Is parse-server a true open-source version of parse.com or is it a new/fork project? #765
Comments
|
As i impressed, the parse-server project is totally different thing than Parse.com. I don't understand the point of releasing something other than Parse.com system when they DO have it already, so why to rebuild something other which is unstable. Any comments? |
|
@grupster, @aryelevin I understand your concerns and we're working very hard to make the open source version of parse as close as the original implementation. You are welcome to contribute to the project when you find bugs, open issues. Know that the whole parse team is dedicated to Open Source now, and this project is a unique chance to talk to all of them. The rate at which the bugs get fixed is impressive. As you point out, many developers don't have the technical know how for developing, deploying, maintaining and scaling node.js/mongodb applications. But there are multiple providers out there that can help you with that, mLab (formerly mongolab) can take care of the database hosting (which is the most critical point) and heroku provides a very simple way to deploy. Parse.com will stay open for business for the next year, within that period, you can smoothly transition to your own hosted solution through this project, that is maybe a great opportunity to learn backend development and improve your skills at the same time! @grupster what features are you talking about? I'll happily implement them for you if you can point out exactly what you need and that is currently missing. |
|
I'm facing issue #432 and #550 (which is basically the same issue) for a while and didn't found a good solution for it, also from the client perspective i couldn't create a workaround for it, so how can i rely on this parse-server thing? |
|
I agree, I would love to see the fully original version fully open sourced. |
|
Yea I couldn't agree more. It seems like you are doing more work than needed by making this a new project. I'd love to get the reliable parse back... |
|
I believe there may be a (many) good reasons to make that project instead of open sourcing the original code base. Deployment has been made fairly easy, and yes some complex compound queries are not supported yet, PR's are welcome. Open source is not about developers working for the rest of the community but a group effort into making great projects. Given the number of high quality contributions I see everyday, the community is embracing the move toward a simple node.js based implementation. I welcome that, and that's why I'm contributing for free many hours every days to this project. I encourage all of you to do the same, it's not complicated! |
|
@flovilmart nobody is underestimating the great effort your team is putting into making this parse.com clone, in fact, is all the opposite, we praise and appreciate your efforts. I just think a clarification is due, in fact, I believe you guys should add a disclaimer to the Readme file saying something along these lines: this is not a drag and drop parse.com alternative, many features are not available, it is not a smooth transition unless users know their way around setting up infra and then debugging the implementation. Personally I have over 10 parse.com apps to port, I know may way around nodejs and mongodb but at this point the effort to try to setup my on parse-server infra and try to make my apps run seems just as much effort as porting to a new BAAS vendor altogether. To sum: the announcement sounds like: hey! we left a copy of the code don't worry; and that is not the case at all! |
|
Another thing is that now I'm not gonna trust any other BaaS, since some day they can also say goodbye, so i prefer to have my stuff in my hands in order to be secured, so thats why I'll use my own server/cloud. I was hoped that the open source parse-server is the same parse.com code, but it is not, so now i have 2 options:
I'll prefer to use parse-server, but if it is not working, and I'll need to fix its own bugs by myself, then i prefer to hire a backend guy and build my own architecture on the server side. Its easy to tell us to contribute in this, but not everyone have the knowledge and time for this, while in-fact we relied on parse.com too much, and now we've worked twice, one time by using parse architecture, and the second time now, we have to build the whatever backend we want to use. parse.com built to let the developer be focused in the product, and get rid of the backend efforts, now it is the opposite. |
|
BTW: i want to declare: I'm really appreciate any efforts the community putting in this parse-server!! :) |
|
@grupster if you know your way around node.js and mongodb then we most welcome your contributions. Please report the bugs you see, write unit tests, we're here to help you have the smoothest transition as possible. The parse team never advertised it as a 'drop-in' replacement, they suggest to keep using parse.com for the time being, migrating the data to a mongodb provider or your own infrastructure. |
|
So 19 months later, does parse-server now offer similar functionality as parse.com, it is more close to being a drop-in solution? @grupster how did you finally migrate your 10 parse.com apps? |
|
@cprcrack I did not migrate to a self-hosted parse-server, I am using the services of these guys: http://www.sashido.io/ So far they have provided an excellent service and support. However, I am not using parse-server anymore for new projects. I am using Firebase. |
|
@grupster that's great to hear, it seems that your choice was a good one! I've precisely just migrated one of my old apps from parse.com to Firebase and really missing some features like background jobs, database objects .increment() or .addUnique()... Even knowing the amount of objects you have in a certain "table" (object) of your database or making a search for a particular "row" is not easy as it was before... Are you happy with Firebase overall for all your future projects? |
|
@cprcrack yes I am very happy. It is not a solution for all needs but works really well for most small/medium sized apps. What I do in some cases, is to program a small node.js server for everything that Firebase does not do. For example, say I need to do full-text search on a given collection, I have a node.js app listening to add/change events and keep an index updated then call that index through a REST API (I suppose you can also solve that with the new Firebase Functions but they were not available when I developed the indexing server). Some other features I do on a node.js app: fire email/sms after certain events, expose a REST API for operations that I do not want to do directly from front-end code (important for validations) and backups. In general, I've learned that Firebase works best if you design your data structure correctly, one needs to stop thinking in relational-database mode ;) You also need to setup Database Rules correctly to avoid malicious or accidental data loss/manipulation. |
|
@grupster just curious. Why wouldn't you use parse-server for the big project ? I personally find it lightweight and well integrated in express, which allows me to use the full potential of the community libraries. (and you can use MongoDB) |
|
First we must agree on what "big" means. I do not use daily-active users or
concurrent sessions for such measure, I use user-base growth. This is my
logic: a project growing the user base normally requires more features o
very particular business logic, using boxed back-end solutions like parse,
may put some limitations on new requirements. I run into that case already
with ACL. Parse's ACL was not enough to accommodate for a particular
use-case, I run into a wall. If I have to fork and patch parse to solve
such needs, or program an additional piece of back-end code, I might as
well program the entire API and have full control over it.
2017-10-02 11:27 GMT-05:00 Olivier Allouch <notifications@github.com>:
… @grupster <https://github.com/grupster> just curious. Why wouldn't you
use parse-server for the big project ? I personally find it lightweight and
well integrated in express, which allows me to use the full potential of
the community libraries. (and you can use MongoDB)
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#765 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ACdHHlZf7UNoJG3Kg_Xm2o_-sw3GuZLDks5soQ8EgaJpZM4HnZZV>
.
--
*Ricardo Moncayo*
Director IT
*Coniski*
ricardo@coniski.com
600-0886/600-0887
www.coniski.com
|
|
@cprcrack Yes! I would say it is much more like an equivalent, and more of course. Feel free to check out the releases and see what's been going on, I think you will find this has moved quite a ways since it's original inception. As for a solution, like most BAAS equivalents you will have limitations inherent in the general 'all purpose' design we have going on here. Unfortunately this isn't guaranteed to work for everyone, and by all means if your solution requires a more customary approach you should explore all options to best handle your case (even if it includes putting up your own api). What we have attempted to do here is create a system, and outlying sdks, that most developers would find helpful to them. For those cases where we're not quite meeting the circumstances we always are open to adding functionality, but we are doing so one step at a time. Super glad to hear that you guys (even if some of you aren't using parse now) are using this or have used this in the past. We hope, that in addition to other such solutions as Firebase, we're adding another option to the table so no one is forced to simply settle for what's available 👍 . Ideally, you'll be able to pick what best suites you considering current and future needs. |
|
@grupster I also meet cases when ACL isn't enough, and can't be enough, when a 'state' value update depends on the previous version....or simply because I'm too lazy to split my class in 2 because of security needs. For all those cases, we have before/after save and before/after find. |
After the announcement I was under the impression that Parse.com was giving away/open-sourcing its most current working copy of its API server, as in: here, take a copy, host it yourself, take your data, point your app to a new URL and you are set.
However, with so many issues raised here, in particular issues related to the most basic features/settings, now I am under the impression that parse-server is not parse.com open source version but rather a new project? A project maybe not even ready for production yet?
The community of parse.com users needs a clarification, in particular non-experts in nodejs and mongodb. There are tons of front-end devs that will be better of moving their back-ends to another simple-to-use BAAS vendor instead of trying to run parse-server. Furthermore, many of parse.com features are not available in parse-server and are not easy to implement either.
The text was updated successfully, but these errors were encountered: