Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed last year.
Improve this question
As the title suggests, I'm building a simple chat app in Node/Express + React, and I'm using socket.io to get real time updates.
However, other than the chatting itself, rest of my app is very basic (registering, signing in, creating rooms etc), and while it's definitely doable to make all those requests just socket emits, it feels like bit of an overkill.
So I was thinking of creating a regular REST API for that stuff and outsourcing it into a separate file, then creating another file just for the chatting function itself, where I'd use socket.io, but I'm afraid that it may seem a bit unprofessional, or leave an impression that I don't know how to use sockets well enough.
Key point: When finished, this app will be posted to my portfolio, so I'm not very concerned about what just "can work", but rather what's the best and cleanest way to do it? Should I switch everything to sockets and use them even for basic fetching of lists and stuff? Or is it better practice to mix the two, and use REST for, well, the rest?
In my experience, To mix the two is the best way to do that.
I did that in many projects and it worked well.
I surely think that modularity is the best way to make code clean and qualified.
So, to use both of regular REST API and websocket is more professional.
Thank you.
Related
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 1 year ago.
Improve this question
avoiding the obvious answer, "Maybe I shouldn't have been given this to work on."
Let's just assume I was given this as a complex 'test of my abilities to learn on the job'.
The app is using serverless framework and I understand the basics of the structure using AWS and where certain things go etc, but I am not used to the structure of the app i was given.
I have a folder for backend, app, and one for 'graph'. I would just like to know where to start? Is the suggested route to user 'serverless-offline' or being that I didn't design this app, should I go straight to plugging things into my AWS, and get it running that way? I know this is kind of a noob question, and regardless I'm just going to go ahead and start playing around with the two options, but I do have a small window of time to figure out how to get this running in a 'Dev' environment so I can give a quote on adding some new React things to the app.
Are you working for a consulting company that advertised you as an "expert" to a customer where you actually have a severe knowledge gap to even approach the project you've been put on?
If yes, you aren't going to get much more information here in a reasonable sized answer than you can easily find using a web search. In fact, your question is so vague that I personally think it's not answerable at all. So, get searching on your own, hopefully you can figure out enough stuff by the deadline that you/your company can "fake it until you make it".
If not, and you are an employee in a normal company, you should have some sort of knowledge transfer process in place where someone who is familiar with the application would tell you at least an overview of how it works and how to approach it for basic changes. Unless this person left the company and now there is nobody in house with the needed knowledge, which is your boss's/company's problem and - if they are a good company - they should give you a reasonable amount of time to figure out all of this stuff the hard way, in which case the answer is - again - get searching the web.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 6 years ago.
Improve this question
Did someone try developing a GraphQL instead of RESTful API? Can someone give real life (not only theoretical) pros and cons. Basically from my research I found that the GraphQL power is to get exactly what you need nothing more. Where with REST APIs, you often have to make a series of requests and you can easily get back more information than you really needed.
Is it worth the time spent on researching and learning GraphQL? Any bugs or showstoppers that get your attention?
This question is primarily opinion-based.
But from my experience:
Multiple requests on a RESTful-API for just one thing often indicates a lack in the API design, namely the needed resource was not available and therefore stuff needs to be gathered from different resources to compensate for this.
A REST-API that could be easily replaced by GraphQL indicates, that the API was in fact a CRUD-HTTP-API, what is considered an Anti-Pattern among REST-Evangelists.
Also worth noting is, that GraphQL puts responsibilty on the client, because the backing API is reduced to be a datastore that just needs to be queried. REST on the other hand enforces the behaviour of the client and therefore reduces responsibility on it. The client gets reduced to be something similar to a browser.
There are cases the one or the other approach would yield better results, but that greatly depends on your situation.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 7 years ago.
Improve this question
I'd like to get the data of a website. I'd like to display the table in an app. Do you have an idea how I could do it? Thanks for your answers!
Usually, you'd want the maintainer of the data you need to supply some API for machine-to-machine communication (a REST JSON web service, for example).
Since you are asking how to display the table in an app:
the easiest way would be to just point an UIWebView that way and go from there.
a more native look might be acomplished by parsing the data. As you included several 'parsing' tags, I guess this is what you'd prefer.
The problem with HTML scraping web pages (what you probably hope to do) is that the data you are looking for and foremost it's structure is prune to changes. If some unexpected changes can easily break your parser.
Thus, if you go for doing that (which might be prohibited by your school or other publisher, especially in germany), try to parse the data on your server and offer an web service for your app yourself. This way, you can react to changes of the structure faster and do not break the app for your users.
Seriously consider asking the school for an API.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 7 years ago.
Improve this question
Yes, while i'm working on node.js, i still love perl, :)
The old web product is based on old perl CGI, i'm looking to the simplest way to fix XSS/Sql injection/etc. web security holes, within a week including testing, :(
So for
Catalyst
Dancer
Mason
Maypole
Mojolicious
which one should i use in the ARM platform ?
Thank you !
You have fallen foul of the primarily opinion-based off-topic categorisation, and your question will probably be closed very soon. However I think it's worth offering a few guidelines here
First of all you should absorb what is written in CGI::Alternatives as it is a reasonable summary of the subject
Next you should separate the HTML generation functionality of your existing CGI code from the interface itself, and consider replacements for each of them separately. If you were to use HTML::Tiny together with CGI::Simple then your code would have to change very little and you would have achieved better partitioning of functionality
Ideally you will move on to one of the many templating systems such as Template Toolkit, together with one of the frameworks, which is the topic of your question. In the end you will need to do a lot of research and many trials to discover how well each framework fits your requirement, in terms of both the feature list and the convenience and clarity of the API
All I can do here is say that I am very fond of the Mojolicious suite and suggest that it may be a good starting point. The API focuses on command chaining in a way similar to Ruby, and there is a Mojolicious::Plugin::CGI accessory which will allow you to execute CGI scripts unchanged during your migration
Note however that all of the frameworks that you mention, as well as several others, will have their proponents. That is why you must make the selection yourself, as such recommendations will be influenced primarily by familiarity, and without your own knowledge of the requirements of your project
Unfortunately I cannot speak to the security issues of the various options, but I hope that has helped a little
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 7 years ago.
Improve this question
I would like to have a URI like this /car/toyota, however I don't want the car resource to map to a database table for example. Instead I would like car to just be resource that is used for information retrieval only (ie. no POST, PUT, or DELETE on it), and /car/toyota/ would retrieve that data somewhere else, say through another REST API on another web server that provides this information.
Is this good design?
This probably belongs on something more like https://softwareengineering.stackexchange.com/
That said, this question depends entirely on the infrastructure of the environment you're making your REST requests on. If you have the ability to control the REST API on the web server providing the information, there's really no reason to wrap that API in another API. All of the call forwarding and potential necessity to translate from one request format to another really just adds un-necessary overhead.
That said, if you're accessing an API that you have no ability to re-format, or if you're accessing an API that you don't want client servers talking to directly, then there's a potential design perk for wrapping a different REST API in your own read only API.
Unfortunately, without having a clear picture of the entire architecture and the problem you're trying to solve, it's pretty difficult to decide if a wrapped API is a good design or not. My only advice is the preferred approach would be to edit the existing API if you can, but that isn't always practical.