That is the best blog post about scaling eva! LMAO
On 08/04/2009, at 3:41 PM, David Turnbull wrote: > > Agreed, http://teddziuba.com/2008/04/im-going-to-scale-my-foot-up-y.html > > On 08/04/2009, at 3:39 PM, Cameron Barrie wrote: > >> I agree with that last point. >> I do loathe the twitter examples in regards to scaling(no disrespect >> meant to anyone BTW). >> I can't think of any framework/architecture out of the box that >> would have, Got that built for them quickly as it did, and not had a >> scaling issue. >> Twitter get's slammed by millions and millions of requests, nothing >> would've handled that load, nor should they have anticipated/ >> architected for that sort of load from the beginning. >> Granted they took to long to come up with a solution, but that's not >> he fault of the language/framework. It's a pretty freakin' good >> solution IMHO. >> >> I had a .Net dev telling me the other day he'd never use Rails since >> it doesn't scale, and pointed out twitter as his example, I asked >> him the last time he wrote a site/service designed to handle that >> sort of load, answer... Never has and probably never will. Like most >> of us. Then of course he also pointed out, that a .NET app straight >> out of the box would've died in the arse as well under that sort of >> load. >> >> Twitter is the scaling corner case or all corner cases. >> >> My opinion of twitter is more along the lines of you want to scale >> Rails, look at twitter that thing is freakin' crazy if they can >> scale that you can scale anything. With some hard thinking and good >> coders you can even make that shit work in Ruby. It's a developers >> best friend after all. :D >> >> C >> >> On 08/04/2009, at 3:26 PM, Daniel Sabados wrote: >> >>> IMHO, if one were building a site that would cater for millions of >>> hits per day before >>> they've even launched it, they've either hit the holy grail of >>> ideas or they're building >>> a site that will host good quality porn. >>> >>> I wouldn't be making any assumptions about high traffic loads early >>> in the game. >>> Therefore, performance or scalability shouldn't be all that >>> relevant. >>> >>> I'm in the "build first, optimise later" camp. One of the things >>> I've learnt is that getting >>> to market first is what counts most. Afterwards, when most killer >>> ideas have been chewed >>> up and spat out, the remaining few are faced with other issues like >>> scalability. >>> >>> As for Twitter, a corner case I'd say. >>> >>> >>> Cheers, >>> >>> >>> Dan. >>> >>> >>> >>> On Wed, Apr 8, 2009 at 3:05 PM, Torm3nt <[email protected]> wrote: >>> >>> Hey Dave, >>> >>> Cheers for the input man, greatly appreciated! >>> >>> My point about development time, depends on what the project is. If >>> it's a small, lightweight project then the quicker you can get it >>> out, >>> more than likely the better. If you're going to be creating a >>> project >>> that is going to attract millions of hits a day, it would be worth >>> factoring that into the technology and framework choice, rather than >>> simply going with whatever is most comfortable. I know I myself have >>> fallen into that trap, and it cost me a lot later down the track. >>> >>> Software projects are very much a build first, optimize later - but >>> if >>> it means redeveloping your project (read: twitter), then it would >>> seem >>> that not enough planning and forethought was provided in order to >>> maximise on the platform of choice. (btw, I don't think twitter is >>> necessarily making the best decision =P) >>> >>> That said, not all software projects are web projects and this is >>> what >>> I was trying to stress (on multiple fronts - obviously I didn't >>> communicate it too well =( ). I mean, using rails to create a >>> reporting tool that outputs to console probably isn't the most sound >>> choice for the job, and on the other hand - embedding HTML within a >>> PHP script wouldn't be the best solution to an enterprise-scale >>> project, either. I've even seen on some forums people wanting to use >>> Rails just so that they have access to ActiveRecord, when there are >>> quite a few solutions out there for such a task, which isn't tied >>> to a >>> full-stack framework. >>> >>> "Seems like a bit of a dead issue to me anyway -- accepted practice >>> is >>> to develop then optimise." >>> >>> Very much like my DRY CSS post - it's more of a thought-provoking >>> exercise, and it helps me map my own current thoughts to paper, as >>> well as hope to do the same for others, as is done via this very >>> discussion =) >>> >>> >>> Cheers, >>> >>> Kirk >>> >>> On Wed, Apr 8, 2009 at 2:48 PM, Dave Bolton <[email protected]> >>> wrote: >>>> I wasn't able to read your blog post before, but I understand >>> more what you >>>> are getting at now that your server is back up. >>>> >>>> So, you're talking about heavyweight in terms of performance. It's >>>> difficult to address your ideas without getting specific about >>> frameworks -- >>>> defensiveness be damned, if you're going to call out heavyweight >>> frameworks >>>> you *need* to get specific so we can discuss. >>>> >>>> The example in your email of a Rails app for database reports >>> seems at odds >>>> with the blog post. Is a reporting app ever going to need to >>> serve so many >>>> pages that the framework is the issue? I think not. >>>> >>>>> [from blog post]: "Development time is generally much more >>> expensive than >>>>> hardware (unless you're serving millions of page view per day), >>> and so >>>>> therefore (depending on the project) a sound and educated >>> decision should be >>>>> made, rather than simply jumping into your favourite development >>> framework." >>>> >>>> Given that development time is more expensive than hardware, then >>> that is >>>> EXACTLY the reason you should use your favourite framework >>> first. I take >>>> your point that when you have a hammer, every problem seems to be >>> a nail, >>>> but also, if you're developing for the web, many problems *are* >>> nails, and >>>> most frameworks *are* hammers. Just use the one you are most >>> comfortable >>>> with. >>>> >>>> Seems like a bit of a dead issue to me anyway -- accepted >>> practice is to >>>> develop then optimise. Are there any frameworks that are really >>> that bad >>>> that performance is going to kill you before your site is massive >>> anyway? >>>> (if caching is used) >>>> >>>> Cheers, >>>> Dave >>>> >>>> >>>> On Wed, Apr 8, 2009 at 12:05 PM, Torm3nt <[email protected]> wrote: >>>>> >>>>> Well there were a couple of frameworks, but I was trying to be >>>>> framework-agnostic for a reason - there are a few that are quite >>> heavy >>>>> and load up a lot of files and libraries, even if they're not >>> used - >>>>> so as to make the job easier for the developer. Plus I didn't >>> want to >>>>> be labelled as a "oh he doesn't like framework x so he mustn't be >>>>> good with it".etc. If I targeted any specific framework, people >>> are >>>>> bound to get defensive of their framework of choice, which >>> detracts >>>>> from what I wanted to convey. >>>>> >>>>> >>>>> Kirk >>>>> >>>>> >>>>> >>>>> On Wed, Apr 8, 2009 at 12:01 PM, Dave Bolton >>> <[email protected]> wrote: >>>>>> - Which frameworks are you targeting? >>>>>> - Heavyweight in what sense? Resources? Concepts? Performance? >>>>>> >>>>>> With respect to Rails, I don't have a problem throwing it at >>> small ideas >>>>>> -- >>>>>> the concepts are easy and standard, and resource requirements >>> are not so >>>>>> onerous to make it unfeasible (in all senses of resources). >>>>>> >>>>>> I guess you could mean heavyweight in the sense of all the >>> things Rails >>>>>> lets >>>>>> you do out-of-the-box, but there's no compulsion to use them >>> all, and >>>>>> I'd be >>>>>> suprised if they changed the resource and performance >>> equations so much >>>>>> to >>>>>> make Rails no longer an option. >>>>>> >>>>>> So, put me in Dr Nic's "scales for the size of the project" >>> bucket. >>>>>> >>>>>> Cheers, >>>>>> Dave >>>>>> >>>>>> >>>>>> On Wed, Apr 8, 2009 at 11:35 AM, Torm3nt <[email protected]> >>> wrote: >>>>>>> >>>>>>> Cheers for your input Dr Nic, >>>>>>> >>>>>>> I wasn't actually specifically targeting rails - rails 3.0 >>> certainly >>>>>>> looks to be much more enticing as far as frameworks goes as >>> you'll be >>>>>>> able to plug and play various libraries together, but not many >>>>>>> frameworks do this =P >>>>>>> >>>>>>> >>>>>>> Kirk >>>>>>> >>>>>>> On Wed, Apr 8, 2009 at 11:24 AM, Dr Nic Williams >>>>>>> <[email protected] >>>> >>>>>>> wrote: >>>>>>>> People talk about "rails doesn't scale" and mean >>> performance. What I >>>>>>>> love >>>>>>>> about Rails is that scales for the size of the project. You >>> can start >>>>>>>> a >>>>>>>> micro project today, and it easily evolves into a bigger >>> project. >>>>>>>> The single-file-contains-my-app frameworks aren't wrong or >>> broken; >>>>>>>> rather >>>>>>>> they take away one of the oft-forgotten but awesome aspects >>> of Rails: >>>>>>>> you >>>>>>>> and I both know where our next model or controller is going >>> to go. >>>>>>>> The >>>>>>>> generators know it. The IDEs/editors know it. >>>>>>>> The heavy-weightedness of Rails will probably become >>> optional as we >>>>>>>> move >>>>>>>> to >>>>>>>> 3.0 and beyond. >>>>>>>> >>>>>>>> On Wed, Apr 8, 2009 at 11:19 AM, Torm3nt >>> <[email protected]> wrote: >>>>>>>>> >>>>>>>>> Hey all! >>>>>>>>> >>>>>>>>> I've recently been musing over the use of heavy frameworks >>> (such as >>>>>>>>> RoR) and how I'm beginning to see (in some cases) them being >>>>>>>>> overused, >>>>>>>>> mostly for the wrong purposes. In one instance I witnessed >>> a Rails >>>>>>>>> application for getting reports on a database. >>>>>>>>> >>>>>>>>> I've written my thoughts on this and would love to hear >>> from some of >>>>>>>>> the more intelligent people in this community, either of >>> their own >>>>>>>>> experiences or even a counter-argument =) >>>>>>>>> >>>>>>>>> http://www.kirkbushell.com/articles/using-the-right-tool-for-the-job >>>>>>>>> >>>>>>>>> >>>>>>>>> Cheers, >>>>>>>>> >>>>>>>>> Kirk Bushell >>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> -- >>>>>>>> Dr Nic Williams >>>>>>>> Mocra - Premier iPhone and Ruby on Rails Consultants >>>>>>>> w - http://mocra.com >>>>>>>> twitter - @drnic >>>>>>>> skype - nicwilliams >>>>>>>> e - [email protected] >>>>>>>> p - +61 412 002 126 or +61 7 3102 3237 >>>>>>>> >>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>>>> >>>>>> >>>>> >>>>> >>>> >>>> >>>>> >>>> >>> >>> >>> >>> >>> >>> >> >> >>> > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Ruby or Rails Oceania" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/rails-oceania?hl=en -~----------~----~----~----~------~----~------~--~---
