But in the end, it is maybe a matter of taste which framework you choose. I like the old Dojo-1 / Delightful.JS-style more, and keeping this up would have been the real unique selling proposition in the market if you compare the frameworks. Besides that, Dojo 1 / Delightful.JS main advantages is its sophisticated, proven UI library with full i18n, a11y support, which has matured over the years.

You could say that the activity on this forum is a metric and it’s not particularly promising - seems to be very little action here.

Sorry, most of the activity is on live chat (Discord) which has a few hundred members asking questions pretty actively.

We don’t track usage strictly, but usage and interest has picked up with the metrics we have (number of people asking questions, pull requests, etc.). In terms of enterprise adoption, we know of some companies that have started adopting modern Dojo and those enterprises seem quite happy with it and with the responsiveness in working through the issues they report.

Paul, I ask that you kindly stop adding drama about Dojo and SitePen that simply doesn’t exist. The team we have working on Dojo today is incredibly strong and you’re ****** on their work open source work which is unnecessary. If you look at the history of frameworks, pretty much every major framework has turnover as people get burnt out or decide they want to work on something different.

Unfortunately no one is going to use Delite/Deliteful because it is not what the world wants today, no matter how many times you mention it. Yes, we still have a fair amount of work to do on the modern Dojo widget system (and we do have Dojo 1 - modern Dojo interop for when our widgets are not there yet), but maybe file issues (or even pull requests) to help improve it rather than dismiss the excellent work being done.

In terms of Dojo leading the pack, it doesn’t really matter, but I will say that Dojo was the first framework to announce and start its rewrite in TS (Angular, Vue, and others have followed), and many of the interesting features we’ve added were also done first (e.g. type safety for class auto-completion, conditional polyfilling, build time rendering, recent advancements in testing)… there’s a lot of great work going on and a lot more great stuff planned for the future, and none of this is about being a React clone.

Regarding Dojo vs. React, React drew much inspiration from Dojo, and modern Dojo does draw inspiration from React and others. It would be foolish for us to ignore the greater JavaScript ecosystem and do everything from scratch without looking at how others solve problems. Dojo strives to streamline and simplify the process of building modern apps in an ecosystem where things have become increasingly complex.

Dylan says, that he does not track usage strictly, but usage and interest has picked up in his eyes with the metrics he has(number of people asking questions, pull requests, etc.).

As of 24/02/2019, the status of pending pull requests on github (and the mentioned programmer’s affilitation on github / google) shows for me as follows:

dojo/framework:
12 pending pull requests: 4 by agubler (SitePen), 2 matt-gadd (SitePen), 1 nicknisi (SitePen), 2 bryanforbes (SitePen), 1 maier49 (probably SitePen), 1 rorticus (probably SitePen), 1 dasa

dojo/meta:
2 pending pull requests: 1 by dylan (SitePen), 1 matt-gadd (SitePen)

dojo/widgets:
6 pending pull requests: 2 by tomdye (SitePen), 1 nicknisi (SitePen), 1 aciccarello (SitePen), 1 maier49 (probably SitePen), 1 schontz

Or better check out for yourself the contributors lists / history of contributions on github for the repsective repositories and check the programmer’s affiliation.

@dylan: I appreciate the work for open source and don’t ****** on the peoples works, but I get quite annoyed about you and your way you communicate with the community, especially about the things you don’t communicate. I perceive this as quite dishonest and this really triggers.

Delitful is a nice rewrite of the of Dojo 1 framework and is much closer to Vue.js, a programming style, I prefer more. Deliteful is good enough for most projects - without any learning curve for Dojo 1 programmers.

I was not aware that it is a crime for a company to actively contribute to open source to push things forward. Yes, most of the active work on modern Dojo has been done by SitePen and I am not trying to hide that, just as most of the work done on React in its early days was done by Facebook, much of the work on Angular in its early days was done by Google. After you reach a point of stability and adoption grows, then other people start to engage.

I would think it would be a positive that SitePen, a relatively tiny company, is willing to invest so much time and resources into Dojo rather than something to attack me about. The Deliteful project that you love is not getting much support at all from the very large company that started it. It’s an absolute shame that Deliteful did not get the support it deserves, but that is not something I can change.

I was really happy with Dojo (the original one), and I was happy to see, that new Dojo (Dojo 2+) is using Typescript, which gives me the chance to get away (partially) from what Javascript is missing.

We are currently investing in a project using Dojo2, but so far it seems, that this seems to be a dead end. You could find tons of information on Dojo, but only very few information on Dojo2. Any idea what I could be making wrong? Otherwise than choosing Dojo2…

Thomas

Thomas,

We’re very active on Discord, we’re working on a revised version of the website with even more documentation which should arrive in June ( see https://github.com/dojo/site/ for progress), and there’s a really nice community website at https://learn-dojo.com/ .

Building adoption, community, and more takes time, but we’re getting there. Contributions and help are always welcome!

-Dylan

Dylan,

thanks for pointing out this information to me.

Does this mean, Discord should be used instead of discourse.dojo.io? If yes, why not deactivate discourse? From what I have read, I thought this is the place where to discuss topics for Dojo2+ and find help.

Sorry if I misunderstood,
Thomas

Hi Thomas,

Keep in mind we’re a relatively small team. We want to be able to provide answers both here and Discord, but there’s definitely value in real-time back and forth interaction which you only get from chat. But some people either cannot access that, or prefer to ask questions in more of a Stack Overflow style environment, which is what this is. So we are striving to get better at answering questions here, but what’s happening is when I end up traveling for work, we fall behind.

So if you prefer this environment, please keep asking questions here.

-Dylan

Dylan,

please do not misunderstand, I never wanted to blame anyone on not providing help/support, it is just that I thought no-one was using Dojo2+ as there is no traffic on the support forum.

Having a chat of course makes it faster and easier to solve problems, but

  1. no-one will ever learn from a solution you are providing for a problem, as no-one will read back in a chat
  2. a chat is an interactive medium, whereas people are living in different timezones, which may provide an issue for interactivity (I’m living and working in a UTC+2 timezone…)

I’ll give Discord a try. Would be happy if it works…

Thomas

I think also a chat style support can be intimidating to newcomers no matter how polite the team. I personally would be much more hesitant to ask noob questions on discord as opposed to discourse.

Great ideas. Have asked a question, and as expected, got no answer. For me it just seems not to work.

In some ways a noob question on chat quickly fades away… the main challenge right now is there are only so many of us. I am probably the only active committer that sees value in this forum, and my travel schedule for work right now is beyond insane so there are days where I just barely have time to delete the emails that come in. :slight_smile:

@Thomas, in general the goal is to make improvements to documentation and examples for all the questions that get asked, rather than making people read back through a forum hoping to find a relevant answer.