Posts by Christian Berger

Put it into a "Nokia Communicator" case

Complete with a wide screen LCD and a standard battery, as well as a "dumb phone" in the lid, so the Raspberry Pi can be in "suspend to RAM" when the case is closed.

Are you sure it's not supposed to be that way?

I mean seriously, with recent versions of Windows it's hard to tell.

This comes from the company...

which redirected all unallocated .com domains to their own advertisement site and used to have a subsidiary in Germany called "Jamba" which defrauded millions of school children by selling them ring tones.

What do all those people do?

I mean seriously you'd expect a service like Twitter to be run by, at most, a dozen people, probably _much_ less. It seems to be far more feasible to just get rid of the idea of selling ads in tweets and just sell normal ads or even go completely donation based.

A lot of money for a short term solution

After all that system relies on an App being installed on your phone... so in 2 years you'll be faced with re-installing it on your new phone, so you'll need to get enough use out of it to make that worth. In 5-10 years you'll be faced with that company probably having moved to something else (e.g. new products or bankruptcy) so you'll slowly have troubles getting that App to run on your new phone.

For something like this to succeed, we need simple and open standards. We need standards which are so simple you could use them with a shell script.

Doesn't seem to be that bad

After all you are surrounded by rather smart people (scientists) who are really doing their job because they like it. Working conditions probably aren't to bad, as most of those stations are evacuated and closed down during the winter.

This is the home page of one of those stations:

http://www.antarcticstation.org/

And here's a German language podcast interviewing one of those scientists on her working conditions:

http://www.wrint.de/2012/01/26/wr040-holger-ruft-an-in-der-antarktis/

She even noted that during the Antarctic summer you can even sleep outside in a tent, which she was actually encouraged to do.

The problem may be in the way we develop software now

Today many people write in languages/environment which try to abstract complexity away from you. Essentially this makes you write hugely overcomplex pieces of software without even knowing.

Adding to this is that many people now grew up on badly designed systems. They never experienced the joy of having a small subset of orthogonal features which add up to something great. That's why they design systems piling feature upon feature.

One should note that it's precisely the same in most companies

It's just that in the IRS, we get to know about it.

Of course the underlying problem is the same for all organisations. In the early 2000s they bought from Microsoft believing that they would continue to develop Windows in a reasonable way. Instead they got XP (which needs more than 2 Gigabytes for the OS alone!), Vista, 7 and 8, as well as their equivalents on the "server" front. All systems which don't offer any useful functionality, but require new hardware and introduce new incompatibilities.

It's not like USB was meant to be used for such things

I mean plugging in devices didn't work even in the early staged demos.

https://www.youtube.com/watch?v=ajmn-_jkpdc

The protocols running over USB are far to complex and ill defined, the hardware makes bit errors very common, which make your bus reset.

Will it at least still work...

...when the cloud service (or your Internet connection) is down?

We are still addressing the wrong problem

Instead of trying to enforce laws made in the 1930s before you could easily tape music and films of radio or TV, or download stuff from the Internet, we should move forward and accept the existence of tape recorders and the Internet.

The question we should ask ourselves is how we make sure artists get a fair amount of money. For musicians that's actually not much of a problem as they make most money from concerts anyway, nearly no musician makes money from record sales.

For films that's harder since all the work is in producing a purely immaterial good, in Germany most movies are financed via tax money or via TV stations, both are models which can still work today.

Considering that "professional" hosters don't exactly have any different tools...

... I'd say his server probably is save. I mean if he has a fixed IP-address to ssh into it, he can even just block the rest of the Internet. And e-mail servers are simple enough to probably not have any serious security problems. The worst that could happen in a DOS.... unless of course he uses a stupid password or something.

Privacy doesn't cost money

Privacy costs effort and comfort. You cannot simply outsource privacy to some company, no matter how much money you pay them.

If you want to optimize something, go for code size...

...both in the webpages and on the browser. More HTML/CSS/JS code, particularly on more domains, kills performance. Many web pages now load _much_ slower than full sized screenshots of them!

More browser code means more bugs and makes it harder to optimize the browser.

Let's look at what are facts here.

She didn't specify what in particular pissed her off. That's a bit of a shame since it leaves things open to speculation, but it's obviously her right.

Her note can be read here:

http://sarah.thesharps.us/2015/10/05/closing-a-door/

Some Slashdot commenter has tries to find the cause of it:

http://linux.slashdot.org/comments.pl?sid=8123533&cid=50664697

Apparently there was some bit of bantering discussing that people need to learn to say "no" to Patches, and that sometimes it might be necessary to "yell at people".

She apparently interpreted this as violence, which I personally find way overblown.

According to her website, Mrs Sharps has "been running Linux since 2003, and I’ve been a Linux kernel developer since 2006". So that's 12 years of Linux experience, no information on how long she's programming was given. Now unless she is a genius, which I cannot rule out, that's not a lot of time to learn how to develop software. I've been writing software professionally since 1997 and I'm slowly starting to write something that could be considered semi-good. So it's very likely that her code just may not have been very good. Most people write absolutely shitty code in their first decades of programming.

The Linux Kernel is one of the most important software projects of this world, maybe not _the_ most important one, but certainly among the top 100. It is vital that this development happens in a rather safe environment. Any change needs to be considered carefully so nothing sub standard gets into the kernel. In a way it's like an operation room at a hospital. Though you may look into some through a window, you may not enter it unless you are medical personnel.

Now imagine how a doctor would react if you were walking into the room during an operation with the expressed intent of messing around with the patient? It would seem very likely that most doctors would certainly get you kicked out, certainly by making one of the assistants push you out. When something greater is at danger, you are allowed to be rude.

Linux faces a new problem. While in the past, not having enough programmers may have been a problem, we now have lots of people who want to write both user space and kernel code. In principle that wouldn't be bad. However just like children, developers need to have spaces where they can learn and fail. In my time this used to be Turbo Pascal and Delphi, where you wrote lots of shitty software. This is where the mountain of really bad legacy 1990s software comes from. Today that proving grounds are mostly apps, but unfortunately also the Linux user space. Just like in the 1990s it was not seen as a good idea to let those people loose at writing kernels or banking software, it's now probably not a good idea to simply let everybody mess with the mainline kernel or other important software projects. You need experienced people there. People who have learned from others.

Actually half of the company I work for...

...acts as an intermediate between companies who want their employees to have cell phones, and cell phone companies. So instead of having to talk to the cell phone companies, they talk to us.

Well so far...

I've seen a project kinda fail because we relied on a supplier which, though having the top most support level, still had worse support than the lowest unpaid (!) support level of the competing dual license product. It also had a lower code quality (we had a source license).

How that would be cool...

except for the "Bundeswehr Universität München" having done pretty much the same thing in the 1980s.

That was rather predictable

I mean seriously, the people working at that company think having a login screen that depends on external Javascript is a good idea. They ignored bug reports for _months_. They work with newbie web developers who apparently pride themselves in their experience (at least the Patreon website does).

But they already found the culprit

http://www.der-postillon.com/2015/09/der-hausmeister-wars-vw-prasentiert.html

It was Hans Böbner, the Janitor. In a recording made in the Caribic he confessed to personally modifying all those cars. Guess his recent 4.3 million bonus was not enough, so he turned on VW.

Why does it even unescape that string?

I mean the URL will be sent verbatim to the server just as it's entered and stored everywhere. There is no reason to turn %20 into a space or anything, let alone doing this multiple times. There may be reasons to do the opposite, for example on forms, but unescaping a string should never be done by the user agent.

It's a battle for the minds of the future

As Microsoft sees their market share erode with no new markets in sight, they as well as many other companies, try their best to capture the minds of the people. The earlier you start, the more likely you are to make someone believe that your particular product is the future.

I see this thing in engineering offices

Having worked on CAD workstations I know how frustrating it can be to have a team hunched over some tiny little screen showing a fraction of a design through a peep hole. With such a screen, you could finally show a significant part of the design and discuss it with others. No more scrolling, just your whole blueprint on one big screen. Everybody can just go there and look at the parts they are interested in. No more printouts just to talk about a design.

Maybe we should write letters to that school

...explaining that though being afraid of something doesn't mean it's bad. Maybe we can convince those people to make the mental leap that something they don't know and they don't understand isn't necessarily something bad. Maybe we could even go as far as to telling them that this letter went over the Atlantic via flying machines called Aeroplanes, which fly like birds. Maybe we could even attach pictures of those wonders of the world explaining that those are pictures and no real things are squeezed into those sheets.

The question is, will it work

I mean VoLTE is VoIP over LTE with _lots_ of complexity added to allow things like international roaming and handover to 3G and GSM networks. Given that today many companies (cough Cisco, cough Mitel, cough 3CX...) can't even properly implement the comparatively simple subset of SIP you need to make everyday phonecalls, I doubt that it'll work reliably any time soon. And even when it works it'll be a security nightmare, as all the miss features of classical SS7 networks have been re-implemented.

This is why we need more European television

If you had a multifeed dish also directed to 19.2°E you could have gotten HR-Fernsehen, a German public free to air channel showing cat videos at night. In fact they probably are the inventor of cat-based video entertainment.

Logical consequence of technical progress

I mean optics and routers have a limited lifespan. After 10-20 years maintenance costs will rise, which means you will want to replace them. Also if you want to get rid of your equipment earlier, there's a decent refurbishing market so you can recover a large fraction of your costs.

Plus equipment costs pale in comparison to the wages of the people maintaining them. A network engineer will easily cost you 100k$ a year if you count all costs directly associated with hiring a person.

The beauty of embedded projects is...

... that when they are well designed they become small, and therefore easy to implement. Plus if you stay below certain thresholds, low memory requirements actually save you money.

Therefore you can get operating systems like FreeRTOS/OpenRTOS which can run with very few kilobytes of RAM.