@Archemar You can rest assured that there are users of the ed editor that do not want to switch from the line editing paradigm to screen editing.
They may not be many, but they are out there... And I've personally used ed to edit files in situations where other editors were not available (albeit not on a day to day basis).
Very recently, past few minutes or so?, a new build was pushed out that changed the location of a question's stats from the top of the sidebar content to underneath the title.
I can't exactly put my finger on why this change feels odd. Perhaps it is because statistical information like that be...
speaking of dog food, the brand I get for my dog recently came out with a squid flavored dog food. My dog seemed to like it but my wife absolutely hates how it makes my kitchen smell lol
@Jesse_b reminds me of an old comedy clip, I want to say it was Paula Poundstone, who ranted about cat food -- "flavors that cats naturally crave, like bison ... when's the last time you saw a cat tearing across the plain and taking down a buffalo?"
Is the mypty3 code valid for modern terminals (e.g. gnome terminal)? Author used cfmakeraw() but after running that program when I use arrow key it shows ^[[A.
@Biswapriyo It depends on if they are more about the language itself or more about the code's use/compatibility with unix and linux. The exception seems to be bash/shell which any questions about which are more at home here than SO
@Biswapriyo perhaps I misunderstood. I thought you were asking about some behavior of it in (for example) gnome-terminal. If you're at the system call level, you're heading for SO. If you're at the behavior/functional level, I think you'd be OK here.
It seems the downside of not using swap is that machine freezes up when it runs out of memory. One would think that millions of man hours of computer research could do a better job of memory management. And I'm just running a browser and an editor here. No processes suddenly grabbing huge gobs of memory.
And it looks like the OOM process fired up (based on the logs), but it doesn't help matters.
Of course, it could be that my computer is just old and doesn't work that well.
I do open a lot of tabs. But I think an OOM ought to be able to figure out there is a problem, and kill the browser before things get out of control. Like the computer freezing. But well, maybe my computer is old. I dunno.
@Jesse_b I did have this idea that programming was just a way to organise thought. At least that's a large part of why I do programming. Very little end product, but a heck of a lot of thinking, tidying up code, refactoring and more thinking. That's usually how I learn stuff.
People who subscribe to the idea of personality types would identify that as a introverted sensing type.
I went all last summer without an A/C, it actually wasn't that bad (not that it gets anywhere near as hot here as it does in India). Your body adjusts to the temperature within a week or two and it mostly doesn't bother you
Don't forget humans have evolved to live on this planet without heat or A/C
Anyone know how to configure the protocol that showmount uses? It's a NFS utility. I was trying to trim unnecessary open ports, and when I cut out port 111 for UDP ( the rpcbind/sunrpc/portmapper port), showmount stopped working.
@Ungeheuer Portmapper can't use TCP. UPD port 111 is necessary for allowing clients to connect to the NFS server, no matter whether the NFS server uses TCP or UDP.
Linux users have a weird notion of "permanence". They ask "how can I change X permanently?" knowing full well that they will have reinstalled the system at least twice in the next year.
I say that, and then I read a "How can I NFS mount /mnt permanently" on an NetBSD mailing list...
speaking of NFS, I figured out why I was getting 50 and 60 second transfer times on files
and why I had triple digit retransmission counts on the clients
After weeks of trying shit, I tcpdump-ed the interface that the server and client were communicating over and watched. Turns out the servers for both RHEL6 and RHEL7 in our lab have been lying to the clients that they can handle 1 MiB reads well.
The servers get hit with 1 MiB read requests, choke, and decide to pass on 65,000 bytes as a reply.
Turns out the the limit before things go to shit is around 500,000 bytes for read requests.
So now the mount options are set to 32KiB, and I've cut down transfer times to about 25-30 seconds for a 200MB file. Not stellar, but good enough for government work.
I also learned that when the manpage says that read request sizes have to be multiples of 1024, they actually have to be multiples of 1024 times a multiple of 32. It's weird. I tried just putting in 60,000 as the request size, and /proc/mounts indicated the read size was 32KiB. Well, if I set the read request size to 65,792 (1024*32*2), that's a valid size that doesn't get changed.