|
I was told that REST style programming is dumb, when RPC exists. Is this at all true? I don't know enough bout either to make that kind of claim, but absolutism in tech is usually rather silly.
|
#
?
Jun 23, 2013 12:52
|
|
|
|
| # ? Jun 2, 2024 16:27 |
|
|
Dirk Pitt posted:I was told that REST style programming is dumb, when RPC exists. Is this at all true? I don't know enough bout either to make that kind of claim, but absolutism in tech is usually rather silly. rpc makes sense when you want to tightly couple two services using different languages or both using c#
|
|
#
?
Jun 23, 2013 13:16
|
|
|
http://www.dadhacker.com/blog/?p=2024 I tend to judge distributed systems by their ability to cope with a fuctuation.
|
|
#
?
Jun 23, 2013 13:31
|
|
|
any such statement made without even nominal reasons to back it up can be ignored
|
|
#
?
Jun 23, 2013 13:32
|
|
|
no true RPC lacks caching and proxying solutions 
|
|
#
?
Jun 23, 2013 13:35
|
|
|
Nomnom Cookie posted:not quite PL but im a rebel I don't think subqueries (or query refinement) with the querystring or a fancier URL is a problem, but if you do ?type=json in HTTP you're bad, and should use the Accept and Content-Type headers to do proper type support and negotiation, given it's already part of the protocol, instead of reimplementing it all yourself. Also ideal RESTful document types should allow for hypermedia and have an explicit 'link' type you can follow in a machine-readable way if required, so that links to alternative resources can be read automatically. HTML is restful (<a>, <link>, etc.), but regular JSON isn't really.
|
|
#
?
Jun 23, 2013 14:52
|
|
|
Dirk Pitt posted:I was told that REST style programming is dumb, when RPC exists. Is this at all true? heh quote:I don't know enough bout either to make that kind of claim, but absolutism in tech is usually rather silly. yep. it's ok the person who said this probably didn't understand it either.
|
|
#
?
Jun 23, 2013 17:24
|
|
|
Dirk Pitt posted:but absolutism in tech is usually rather silly. so you admit there are reasons to use node.ks
|
|
#
?
Jun 23, 2013 17:26
|
|
|
Dirk Pitt posted:I was told that REST style programming is dumb, when RPC exists. Is this at all true? I don't know enough bout either to make that kind of claim, but absolutism in tech is usually rather silly. rest is for when you want to serve a resource like an image or another file. rpc is for when you want to perform a remote method as if it were local. you could build your own custom rpc on top of rest but that would be silly when there are already standardized rpc protocols.
|
|
#
?
Jun 23, 2013 18:03
|
|
|
Shaggar posted:rest is for when you want to serve a resource like an image or another file. rpc is for when you want to perform a remote method as if it were local. you could build your own custom rpc on top of rest but that would be silly when there are already standardized rpc protocols. It's funny you say this because I was told this in the middle of a discussion about why we are wrapping webapi references and using them in a rpc style manner, when we could just use a REST endpoint like webapi or service stack and not spend a week writing visual studio extensions and t4 templates. All this from the "fragility" of wcf and silver light (not my solution)
|
|
#
?
Jun 23, 2013 18:18
|
|
|
if ur using c# or java they both have service frameworks (wcf and cxf respectively) that are designed to allow you to extend existing interfaces across the wire with little effort. the developer doesn't have to think about anything other than "wheres the endpoint?" or maybe "what is my security config?". and both of those are configured outside of code too, so really you just replace a local implementation of an interface with a remote implementation by swapping a DI object and whammo blammo its done. no worrying about serialization cause its being handled by the runtime. its always painful to see a REST "api" that someone makes for a thing and find that there are 3 or 4 different versions and they have prebuilt clients for the old ones but not the new ones so have fun writing all that serialization code urself. oh wait they hosed up the documentation and/or left out features so it doesn't work the way their docs say. I think wcf has a wsdl type deal for rest, but idk if there are any real standards. that would probably go a long way to making rest rpc usable, even if the entire point of most rest apis is to not have well defined data + syntax.
|
|
#
?
Jun 23, 2013 18:32
|
|
|
The writing my own service is the scary part about what some developers like to do. I would rather work on WCF channel management, rather than re-write my service in an entirely different framework, and then write my own wsdl.exe to generate my new rest "methods" that are rpc. Anyways, that's my story, thanks for reading. Any one have a good link for rpc vs. rest? Just so I can get some better understanding and rebut statements like "REST is useless when you have rpc".
|
|
#
?
Jun 23, 2013 18:37
|
|
|
REST and WebSockets
|
|
#
?
Jun 23, 2013 19:15
|
|
|
Can I have REST ful WCF service use TCP instead of HTTP for transport ?
|
|
#
?
Jun 23, 2013 19:16
|
|
|
RESTful UDP: a Live Framework Feature Request
|
|
#
?
Jun 23, 2013 19:17
|
|
|
i looked at go again yesterday and it's p cool. a lot better than the initial release, starting to live up to "c done better" still no generics, but you get generic map which is what everyone really wants anyway
|
|
#
?
Jun 23, 2013 19:22
|
|
|
Imagine walking up to a large-screen Mesh-enabled device such as an XBox, Microsoft Surface, or public kiosk. You pair your Mesh-enabled smartphone and project its apps and data onto the big screen, with your smartphone acting as the data entry device. Expanding on this scenario, imagine a game that takes advantage of the smartphone’s accelerometer and camera, turning your phone into a high-powered Wiimote with the entire touch screen used for control surface. You might want to attach a wrist strap… Imagine the cool apps you could write if a group of people shares real-time GPS data from smartphones and carputers. This feature would be useful for more than just extending the capabilities of smartphones. You could remotely control media playback, chat with people, push real-time financial data, and build a variety of interesting distributed apps that are designed to run in real-time across a mesh of devices, aggregating specialized device capabilities into a single composite experience. Cross-platform support exponentially increases the possibilities and relevance of the mesh. You can imagine special-purpose devices whose entire reason for existence is to be plugged into the mesh to supplement apps and user experiences. This is true even without real-time messaging, but this capability is crucial for enabling the most seamless composite device experiences.
|
|
#
?
Jun 23, 2013 19:22
|
|
|
imagine a cold steel blade slowly being inserted through your sternum
|
|
#
?
Jun 23, 2013 19:32
|
|
|
imagine 4 mesh-enabled devices on the edge of a cliff
|
|
#
?
Jun 23, 2013 22:15
|
|
|
meth-enabled devices
|
|
#
?
Jun 23, 2013 23:00
|
|
|
Microsoft announced on December 13, 2012 that Windows Live Mesh would be discontinued on February 13, 2013.[2]
|
|
#
?
Jun 23, 2013 23:04
|
|
|
not that mesh, a different mesh
|
|
#
?
Jun 23, 2013 23:08
|
|
|
mesh-back smartphone
|
|
#
?
Jun 23, 2013 23:09
|
|
|
MononcQc posted:I don't think subqueries (or query refinement) with the querystring or a fancier URL is a problem, but if you do ?type=json in HTTP you're bad, and should use the Accept and Content-Type headers to do proper type support and negotiation, given it's already part of the protocol, instead of reimplementing it all yourself. this is woefully inadequate, for example this one is often implemented completely wrong or not at all: Accept-Language: en-gb or, saving images as lossless GIFs or PNGs is a good idea and displaying images on the screen taking full advantage of 8-bits of transparency and 60Hz refresh rate is a good idea. yet setting these perfectly reasonable headers is not supported: Accept: image/gif;animated=yes, image/png Accept-Encoding: image/apng;fps=60
|
|
#
?
Jun 24, 2013 00:48
|
|
|
PleasingFungus posted:Microsoft announced on December 13, 2012 that Windows Live Mesh would be discontinued on February 13, 2013.[2] http://www.joelonsoftware.com/items/2008/05/01.html Heh. Spolsky posted:And now Ray Ozzie's big achievement arrives and what is it? (drumroll...) Microsoft Live Mesh. The future of everything. Microsoft is "moving into the cloud." Dirk Pitt fucked around with this message at 02:07 on Jun 24, 2013 |
|
#
?
Jun 24, 2013 02:05
|
|
|
sounds like iCloud tbh if iCloud stored my email too and not just contacts, photos, browser history, bookmarks, calendar, notes and documents. as it is i have to log in to iCloud *and* log in to google apps *and* log in to gmail stebe come back and save me from this usability nightmare 
|
|
#
?
Jun 24, 2013 02:11
|
|
|
Nomnom Cookie posted:sounds like iCloud tbh if iCloud stored my email too and not just contacts, photos, browser history, bookmarks, calendar, notes and documents. as it is i have to log in to iCloud *and* log in to google apps *and* log in to gmail you remember mobile.me? this was stebe's achilles heel
|
|
#
?
Jun 24, 2013 02:27
|
|
|
hiring ray ozzie as chief architect was a mistake but joel remains ignorant, he would never say anything good about live mesh because he believes the idea of network file systems is fundamentally wrong
|
|
#
?
Jun 24, 2013 02:48
|
|
|
if he'd been around when OLE was in planning stages he would have panned it for trying to solve the cross-app integration problem universally rather than just for microsoft's own apps. greg whitten = big dumb architecture astronaut who doesn't accept reality
|
|
#
?
Jun 24, 2013 03:03
|
|
|
well the result was COM so maybe joel is on to something
|
|
#
?
Jun 24, 2013 03:11
|
|
|
COM does what it says on the tin
|
|
#
?
Jun 24, 2013 03:15
|
|
|
what does it say on the tin
|
|
#
?
Jun 24, 2013 03:16
|
|
|
"make C++ usable between shared libraries"
|
|
#
?
Jun 24, 2013 03:17
|
|
|
a framework for software objects to communicate through programming language-independent binary interfaces
|
|
#
?
Jun 24, 2013 03:20
|
|
|
Gazpacho posted:hiring ray ozzie as chief architect was a mistake but joel remains ignorant, he would never say anything good about live mesh because he believes the idea of network file systems is fundamentally wrong I completely agree with you. I just so happened to read that Spolsky post a few months ago and wanted to share. I'm no Spolsky fan.
|
|
#
?
Jun 24, 2013 03:21
|
|
|
Gazpacho posted:a framework for software objects to communicate through programming language-independent binary interfaces and here i thought the tin said HRESULT IClassFactory2::CreateInstanceLic(IUnknown *pUnkOuter, IUnknown *pUnkReserved, REFIID riid, BSTR bstrKey, PVOID *ppvObj);
|
|
#
?
Jun 24, 2013 03:29
|
|
|
Suspicious Dish posted:and here i thought the tin said HRESULT IClassFactory2::CreateInstanceLic(IUnknown *pUnkOuter, IUnknown *pUnkReserved, REFIID riid, BSTR bstrKey, PVOID *ppvObj); if u want to communicate between languages then use a socket
|
|
#
?
Jun 24, 2013 03:39
|
|
|
yes let's push messages and data through a byte-sized straw even when there's no hardware reason to do so
|
|
#
?
Jun 24, 2013 03:46
|
|
|
u can also use a datagram socket
|
|
#
?
Jun 24, 2013 03:51
|
|
|
|
| # ? Jun 2, 2024 16:27 |
|
|
windows has datagrams?
|
|
#
?
Jun 24, 2013 03:55
|
|