Home – New Forums Tech talk What do you wish you knew about design/development/websites?

  • This topic is empty.
Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #987546
    LuchiaBloomfield
    Member
    • Total posts: 117
    Up
    0
    ::

    I’m interested to know what the average business owner would like to know about my field of expertise (design and development).

    Do you have any questions or do you wish there were articles about certain topics?

    #1163284
    martin.firth
    Member
    • Total posts: 259
    Up
    0
    ::

    Hi there,

    A very open-ended question indeed! Which may be the reason behind your lack of replies (until this one).

    There are also quite a few web gurus on the site, and especially in this forum, which might also account for the quiet, following your post.

    My recommendation would be to keep an eye out for threads in this section, and to journey into the review area. You’ll find people hungry for information, and other developers to bounce your assertions off of.

    Having said all this, I won’t leave you disappointed and with nothing to answer. I’ve had a look through your website to help me dig for a question: Why host your site’s images with TinyPic, rather than on your own server?

    Happy Easter, and welcome.

    #1163285
    LuchiaBloomfield
    Member
    • Total posts: 117
    Up
    0
    ::

    It’s running on Heroku and I’m not completely sold on it which is why I’m not paying for extra dynos until I find something I like more :)

    #1163286
    Greg_M
    Member
    • Total posts: 1,691
    Up
    0
    ::

    You may find you get a better result at Heroku if you use Unicorn as the app server, it’s multi threaded, unlike Webrick which is the Heroku default and perhaps not that great in production.

    Don’t know if you’ve tried it but https://www.openshift.com doesn’t go to “idle” mode like Heroku. It also uses Phusion Passenger as the default app server rather than Webrick.

    I don’t find it as sweet to use as Heroku, but they do give you pretty good resources for free, plus a range of DB’s.

    Cheers

    #1163287
    Greg_M
    Member
    • Total posts: 1,691
    Up
    0
    ::

    Actually I did have a question about the design process.

    Is inspiration, or design training the primary driver? And do you (or other designers for that matter) have set steps to get started on a new project?

    I find the web design process intriguing. In part because I can code, and can usually tell ugly from brilliant, but don’t always understand why it’s brilliant.

    Using something like Bootstrap, or Foundation I can usually avoid a total dogs breakfast, beyond that I struggle.

    Guess I’m specifically asking you the question because often I see rehashes of CMS templates here.

    As a Rails user I’m assuming you pretty much have to start from scratch … or at a minimum build your own templates, seeing as you can’t just sneak off to Theme Forrest, buy one and tweak it to suit.

    Cheers

    #1163288
    LuchiaBloomfield
    Member
    • Total posts: 117
    Up
    0
    ::

    I guess the main positive for Heroku is that it’s dead easy to deploy to and you don’t get that anywhere else. There is another one that I can’t think of the name right now but it’s a lot cheaper with similar features but does require a proper deploy every time you want to make changes.

    I’m looking at ways to optimise for Heroku and I saw Unicorn but haven’t had a chance to do anything yet due to client work being full on at the moment. I’ll take a better look at it, thanks for the suggestion.

    I do build from the ground up every time but I generally don’t build until I know what it looks like. If I’m doing client work it begins with their ideas and what they like as well as competitor analysis and market research to see what everyone is doing in the field. Then I’ll go away and sketch (and sketch and sketch and sketch) to try and create something that fits both what the client wants and what I think will work best to achieve their goals.

    I probably do about 3-4 revisions in sketches in total, then depending on the client I will either prototype/mock up in Photoshop or start building immediately. If the client is picky and a bit of a pixel pusher, I always go to Photoshop because then they feel like they can “see it” and be a part of it. If my client is more relaxed, I normally start building and they are happy to see progress, even if to begin with the project isn’t exactly the way we sketched it out.

    I guess my main drive for inspiration is competitor analysis and market research to be honest. If I can find some examples of what doesn’t work and iterate them to make them better for my client, I normally get a pretty amazing piece that stands out in it’s industry. I sometimes get hindered by client ideas, but generally I can persuade them to follow my lead. I’ve never had a client that hasn’t gleefully enjoyed taking bits and pieces of their competitions websites and improving them for their own.

    It’s a whole different story if the work is my own project. I normally sketch what I feel like and get to be creative as possible. Even if I don’t necessarily know how I’ll make it work in terms of developing and coding it, I can push and test boundaries. I normally err on the side of caution with that for client work, only because there are deadlines to be met and I’m not always sure I am capable of debugging a complex feature to make sure I meet dates.

    Wow, that was a long post. Hope that covers everything!

    #1163289
    Greg_M
    Member
    • Total posts: 1,691
    Up
    0
    ::

    Thanks for the reply.

    Good to know that a good old sketch is at the heart of the process.

    Given my ability with pencil and paper I think I’ll leave it with the pro’s, and one look inside Photoshop cured me for life.

    When you get some time (good that you’re busy), Unicorn is straightforward … install gem … plus a couple of cut and paste files from the Heroku doc’s.

    The reason I mentioned Openshift is that it uses a Git deploy method very similar to Heroku. They leverage off AWS in a similar way (just not so polished IMO).

    I’m using both ATM, Openshift for Rails 3.2 for a client prototype, so it won’t go to sleep if they want to see what’s happening. Plus one for myself at Heroku.

    Might need to drop you a line down the track … designers that don’t get lost in the back end of Rails are a pretty rare breed :)

    #1163290
    LuchiaBloomfield
    Member
    • Total posts: 117
    Up
    0
    ::

    You’d be surprised how bad my sketches are!

    I’ll definitely have a look into Openshift – I do like Heroku but it’s just so expensive when you want to scale it. Thanks for the input, by the way – sometimes it’s hard to find anyone to talk to about Ruby/Rails. I feel like I’m surrounded by PHP developers!!

    #1163291
    Greg_M
    Member
    • Total posts: 1,691
    Up
    0
    ::
    LuchiaBloomfield, post: 188929 wrote:
    You’d be surprised how bad my sketches are!

    I’ll definitely have a look into Openshift – I do like Heroku but it’s just so expensive when you want to scale it. Thanks for the input, by the way – sometimes it’s hard to find anyone to talk to about Ruby/Rails. I feel like I’m surrounded by PHP developers!!

    Even worse, their all WordPress/PHP developers.

    Openshift allocate a certain no. of “gears” (read RAM) for free (3 of), you can use each gear for three separate projects, or use the gears to scale an app or site.

    1 gear will run Rails OK, without idling, add a gear for traffic scaling and you’d have most situations covered. It does get about as expensive as Heroku once you get over the 3 gear limit (at that point a VPS starts to look cheaper-but you have to maintain it).

    Typically I use it as a test server, but I have had several Sinatra sites (client) running on it at different times with no issues.

    I actually prefer using Sinatra to Rails for small sites, but you can’t beat having all the Rails stuff there if a site is likely to grow in functionality.

    Most of what I do/done is speculative stuff for myself (not clients)… so Heroku is fine, if I actually find a winner, I think the cost is bearable when compared to the hassle of maintaining a VPS (I did do this for a while).

    If you’re building for clients, maybe a VPS is worthwhile … getting your head around deployment is a pain though (how Heroku captured the market in the first place).

    I’d stick to building sites that are so awesome, that clients wear the cost of Heroku, 1 dyno is still comparable to a VPS when you add up all the maintenance and security issues.

    Using a Cloud service like Heroku, your clients get the added bonus of Rails security, low maintenance, seamless scaling and a heap of industrial strength add on services … I think that makes conventional shared hosting look pretty sad as an option for a serious business.

Viewing 9 posts - 1 through 9 (of 9 total)
  • You must be logged in to reply to this topic.