All true. But to be fair, what really ticks some people (bit me included) is how the systemd folks decided to piggypack all the other things, what they later have started to call "coreOS" with the original "systemd". Sure, they are all somehow related to the process of bringing the system up and maintaining daemons and almost all can be left out if decided. Given it's essentially a Redhad backed project, it will come with it all and I suspect Debian leaves most of it out.
Which all good the either way. The Linux ecosystem have survived from worse misteps if mistep this be.
I wonder if it could be circumvented by setting something like Gnome terminal or krusader as the login shell instead of desktop environment's session manager. Maybe then a new program would be opened in its own window and not inside that hideous box. Or, it won't open at all. I think Xrdp doesn't do rootles.
How have you configured Xpra, do you run it all in the server or tunnel Xpra session over the tunnel? I tried to do it all in the server end when it was all new and flimsy (because the receiving end was a windows machine with Xwin server), but didn't get it running.
And it used to do a lot more, like load executables, and be a print server.
On the other hand, it does not do much of what its original strong points were: graphic primitives, no one uses those any more, only raw bitmaps prepared by various external libraries; fonts, people use Pango, device drivers, they are now in the kernel; remoting, now done by RDP and VNC and the list goes on, hence why Wayland is now actually viable.
You mean it needs to be separately enabled and or installed?
You know, for the past 20 years, all major X-distributions have come networking disabled by default. Yeah, its there, you may be able to (I've always been) enable it, but for sure it's not ideal. Lack of any model of sessions or the huge round trip delay of modern applications makes using it over the public internet pretty miserable. Plus one needs to set up an SSH tunnel for it separately, because X doesn't do real security either....Or they could just use Xvnc, like folks have been doing about equally long.
Also, DRI2 and various other common extensions don't work over network.
Also, Knuth hasn't yet finished "The Art of Computer Programming".
(Publish the fifth issue and update 1-3.) At this phase, he'll need to break the oldest human alive record several times over...
It's just too much work. If it gets finished, it won't be by him.
Then I remembered a friend who has gotten into these vapes/e-cigs or whatever they are called. The thing is, he mixed his own juice, because ordering raw nicotine fluid is apparently a lot cheaper than premix....Now just wait when his head starts to swell.
IEC 61131 is the automation guys bread and butter, but apart from it, every machine smarter than a induction motor seems speaks it own variant of ugly QBASIC.
Learn to grok those, get good, have a proof of it and there are plenty of jobs.
I mean every guy can do simple "put an input in in these circumstances" like things in them, but dare to use the buggy and feature-depleted networking/gui/string manipulation utilities the programmers put in probably thinking no-one is going to use them anyway, you need some good wits, because the thing is literally a black box, sometimes you load the thing in with FTP or serial and the blinking lights at IO board is all you get to show that the thing is doing what you inted to. No debugger, though there will be bugs - hardware bugs! And bad, insufficient, badly translated documentation, no user community to get help from; maybe a manufacturer hotline which will tell "gee, I dunno" if you are lucky and privileged.
PowerShell is not like the shell languages. It tries, but fails subtly the moment one tries to use something not.NET based, or anything not specially wrapped into a cmdlet. Pretty much fails the original promise of shell languages, the ability to pipe together programs that don't necessarily know anything about ech others beforehand. I don't like the attitude of everyting's.NET, raw data is irrelevant.
Say, try to manipulate SQL dump with binary data in it in the normal PowerShell way, it gets corrupted due carriage returns helpfully being added in. There are workaround though, just have to be super careful to not use THE PIPELINE for it.
Anyway, that's what bite me recently. Sorry for having to take my eruption, o random stranger.
As a person who uses C-preprocessor extensively, I need my capslock key.
Jokes aside, I often program these industrial robots where basically every variable is a macro-define on bare memory address. The macro system doesn't know about tokens, but thankfully thinks case matters, so I use all caps on my variables to reduce name clashes.
Indeed, if you took mechanical/industrial engineering of some flavor, surely you saw the video of people using the most imaginative ways to circumwent the safety measures for what ever reasons they have. Insane stunts like people walking inside a large cardboard box past the safety scanner in order to fool it to think that they are product.
The robots we provide have a small key on the servo off / teach / auto switch. Very handy.
(The electrician has his own locks of course for the stuff that is not powered from the robot controller cabinet.)
Out of habit we still avoid working inside the envelope of a moving robot, how ever unlikely it is that the thing goes accidentally to auto (you know, deadman switches don save your ass then). I can think of a few scenarios, but all of them require at least two misguided steps. The easies would probably be mixing the symbols for automatic and manual operation, so the cabinet switch and also the TP switch are both in "auto". The guy goes inside the envelope to start teaching (safety gates are disabled, naturally), hits "servo on" button on the TP and off the robot goes.
There has been a lot of talk about how division by zero should be interpreted in mathematics, with real numbers; the concept of limit has been touched.
What I'd like to hear would be some proper math guy's interpretation of this issue, how pure integer maths (discrete math?) treats this? All I personally can think of would be the a = divider × quotient + remainder thing. We know that anything times zero is zero, which leaves a = remainder; quitient may be any value, so strictly zero is not incorrect in this regard.
Doing it the error checking in shitty way around, that is you expect some result other than zero and check for that.
In that regard the result of zero integer division is kind of logical: diveder gets too big: zero, diveder gets too smal: zero again, albeit this happens only on one single case: when divider is zero. It essentially says "integer division out of bounds".
Ensure that the physical hardware that you are going to use to run your development tools on will be available as surplus for the service life of your system. Have at least one backup machine any time for quick swap on hardware failure, maintain multiple copies of all your software which you refresh periodically.
I think x86 wont be a bad choise, and Iäm fairly sure you can buy DVD's in 25 years from now. x86 is about at its end of life now, so what ever you buy in the next 25 years, will most likely be from this time period.
Personally, I suspect split 100 or 110 volts vith neutral : close enough to run 48V equipment and also 120V AC equipment that doesn't care much about its voltage.
The real win of the 400V system, btw is that you can travel about 4 kms from the transformer, which gives much more freedom designing the medium voltage network. The transformers are all three phase to keep the inbalance off from the medium voltage side (much less transformers, less chance to do blancing there).
Actually, us Europeans hardly ever use bigger than 16 A breakers for single circuit. If it's a major appliance, it will have three phase supply, with 16 A three-phase breakers (some circuits may opt to 10A, due main breaker "overbooking" rules). Usual house fuse size these days is 35A, I believe apartments have 25A.
In the bad old times you might have seen stuff like 25 A main fuses for electrically heated home, or two 10 A circuits with ton of branches overbooked from a single 16 A main fuse in an apartment.
Slashdot Mirror
This is how the year of Linux desktop finally happens!
It's just planet Nibiru.
Poor english skills?
Didn't you mean to say poor Chinese skills? Those US workers, all cocky and don't even speak Chinese...
All true. But to be fair, what really ticks some people (bit me included) is how the systemd folks decided to piggypack all the other things, what they later have started to call "coreOS" with the original "systemd". Sure, they are all somehow related to the process of bringing the system up and maintaining daemons and almost all can be left out if decided. Given it's essentially a Redhad backed project, it will come with it all and I suspect Debian leaves most of it out.
Which all good the either way. The Linux ecosystem have survived from worse misteps if mistep this be.
Hum. That's true.
I wonder if it could be circumvented by setting something like Gnome terminal or krusader as the login shell instead of desktop environment's session manager. Maybe then a new program would be opened in its own window and not inside that hideous box. Or, it won't open at all. I think Xrdp doesn't do rootles.
How have you configured Xpra, do you run it all in the server or tunnel Xpra session over the tunnel? I tried to do it all in the server end when it was all new and flimsy (because the receiving end was a windows machine with Xwin server), but didn't get it running.
Wow, what a good round-up and start collecting karma. You should sign up and I'd mod this message up if I had points.
And it used to do a lot more, like load executables, and be a print server.
On the other hand, it does not do much of what its original strong points were: graphic primitives, no one uses those any more, only raw bitmaps prepared by various external libraries; fonts, people use Pango, device drivers, they are now in the kernel; remoting, now done by RDP and VNC and the list goes on, hence why Wayland is now actually viable.
You mean it needs to be separately enabled and or installed?
...Or they could just use Xvnc, like folks have been doing about equally long.
You know, for the past 20 years, all major X-distributions have come networking disabled by default. Yeah, its there, you may be able to (I've always been) enable it, but for sure it's not ideal. Lack of any model of sessions or the huge round trip delay of modern applications makes using it over the public internet pretty miserable. Plus one needs to set up an SSH tunnel for it separately, because X doesn't do real security either.
Also, DRI2 and various other common extensions don't work over network.
Also, Knuth hasn't yet finished "The Art of Computer Programming".
(Publish the fifth issue and update 1-3.) At this phase, he'll need to break the oldest human alive record several times over...
It's just too much work. If it gets finished, it won't be by him.
But still no "advanced macros, non-blocking IO or much of synopsis 9 and 11"...
Thank you, I didn't know this.
I was wondering, how anyone would inhale water.
...Now just wait when his head starts to swell.
Then I remembered a friend who has gotten into these vapes/e-cigs or whatever they are called. The thing is, he mixed his own juice, because ordering raw nicotine fluid is apparently a lot cheaper than premix.
IEC 61131 is the automation guys bread and butter, but apart from it, every machine smarter than a induction motor seems speaks it own variant of ugly QBASIC.
Learn to grok those, get good, have a proof of it and there are plenty of jobs.
I mean every guy can do simple "put an input in in these circumstances" like things in them, but dare to use the buggy and feature-depleted networking/gui/string manipulation utilities the programmers put in probably thinking no-one is going to use them anyway, you need some good wits, because the thing is literally a black box, sometimes you load the thing in with FTP or serial and the blinking lights at IO board is all you get to show that the thing is doing what you inted to. No debugger, though there will be bugs - hardware bugs! And bad, insufficient, badly translated documentation, no user community to get help from; maybe a manufacturer hotline which will tell "gee, I dunno" if you are lucky and privileged.
PowerShell is not like the shell languages. It tries, but fails subtly the moment one tries to use something not .NET based, or anything not specially wrapped into a cmdlet. Pretty much fails the original promise of shell languages, the ability to pipe together programs that don't necessarily know anything about ech others beforehand. I don't like the attitude of everyting's .NET, raw data is irrelevant.
Say, try to manipulate SQL dump with binary data in it in the normal PowerShell way, it gets corrupted due carriage returns helpfully being added in. There are workaround though, just have to be super careful to not use THE PIPELINE for it.
Anyway, that's what bite me recently. Sorry for having to take my eruption, o random stranger.
As a person who uses C-preprocessor extensively, I need my capslock key.
Jokes aside, I often program these industrial robots where basically every variable is a macro-define on bare memory address. The macro system doesn't know about tokens, but thankfully thinks case matters, so I use all caps on my variables to reduce name clashes.
Indeed, if you took mechanical/industrial engineering of some flavor, surely you saw the video of people using the most imaginative ways to circumwent the safety measures for what ever reasons they have. Insane stunts like people walking inside a large cardboard box past the safety scanner in order to fool it to think that they are product.
I'd mod you up if I had points.
The robots we provide have a small key on the servo off / teach / auto switch. Very handy.
(The electrician has his own locks of course for the stuff that is not powered from the robot controller cabinet.)
Out of habit we still avoid working inside the envelope of a moving robot, how ever unlikely it is that the thing goes accidentally to auto (you know, deadman switches don save your ass then). I can think of a few scenarios, but all of them require at least two misguided steps. The easies would probably be mixing the symbols for automatic and manual operation, so the cabinet switch and also the TP switch are both in "auto". The guy goes inside the envelope to start teaching (safety gates are disabled, naturally), hits "servo on" button on the TP and off the robot goes.
To be fair, you were much more patient sifting the search results for relevant context back in the 90s, now one seldom goes to the second page.
There has been a lot of talk about how division by zero should be interpreted in mathematics, with real numbers; the concept of limit has been touched.
What I'd like to hear would be some proper math guy's interpretation of this issue, how pure integer maths (discrete math?) treats this? All I personally can think of would be the a = divider × quotient + remainder thing. We know that anything times zero is zero, which leaves a = remainder; quitient may be any value, so strictly zero is not incorrect in this regard.
Doing it the error checking in shitty way around, that is you expect some result other than zero and check for that.
In that regard the result of zero integer division is kind of logical: diveder gets too big: zero, diveder gets too smal: zero again, albeit this happens only on one single case: when divider is zero. It essentially says "integer division out of bounds".
What about the old fachioned way:
Ensure that the physical hardware that you are going to use to run your development tools on will be available as surplus for the service life of your system. Have at least one backup machine any time for quick swap on hardware failure, maintain multiple copies of all your software which you refresh periodically.
I think x86 wont be a bad choise, and Iäm fairly sure you can buy DVD's in 25 years from now. x86 is about at its end of life now, so what ever you buy in the next 25 years, will most likely be from this time period.
Personally, I suspect split 100 or 110 volts vith neutral : close enough to run 48V equipment and also 120V AC equipment that doesn't care much about its voltage.
The real win of the 400V system, btw is that you can travel about 4 kms from the transformer, which gives much more freedom designing the medium voltage network. The transformers are all three phase to keep the inbalance off from the medium voltage side (much less transformers, less chance to do blancing there).
And what I meant to say before getting carried away was that 16 A breaker is rated for about 3.8 kW and 10 A for 2.2 kW.
Actually, us Europeans hardly ever use bigger than 16 A breakers for single circuit. If it's a major appliance, it will have three phase supply, with 16 A three-phase breakers (some circuits may opt to 10A, due main breaker "overbooking" rules). Usual house fuse size these days is 35A, I believe apartments have 25A.
In the bad old times you might have seen stuff like 25 A main fuses for electrically heated home, or two 10 A circuits with ton of branches overbooked from a single 16 A main fuse in an apartment.