I have one minor point to do with the design of the "compressed" stories. The left hand curve makes it look like a footer of the full-text story above it. Of course, if you get two or more "compressed" stories, it rapidly becomes obvious, but with only one, they start to feel like they're part of the story above, instead of separate entities in their own right.
The "snobbery" is because most of the new builds look bloody awful. Poor architecture, crammed together, badly designed in just about every way that you can think of. Yes, they may well comply with the building regulations, but that doesn't make them aesthetically pleasing.
It doesn't help that the UK has such restrictive planning departments. Anything that looks even vaguely different is frowned upon and unlikely to be approved. Look at some of the house designs in use in Europe. There's real innovation there, unlike the UK.
The only problem with using Python, Ruby etc is that as they are not strongly typed, you will not find out some problems at compile time and these problems will only show up at runtime, unlike in Java.
That's why you have unit tests. And once you start writing them, you'll find that they catch a lot more bugs than static typing ever did.
I've been using Perl/Python for over 10 years and I can count the number of times static typing would have caught bugs of mine on one hand. I find it thoroughly overrated.
yet, python tends to have its own set of problems, mainly stemming from the fact that NO type checking is done at startup time and so you have to start your program again and again, everytime only finding/fixing one small typo!
As opposed to starting your compiler again and again, everytime only finding/fixing one small typo!
To say this shows a lack of real world database experience. Whilst it's simple to change the connection to point at a different database using DBI, once you're actually connected the SQL that you have to use varies in many subtle and incompatible ways.
Probably the first one that everybody comes across is the difference in the integer primary key. In MySQL, it's auto_increment, in PostgreSQL it's a serial datatype with a backing sequence. If you want to know the primary key value after creating a new row, it's accessed in different ways. And this is just the tip of the iceberg.
Thankfully, because they're all based on a common standard language (SQL), it's possible. It's just still a lot of very hard work. But it's not impossible.
If it was easy, you'd see many, many more open source projects supporting something other than MySQL (which bugs me as PostgreSQL user:-)
If you do want to go the XHTML route, take a look at mod_xhtml_neg, which serves things up correctly using content-negotiation. It works rather well for me. The added strictness is sometimes annoying, but it does keep you honest.
The point about the metabase being xml is very, very important. How many people keep their apache config files in version control? Lots (the sensible people). How many people keep their IIS configs in version control? I don't know, but I'm betting it's a tiny, tiny percentage of the user population.
Version control is essential for systems administration. You need a good, working "undo" button. That's what version control gives you. But VC works best with text files, not the registry. So switching to XML config files will give IIS admins a chance to bring their practises closer to those used by Apache admins (and the rest of the Unix sysadmin world).
That's just the point. I really like mod_perl; it's just that some of the architectural decisions about it aren't quite right. They're great for performance, but not deployment.
The trouble is that you can work around all of this, it's just not simple. PHP makes it simpler, despite all its other warts.
I'm a Perl programmer. I love Perl. But there are a couple of things that PHP really, really does better than Perl (well, mod_perl in particular).
Deployment. Getting stuff into mod_perl isn't as simple as it could be.
Shared Nothing. mod_perl keeps stuff around between requests. This can be useful, but also limiting. In fact, it's been the cause of some of the weirder bugs I've seen. And it stops you deploying two copies of the same mod_perl application into the same server easily.
Disclaimer: I'm not a PHP programmer, the above is just what I think happens.:)
Of course, none of this matters for a big site like slashdot, or the ones I develop. But for small people who are just "playing around", it can be a big inhibitor. And that's one reason why I think mod_perl has lost mindset to PHP in the past.
But in emacs, it's as simple as C-x v v. Really, it's almost as quick as saving the damned thing. Then your problem switches to "how do I get people to write coherent log messages".
Uploading the code shouldn't be a developer problem, even though it usually is. There aren't enough people who can distinguish between deployment and development.
Slashdot Mirror
In fairness, setAttribute() is part of the DOM standard, whilst innerText isn't, so I wouldn't be criticising Firefox here.
Don't fall into the classic trap of "Let's rewrite it and make it better". One rarely does, particularly with an installed base of several billion.
-Dom
Even better, how do I change my retina and fingerprints when it is stolen?
-Dom
Actually, it's not about time. This article has nothing do with clock synchronisation, which left me very disappointed.
Good idea, though. Thanks!
-Dom
It doesn't help that the UK has such restrictive planning departments. Anything that looks even vaguely different is frowned upon and unlikely to be approved. Look at some of the house designs in use in Europe. There's real innovation there, unlike the UK.
-Dom
Now if you'd asked me about testing with databases, that's a bit harder. :-)
Getting back on topic, the rails support for testing is a real breakthrough. Well worth taking a look at how it handles things.
-Dom
-Dom
That's why you have unit tests. And once you start writing them, you'll find that they catch a lot more bugs than static typing ever did.
I've been using Perl/Python for over 10 years and I can count the number of times static typing would have caught bugs of mine on one hand. I find it thoroughly overrated.
-Dom
-Dom
As opposed to starting your compiler again and again, everytime only finding/fixing one small typo!
-Dom
Probably the first one that everybody comes across is the difference in the integer primary key. In MySQL, it's auto_increment, in PostgreSQL it's a serial datatype with a backing sequence. If you want to know the primary key value after creating a new row, it's accessed in different ways. And this is just the tip of the iceberg.
Thankfully, because they're all based on a common standard language (SQL), it's possible. It's just still a lot of very hard work. But it's not impossible.
If it was easy, you'd see many, many more open source projects supporting something other than MySQL (which bugs me as PostgreSQL user :-)
-Dom
-Dom
That would be *perfectly well-formed* as opposed to perfectly valid.
I do that on some of my sites and it's a real pain seeing that mozilla "xml parsing error" when I get it wrong...
-Dom
I usually find it's simple right up to the point where I have to bend it in some new and unknown fashion...
-Dom
Version control is essential for systems administration. You need a good, working "undo" button. That's what version control gives you. But VC works best with text files, not the registry. So switching to XML config files will give IIS admins a chance to bring their practises closer to those used by Apache admins (and the rest of the Unix sysadmin world).
-Dom
Or just press Ctrl-Space.
-Dom
The trouble is that you can work around all of this, it's just not simple. PHP makes it simpler, despite all its other warts.
-Dom
Disclaimer: I'm not a PHP programmer, the above is just what I think happens. :)
Of course, none of this matters for a big site like slashdot, or the ones I develop. But for small people who are just "playing around", it can be a big inhibitor. And that's one reason why I think mod_perl has lost mindset to PHP in the past.
-Dom
-Dom
Would you Adam & Eve it?!?
I suppose I should give it another go at some point. It'd be a better match than CVS in Eclipse for sure.
-Dom
-Dom
Uploading the code shouldn't be a developer problem, even though it usually is. There aren't enough people who can distinguish between deployment and development.
-Dom