Express needs a collection of first party middlewares for logging, strong params, cookie-jar and etc with extensible ORM ActiveModel for more than regular SQL, e.g RethinkDB and MonogDB.
Is Express.js dying?
61–70 of 77 posts
Re: Is Express.js dying?
#62Earlier quoted context omitted.
I think it is a fair comparison, because I think Linus WOULD loudly (and with much profanity) fork it and go back to getting shit done rather than hit slow-mo on a multi-car highway pile-up. EDIT: After further thought, I do see your point. My comparison was purely hypothetical and has some logical flaws. I'm literally guessing at how he'd handle losing control of the reins. My takeaway was that Doug (who, I should c…
Forking a project as significant as Express isn't easy; you need help from many people to get it going. You make it sound as simple as clicking a button. (Linus may pull it off easily because he's Linus, but that proves nothing.) This "soap opera" isn't of Doug's making. What he's saying seems very fair to me: he's willing to work on it if it's a community effort, and not under a corporate banner. Throughout the thre…
I mentioned the dialog in IBM's back channels because Doug himself was pushing a participant there once he got some traction in that more appropriate venue: "Now, please stop contacting me as a side channel without the context of the conversations I have been having with IBM. Please get involved in the existing IBM channels I have, so you have context here."
You are correct: Doug does give excellent reasons for supporting the project under the expressjs org, the same reasons brought up back when the github ownership of the project changed (as in the issue you linked). The public debate only served to confuse and frustrate watchers, and that is the "soap opera" I refer to.
It's unfortunate that sometimes such public drama is necessary to effect change, but as evidenced by the issue raised when the repo was taken over originally, public drama wasn't going to effect any change here. It now serves only to foment ill-will and promote confusion, regardless of who is "right".
You are also right that we should be thankful for OSS contributors' efforts. I am thankful for his. For StrongLoop's too, in fact. I am, however, not thankful for the wasted energy in maintaining the debate on a github issue, when the debate didn't do any good the first time 'round. I am not thankful that politics have derailed the work.
EDIT: I think you misunderstood my "be the best..." comment. I was merely suggesting that Doug would find himself less stressed if he was able to focus on doing what he loves and not playing politics.
Re: Is Express.js dying?
#63Earlier quoted context omitted.
I think it is a fair comparison, because I think Linus WOULD loudly (and with much profanity) fork it and go back to getting shit done rather than hit slow-mo on a multi-car highway pile-up. EDIT: After further thought, I do see your point. My comparison was purely hypothetical and has some logical flaws. I'm literally guessing at how he'd handle losing control of the reins. My takeaway was that Doug (who, I should c…
It's an issue in an issue tracker. Nobody's blogging about it, or doing interviews in the media. I would consider this a reasonable level of drama for an issue tracker conversation. Again, I think Doug is probably pushing this in public (though not dramatically so, again it's an issue tracker which normally would have a couple dozen readers) as a final last ditch effort to get IBM to act right. Nothing wrong with tak…
That's a fair implication, that I didn't give Doug all the credit he's due as a "head down get shit done" kinda guy, so let me correct that: Doug is an amazing talent, and it's a raw deal that he got sucked into the politics of this; I hope he's able to move on (FWIW I think his name carries enough weight that a fork would get attention) and we can all forget the soap opera that happened on that thread.
Re: Is Express.js dying?
#64It's simple. Everyone has switched to either Koa (designed by the team behind Express) or, more likely, a client-side framework like React.
I use express because it is the most barebones solution before vanilla http/https. The middleware API is trivial and relatively agnostic (most just expect node-style req/res objects and a `next(err?)` callback). The router does what it is supposed to and allows nesting.
However I would likely chose something different if my needs were different. If I wanted to implement REST APIs in Node I would use Hapi. If I wanted to build more complex web apps I might look into Koa (although I don't think I'll ever move back to classic server-only rendered apps).
Express is my JS equivalent of Ruby's Sinatra and Python's Flask. Nothing more, nothing less.
Addendum: I wouldn't say "everyone" has switched to Koa. Koa was handicapped by its conceptual overhead for most ordinary express users in its early days (relying on a combination of generators and promises -- both not yet sufficiently widely established and understood technologies at the time -- to build coroutines, which even fewer people understood). It has improved and the JS community seems to have caught up (just look what's going on with the Redux ecosystem -- "thunks" and "sagas" (essentially built around coroutines) are beginning to see some traction). But it didn't initially deliver the same "read this tutorial, start building apps" experience express had.
Re: Is Express.js dying?
#65Well here is the answer from the guy who is still maintaining it: https://github.com/strongloop/express/issues/2827 The link is from the fifth comment. When developing in node it makes a lot of sense to use lots of different modules for the one project. At work I have a good dozen or so, it really helps with iteration and testing.
> guy who is still maintaining it
without commit access...
Re: Is Express.js dying?
#66people still use Express? Koa is where it's at.
Re: Is Express.js dying?
#67It's simple. Everyone has switched to either Koa (designed by the team behind Express) or, more likely, a client-side framework like React.
> Everyone has switched
4 Million downloads per month for Express [1] (vs 60K/mo for Koa [2]) .. doesn't look like everyone is switching.
[1]: http://npm-stat.com/charts.html?package=express&author=&from...
[2]: http://npm-stat.com/charts.html?package=koa&author=&from=201...
Re: Is Express.js dying?
#68Earlier quoted context omitted.
Here's some history, from what I've read. The original discussion of the transfer of the Express repo's ownership to Strongloop occurred here in June 2014: https://github.com/strongloop/express/issues/2264 Back then, many in the community were surprised by the sponsorship (sale?) of the project and called for the repo to be transferred to the Expressjs org instead. So the question of ownership has been a long-running…
This is one of the pitfalls of contributing to an open source project, sooner or later (assuming the project gains traction) some company will acquire the project for what for them is peanuts and in the process they'll do what they can to keep the current project the one people will flock to. That's why we have MariaDB and a whole raft of other projects that are technically closely related to the original and maintai…
I don't have any type of contract around Express; only the interest in providing a stable project to the community.
Re: Is Express.js dying?
#69Earlier quoted context omitted.
I actually have to agree with this. It seems like some semi-political debate about how one guy left and they don't like the direction the project is headed, but it is clearly not dying. I've flagged it for mod review, I'm not sure if this should be on the front page, it doesn't seem like it meets the requirements for HN.
That one guy is the only one currently contributing code (a slight exaggeration, but very slight). That info by itself seems HN worthy as many of us have likely used Express as the "safe" choice when starting a new project. Time to rethink that...
Perhaps, but I can assure you that Express 3 and Express 4 are completely safe. The interface set up by Express is depended on by far too many companies and other frameworks for that to not be the case -- for instance, even though Express 3 is officially unmaintained by Strongloop, the Sails core team has committed to taking over critical patches.
As for the future of Express: I've only had the opportunity to get together with Doug in person once, but we've talked about this a few times in the past. I can guarantee you the last thing he wants is drama, so I'll leave the ball completely in his court there.
I will say this though: keep an eye on Doug's federated projects (jshttp and Pillar). Whether or not the various pieces are released as Express 5 or not, the _code base_ is still improving and becoming increasingly standardized, regardless of the monicker. More on that: https://github.com/balderdashy/sails/pull/3235#issuecomment-...