I don't really have much experience with robocopy, but explained here: http://blogs.technet.com/b/filecab/archive/2008/07/31/robocopy-mir-switch-mirroring-file-permissions.aspx you might have to set /MIR and /SEC parameters and /SECFIX if there are already files in the target that haven't changed.

All right, I figured out why it doesn't work - and I should have been slightly more observant and noticed that before - the permissions are inherited. Damnit. Oh, well, at least that means I don't have to bother.

Do you have the source code still by any chance?

Curiously yes, although this is all written when I was a young teenager back in 95 or something of that time. It's always appalling to look back at your stuff and think how bad you were at one time.

Keep in mind when looking at the sources I used a custom assembler I hand wrote over 3 years, although very little of it would need to be changed in order to work with a proper assembler.

(Hmm if you want the source I can supply it, it doesn't seem to hold formatting here well)

Rather nice thread here. Might as well introduce myself.

I started fiddling with computers back when my family got a C128 (only ever really used the C64 mode). Taught myself BASIC through reading various programs and a few books with BASIC programs you could type in to get simple games. Must've been around 5 at that time, know I knew Basic by third grade, as I recall having coded a program to help me study English glossary...complete with me making spelling mistakes in the DATA structure used to store the questions it'd ask me.

Back then I dreamed of growing up to code amazing games, not for telling stories or time wasting. I loved games for being quirky and often asking you to learn something new to get anywhere, be it a completely new concept, or just new rules for a formula you know. 'course, seeing a magical world (in 2D) sure helped a lot.

While my mother was studying while I was in about grade 8-9 I borrowed her Turbo Pascal book, ended up doing her homework in that subject because I thought it was fun.

Got a formal education (short thingie) mainly focused on Java programming against SQL databases, rudimentary understanding of a computer architecture and networking.

Ended up in an IT admin job managing servers, coding applications, scripts and homepages, doing support, troubleshooting and fixing basic hardware issues and managing networking. Through mergers and outsourcing it's now primarily end-user support and primarily C# and HTML/Javascript/CSS coding when there's no support or maintenance to be done.

It's probably for the best I don't do much coding anymore, coding wise my mindset is still stuck in the 80s and I'll often find myself making weird code to save a couple of bytes of memory (memory doesn't grow on trees, if we get enough data in here we might be using *gasp* 5 kb more than needed!!!)

Kinda dropped the idea of game programming. I clearly lack the skills, and game development is a long way away from the whimsical anything goes, knock it together in a weekend to see if it'll run, mentality of the 80s.
Still have a few notions I'd love to see made (Someone who does game jams, why not make a prototype of a turnbased team based FPS game. Team takes turn at the same time having a timer. Moving uses time faster than standing still, shooting uses large chunks of time (depending on your weapon), and you can "overwatch" to save time for interrupts.)

So, yeah, my IT prowess is Jack of all Trades, Master of none. Kinda like Wikipedia, only with even less citations.

That kind of mindset is still relevant even nowadays. Someone has to program an entire OS and cryptographic library into the chip credit card you are using (assuming you are using one). It might have as little as 1KB of RAM and 16KB of ROM to work with.

Any fans of functional programming here? We did (extracurricular) classes in Scheme in college in Poland and I loved it. It was just a group of students meeting after school. Then for my BSc final project I created a physics engine in F# (which wasn't as much fun, probably because unlike the Scheme stuff which was just meetings after classes, this one I had to do with a time limit and documentation).

I could never see it as viable for commercial applications however, so I'm wondering if anyone here actually used a functional language commercially.

Oh I'm well aware of that. But lacking the skills otherwise means the mindset usually affects homepage backends, automation scripts and data maintenance services, items which rarely benefit from saving a few kb, and which it would be better to spend the resources to make sure the code is readable and with less risk of memory being in invalid states.

Never learned ASM level programming, which is about where I'd expect you to need to be if you're coding things to fit on a 16kb rom with 1kb ram

Not necessarily. One can do it in C, and one really should, otherwise the code is pretty useless when the next generation has to run on a different CPU from a different manufacturer.

That made me laugh xD. Nice one.

Then you should become a web developer. Many people use their smartphones to browse the web and most of them have a slow AND capped connection. Your site is one of the few that works well for them? Then your site is good and all the others are crap ;) And loading speed is an important factor for SEO (search engine optimization) too. Light weight websites will load pretty fast and Google loves fast websites.

Except that, every now and then, you may need to do something that you can't do in C. For example, on an x86 CPU, accessing I/O ports requires special instructions, and the code to do so has to be written in assembly. (Of course, using C wherever possible and ASM only when necessary is still good programming practice in this context.)

Note that this is not necessarily the case on other CPUs; sometimes, such ports are mapped to memory addresses, allowing you to simply convert an integer to a pointer and access the port that way. One nice thing about C is that you can actually do this sort of thing if you need to.

Incidentally, this can also be done in C++ (there's even a feature called "placement new" that lets you construct an object at an arbitrary memory address) and Rust (if you use the "unsafe" keyword and mem::transmute).

I know you mentioned earlier working with SEO stuff, if you've got the time and inclination I would love a kind of basic guide to SEO as I've never quite been able to grasp how to apply SEO when building websites and most of the websites that come up when you search seem to be trying to sell you things rather than teach you...

As a few people expressed interest I'll put together a little guide of how I got my script up and running for GoG (which may in turn help anyone wanting to make userscripts for other sites) at some point.

First off: When I say search enigne optimization, I mean Google optimization. I build my websites for the German market and Google search has more than 90% market share in Germany.

The main reason why no one tells you anything about SEO is, that there's no Big Secret. There's no "Do this and you'll get that." stuff. SEO is a collection of a lot of stuff you can do and a lot of stuff that you can't do (you can try to fake it, but... say bye to your ranking if you get caught). So... Knowing the basics is exactly all you need to know about SEO (if you don't want to compete in markets like private health care, credit cards, or other billion dollar businesses). Trends and tricks come and go, but the basics are the basics and will always stay. And they're something that you can learn directly from Google. They have a very useful Search Engine Optimization Starter Guide. Everything else is about the content and users of your website. And that's nothing that a SEO agency could change.

One of the most important things about SEO - and that is really hard to fake - is to have users who like your site. How does Google know if people like your site? Well... People do share everything on Facebook. Your site gets mentioned there a lot? People like your site! The same goes for forums, blogs and comments. And Google observes how users react to your site when accessed through their SERPS (search engine results pages - the stuff you see after doing a search ;P). If someone clicks your site in the SERPS and returns to Google to click the next result after only a few seconds, your site wasn't relevant to him. If this happens a lot, you'll lose some positions with this keyword. But if someone stays on your site very long (compared to the other sites listed together with you) and even navigates to other pages of your website, then Google thinks that the user likes your site and finds it relevant to his search query. This is one of the most important (if not THE most important) factors, because you'll have a really hard time to fake a few hundred "happy users" a day. Especially since Google knows your IPs and cookies...

Now you'll only have to learn how to write (define and use keywords - see Google's guide from above) and how to arrange your content to keep people on your website for as long as possible. You'll have to learn how to balance search engine optimization (get users to stay on your site and to navigate a couple of other pages on your site) and user experience optimization (don't spread the information over too many pages just to satisfy Google!). A good way for me is to give a short overview on the entry page and more detailed information on a couple of subpages. To give a short example: You have a website with an overview about... gaming events (fairs, e-sport tournaments, etc.). When you enter the page, you list all the event names, the dates and the cities where they are. If you click the event, you get a page with all the information about the event. The exact time, the exact adress, where to get tickets, etc. On this page you can link to some general information about the city (you build this page once and can reuse it for every event in this city) that could be relevant for your users. Hotels (maybe you want to sleep somewhere), possibilities for park and ride (some people don't want to drive around in foreign cities), restaurants and other stuff that could be of interest. That's a lot of information (user stays on your site while reading) arranged in a logical structure (not torn apart for SEO reasons) over three pages (animating users to browse your site). It "forces" your users to do what Google sees as a signal that they like your site and it gives your users a lot of information.

Finding the balance there is the most difficult part about SEO. If your site's good, people will start to share it in Facebook (and elsewhere) and do your off page optimization (something that shouldn't be your business anyway, if you ask Google) for you. The rest is just technical stuff like clean and semantic use of HTML (HTML5 is awesome for that!), mobile-friendly layouts (no 12px font, enough room between links, light weight code), good server speed and so on.

If you stay clean (no SEO "tricks"), you'll not have to worry about algorithm updates. Google isn't stupid anymore. They'll notice if you change things on your website whenever webmaster forums fear the end of the world (which happens with every single algorithm update). Just keep your site clean with good and semantic code, create good content, think about good keywords (use them in your text) and try to make users stay on your site. Avoid webmasters who want to trade backlinks, don't spam around and be careful with any SEO tricks. Content is king.