I am in contracting staffing, so I see everything as people and hours, which seems to be similar to how you are thinking about it.

It sounds like it may be an awesome opportunity to get this client to pay for an added position (support engineer) and add an in-house dedicated support to your product.

I would consider what it will cost you to pay that person and multiply that number by 1.50 to give yourselves a margin for profit.

Maybe you should pitch it as a dedicated support individual for the client, and see what their ideas are.

We have some people in mind we'd love to have on our team as full time support, you're right that it would be an amazing opportunity to have him on our team as fully time support for that contract + extra time spent doing our support. Pitching is as a dedicated support individual is an amazing idea thanks!

I believe our product should be saving this company a lot of money for various reasons which leads me to believe we can probably negotiate a higher rate than 1.5x. I imagine wages for our full time support person would be in the region of $60k/y if we also include various expenses for having them. So 1.5x would bring us to $90k. We'd also need to be doing a lot of programming work for this contract I imagine, which is skilled and takes time and the support person wouldn't be skilled enough to undertake. It's also going to be sacrificing the progression of our business as well as we would need to divert time/resources to them which is a risk as all the work we would do for them would be specific to them and not useful for the product in general. This makes us think it's only going to be worth it around the $200k/y range. And I'm not sure if that's a figure that would be offensive as it seems very large and am not really sure what price ranges contracts like this fall under.

If we add the fact it's going to be a 3 year contract as well this makes it even riskier.

$200k/year is not expensive for some things.

They are getting dedicated support and specific customized programming, you can host webinar's and trainings for the client's developers, etc. There is a lot you can offer over the course of 3 years.

My dad was at an escalated support position with a big company and a client was implementing a new product and wanted to do it over the weekend to be as seamless as possible, so they purchased dedicated 24/7 support for 1 weekend and they were paying $10k/hour for it.

I want to create the set of graphs with n vertices and k edges in mathematica, any ideas how I would do this?

First stage of my ventures into Rubygame...an artist I am not



Dunno why the player transparancy doesn't work, probably did something wrong in the Gimp. Already looked into Incscape Vector graphics tutorials etc and will fiddle with that a bit.

I think I'll center the verbs a bit and refactor the code before moving on. The way the inventory is drawn is currently rather ugly/unDRY etc.

[I did smth similar a while ago in PyGame but not TDD and I just decided to throw the code away because I couldn't be bothered reading through it again...+wanted to try Rubygame/Ruby anyways]

Gull

We have a client that pays us for two dedicated developers at work. I don't have the details of the contract (and wouldn't share them here anyway). I do know that they pay enough not only to cover their salaries, but also a dedicated team lead (project manager), and leaving a profit at the end of the day.

This is a major player in their industry, and you've certainly heard of them. Those big corporations are willing to pay significant money for you to provide a solution to their problem, and they pay a premium for the ability to just drop it on your plate.

Determine exactly what their needs are - be it a retainer or on call, paid by the hour support - and quote appropriately. If it's a retainer, I would absolutely propose a monthly amount sufficient to cover a salary (full cost of employment, not just their pay check) plus profit.

I wouldn't be afraid of chasing them away - if it's too much, they'll say so, and you can negotiate from there.

Thx, already found the problem a couple of seconds after just using your dude. I used "transparancy from color" or whatever it's called which somehow left some grey. Needs to be redone in vector graphics soon anyways. But I did plug in your dude and it made a test fail...weee (I test that it is size 60x120)

I'm also missing a "Look at" verb heh...I think use,take,look,talk is the minimum. I might change the entire inventory/interface later anyways (right click to rotate through icons for the verbs, small icon inventory a la broken sword)

Rspec:

So uh I do existance testing fairly often and usually start testing classes with something like
Anyone bothered by the fact that it's a inventory .. guess I could just write a/an don't really want to write a [a,e,i,o,u] check wrapping around it
NOTE: I do prefer explicit do...end with my it over curly braces for readability

Anyone doing it differently?

i mean, it's ruby, so surely you can just do something like:

?

also i'm (obviously, given the above) not much of a rubyist but curly braces would definitely look awkward to me. do/end seems more natural.

I know there was alot of discussion about Registrars around the SOPA issue...was there ever a consensus about which registar is best?

What about for basic email hosting? Cheap is good, but I definitely would rather pay a little more to get a reputable/stable company. Definitely can't have something so cheap that they aren't reliable or performance is terrible.

Thanks!

namesilo w/coupons

gmail offering morse code input instead of normal keyboard

https://mail.google.com/mail/help/promos/tap/index.html

srsly ?

or did i get april fooled

nope.

yup.

relatedly, check out quest mode in google maps!

April fools was good a few years ago when it wasn't yet another dumb Internet meme. Slashdot is unreadable today, google gonna google, at least it's Easter so the religion haters can have some fun. </oldManMode>

"has #{attribute}" bypasses having to do any word fixing.

Anyone do any programming related april fools? We did one http://www.scirra.com/forum/r84-bork_topic50733.html
(Goes away when you reload the project )

Also anyone else got much experience with writing AJAX content pages? (eg: https://chrome.google.com/webstore/category/home , not simple AJAX but entire pages) I've never really done it before and it's really fiddly to get everything right (also writing a store)! I'm in two minds as to if this sort of page loading is the future or not. It probably is, but I don't see it sweeping the web quickly as it is difficult to execute. Pages will appear to load a ton quicker which is the main benefit, but to get it to feel like a 'native page load' is tricky, Chrome Webstore does it very well though.

We should probably start a thread on this. Been looking into the various js frameworks for this kind of thing, and ember.js looks like the most impressive so far, even though backbone.js had the most users at the moment. I think I'm going to spend some time learning ember soon. One thing I can say: doing client side heavy webapps without one of these frameworks is a fool's errand.

I really don't like relying on frameworks for things like this, you might be right with heavy webapps but for 'simple' page content loading the actual code is quite simple but just fiddly.

Edit: By fiddly I mean making sure the actual pages and things like SEO don't get ruined

There are too many browser incompatibilities to deal with to make ajax'ified pages the next big thing and by the time all of them are ironed out we won't even be using http anymore to serve content (or even access sites perhaps).

Even that Google site falls back to hash urls for browsers that don't support pushState, and pushState itself has awkward side effects on the browsers that support it.

You also have to build 2 versions of your site if you want to support people without JS or have those precious dynamic pages properly indexed by all search engines (not just Google).

The future of web applications will be web sockets and it's not even close. The only problem is it's not fully supported by all grade A browsers and the servers that serve our pages are heavily geared towards dealing with http requests/responses and the notion of "pages" (for caching), not an open connection where data is streamed to/from a client.

Personally I would build normal sites, spruce it up with javascript, and optionally have ajax content layered on top.

If you want legit real time bi-directional streaming content that's not meant to be crawled/cached then I would say fk everything and use web sockets. If the user doesn't have a compatible browser then tell him to download one that supports it.

Edit:
http://www.youtube.com/watch?v=hdTxeR90_1E
This is a pretty decent video to watch, as it's slightly related to what you're talking about.

I am using namecheap.com as a registrar, Amazon Route53 for DNS (because I load balance behind EC2 this is really my only option, but I am very happy with the service)

For email hosting we use google apps. I really have to say that google apps is fantastic. Very straight forward SPF and DKIM setup and I can check the emails automatically from my existing gmail account.

Shoe Lace,
Even with web sockets you still need a UI in js, no? What are you going to use to build that?

I cannot stress this enough: learn ember now.

I've written a number of single page JS apps over the last year. At first a couple in backbone and most recently one really big on in Ember. I don't see myself really ever wanting to write another backbone app again, ember is just that great.

Ember.JS is really impressive and it makes writing code fun. There are a number of awesome features it has, but the two that really stand out are the smart binding system and the computed properties.

I'd love to see a thread get started around client side js frameworks/single page apps.

Exactly. Use the websockets to send data back and forth between the server and browser. Then use a client side JS framework to maintain state on the browsers end. Suddenly you'll see the importance of MVC and how hard it is to maintain without a good framework.

Shoelace does bring up some good points. Push state is very awkward right now when dealing with older browsers. Luckily many of these frameworks have great push state fallbacks. See backbone's implementation of this for a really simple (and good) example. It is still ugly having all these #! urls though.

The big problem here is SEO. We don't know how well google is at running these apps and if/when it will index them. A trick I use is I have a url: www.example.com/user/123 and when you request that it will generate the page for user 123 and render it, and then start the JS application with the state of user 123. This allows search engines to actually request real urls representing different states of the application. The problem here is this sort of idea quickly becomes an awful cluster**** of complexity. It is certainly the biggest pain point when writing these apps.

I know of a few developers that don't bother with the search engines or pleasing them. I think this might be the right approach because if these websocket/backbone/ember apps get more and more popular on the internet then google is going to have to learn how to index them.

If you're writing a webapp, SEO is irrelevant. You need to SEO your content pages, not your application, and I personally don't see the point of doing heavy AJAX loading on content. It's a hell of a lot of effort and hassle without really adding anything to your site