Web Applications Should Be Compiled
ediblepet.net
ediblepet.net
Bloggers, please try a kinder, gentler tone, one that suggests you and I are colleagues. One that suggests you have discovered some neat thing and won't I have a look at it and isn't it interesting to play "what if" games with it and maybe, just maybe I might experiment with it.
Looking back on my own posts, I would say the strongest ones are those where I found a more reasoned voice. An essay can be strongly opinionated without being imperative.
I guess this is just a reaction to the froth+hype. Both sides need to step away from the keyboard, and calm down.
What have people got against enthusiasm, any ways? There's some kind of weird puritanical thing going around programming culture that seems to view youthful exuberance as some kind of a sin, like wearing buttons on a frock coat. Yeah, some twenty-somthing year-old has discovered something way-cool really neat amazing thing while hacking on a macbook in the café. Isn't that exactly the attitude you're supposed to have in your twenties? How would anything ever change if people didn't get super-excited beyond all rational limit?
Can you imagine Wozniak and Jobs trying to blog about microcomputers these days? They'd get flamed as hippies fiddling with toys :-)
Why stop? It's a great attitude to have forever.
The ruby and rails fervor never read like genuine enthusiasm. The pro-rails bits were always framed with negative comments about how php and java sucked, or how stupid business people are. Even DHH has a lot of this in his blog posts and talks. I thought it read more like an abuse victim or recovering addict who found a religious support group. It was always a bit manic sounding.
I'm a bit cynical but I think a critical mindset is the bane of the engineer, or really any maker. If people are zealously gushing about their framework, something dark is lurking underneath...
I wish all I had to do to be happy for the rest of my life was to never use a computer again.
Apart from all the hand-waving and generalizations, this guy doesn't sound like he's ever really a webapp under serious load.
I'm not a Ruby guy, so I don't know how things fly with that, but I've never once had a problem with a webapp being slow because of the language it was written in. Bad database queries, unnecessary assets being loaded, definitely. But CPU-bound limitations caused purely because of the cost of interpretation of a scripting language? Not so much...
Example usage: "Yes, he can write well, but has he ever really a webapp?"
You're correct that database queries and resource bottlenecks are the usual problems for web application scaling, but even taking that into account, I've had more than one problem that was caused by the simple fact that Ruby is a dog of a language. Python is better, but only just so. Perl is faster than both, but still nowhere near as fast as a C++ program.
But let's ignore execution speed, since one can make a reasonable argument that even the slowest scripting language can saturate the network I/O on modern hardware. I think I'd still consider using a compiled language, because the type-safety and compile-time guarantees of a statically typed language would go a long way toward eliminating a huge class of silly bugs that we deal with on a daily basis. That's developer time in the bank.
Not all compiled languages are statically typed.
> That's developer time in the bank.
And the time spent in the edit, compile, test, run, cycle? I'd think there's more developer time in the bank if I can edit, test, run.
The time spent compiling is a red herring. I know that I've spent dozens of hours chasing down weird runtime bugs in Ruby that a C++ compiler would have caught for me during static analysis. That time more than makes up for the 15 minutes a day I might spend waiting on a compiler (I waste more than 15 minutes a day on coffee breaks!)
My experience (and most everything I've read) says that the development speed is generally shorter for scripting languages, all things considered.
You seem to contradict that? References?
Now if he were to propose writing JUST a web application in C that has a custom HTTP server designed from sockets up to HMTL... you can streamline the sh*t out of that.
Much luck to him. Success or failure, he'll still probably be learning...
I had that basic idea when web app frameworks starting becoming popular. So I implemented a web framework which took a 'description' of the app written in perl, and generated C++ which compiled into a single exe file which served the app with a built in http server. Small bits of business logic were written in C++, but basically it was done by the perl. All templates, internationalisation, everything, compiled down to a few database queries, string concatenations and the odd if statement.
I then build a couple of production sites in this, one on my own, and another with a few other developers.
Conclusions from this:
1) It's really really fast and efficient, and the single exe model for deployment is lovely.
2) It's a pain to implement such a framework, and a pain to use it.
3) I'm never using perl again.
4) Life's too short for this kind of game.
The code is here: http://www.boxbackup.org/trac/ticket/6
You should give it another try sometimes. Take a look at Catalyst (the web framework), DBIx::Class, Moose and POE.
Everything computer-related sucks, you just have to choose something which sucks the least where it matters for your project. And familiarity counts for a lot too.
I have worked with a project for the past several years that, by necessity not choice, required a web application framework written in C (it's an embedded platform) and I can tell you it is pure hell. The most mundane and simple tasks, often trivial one-liners in scripting languages, require many lines of error-prone C code even with a fairly rich set of libraries available. Working in this environment, you'll quickly realize why the rest of the world doesn't do it this way. Thankfully, this project is finally moving away from this architecture and I couldn't be happier about it.
It's also a pretty good scripting language on its own merits, and has replaced Python as my default language. (Unlike Python, its design expects you to know at least a little C, but in exchange, it avoids cluttering the language with things that C already does well.)
It's not clear from the post which end of the spectrum he or she works with. (I don't work with embedded systems professionally, I've just tinkered with OpenWRT, cell phones, etc. as an enthusiastic amateur.) It would be usable in the same niches that SQLite is, for example.
What would be an example of this? I'm mostly a C and Perl person, and once a library is available I've never noticed cases where something was much simpler to write in one than the other. That library might take more time to write in C than Perl, but once it's written I find it just about (less than a factor of 2?) as easy to write the application logic in C as in Perl. My instinct is that if you're writing 'many lines of error-prone C' to do something simple, there's probably room for improvement somewhere.
Having said that, choose the right tool for the job.
edit: Downmod me all you like. A scripted language wouldn't handle 2k req/sec on a couple of VPSs.
People often seem to think a db is the only way to write any webapp, and that everything must depend on the db. That's bad architecture.
In the case of Mibbit, db access is only done for writes, lazily, in the background.
The bottlenecks for me, are CPU, network, and ram. And yes, you can throw servers at it, if you're frivolous and don't intend to make any money. I decided to go the other way and be profitable ;)
As far as relational DBs being the problem, that's more a matter of the free ones not cutting the mustard. Oracle and DB2 can handle jaw dropping load in the hands of an expert.
For the average web app, it's a lousy point, really (and, as pointed out by others, it's a troll/satire).
The fact that you take it seriously is a little worrying tbh. Are you feeling ok? :-)
Maybe because they don't get past the scaling issues ;)
Why would I want to run and manage and design for a server cluster just because each and every Rails process needs to load the entire language runtime, the framework and all the libraries using 100MB of RAM for no application functionality at all? Not doing all that work is premature optimisation?
Perhaps a new type of web framework is needed to address these breeds of apps.
Web front-ends scale very easily. You can just add more web heads and load balance between them. Databases are much harder to scale since you need to keep data consistency and can't just throw more boxes at the problem.
So, you don't need to be computationally efficient. However, you do need to develop fast. You need to be able to create features before your competition. You need to be able to make your system more reliable and bug-free than your competition. And that's where Ruby (and similar languages) helps you.
Sure, you and Linus have a point that selection of language can keep out the riff-raff (Linus Torvalds having a huge rant against C++ simply because of the programming community), but it also means that you're missing out on the good parts of the community. Ruby programmers tend to put a lot of thought into user experience. While you've ranted against testing, I don't think any IT manager would agree with you there. Testing substantially improves one's ability to create new features because you can easily spot anywhere you've broken something. And that leads to better quality applications too.
There's a difference between speed and scaling and if something scales fine, that's ok. Look at it this way, array indexing in a slower way to go through an array in C than incrementing a pointer. And if you had a lot of code you could go through and change all instances. However, it wouldn't be worth your time. Such an improvement wouldn't net you any real benefit while you'd be neglecting improving your product.
Your product is what matters.
You just can't do that in compiled languages, and the compiled language of choice for the web-- Java --isn't known for blazing compilation speed. You'd probably also have to jar it up and copy the damn thing over, whereas psps could be modified in-place. I think this is something that compiled folks like me simply failed to grasp: all the things about scripting languages that make them orders of magnitude faster for developing, especially for the web.
When I do Java webapp development in Eclipse, it compiles and hot-swaps my code in the background without requiring any restart or redeployments. Works pretty well.
I do have to restart occasionally because the hot-swapping has limitations, so that does slow things down a bit, but when you consider the time saved by having good compiler integration, then I don't think there's a big difference in productivity.
That said, I don't need no stinkin IDE integration for .psps. Being able to log in remotely and edit something with vi is kind of handy.
I think the reason there isn't a "C web development community" is that programmers are more than willing to trade time for performance. Safe, readable code can be written in high level languages in a snap. You can see your ideas happen very quickly. If you run into performance problems (which I think is safe to say is a fairly rare occurrence for the average web developer who produces decent code, in any language), you can a) refactor or streamline what you already have or b) start fresh, lessons learned, in something lower level.
An aside: the site http://halfbakery.com is implemented entirely as an Apache module.
The old blog software, Typo, is nowhere near as good as Wordpress. I only used it because it was written in Ruby on Rails and I am a Ruby on Rails coder. This is a very poor means of choosing software.
SSP results in web pages as compiled machine language.
The HTML file is translated into a new Smalltalk method, the tags become Smalltalk message sends, and the HTML between the dynamic tags just become accesses into an Array. Because Array access is optimized and JITed as machine code, the only message sends are due to the dynamic tags. When the webpage/method is JITed, you have a web page compiled into machine code.
But what I don't agree is using just C. C is too low level and is a horrific unfit task for web development. Instead C++ would be a good fit (and writing in C mixed in). The STL/Boost provide a wonderful abstraction layer that does pretty much anything you throw at it with ease, plus there are lot of other libraries for anything else you can ever think of. I've written several web applications in C++, and it was actually very much like writing in Python/PHP/Ruby and at time prefered writing it in C++ after writing a poc in a scripting language. They were just a little bit less readable but very clean, but the biggest plus the the multiple magnitude performance increase of the application it self with FastCGI. (all these applications were serious number crunchers and had lot of CPU and Memory Bound problems).
tldr; C for web dev painful, C++ with STL/Boost good/easy/fast/clean.
Also I've heard of wt, a C++ web framework although I've never used it http://www.webtoolkit.eu/wt
First, this isn't a troll. The guy posted a blog post on his own blog. He didn't post off-topic content to a thread and didn't try to provoke an argument in an otherwise civil conversation. The only possible way that you can characterize this as a troll is that he uses some crude language, and says some things that are unpopular on HN. But say what you will about the content, he didn't try to make you read it.
Second, the guy has a point. I've had a rant like this building in my craw for a long time, and this guy nailed nearly every point I've wanted to make. From a software engineering perspective, web development is in the dark ages. What's worse, it's a luddite process -- a totally self-made exile from the benefits of 50 years of progress: Type safety is shunned. Language efficiency is ignored. Data structures beyond the hash table are exotic and rare (and worse, I've met more than one "professional" web developer who couldn't tell me the difference between open and closed hashing!)
I understand very well that programmer time is more valuable than computer hardware, and that the speed/development time trade-off is often biased toward development time. But if you ask me, what was once an intelligent guideline is now becoming a cult. It is possible to do rapid development in a compiled language, and there are some really good reasons -- beyond execution speed -- to revisit the idea.
Sometimes the truth is very simple. It can be considered a troll because it is in the tone "if you don't do as I say you are an idiot, like everyone else".
Second: not all scripting languages are interpreted on the fly. In fact, many of the most popular frameworks load objects once and keep their compiled forms in memory. Even for languages that run under VMs, some of them run under JIT that, in the end, results in machine-language-like times.
Third: I have profiled a lot of web applications and most of the time they are either waiting for the pipe to deliver a request or waiting for the database to deliver the data. Very few web applications I know of do heavy processing. For those, I agree compiled code may help.
and, finally, by the time the guy finishes his C code, an equivalent competing application that uses Rails will be approaching the millionth registered user.
Assertion: "Some scripting languages are as fast as compiled languages."
No, not really. I challenge you to prove that statement with a non-trivial benchmark.
Assertion: "most of the time, web apps are waiting for something external to respond"
This doesn't survive a few seconds of thoughtful analysis. If a web app is spending most of it's time blocking for I/O, it's poorly designed. There's no reason that it can't be handling other requests while waiting for a DB response or a pipe. And once you've allowed for that possibility, you're back to a situation where you can handle more requests with faster software.
Empirically, this is obviously true: languages like Ruby and Python don't hold a candle to Java when it comes to web application speed. We choose them because of developer convenience, nothing more. Which brings us to...
Assertion: "C programming is terrifically slow for development. <insert snarky comment about Rails applications with a million users here>"
Ignoring the fact that there are other compiled languages than C (that are probably better choices for web development), I can tell you from experience that by the time your Rails app reaches a million registered users, 80% of your week is going to be spent either working around the myriad scalability flaws in Rails and the dog-slow nature of Ruby, or installing hardware to accommodate the load. Scaling Rails is a nightmare.
Now, it may be true that you can develop that Rails app a bit faster than you could a webapp in a compiled language, but it's a wild exaggeration to suggest that the difference is more than a few weeks of skilled developer time. It's a fallacy to imply that all compiled-code projects end up in death-march hell. Truth be told, the real reason that web apps can be written so quickly is because they're easy, and they tolerate a lot of development error. So long as there are good string processing facilities, the choice of language is a fairly minor point.
Concurrent request processing most likely belongs in the server, not in the application.
"Scaling Rails is a nightmare."
For someone who complains of "well-trodden ground" in others' arguments, you certainly seem to have a deeply-rooted belief in the myth that particular languages and/or frameworks "scale" or "don't scale".
Rails has terrible support for database replication (we've had to roll our own). Its caching architecture is laughably bad, and the implementation was/is buggy (again, we rolled our own), and there's next to no support for the kind of robust page cache expiration mechanisms that are necessary to run a large, dynamic site (Observers are only a start). Rails is a memory hog, takes too long to start, has poor translation and locale support, and simply passing a request through ActionController adds huge overhead that cannot be escaped.
And even though you don't want to hear it, Ruby is slow -- even after working around all of the above, we still had to implement a fast page caching layer just to work around the general, all-purpose slowness of Ruby.
But like I said, write me off if you want.
I find this similar to Woody's fear in the following exchange:
Cliff Clavin: Boy oh boy, talk about your lucky days, huh, I just found twenty smackers back there in the pool room. And to celebrate my good fortune, I'm going to buy a round a drinks for all my friends.
Carla Tortelli: What are you going to do with the other nineteen bucks?
Sam Malone: Woody, that could be your twenty dollar bill, the one you lost.
Woody Boyd: Hey yeah, you know I was in the pool room earlier.
Cliff Clavin: Hey, wait a second. Let's be fair about this. Give me the serial number.
Sam Malone: Oh, come on man. Nobody knows serial numbers.
Woody Boyd: L21886119B
Cliff Clavin: [hands the bill to Woody] That's amazing.
Sam Malone: How did you do that?
Woody Boyd: I memorize the serial numbers on all my currency.
Sam Malone: Why?
Woody Boyd: For just such an occasion. Tell you though, I pray every day I don't get rich.
http://www.imdb.com/character/ch0003010/quotes
Not quite a perfect analogy, but my point is worrying about scaling is a very good problem to have, and is likely to be accompanied by the resources to fix it. Twitter is the poster child for Rails scaling problems, but they re-wrote the parts they needed to write, and people kept using it because they liked the service even when they did have scaling problems.
In any case, at this point I think it's fairly safe to say that I have some direct experience scaling Rails, and it's not as easy as the hand-wavers suggest. It's certainly worth a bit of work early on to avoid the nightmares that come with trying to scale something like Rails later on.
Not to be too silly, but if someone wrote a web framework in C++ that was 10% more difficult than Rails, I'd still consider it. The distribution of suck for Rails is exponential over time, so the expected value of suck is much higher than it initially seems.
I guess I wasn't clear enough. I meant they rewrote the parts that didn't scale in another language. Scala, I believe. Is that incorrect?
So, my point then is that writing it in Rails to quickly get it in front of users and quickly iterating on features turned out to be worth it, even though they had to rewrite large portions in another language once they had a lot of users. Do you disagree? I'm not a Rails developer, but this is my understanding of how Twitter development played out, and it seems to have worked very well for them.
I'm curious, if you would accept 10% more difficult, would you also accept 20%? 50%? I'm wondering just how exponential the suck can get.
If your request-handling logic is responsible for concurrent requests, your application (and the platform it runs on) is poorly designed. This belongs to a thread/process dispatch mechanism in the platform, not inside your app. When one thread waits for the database, I am sure the processors can be busy with other requests.
"Scaling Rails is a nightmare."
No. It's not. There is little reason for it to be. While I am no Rails expert (I use Django and was a Plone guy not that long ago), the scalability issues most people have happen because of RDBMS concurrency much more than any problem with Rails itself. Like I said before, that time spent waiting for the database to respond is time spent in the database working. It is incredibly rare to have a front-end having heavier usage than the database that feeds it.
If you want your applications to scale, design it around the database limitations. Use C, C++ (I would prefer ObjC), Java or Assembly if you want, but faster request processing won't get you past database scalability issues.
Using C* or Java will only push back your release date.
Also you might want to look at the profile of the guy who said that rails scaling is a nightmare. I wouldn't use the word nightmare as I'd prefer to think my professional life story wasn't written by Kafka. But, I don't know anyone who has scaled a big Rails site and thought it was a pleasant or fun experience. Most guys who say that it's no big deal haven't really done it.
The beauty of the ORM is that you can replace the underlying layer without disrupting the code running on top or add fancy things like local memory caching or a networked memcached store. I insist that if your application was built from the start with this kind of issue in mind, it will be easy to scale it if and when the time comes. Until then, you are solving a problem you don't know you have.
This gets harder the lower-level your language gets.
(Or you can use an event model; the point is, if your server is spending all of its CPU time blocked while under load, then something is wrong.)
That evented server is also faster than the threaded version of same.
What's the compute task you think a program is bottlenecking on in an application that does nothing but shovel data out of a database into a socket based on a btree key?
I suspect that you're thinking of something like a lightweight proxy server, that waits for incoming requests, then quickly hands them off to an app server that does most of the work. That's trivial. If you just profile just the webserver process, then yes, you would expect that most of its time would be spent in wait. But while the proxy server is blocked on I/O, the app server is doing a ton of other things for any non-trivial page rendering.
You can then ask, why bother writing in Ruby at all, and i think you'll find the answer to be, it's not anywhere near as much of a pain in the ass as writing C is.
The elitism argument is bullshit. We don't need fewer people programming. We need better ways to tell if someone knows what the fuck they're doing.
In my opinion the tools we use and develop should be as accessible as possible. Anyone should be able to pick them up and get what they need to do, done. Good luck doing that with C.
> So this is why I’m writing a web framework in C.
> Grow the fuck up...
> finding Rails developers is fucking impossible...
> ediblepet.net is proudly powered by WordPress
Only recently they decided to change it to a must robust Erlang based implementation - http://www.37signals.com/svn/posts/1728-nuts-bolts-campfire-...
Arguably, all the languages that show up down near C in the Shootout would be fine languages to write webservices in, since nearly all of them are as fast as C without the eXtreme danger.
Some web services are "shovel strings from the DB out to the user", like CMSs, but for those that aren't, this isn't a terrible point.
(It's easy to forget when you're the type of person who hangs out on HN, but for most people, compiled == C, or C++ if you're lucky.)
Granted, it's not from the most unbiased source, so take it with a grain ;)
The point is - for 99% of people it's not the language, it's the I/O
Although I understand the author's position, it is obvious this is a tech person and not a well-rounded business person. The thing about programming is it's not all about tech.
A program typically serves one purpose: to make a company money. This requires more than just pristine tech. First and foremost it requires time to market so that a company can stay competitive. It also requires the ability for more than just a handful of people to be able to write it for expandability and, God forbid, staff changes.
C is fast. But slow to market. And slow to pick up a program. And slow to test. That's why there is so much WYSIWYG.
Don't get me wrong: there needs to be a lot more code behind and a lot less script out there, but in the end coding is a balance. Anytime anyone says it's all this or all that, you know the position is not sound. It's part speed, part usability, part scalability, part development speed, part readability, part expandability, part pretty pictures, etc. Balance in all things.
Good bit of satire. I wouldn't call it a troll because it's funny and just unbelievable enough, rather than inflammatory and merely misguided.
I like not having to worry about buffer overflows and memory management and such.
That said, I'll be interested to see what he comes up with.
No matter how many rps your app can serve, if there is no app you get no requests to serve...
Perhaps unsurprisingly, it shows no awareness of SQL injection, buffer overflows, or concurrency. These seem like things a web framework might need to deal with...
Plus there's the whole portability thing, which is entirely pissed away by C compilation (yes, arguably you can write generic C, but there are fundamental differences (file i/o, sockets, etc.) between different OS's, and restricting yourself is dumb)
Pre-compilation of scripting languages isn't necessarily a bad thing though.
And this guy is no C expert. He's just got a golden hammer. http://en.wikipedia.org/wiki/Golden_hammer
+1 for "academic idiots"
+1 for "improve our democracy"
+1 for "circle jerk conferences"
+2 for "scumbag just oozing hubris"
I just hope his c isn't as poetic as his prose.
I'm not sure I'd throw away the scripting languages entirely, but making it easy to integrate c or c++ into these languages in this manner is probably a good idea. It's not that hard to write very highly optimized c++ and it can have a huge impact on performance... at least that has been my experience.
While the article was inflammatory, I think the guy has a valid point. He just needs better people skills ;)
Does he have performance problems? Is it cost effective to solve those performance problems using C? What is the ROI?
Has the performance problem been analyzed to a degree that its been identified that its the web framework? (Most of the time a decent database index will help)
Go after the lower hanging fruit first and make decisions on quantitive data analysis.. not just your gut.
Disclaimer: I don't actually work on web apps, and therefore have no experience with these things.
It is done occasionally though; last I heard, large portions of Amazon.com were written in C.
But for most web apps, rendering HTML pages from script is not the bottleneck, it's the interaction with the database, and the solution is intelligent caching to reduce load on the database. The tools to mitigate these problems are already written in C (memcached, etc).
Luckily rendering scales just by adding more servers.
Rewriting templating code in C though rarely adds more speed, most dynamic languages tune their string handling as much as possible.
Also, utilizing http caching mechanisms are a (quite underused) great solution for lowering load.
It is highly unlikely that happens. Once you lower your DB issues, if the application is responsive enough you start working on other features. By the time you're done the application got new users and the DB is a problem again ... it's a cycle :)
Fortunately HTTP servers do scale by caching and load-balancing.
The author of the post probably never worked in real-world conditions. You first have to have users and real needs for scaling, and that doesn't happen unless you deliver a useful application first, and you also have to deliver it pretty fast, otherwise you might lose to a competitor with the same idea.
Not really. Sending and receiving data to/from the browser becomes the bottleneck.
If it takes 50ms to receive the query from the browser, 0.1ms to retrieve the data from the cache, 10ms to render it, and another 50ms to send it back to the browser - the bottleneck is now network communications with the browser.
There are, unfortunately, limits on how much that can be improved.
An example - from messaging middleware, not web frontends since it's somewhere I do have proper data:
I did a messaging middleware system in C a couple of years back. It's easy. I've done it before. After a while needs were changing and I decided to try writing the replacement in Ruby. It was a lot quicker to write (1/10th of the code size for roughly the same feature set).
We were handling 3-4 million events a day, which could easily result in 2-10 messages each.
The Ruby replacement spent 10% of a single 2Ghz Xeon core handling those messages. Of that 90% was spent in the kernel handling system calls that would've been roughly the same in a C/C++ version.
Even if the Ruby part of the equation was 10 times slower for the userland part, the most we'd have saved for our load was 0.9% of a single core. The most we'd save overall, regardless of scale, would be around 9-10%.
Doesn't mean there aren't things for which a language like C is better suited - I did my MSc. thesis on reducing error rates for OCR, and some of the image processing I did was prototyped in Ruby, but then translated to C for the final version because processing literally would've taken days for things that took less than an hour in C.
But scaling is very often constrained by other things than the performance of the laguage you write your app in. Especially in the web space where most interesting apps are database backed.
If your DB gets maxed out, having a hyper-optimizing compiler is not going to help you. If your app isn't designed to be scale across multiple servers (or at least cores - these days you can rent managed hosting on a 24 core server for less than $1k/month) you could very likely be screwed in ways that no compiler will save you from.
The extra developer time spent doing web development in low level languages is unlikely to be worth it unless you're at a massive scale.
If you can fit in 24 cores with a language like Ruby, your best case saving from switching to a compiled language these days is around $10k/year on hosting if you're ok with a single server, and about $20k if you want redundancy.
That doesn't pay for a lot of wasted developer time. Most web apps can fit in that, and for the lucky (or unlucky, depending on reason) that need to scale beyond that, the CPU time spend on the actual web app is likely to be a relatively small fraction of the total CPU time spent (on databases etc.) - architectural decisions is likely to be a far more important cost driver than the language of the web frontend.
As for developer availability - good luck to him finding good C programmers, and even more so finding good C programmers who aren't expensive and knows web development.
Are you saying it's difficult in general to find good C devs? Because it would seem that's the case for C++, but not for C.
However, most website projects are business-driven, which means that you are constantly changing and refactoring stuff as misunderstandings get cleared up, design flaws emerge, business models change, minds change, etc etc.
Business software projects are not just engineering, and if you treat them as such you will endure pain.
The analogy with a house is odd - if a house was purely an engineering problem, then they would probably all be built out of precast concrete or something, not "wood and steel". Aesthetics, tastes, budgets, and tradition all have an impact, and nearly every house is unique, so you use the "best tool for the job" - the one that is the easiest to customize on-site.
When requirements change weekly and by the projects end your original project description is only useful as a piece of history - then using concrete would mean lots of demolition work in the process.
This means that doing the iterations with minimum time and effort are important than worrying about scalability and performance.
You can always throw hardware at the perfortmace problem. If you get enough traffic to become worried about performace, that itself is the mark of success. But to become successful, first you must build the app and get it live somehow instead of worrying about performace issues or wasting months coding in some very low level language!
No, you can't. That's what scalability means. If you don't have scalability, then additional hardware will only take you so far before you begin wasting it.
If you build your app such that the webservers can act independently, then by using various methods of load balancing you can just throw hardware at the problem.
However, this just usually leaves the database as the single point of contention among all webservers. There are various ways to reduce load on the database server by some constant factor (memcache, query caching), but in the end database load will be roughly linear with site load (requests/time). You can't just add SQL-based DB servers and get linear speedup like (properly thought out) webservers.
If you think your site will get so big so fast you won't have time to hire engineers to figure out the scaling problems before they become unmanageable (for example, facebook got huge, but it took a few years, so they had time to raise money and adapt their stuff, although they were pretty slow for a long time), then use appengine datastore or simpledb or some other db that promises to scale really well (this is a really hard problem, which is why there aren't a lot of open source things that do this. Note: couchedb can't scale writes, so I don't count it).
My personal thinking is that one big properly designed database can get you pretty far, and database clustering can get you far enough. If you find it pleasant, putting your app in appengine would make you pay the price of scalability a little at a time starting right away, instead of letting you push it off (but with a huge performance wall that you won't be able to get over without basically redesigning much of your apps internals).
PHP (and Rails/Django, if you ignore their ORMs) encourage you to store state on some collection of backend servers and communicate via RPC to retrieve that. This is bad for performance, but good for scalability. The front-end is just a dumb intermediary that converts data into HTML, so if you need to scale, just add more of them. Scaling the backend is more challenging, but at least you can do it without dragging all the front-end code around.
But if I could do it where nobody was watching, I would use a different compiled language, Lisp (SBCL) that would get good enough performance and get to market before the manly C/C++ coder.
So on a semi-serious note, why is (apparently) nobody going for the real performance kill, which is to write a framework in hand-coded assembler? Would it be twice as fast as C?
fixed that for you ;)
Isn't that exactly who is writing to? Doesn't he WANT people to pontificate about his post and change our thinking?
So why is he mocking his intended audience?
Personally I'd go with perl - as thats the language Im most comfortable with.
You could come pretty close to simply rewriting a scripting framework in C++. I don't know what you can get done in C. It will be so dependent on programmer conventions it will be hideous.
People actually have made some damn good C++ frameworks. See asio. What are the C ones? Any complex C system devolves into a mess of macros, casts, and naming conventions.