Since 7 reasons I switched back to PHP there seems to be a trend of Rails-bashing articles, epitomized by this one which is a fine example of the form until it advocates ASP.NET. Through it all, I still haven’t heard of a startup or web service that failed or succeeded due solely to its web framework or language. These articles are like the celebrity gossip stories of Web 2.0, complete with ad hominem attacks, and just as useless. Hacker News tends to be a fairly high signal source of discussions actually relevant to startups.
Rails has been most useful to me as a philosophy. “Convention over configuration,” “only code things once,” etc. I find that most of that stuff is completely language-agnostic.
Amen matt!
A now-offline Ruby wiki mentioned Rails along with a statement that reads like: “When it comes to programming languages, we’re all in the fashion business”.
If Paul Graham was 15 years younger, founded Viaweb using Lisp on Lollipops (LOL), our arguments wouldn’t have been much different, don’t you think?
Ruby/Rails community are becoming like a cult, except from the original Japanese Ruby community before the western world bastardise the whole attitude.
Meanwhile, Mac community is also another cult.
No wonder there are so many Rails developers worshipping Mac. Oooo, TextMate is all I need.
Cult, baby. Cult!
I think a lot of it has to do with inflated expectations — “I can take my Mediocre Idea ™ and thanks to the magic of rails become an InterWeb billionaire!”
I like PHP because it’s quick and dirty and when I absolutely need to get something done right now I can just do it. Rails starts a little slower (for me, at least), but the speed picks up after a while.
I’m off now to search for a getting started tutorial for Lisp on Lollipops. Now that’s a framework I can get behind! 😉
Colin Brumelle has taken to saying that what’s important in picking a language or framework now has less to do with the technology, syntax or features but with the community that maintains the language.
I think this is true and one of the enduring reasons why WordPress continues to excel. It’s not that it’s written in PHP, though that does seem to lower the barrier to hacking and to hosting, instead it’s the quality of community that have gathered around, and stuck with, WordPress.
Matt,
I’m sure you’re aware but the author is referring to ASP.NET and not the original ASP? If you are aware that he was referring to ASP.NET, out of interest what makes you make a statement like that?
Al.
For every post like Rob’s, there are thousands of posts that mindlessly bash Microsoft tech for no better reason than it came from Microsoft.
It may be the case that no startup or project ever succeeded or failed because of the language or framework it uses, but that’s more a function of the fact that it’s nearly impossible to measure developer productivity in an accurate, granular, and reproducible fashion.
Matt, you should understand that there is a difference between “ASP” and ASP.NET which is what Rob Conery is advocating in his post. They’re two totally different things.
I think the 7 reason article shown as the first link in this post is a far cry from “bashing Rails.” I read the article and believe the author was very clear that he loves Rails but didn’t need to dump PHP. After 2 years of trying to convert his site from PHP to Rails, he learned a lot and finally realized that he could do everything he wanted with PHP and learned how to make it look nice by spending 2 years entrenched in Rails. His article was very fair to Rails and no where did he “bash” it.
That being said I agree that some of these negative articles are nothing more than “celebrity gossip.” And, I’ve grown very tired of articles or comments that assert one language isn’t a “real” programming language or that one language is vastly superior to another.
Programming languages are mostly about personal preference. Some do tend to outshine some in certain areas but for the most part they can all do the same thing — they just do a little different from each other.
Its like that old saying on the internet (I’m paraphrasing here):
“Arguing about programming languages is like winning the Special Olympics. At the end of the day, you’re still retarded”.
I investigated Ruby/Rails – even read two books on the subject, and realized – I can just use CakePHP, have the benefits of a RAD and already know the syntax.
Apples and Oranges people.
Hi Matt. If you talk to some of Mr. Sivers’ employees at CD Baby, you’ll find that they HATE his new system, and it’s been a huge disaster for them, and their customers. Only in his own head, located halfway across the world from his own small business, has PHP been an inarguable success.
Jeffrey, my intention wasn’t to kick off a debate just like the one I was criticizing. I’ve updated the entry to say .NET, and my intention with the comment was to point out that generally when people make that type of rant they advocate Python or PHP.
Raymond, that’s a pretty serious suggestion to make, do you have a link or proof to back it up?
I think articles like these are slightly FUD. The title states “Rails Bashing” and the main idea behind the article is the “7 Reasons I Switched Back to PHP”. Great. Except that the article linked doesn’t necessarily bash RoR, it in fact states that with a clean slate RoR is a great framework. This article does nothing to show that a migration from an existing (might I add bloated) PHP app (i.e. cdbaby) didn’t convert well — but realize that users needed to be supported in the transition, documentation probably needed to be revisited and there was going to eventually be a hard cut date which, with some good process and project management, is very doable. I, personally, think that the project may have lacked some good direction and/or management. That being said RoR would not be where it is today if it wasn’t great. I’ve dabbled in PHP and RoR and find RoR with the upper hand in the end when I’m looking for enterprise class scalability. I’m not saying PHP can’t get there, and someone mentioned CakePHP (which does a fine job of modeling after RoR) which is great. In the end though, this article lacks proof and in my book ends up being FUD. Referencing “Hacker News” gives no extra merit — how about look to what large corporations are moving to in the background. Look at EngineYard and how successful they have been in the enterprise space. You need to look outside the little box you setup in this article and realize there are companies out there with more than 10 people. Again, flagged as FUD in my book.
One final note is that when your comments are longer than the main article I think it might show that you may be lacking in depth. Technically and in originality.
“Through it all, I still haven’t heard of a startup or web service that failed or succeeded due solely to its web framework or language”
Oh Really? Projects don’t fail because of their selected platform? Nice statement – not true at all.
Tell you what – do something constructive and tell me how I “bashed Rails”. All I did was quote those who’ve had a hard time, and point out how DHH is basically ignoring all of you. I ask the hard questions – you interpret that as you like.
Or don’t, and feel better about yourself. Nice post though. No really…
windexh8er, you’re welcome to compare the traffic of WordPress.com, which I run, with any sites you’re involved with. I think it proves enterprise-class scalability. Here are handy comparison links:
http://www.quantcast.com/wordpress.com
http://alexa.com/data/details/traffic_details?url=wordpress.com
http://siteanalytics.compete.com/wordpress.com/
Of course if you read my short entry closely, I make no claims or suggestions about Rails or PHP or ASP.NET, quite the opposite, I think you can create something fine in any. That’s the whole point. More sites fail because they don’t create a compelling experience than fail because they couldn’t handle all the users beating down their door.
I think just the fact that the article is entitled “Rails Bashing” puts a pretty concise spin to the article.
I do agree, however, that a compelling site will always do well. However the links you’ve posted still don’t tell me business grade sites. These sites are all externally facing end users reading articles, news, etc. I’m talking enterprise class in terms of business critical, internal sites. If AskMen.com goes down, nobody really cares — it’s just an irritation. If my B2B analytics portal goes down that houses information for investing millions of dollars in transactions are on the line. There’s a class difference there.
Yes, languages are irrelevant. Any good programmer should know that. Rails is a tool not a lang. PHP has lots of good tools too (see PEAR libs, WordPress, etc).
I know a plethora of langs, and PHP is still the lang I use if I need something done NOW. However, if I need something more complex or something that will be around longer, I use Rails. I still do all my utils for Rails sites in PHP however, such as cleanup tasks, internal search spiders, etc.
As an aside, at work (a decent size company) we switched from PHP to Django because the lead devs didn’t like Rails’ memory/speed issues but needed a similar tool.
So there you go. Didn’t we hear all this from the Perl devs when PHP came out too? Hrm.
I don’t want to hear any rails bashing. The next version of rails is out next week(?) and it’s gonna be sweet. Don’t hate the framework, hate the ninjas who want to destroy it.
I wonder why most commenters are attempting to read this post’s palm, or it’s tea leaves and whatnot.
This is a pretty brief and easy to understand jab at advocacy in general. Though everyone has their own bias (Matt included) it was worth pointing out that there has been a dramatic increase in language fanboism — an exercise in spectacualr time-wasting.
The point is, if you’re a carpenter and you work best by hammering nails in with a saw, nobody should really complain as long as the nails are in straight, and your masterpiece stands at the end of it.
As Rami Kayyali pointed out above, fashion really shouldn’t enter into it, unless you’re more concerned with others thinking you’re cool, than you are with achieving results.
Unfortunately, the former is far too common.
Choice of framework/language/platform etc should be made based on hard data relevant to the project’s desired outcomes. If that’s not readily available then by definition, one choice is as good as any other.
There have been a few startups where the choice of platform used had a great impact. The obvious example is probably Paul Graham’s online shop thing, which did very well out of being written in Common Lisp.
On the other hand, Reddit probably only survived because its authors were able to migrate it very quickly _away_ from Lisp; they, unfortunately, started at a bad time, when there were very few stable portable libraries available in their product domain.
And, of course, Twitter encountered serious scalability problems early on; I suspect that these can at least partly be blamed on their platform.