15 years ago wired networks were the same.

True, but the summary was talking about Google and Microsoft.

And things always get larger so even if 4GB isn't a problem now it might be pretty soon. For example our (front-end) web servers has 512MB in 2003 but now the smallest we have is 2GB and I reckon 512MB wouldn't be enough any more (although I couldn't say off the top of my head why).

Although 2GB < 4GB, who knows what will be standard in 10 years.

Yes !! I don't get why this isn't mentioned more ... 4GB will not be enough for desktop computers pretty soon, and certainly isn't enough for servers.

I suppose so, but if the process is "Firefox" then that's still a problem.

And ways around it, I mean I reckon that could be OK if you have maybe RAM being 2x or 4x the address space (although for sure not fun for developers; but OK fun for developers is irrelevant for the users..) ... but I mean imagine a 16-bit 64KB address-space with a few MB of RAM, you'd spend all your time switching pages in and out of address space .. I mean the same would be true if you have a 32-bit 4GB address-space with e.g. 128GB of RAM..

Yeah I dunno, maybe memory availability will not expand as fast as I imagine it will, or maybe ARM will address the issue (as x86 and others have), but if ARM doesn't address the issue, it's going to be a problem one day, the only question is when.

Is ARM 64-bit? By which I mean, can an OS create a process of larger than 4GB in size? (Or can it even use more than 4GB of memory in total?)

I was on public transport the other day, there was a leaflet advertisement from a consumer electronics high-street shop, selling TVs, DVD players, PCs etc, and there was a €500 PC (US$ 700) with 8GB of memory. So assuming in 2 years, every (Intel) PC you buy off the high street has say 8-16 GB of memory, ARM computers (90% of PCs according to the guy) are going to look pretty stupid with a 4GB limit.

But maybe I'm wrong, maybe it's totally 64-bit? But Wikipedia doesn't seem to think so, and I can't find much evidence one way or the other.

And it's not the case that "nobody needs more than 4GB" either, a) every PC will just have that much memory, and b) Because of that, programs will start to use that much memory. i.e. there is often a speed vs memory trade-off, and if everyone has GBs of memory, there's no point making your program run slower to fit in 100MB of memory.

And further, "nobody needs more than 4GB", well I mean in a way nobody needs more than 100MB I would say (10 years ago my desktop machine had 128MB, could do word-processing, internet browsing, etc.) But have fun running modern software on a 128MB computer! The same will happen to 4GB computers.

Anyone know?

Yes I totally agree, that's one reason why I'll never buy a Mac. (And I'm sure Apple aren't going to change any time soon, as all their users are presumably used to their positioning).

CTRL should clearly be on the left, that's where you expect it to be from normal keyboards. If there's one key that's flexible where you put it, it's the FN key, that doesn't have an equivalent on normal keyboards so there is no expectation.

But surely if you readvertize and leave the sync off, then data inconsistencies will start to occur? (e.g. a modification to the one database, and a different modification e.g. to the same row in another database). How are these inconsistencies then reconciled?

"The issue is when the systems designed to create redundancy actually cause the failure" - exactly.

For example we had two Oracle systems (hot-cold) and one disk array connected to both systems. The second Oracle was triggered to start automatically when the first Oracle died. One time the second Oracle thought the first Oralce had died and started, even though the first Oralce hadn't died. (We never knew why it started.) Then we had two live instances writing to the same set of data files, and not knowing anything about each other - not good.

I'm not saying redundancy is bad, but it has consequences, and one of those consequences is complexity which can introduce its own downtimes.

Were you running relational databases? What did you do about schema changes?

(i.e. presumably if you were running relational DBs then there would be one big data set which would be shared between all three sites; you couldn't e.g. deadvertize one site, change the schema, then readvertise, as then the schemas would be different...)

But surely easier than using TeX, however.

http://googledocs.blogspot.com/2009/03/just-to-clarify.html

"They've been doing MVCC now for almost 10 years"

Oracle introduced MVCC in Oracle 4, released 1984 (25 years ago)

What's your proposed alternative?

Most people in Germany and the surrounding area use Xing.

Not necessarily.

You assume that payment is needed to view the site (in which case URLs being sent to random people who hadn't paid wouldn't work.)

However there could be other alternatives, e.g. pay to be able to upload something, then it's available for everyone, i.e. combining payment and being able to view stuff.

(I mean for sure that has other consequences as well, but I just wanted to point out that payment doesn't have to imply non-emailable URLs)