So, I've recently looked into JSP and Java servlets for server-side code, and while I'm not particularly thrilled at all to be working in Java again, I was wondering what /prague/ thinks of it (minus the usual PIG DISGUSTING stuff).
For delivering ``Web2.0'' shitpages, I have to say I'm actually liking this more than any other server-side crap I've used so far. This is probably because that was Python and PHP and thus slower and fucking ugly. (No way I'm gonna touch Ruby either, and node.js can fuck right off.)
I recommend ASP.NET if Windows servers are not a problem. It seems to be more actively developed and used than JSP. I'd say anything but PHP or FIOC is a good choice.
>>7
I'd rather not be stuck using only Windows servers though. At least you can run Tomcat on both Windows and Linux (and probably OSX but fuck that).
Name:
Anonymous2012-07-29 5:46
>>9
Sun has some ASP stuff, but I don't know about ASP.NET. I suspect Mono can run it. I suggest using Haskell for this sort of junk, unless it's performance sensitive, in which case I suggest using C.
>>11
Seriously thank you for this, I will need to look into it. If I'm writing server code that ISN'T for the regular web / will be hosted with a generic host then I will probably give this a shot. Have you used it yourself?
>>24
As far as I'm concerned it's a lot of buzzword crap for something that already exists, namely using various HTTP requests (GET, PUT, DELETE, etc.) to modify or retrieve data. It all lies in marketing...
>>25
I've always felt this. I think it's good to have an explicit term to refer to it, though, but it definitely gets tossed around an insane amount.
Name:
Anonymous2012-07-29 17:40
>>24
It has a formal definition you can look up, but in common use I typically hear it used to describe a web service that doesn't use SOAP data. That's a natural result of the .NET framework's WCF making use of it in that way, however, and doesn't really have much relation to its real meaning.
Name:
Anonymous2012-07-29 21:31
>>23
The nice thing about Laravel is that it is a really small framework. Much nicer than working with something xbox huge like rails.
>>24 >>25
Actually, REST refers to this one dudes paper he wrote about how the internet should work. Now everyone just bickers over every little detail basically.
Name:
Anonymous2012-07-29 21:48
>>29
Laravel: 229 PHP files, 18368 lines of code. I won't call that "a really small framework" by any stretch.
Name:
Anonymous2012-07-29 22:13
>>30
The bizarre thing about it is it has more lines of comments than code.
>>24 >>25 >>27 >>28 >>29
It actually means ``open a socket and receive GET, DELETE, HEAD and POST even if your program is a fucking calculator"
Name:
Anonymous2012-07-29 22:18
My guess is that an HTTP API would be too ambiguous at a time when there was SOAP (``Simple''), WSDL and XML-RPC. HTTP also has excess notional baggage of (like being coupled to HTML... its the ``Hypertext transport protocol'') that a made up name didn't. Finally it perhaps helps to get through to the enterprise to formalize it with some comp sci worded acronym. So REST it was.
WebDAV is another one that is basically HTTP with extra verbs. UPnP reuses HTTP syntax in UDP packets. You could say that there is a HTTP family of protocols.
Anyone using anything but GET and POST is overdoing on their bloated service. All other requests (even HEAD, since you can just close a GET after the headers) are redundant and part of the reason why HTTP is so bloated nowadays.
>>34,37,38
The reason for HEAD is to allow the server to stop after the headers. With a GET, by the time your application sees the end of the headers, the data after that has probably already arrived in your NIC's buffers, and the server sent a lot of body that's still travelling over the network. When the client closes the connection the server is still sending. Only until the FIN reaches it will it stop.
Also closing the connection doesn't work with Connection: keep-alive.