New website idea, just need the right team of people!!

awec

IrishExpat wrote: »

I wish I could thank twice. All of this talk around 'the cloud', yet it's been reduced to a buzz-word with the lack of knowledge on the topic.

At the same time, the lack of knowledge around the cloud is leading people to think that it's more complicated than it really is.

The major cloud providers have all changed massively over the past year or so, getting off the ground with the cloud (e.g. aws or azure) is unbelievably easy.

awec

jimmycrackcorm wrote: »

You'll also need an architect. To support any growth in users then your system needs to be scalable which means designing around something like Amazon AWS or Microsoft Azure. You can start small and grow as required provided your design facilitates growth.

I'd also caution about using an open source solution as people mention. Not that they aren't good but that if you need to significantly deviate from their design and model then it will be very difficult. But as a proof of concept to get up and running it would be fine.

Colonel Panic wrote: »

Using AWS or Azure is a pretty compelling alternative to your own data centre not that you especially need to design around them, unless you plan to leverage some of their more interesting features.

They don't manage anything, they just offer metered services that you do what you want with.

Actually I was wrong, cloud was mentioned twice in the thread. OP - this is good advice.

I'm not sure about AWS, but I know Azure has the Azure Web Sites feature which is basic web hosting in the cloud with the scalable cloud benefits. I'd be surprised if AWS didn't have their own version of it.

Pay for what you use with instant scaling and instant worldwide delivery makes it a good choice for someone with an idea who isn't sure if what they have will be hit or miss.

ChRoMe

awec wrote: »

It is that simple.

That's part of the point of it, simple scalability.

Can you explain what isn't simple about it?

Throwing more hardware at a problem isint going to solve all scaling issues. Its incredibly naive to consider that to be the only possible bottleneck.

awec

ChRoMe wrote: »

Throwing more hardware at a problem isint going to solve all scaling issues. Its incredibly naive to consider that to be the only possible bottleneck.

Er, more hardware?

It's all virtualised, there is no throwing more hardware at anything.

What problem/bottleneck can he have that would be better solved by traditional hosting as opposed to AWS or Azure?

ChRoMe

awec wrote: »

Do you understand how the cloud works? It's all virtualised, there is no throwing more hardware at anything.

:eek:

Jesus wept..... I really hope you are not employed somewhere writing code.

awec

ChRoMe wrote: »

:eek:

Jesus wept..... I really hope you are not employed somewhere writing code.

I do this for a living. This has nothing at all to do with code, no idea why you're bringing that up.

Try again.

Are you saying the cloud isn't virtualised? You think that when you scale in the cloud some guy goes and plugs in more servers for you? Ha!

I'll ask again, what are the bottlenecks that traditional hosting solves that aws or azure doesn't?

Colonel Panic

What do you do exactly?

If the software doesn't scale, spinning up another VM won't make much of a difference!

awec

Colonel Panic wrote: »

What do you do exactly?

If the software doesn't scale, spinning up another VM won't make much of a difference!

Of course it won't. You still have to write your software to scale, nobody has denied that.

My point is that it is MUCH easier and MUCH cheaper to start of using AWS or Azure, given that scaling in future if needed is much faster.

Better to do that than have to arse about with traditional servers and all that sh!te to be able to handle the extra load.

Do it correctly at the start and your scaling will consist of you clicking 3 buttons.

The advantages are there, plain as day. A lot of people are afraid of it because it's new to them.

I'm a software dev, for one of the major tech companies. Not that that matters.

ChRoMe

awec wrote: »

This has nothing at all to do with code, no idea why you're bringing that up.

awec wrote:

You still have to write your software to scale, nobody has denied that.

These comments ,at least appear, to be contradicting each other.

Colonel Panic

It does matter what you do if you throw in the "I do this for a living" claim. It's pretty tiresome. There are plenty of professional devs here, some need to shout it from the rafters, others don't!

Your original post was a bit hit and run. You didn't really say much apart from the cloud solving infrastructure plans he doesn't have yet!

As it happens, I agree with out about using cloud based services, but even using something like AWS doesn't take all the pain out of the fact that even for a small application, you could be looking at a few virtual servers running an assortment of operating systems and services.

The Corinthian

awec wrote: »

It's all virtualised, there is no throwing more hardware at anything.

Actually there is; the 'Cloud' is simply a product of hiding this process from the customer so that it appears seamless.

Or did you think virtualization meant that it's all in some magical dimension where bandwidth, CPU and disk space just expanded magically when needed with no physical presence?

Ultimately, it may not be a guy plugging in additional servers for you, but a program that effectively does that by giving you the resources of additional servers - the cloud is simply the culmination of software automation in the hosting sector that was already visible to anyone dealing with Rackspace twelve years ago.

In this regard, if the pricing model is good, using the 'Cloud' can be very beneficial to a start-up company (esp. under Google's old pricing model), but there are many cases where it's a waste of money.

What problem/bottleneck can he have that would be better solved by traditional hosting as opposed to AWS or Azure?

Those that are the product of inefficient code, for example.

awec

Colonel Panic wrote: »

It does matter what you do if you throw in the "I do this for a living" claim. It's pretty tiresome. There are plenty of professional devs here, some need to shout it from the rafters, others don't!

Your original post was a bit hit and run. You didn't really say much apart from the cloud solving infrastructure plans he doesn't have yet!

As it happens, I agree with out about using cloud based services, but even using something like AWS doesn't take all the pain out of the fact that even for a small application, you could be looking at a few virtual servers running an assortment of operating systems and services.

I don't. Seriously, I don't. As I said, it doesn't matter, the beauty about this is that you don't need to have some super degree or some super job to understand what goes on.

However, if you read back, I had the old "people don't know what the cloud is", "buzzword" etc thrown at me in a thinly veiled attempt to make out like I'd only heard about it in passing and didn't know what I was talking about.

That got my back up, but if I was rude or annoyed anyone then I apologise for that.

Anyway, I think this particular sub-discussion has been done now, we'll only end up confusing the OP.

kippy

Most important person in your "team" initially, will be your bank manager.

ChRoMe

awec wrote: »

didn't know what I was talking about.

As The Corinthian has pointed out in his post, your subsequent responses don't really support you having a body of knowledge. Throwing more hardware at a problem is EXACTLY what AWS does.

Colonel Panic

ChRoMe wrote: »

As The Corinthian has pointed out in his post, your subsequent responses don't really support you having a body of knowledge. Throwing more hardware at a problem is EXACTLY what AWS does.

That's a bit of a generalisation. I mean, that's what scaling is. If you have some front facing web server acting as a proxy for multiple services in the background that all access some database cluster with Memcached on multiple servers sitting in between the DB and app, then scaling the application is a matter of throwing more (virtual) hardware at the problem.

Of course the software needs to support this as much as it does smart development practices like handling requests asynchronously in multiple threads, for example. I've had miserable experiences with monolithic pieces of software in environments like this.

It's fine to be dismissive of the Cloud term, but don't knock being able to bang out an EC2 instance in no time, run some scripts, edit a config file and see the load on your DB server drop right before your eyes.

The Corinthian

Colonel Panic wrote: »

That's a bit of a generalisation. I mean, that's what scaling is.

Indeed, there's certainly a lot more to Cloud services than flexible scaling, but what ChRoMe and I did was correct the erroneous claim when scaling it did not involve "throwing more hardware at anything". Essentially it does do that.

It's fine to be dismissive of the Cloud term, but don't knock being able to bang out an EC2 instance in no time, run some scripts, edit a config file and see the load on your DB server drop right before your eyes.

I certainly wouldn't be dismissive, but this tangent in the discussion came about only when someone did the opposite and claimed that "Cloud hosting solves your infrastructure problem". It might, but so does buying your own server farm - doesn't make it the right solution.

As for the OP, I think it's clear that they're either in the very, very early stages of an idea (far too soon to even contemplate a team) or they were just tyre-kicking. So I suspect, unless they return or we end up on another tangent, this thread has run its course.

clint_silver

Starting up with the cloud is a no brainer for me. If he has performance/scalability issues, more resources from AWS may solve at a cheaper cost than buying hardware and renting landscape in a data centre somewhere.

If OP has scalability problems, the cloud may be the answer, it may not, it depends on many twists and turns his architect and developer take, but either way I think the argument here is several steps down the line from where OP needs to be.

To get it off the ground, Id suggest discussing with a Professional web development company. Any of the big guys will talk to you briefly to discuss what you need but only a superficial level before advisory costs are agreed upon. It would then be well worth it to pay for a days consultancy. They will discuss the challenges ahead of you more professionaly.

Remember facebook was an idea (debatable it was his, but hey ho) from a guy who also able to do up a prototype.

fergalr

clint_silver wrote: »

To get it off the ground, Id suggest discussing with a Professional web development company. Any of the big guys will talk to you briefly to discuss what you need but only a superficial level before advisory costs are agreed upon. It would then be well worth it to pay for a days consultancy. They will discuss the challenges ahead of you more professionaly.

I strongly disagree with that advice.

As I said, I believe the OP needs to do a lot more domain research, and understand what they want to build, and the process of building startups.

Then the OP needs to try convince a technical co-founder to come on board, rather than trying to pay consultants to build this.

ChRoMe

clint_silver wrote: »

Remember facebook was an idea (debatable it was his, but hey ho) from a guy who also able to do up a prototype.

Not to mention being in one of the most exclusive 3rd level educational institutions in the world and therefore had a near limitless access to capital to start the idea.

Colonel Panic

The Corinthian wrote: »

Indeed, there's certainly a lot more to Cloud services than flexible scaling, but what ChRoMe and I did was correct the erroneous claim when scaling it did not involve "throwing more hardware at anything". Essentially it does do that.

I certainly wouldn't be dismissive, but this tangent in the discussion came about only when someone did the opposite and claimed that "Cloud hosting solves your infrastructure problem". It might, but so does buying your own server farm - doesn't make it the right solution.

As for the OP, I think it's clear that they're either in the very, very early stages of an idea (far too soon to even contemplate a team) or they were just tyre-kicking. So I suspect, unless they return or we end up on another tangent, this thread has run its course.

I get what you mean, I find it hard to resist smirking at Cloud as Golden Hammer suggestions myself and I'm actually a little wary of it having been put in an awkward position developing apps for cloud based platforms in the past where the platform and app didn't gel well together.

I think the debate is much ado about nothing really, I got the impression ChRoMe was just dismissing the concept of being able to scale with hardware as a Bad Thing when this is far from the truth.

ChRoMe

Colonel Panic wrote: »

I think the debate is much ado about nothing really, I got the impression ChRoMe was just dismissing the concept of being able to scale with hardware as a Bad Thing when this is far from the truth.

Far far from it, my post was a response to awec. Apologies if that wasnt clear.

The Corinthian

fergalr wrote: »

As I said, I believe the OP needs to do a lot more domain research, and understand what they want to build, and the process of building startups.

Agreed. This is a business he's thinking of setting up, not a college project, not a hobby. While early technical input can be important, he's at so early a stage that he should be looking at if it's a runner first, before he starts looking at how it's a runner.

Colonel Panic wrote: »

I think the debate is much ado about nothing really, I got the impression ChRoMe was just dismissing the concept of being able to scale with hardware as a Bad Thing when this is far from the truth.

Not at all. Hosting a solution via the cloud can make lots of sense in many circumstances. ChRoMe and I were more dismissing the portrayal of the cloud as some sort of answer to all hosting needs and what appeared to be very flawed understanding, on the part of awec, of what the cloud even is.

If its scalability was not based on the addition or subtraction of hardware/bandwidth resources, what does he think it manages this? Pixie dust?

Colonel Panic

Usually that sort of assumption is reserved for management!

clint_silver

fergalr wrote: »

I strongly disagree with that advice.

As I said, I believe the OP needs to do a lot more domain research, and understand what they want to build, and the process of building startups.

Then the OP needs to try convince a technical co-founder to come on board, rather than trying to pay consultants to build this.

Maybe I wasnt clear. I was suggesting he speaks to a development company to get a better idea of how much will be required both in work terms and buidget terms. We can throw best practice testing and development terms in for pages of this thread but if he seen a properly built project plan draft in front of him he might save himself hundreds of thousands by knowing its not a runner.

It seems the ideas are there already, OP is on here looking to see whats needed to take it to next stage. We all know a massive budget is required. He wont really grasp what until he sees the scope. Any pro development company will know the budgetary scope. Lets say we say a PM, an architect and 3 developers. 3 months to prototype. Even if we were close on the money, theres no way the OP is going to be able to find those people so a development company is the way to go. So he goes to company X. Tells them what hes looking for, they say figure X, he then has to go investors looking for it.

Unless OP has 250k wasting away in the back phoca, he needs investment. How does he know how much until he talks to someone?

The Corinthian

clint_silver wrote: »

Maybe I wasnt clear. I was suggesting he speaks to a development company to get a better idea of how much will be required both in work terms and buidget terms. We can throw best practice testing and development terms in for pages of this thread but if he seen a properly built project plan draft in front of him he might save himself hundreds of thousands by knowing its not a runner.

Researching the development needs of the business is certainly something he would need to do - just not yet. First, I really do think he needs to answer a far more fundamental question, which is whether the business will make sufficient money to be viable, and if so how.

For example; we know it's a social media Web site, but that's about it. Will it seek to make money on subscriptions, advertising or both? Or does it have another business model in mind? No doubt he has some ideas on this, but he would want to research their viability before deciding upon it, and by his own admission he hasn't.

Now, he may need to consult someone to do this research, but it won't be a developer or development house, as these may know how to build it, but are rarely around to see if it makes any money. Instead, he'd might need to contact someone more specialized in advertising or e-commerce consultancy - and I don't mean (just) the development thereof.

Until he has a business model pinned down, there's not a lot of point talking to developers. Sure developers can give him lots of ideas on what he can do technically, but not commercially. In fact, how can he even assess what team he needs if he doesn't now if he'll need, say, an on-line advertising specialist or an e-commerce expert?

It seems the ideas are there already, OP is on here looking to see whats needed to take it to next stage.

If you read what the OP wrote, you'll find that a basic proto-idea is there, that's all. I don't think he has any revenue model researched, and as per above, it's a bit pointless trying to decide who's on your team if you don't yet know if you'll need e-commerce or marketing skills on-board yet.

So it's currently only ready to go to the next stage, if you mean step one.

I think he's realized this and either has gone off to do this, or has given up, as he essentially admitted it and stopped posting.

PrzemoF

jimmycrackcorm wrote: »

[..]
I'd also caution about using an open source solution as people mention. Not that they aren't good but that if you need to significantly deviate from their design and model then it will be very difficult. But as a proof of concept to get up and running it would be fine.

Interesting... So, do I read it correctly that a closed, proprietary system would be easier if you want to deviate from the it's original design? :eek:

ChRoMe

PrzemoF wrote: »

Interesting... So, do I read it correctly that a closed, proprietary system would be easier if you want to deviate from the it's original design? :eek:

Heh, I'd imagine that he is referring to rolling your own from the start

The Corinthian

ChRoMe wrote: »

Heh, I'd imagine that he is referring to rolling your own from the start

A 'closed, proprietary system' is always easier to make changes to if you wrote it in the first place

ChRoMe

The Corinthian wrote: »

A 'closed, proprietary system' is always easier to make changes to if you wrote it in the first place

You haven't worked with some of the developers I have then!

Giblet

I think scalability isn't a concern at this point. I'd get to writing some documents.