Slashdot Mirror
Facebook Is Shuttering the Parse Developer Platform (cio.com)
itwbennett writes: In a blog post yesterday, Facebook announced it is shutting down the Parse developer platform as of Jan. 28, 2017, giving developers a year to move off its hosted services. This comes as a bit of a surprise, considering that just last month, Parse launched a set of new tools to help developers work with Apple's watchOS and tvOS last, and at the time, Parse Product Manager Supratik Lahiri promised more updates in the future. Developers who don't want to rewrite their applications to work with a new back-end service provider can follow a migration guide from Parse to make their applications work with an independent MongoDB instance and a new open-source Parse Server that's running on Salesforce-owned developer platform provider Heroku.
... another web API bites the bullet and all the kids will have to go learn yet another flash in the pan interface to connect to some moronic social site to scrape bullshit data to pass to an app they can sell to idiots.
you might care, if only we had https://www.google.com/search?...">ever heard of Parse. At this point it seems pointless to go
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
This is why you should avoid corporate controlled APIs and languages. Things like Rust, etc can just get pulled away at the last moment, leaving you without any future path and support. Lesson learned.
This is the technological equivalent of obsessing over some Kardashion boob job.
When Fascism comes to America, it will call itself Anti-Fascism, and tell you to give up your guns.
Rust will be dead in 5 years, just like Ruby and a long list of other tossed away ridiculous languages. C++ forever! Rust is about as Open Source as Android is (meaning "not really" in practical terms).
I really can't see doing much heavy lifting with scratch...
https://scratch.mit.edu/
It does look like kids might have some fun with it.
You have the right to remain sentient. If you give up the right to remain sentient, you will be elected to public office
It ends when the hipsters are kicked out of the industry. They've ruined everything they've touched. They ruin UIs that were good. They create awful frameworks. They prefer to use the absolute worst programming languages, like JavaScript, for everything. They took git and centralized it on GitHub. They're too lazy and/or dumb to learn SQL, so they use persistent hashtables for storing data, and query it using imperative JavaScript code. It's nothing but idiocy and disaster when it comes to these people. It doesn't matter if it's an 18 year old hipster or a 30 year old hipster or a 55 year old hipster. They all need to find a different occupation.
Let me guess. You work on Javascript?
Like it or not, C++ has been an industry cornerstone for the past 30 years. And it will likely remain that way 30 years from now.
This is very true. It is kinda sad to see the current state of affairs over at Mozilla, while we're at it.
LOL! C++ is indeed the future!
Everything important today is written in C++. LOL! Even Rust's implementation depends on C++, since it uses LLVM which is written in C++! LOL!
It's the same for pretty much all of the other major competitors to C++ out there. The JVM, the .NET CLR, the major JavaScript engines, Ruby, Python, Perl, PHP, Lua, etc., etc., etc., all use C++ or C. C, of course, is just a stripped-down version of C++ these days.
The future has never looked brighter for C++. It's improving at an astounding pace. People are finally realizing that they were wasting their time using other languages.
Maybe you're not smart enough to have realized this fact, but others have. While you're sitting there with your thumb up your ass, crying about how Parse or whatever SaaS nonsense you're dealing with just shut down, the rest of us are using C++ and running our software on our own hardware, getting real work done. You Rust users are panicking; we C++ users are busy accomplishing.
And it will likely remain that way 30 years from now.
Prediction is hard, especially those about the future.
Not really. It is simple - there's no alternative offering the same level of performance, support and established userbase. Those things aren't built over a weekend.
Is the same reason C has stayed relevant for almost 50 years. Nothing else covered the "portable assembler" role as well.
Person: Who are you sir?
Shutterer: I am the shutterer
Person: What do you do?
Shutterer: I shutter
We'll make great pets
Rust is a product of Mozilla, which as we know has had some tough times lately. They lost their Google sponsorship, and have had to settle for Yahoo instead.
Wrong, wrong, wrong... Moziila chose Yahoo over Google, presumably the deal was better. And looking at it form the standpoint of the Mozilla mission it might just enable more competition on the search market.
Rather, I think it's a Mozilla project with source code that's publicly available.
Nope, even the Mozilla people working on this is very attentive to make sure they aren't tried too much to Mozilla infrastructure.
So I wouldn't worry about this.
My only concern with rust is that it's too complicated and encourages too much over-engineering.. Maybe I just haven't really gotten good at writing it yet. But that is definitely my major concern.
What the fuck? Are you writing C++ code like it's 1987? You do know that the language has evolved, right? If you use RAII and modern C++ techniques, you can develop massive software systems without dealing with raw pointers even once, for example. You get the safety of Rust, but without the many downsides of Rust.
Rust is a lot like Ada. It's all hype, and much less substance. Whatever small amount of safety you might get over C++ is totally outweighed by the problems that Rust and Ada have that C++ just doesn't have. Those who might benefit from using Rust or Ada will just do what they've always done, and use a safe subset of C++ instead.
Everything you wrote about Ruby is wrong, too. The standard library is quite lacking. That's why you need to use so many goddamn Ruby gems in order to do anything remotely useful. And if you'd ever actually dealt with Ruby gems, you'd know how goddamn awful they are to work with, since the Ruby ecosystem is fragmented and total shit. Ruby on Rails alone changes so much between major releases that upgrades become a big disaster. Each major release of Ruby on Rails essentially is a new framework, they just all have the same name!
You should be careful about using insults like "fanboy", too. Name-calling like that could put you in violation of the Rust Community Code of Conduct! The Rust Moderation Team could then come after you.
Parse is hugely useful as a way to "assume" a back end. Any good NoSQL database does most of the association work during the write task; in this case on mobile client. With a few housekeeping tasks running on the parse server you can actually make a clean, fast, scalable service but obviously it's also great for prototyping. I think Facebook dropped Parse because they are not one of the big four making $$ in the cloud and Parse is a SaaS; it runs great on Google cloud.
Parse is/was a service that allowed people who knew next to nothing about server programming to cobble together a backend for a mobile app (or other app that can make http requests).
So in other words it was a tremendously useful or harmful service depending on your level of cynicism. It is of course hard to monetise a solution like this, since any app that becomes highly profitable will attract developers that know how to build a proper backend and then that app will migrate away and stop paying monthly fees.
It's not so bad as long as you stay close to the intent, where you want to do something like facebook, or a blog, with frequently added/updated items and comments and such. So if you're writing a little half-assed blog type thing it would be pretty easy to make it "reactive" to people adding new posts and comments and things. But try and apply it to something like what I was tasked with, which was a little search/modify/create/delete thing it was a complete disaster. YMMV
while [ 1 ]; do echo -n -e "\xe2\x95\xb$((($RANDOM&1)+1))"; done
People keep bringing up RAII as if it's some kind of panacea. It's a useful tool, but it is no panacea. Sure, you allocated an std::vector on the stack to take advantage of RAII, but if the memory allocation in the constructor fails, it will throw an exception. Are you handling that exception? Because I've literally never seen C++ code that handled std::bad_alloc exceptions.
Do you do a push_back() operation later on? Because that can also throw an exception, and if you have a bunch of push_back() statements in a loop, and you get an std::bad_alloc exception halfway through, you are going to end up with a vector that is stuck in an intermediate state. How do you rollback the changes? RAII can't help you solve that problem.
And compound this with the fact that your compiler can't even give you hints on what exceptions you need handle because basically no extant C++ code (including most of the standard library) comes with proper throw specifications. There is literally no way for you to know that you missed handling that std::bad_alloc exception until some bug rears its ugly face.
And when it comes to Ruby having an anemic standard library, let's look at the libraries that are included with a vanilla install of Python versus a vanilla installation of Ruby:
Decimal arithmetic:
Python: decimal, Ruby: BigDecimal
Templating:
Python: string.Template, Ruby: ERB
Logging:
Python: logging, Ruby: Logger
Compression:
Python zipfile, Ruby: zlib
Argument parsing:
Python: argparse, Ruby: optparse
XML:
Python: xml, Ruby: REXML
Encryption:
Python: crypt, Ruby: openssl
Again, all these modules are included out of the box. No RubyGems required.
The obligatory Obligatory XKCD
And, for "real programmers write in....:" The Story of Mel.