« first day (1245 days earlier)      last day (3700 days later) » 
02:00 - 16:0016:00 - 00:00

2:44 AM
@derobert that's true. I just have a couple typos, small errors, etc to fix
 
2:58 AM
Fun:
5
Q: Bash Script to repeat every word in a line?

CristianI have a string like: dog cat bird wale And I want to get dog dog cat cat bird bird wale wale All the words are in the same line. Any idea?

 
3:44 AM
meh, I know the answer to this askubuntu.com/q/430144/169736, I'm asking for it to be migrated
 
4:25 AM
so much spam in the suggested edit queue....
 
mm? just one...
 
5:25 AM
any kernel programmers in the house?
 
5:44 AM
hah!
@gideon no. just ask your question; if you just ask to ask it's useless to anyone who reads the backlog later. 13.4k, you should know this
also, you know how GNU/Linux people always say that the only part that can't be updated rebootless is the kernel? well, no.
 
6:04 AM
@strugee ah yes, I was about to ask but got sidetracked.
@strugee wow :)
I was looking at the source code for drivers in linux
I'm especially interested in NIC drivers, while looking at the Realtek8069 driver most of it didn't make any sense.
and I mean, what the code was generally doing. I read about Linux modules and PCI driver api and stuff, I also looked at Linux device drivers 3 and it talks all about the linux api.
But how does one generally know what to say to an ethernet card? How does one know that to start transmitting one needs to do the following:
The network cards datasheet talks generally about the hardware, but now what needs to be done generally with an ethernet card where do I find such details?
 
6:38 AM
Morning
 
@JennyD morning
 
 
2 hours later…
8:49 AM
@derobert Odd. Chromium version? So, you slide back and forth on the lower bar and the upper back shifts around?
 
9:03 AM
That should be upper bar, sorry.
 
 
2 hours later…
10:45 AM
@FaheemMitha, about that "going back and reading documentation and realising it is crap" - that is a good thing. If your old stuff was still perfect it would mean you had learned nothing in the meantime.
 
 
2 hours later…
12:22 PM
@JennyD I suppose that is a positive way of looking at it. :-) BTW, what happened to your getting to know you answer? Don't mean to be a pest...
 
I spent most of this week puking my guts out so didn't really have the energy to write it... Hopefully this evening or tomorrow
 
Actually, after I have written something, I'm often surprised when I go back and read it later how much it can often be improved. One gets a particularly large dose of this when writing research papers, where, if it isn't in good shape, the reviewers tear you off a strip. Or simply refuse to read it.
@JennyD Oh, sorry you were sick. Flu? Some unknown virus?
I think the looking back and discovering crap thing may simply reflect the fact that you are looking at it with fresh eyes, and after you have forgotten about the thing a bit.
When one is initially writing stuff, and it is fresh in ones mind, it is very easy to think what you are writing is clear, when often it isn't, and one is simply not seeing what one has actually written with the eyes of someone who does not know what it is about.
 
@FaheemMitha Some virus I guess
@FaheemMitha Also this:
 
@JennyD Ok, I hope you feel better soon. If you are in Sweden, hopefully you have a good health care system to take care of you.
 
@FaheemMitha It's not as good as it used to be... but yeah, I'm better now. It was just a matter of letting it take its course, and while I do lose money from being off sick, it's not too bad when it's this short a time.
 
12:30 PM
@JennyD They don't give sick leave in Sweden? Shocking.
 
@FaheemMitha No money the first day, after that it's 80 % but it's capped at a level way below my own pay. I have extra insurance to cover the rest.
 
@JennyD Ok, sounds like it is better than nothing. It still sounds to me like Sweden is way better than most places on the planet to work in. I mean, you should check out India. Horrific. And it gets a lot worse than India.
Well, Ok, worse. I don't know about a lot worse.
 
@FaheemMitha Oh, it definitely is - but the current gov is working hard on increasing inequality and selling out all the stuff we own together. While still good on a global scale, we're falling behind our closest neighbours and we're getting people asking other people for charity because they don't get enough money to pay for food. That was unheard of ten years ago.
 
@JennyD Oh, I'm very sorry to hear that. I guess nowhere is safe these days.
 
I've been driving out to poor families delivering food when the one adult in the family has been injured or sick and unable to work. THat bloody shouldn't be necessary in this country and it drives me furious
Pardon the language.
I really hope they are voted out this year
 
12:35 PM
Can't you get rid of the govt that you have? Does it have much support, or is it one of those crazy coalition deals?
Incidentally, are you Swedish, or just work there? You don't sound like a foreign speaker.
 
@FaheemMitha It's four of the five right-most parties (the one not included is the overtly racist one). Current polling says they'll be voted out but the election isn't until September.
 
@JennyD Eww, nasty.
 
I'm Swedish, but I read a lot in English and have a lot of English and American friends so I get a lot of practice
 
@JennyD Ah, Ok. Yes, most people don't guess I'm Indian either.
BTW, if anyone has thoughts about this question...
9
Q: Why many talented scientists write horrible software?

ThanatosI am a software engineer and I have been working for years with people hired from academic background. Many times I've noticed that (even otherwise brilliant scientists) produce code of extremely low quality (unless their background was precisely Computer Science). Since those people are very go...

 
@FaheemMitha But a lot of Indian people speak and write English very well - only many of them do so with their own dialect which westerners think is "wrong".
 
12:38 PM
This got closed, then reopened.
 
@FaheemMitha I think it should stay closed.
 
@JennyD I don't think I agree with that, actually. Maybe a few, highly educated ones. Bengalis, for example, have a highly literate culture, and they tend to speak and write English quite well.
@JennyD Why? To be clear, it was just reopened. Please don't cast a close vote.
 
@FaheemMitha I won't, I'm not on that site.
@FaheemMitha I don't meet the ones that don't, of course, which skews my sample
 
Presumably they speak and write Bengali well too, but I wouldn't know about that.
@JennyD Well, no reason you shouldn't.
 
@FaheemMitha It's discussion, and it pretty much answers itself - the academics know their topic but they don't know how to do stuff with methods and languages they haven't studies. May as well ask why they don't build good houses.
 
12:41 PM
People don't necessarily get jobs based on their command of English. I've met a fair number of Indian researchers with a fairly dodgy command of English.
@JennyD Mmm. I guess I just like discussion. I think there is useful stuff to say about it, though.
 
@FaheemMitha I mean, the Indian people I meet is usually via work online or through forums where the going language is English. So obviously I will not meet people who don't know enough English to get there in the first place.
 
@JennyD Right, they know enough English to get by. But that does not necessarily mean a native level command of English.
 
@FaheemMitha There are, and discussion is interesting - but is this the place for it? It's like all the "best practices"; "how would you go about designing X", and so on. They are interesting but they don't work well in a QA forum.
@FaheemMitha True. But not necessarily worse than that of e.g. French people. (or american, hurr hurr hurr :-)
 
@JennyD Yes, well, I suppose I'm really not all that wedded to the concept of a strict QA forum. Like others on SE seem to be.
 
Why can you post a bounty you can't pay for?? unix.stackexchange.com/questions/118170/…
 
12:44 PM
And I have to admit I don't have a native level command - there are entire areas where my vocabulary is very limited. Stuff in the home, like cooking implements and so on...
 
@JennyD True. Native English speakers don't necessarily speak English well.
 
@FaheemMitha Well, it's what SE is designed for... but I too feel limited by it sometimes.
 
I remember in my neighborhood in Durham, someone posted a paragraph long sign warning about burglaries or something. There were 5 different elementary English errors in that note.
@JennyD Let us think outside the box. :-)
I think my favorite was "brake" when it should have been "break".
 
@FaheemMitha The worst thing about seeing bad English all over the place is one gets used to it and then it's hard to remember what is right because you always see it wrong...
 
slm
@FaheemMitha I voted to close it
 
12:47 PM
@slm Tch. I guess I made a mistake posting it here. :-)
@goldilocks I'm surprised that is possible.
 
slm
Seriously how is that even answerable?
I'm not disagreeing w/ the premise of the Q but it's not going to go anywhere productive
 
@slm There are two answers. :-)
 
slm
yes and 34 comments
 
@FaheemMitha That person sounds like an absolute twit. The "unless their background was precisely Computer Science" kind of renders the whole question absurd. Then posting it on the "Academic beta" site as your only question...
 
@JennyD There are some particularly horrible examples out there. My bete noire is the "revert to you" so popular in India. Makes my teeth grind together.
@slm Yes, but good, fun comments.
@goldilocks Well, I can see you guys are not inclined to be sympathetic. :-(
 
12:50 PM
@FaheemMitha It used to be popular in the UK some 15-20 years ago. Apparently it's not wrong as such, just less common. But I hate it too.
 
slm
yes fun, and i have a feeling they're seeing where it goes before closing b/c if that don't allow that Q then what's the purpose of that SE site, I think that site should just be closed anyway, it seems like a stupid niche
I agree w/ gold
someone massively edited the Q too to try and make it less slanted
 
@JennyD There is a discussion on english.sx about it. Apparently it was once correct English usage in a legal context or something. They think maybe it is that India is 50 years behind or something. Wouldn't be the first time. :-)
 
@FaheemMitha To who? People who's major activity on S.E. (look at the account) is posting a sort of pointless screed? Honestly -- do you think the poster even cares what the answer is? S/he's just ranting.
 
I mean, India might have the sexual revolution any century now...
 
slm
the issue is the Q has already taken a position, and how would one even attempt to answer it? How many open source projects are done very well that come out of academia
 
12:53 PM
@goldilocks To the poster. He/she obviously has feelings about it. As I do.
 
@slm I don't think it's possible to salvage it - because the question is "why are they doing it wrong" which has only one answer and that's "because they are people doing stuff they're not experts in". Everything else is just verbiage. (Not that I'm bitter after having had to deal with that kind of software cough)
 
@slm Not that many. Name some examples.
 
slm
it's a fruitless debate, we can find man datapoints on either side, there isn't any really good metric I've ever seen to say this code is better than another, you have to go through it line by line and when you see good code you generally just "know it"
 
It is actually a pet topic of mine, so yeah, I like to see it aired.
 
@FaheemMitha Erlang. qmail - arguably, because it did a good job for its time even though DJB is a... special person.
 
slm
12:54 PM
There was a Q here the other day about VMD
 
@JennyD djb is pretty special. quite, quite, unrepresentative of the average academic.
 
among other things, he is a software expert of sorts. though apparently generally hated in the free software world, as far as I can tell.
 
@FaheemMitha I like qmail. I ran the email servers at one of the 3 largest ISPs here on qmail and erlang.
@FaheemMitha He can be somewhat abrasive. And that was my understatement of the month.
 
celeb alert. I was a student at UIC briefly, and I met him while i was there. Before he was famous.
 
12:55 PM
@FaheemMitha ooooooohhhhh! swoons
 
I could have taken one or more of his classes, but I don't think I did.
@JennyD heh
 
@FaheemMitha :-)
Wasn't emacs originally from academia too? And now it's an entire OS!
 
@JennyD erlang, i dunno. didn't that come out of the telecomm industry or something?
 
slm
There's are very good libraries that Lincoln Stein has developed in Perl
 
@JennyD Only very very tangentially. That was a product of the old AI lab
 
slm
12:57 PM
He's been doing these for years, he's also the original author of most of the CGI libraries too
 
Those people were hackers. they despised academics.
 
@FaheemMitha You're right; it was Ericsson that commissioned it to run the AXE stations. but it was built by academics, IIRC.
 
slm
this is my point, we can sit here and cite examples on both sides, but we can never answer that Q b/c the Q doesn't contain the seeds on how to actually answer it
 
@FaheemMitha Working with people who won't do things properly is frustrating. Beyond that, good code is good code and bad code is bad code. I'm taking CS and while it should diminish the possibility, I have no doubt lots of people coming out of that still write bad code. But that person seems to be dumping it on a particular lap. It seems very odd to me to say, "I have been doing this for years with these people" and yet have to start an account on a web site in order to ask the question.
 
Although I may be wrong - I use it, I don't study its history...
 
12:58 PM
@JennyD Oh, I dunno. I'm not saying there aren't counterexamples.
 
slm
@goldilocks - that's 100% dead on
 
@goldilocks I think it is just an expression of frustration. Academia is a particularly toxic environment if you are interested in software.
 
slm
I've been writing s/w for years and it's a craft that you have to develop techniques for over years, each project you apply learnings from previous ones
3
 
@JennyD This is a good point. That's half Richard Stallman, who's degree is actually physics.
 
as something that actually works, as opposed to a sketch of something that is just good for making a few graphs or generating some numbers.
 
12:59 PM
@goldilocks Agreed. The thing is very few people get everything right the first time. A mathematician may get the algorithm right but if they don't know anything about software development, why should they be expected to do it right?
 
Actually, a better academic counterexample is R.
 
@FaheemMitha And a non-academic example of utter crap is PHP
 
Though I think it is kind of crap, it is still 1000 times better than most of the stuff that comes out of academia.
 
slm
and yet it powers a fair amount of things we use on a daily basis
 
@slm This makes me scared and unhappy.
 
slm
1:01 PM
@FaheemMitha The prob. w/ this statement is it is only grounded in your experiences, not in tangible facts that we can measure and track
 
@JennyD Very true.
@slm Unfortunately. But that is a separate discussion.
@slm True. But lots of other people say the same. And I have seen a fair amount of academic code.
 
@FaheemMitha Basically there is no correlation between being good in academia and being a good coder, much less developer.
And why should there be.
 
@JennyD I do sometimes notice an attitude from science grads online in places like this where they seem to assume coding takes much less study and practice than it actually does -- implying I'm a mathematician, how complicated can this really be? lol But since there's no context it's impossible to say this is a trait of science people generally; they could also be stunningly mediocre at science.
 
@JennyD Because bad code means bugs. Which can compromise your results.
 
slm
@JennyD agree
 
1:04 PM
@goldilocks I think this a definite meme/theme among math people.
 
@goldilocks Because they are used to being the smartest guy in the room, which makes them unlikely to realise that they are only smart in a very limited area.
 
slm
@goldilocks I've gotten that from pretty much every manager too
the attitude is..."aren't coders just interchangeable?"
 
Math people in my experience are smart, lazy and arrogant.
 
@FaheemMitha Which, in fairness are good characteristics for a coder to have. But they're not sufficient, one also needs experience.
 
@JennyD smart and lazy yes, not sure about arrogant.
 
slm
1:06 PM
@goldilocks Ah it's just loops and counters, how hard can it be?
 
@FaheemMitha I think the canonical three are "laziness, impatience and hubris". Which isn't quite the same, you're right.
 
@goldilocks "I'm taking CS". You're a student?
 
@JennyD I am not a particular fan of coding in PHP, but there is nothing wrong with it as a technology. Wikipedia would not exist without it. There may be a disproportionate amount of bad PHP code around, but that's because it's primarily used in web-dev and you have a lot of self taught hacks there.
 
@JennyD not sure about impatient either. takes a lot of time to write good software.
 
@goldilocks Wikipedia as a concept is not dependent on language.
 
1:08 PM
@goldilocks I disagree. I think it is awful. And lots of people agree.
 
And there are enough things wrong with the actual language even without all the crappy self-taught hacks.
 
I remember the first piece of PHP code I saw some years ago. I sat and looked at it in disbelief for like 20 min. I couldn't believe how ugly it was.
 
@FaheemMitha I'm a few things at the same time. I have the core CS curriculum but no degree in that yet. My original B.A. is a joint major in English and Humanities.
 
@goldilocks Oh. So, a polymath, eh?
 
I don't have a university degree. I'm a mostly self-taught hack :-)
 
1:10 PM
hey guys, trying to recover a VM by removing root password, but having trouble remounting in rw using "#mount -o remount rw /"...tips?
 
@pinkunicorn Error message?
 
I get the help, as if it was an invalid command...actually I don't really know mount very well, just a command I've used before from a tuorial
 
@JennyD True, but wikipedia as a reality is written in PHP. I believe there have been discussions about porting it (mediawiki) to something else, but doing that won't save anybody any work. Ergo, PHP has served a purpose in reducing the required labour, which is to its credit. That would be the major reason smart people use PHP (it's easy). Which is also sort of the reason dumb people use it, and the later give it a bad rep.
 
@goldilocks Yes, but Stallman was never an academic. he has a degree yes. Like a zillion non-academics out there.
 
@goldilocks You are saying that if PHP did not exist nobody would have used another language to write it? I find that... unconvincing.
 
1:14 PM
@goldilocks I don't really understand why PHP has so much traction. You'd think anyone would take one good look at it and run away screaming. That was basically my reaction.
 
@pinkunicorn You're missing a comma. It should be mount -o remount,rw /
 
PHP is powerful and somehow neutral these days...but people have reasons to hate it because it's still inconsistent
 
@FaheemMitha Sure, and lots of people think "Python is the coolest dude". So what? PHP has appropriate and inappropriate uses IMO. Also, I do not think it is significantly different than the other dynamic OO langs (perl, python, ruby), although personally I'd rank it #4 on that list.
 
@goldilocks I disagree. It is to programming what mysql is to databases - popular because it doesn't require its users to understand the concepts they're using, unstable and full of security holes.
 
@goldilocks There are like 1000 posts on the net, pointing out what a terrible design it has. some with expletives.
 
1:16 PM
I think javascript, through recent things like node, mongo and so on, could easily become the new 'Esperanto' of web programming, but until then PHP is a close as you can get
 
not the case with python, for example
 
@FaheemMitha He's had a chair at MIT since forever. No sure if he teaches tho, or publishes in his field.
 
@goldilocks You're mistaken.
 
@pinkunicorn Except for the part where they are used in entirely different ways...
 
ugly, sure but still cost-effective
 
1:16 PM
He has no formal academic appointment at MIT or anywhere else.
He used to have an office in the old AI bldg, but I'm not sure if he still does.
 
all languages have their base, java has a VM, c has compilers, perl has an interpreter
 
@pinkunicorn PHP is strictly server side, whereas js (other than node) js is client side. Things like mongo mainly serve up js, although I think you can use it to write little db routines there too.
 
He used to be a MIT AI lab staffer till he resigned. That is as close as he comes to academia.
Anyway, off to do a little yoga. Take care, guys.
 
@FaheemMitha AI eh? Huh. Wonder why he resigned -- actually best if I don't speculate there, lol.
 
popularity of JSON and service/event-based systems are making JS more popular by the day
I love PHP, it has thousands of use cases that you can do in a few lines - like writing a crawler for example. A nodejs webcrawler isn't a great plan, since you usually need a synchronous flow
 
1:22 PM
@pinkunicorn Yeah, again, it's totally ubiquitous client side. It's essentially mandatory client side. JSON is not strictly a js thing although it's used there (I think its a derived subset of YAML, in fact). I'd say js is a much worse choice server side than PHP, except in so far as it can be used to better integrate with client code.
 
I thought it was crazy at first too, I'm paid to write PHP but node + express is really powerful once you get the hang of it
 
Yeah node is pretty awesome thing.
 
php+apache or nginx+fastcgi implementations are much harder to configure safely.
Anodejs app ONLY answers requests that are explicitly listed
 
@pinkunicorn Ruby FTW...
 
Here we go: en.wikipedia.org/wiki/YAML#JSON Remember, javascript uses JSON, but JSON is not javascript.
 
1:26 PM
^ yes I know JSON is independent, it's just very convenient for javascript, whereas php serialized for example is not very nice across platforms
 
But you and I do very different things. I've spent a couple of decades now working with systems that must not be down, so it's natural that I would use a different set of tools than someone who just does websites.
 
@FaheemMitha Hmmm, you're right. csail.mit.edu/user/888 Dunno why I believed he had a permanent chair.
 
aha, completely agree: crashable servers take some getting used to! it's a shock when one error kills your whole site! but that can be seen as a good point: it forces you to write superstable code and you have assurance that no undefined behaviour will do damage
 
@pinkunicorn Oh dear. We are not even talking the same language. I suppose you missed the part where I mentioned using erlang, or possibly you don't know what it is.
@pinkunicorn But. Under those circumstances you are using PHP? Because that language never changes...
 
erlang for db routines?
 
1:30 PM
@pinkunicorn Have you by any chance heard about couchdb?
 
@JennyD You can write little js routines for that too.
 
@goldilocks ENOPARSE
 
In fact I think that's more the purpose, it's just implemented in erlang.
 
@goldilocks Yes, it is. Because it needs to not be down. Like I said.
 
yes, I like it a lot, but mongo gives you more features out of the box...like you said, people take what they can use quickly
 
1:32 PM
@pinkunicorn Fast, cheap, good. Pick at most two.
 
^^the reality is you must pick...!
 
@pinkunicorn Oh, you can usually get two of them, but never all three.
 
where cheap also equates to means "you can find DEVs who have some experience with it"
 
@pinkunicorn and they are plentiful enough that you can pay them peanuts, and still you expect them to not be monkeys
(not "you" personally, "you" as "generic boss")
 
LevelDB and derivatives vs MongoDB is a good example...LevelDB is superfast but you have to write everything, Mongo has much more flexibility but is much much slower
 
1:34 PM
Haven't actually used mongo. Couchdb I thought was clever but TBH I'd think twice about choosing it for anything. There is a lot of awkwardness in it too, you almost have to use python for certain things...maybe this has improved.
 
some parts of mongo are disgustingly un-intuitive for me, but still clean. There's "dead weight" (=possibly unnecessary features) in the binary but the typical operations are clean and hard to use wrongly
 
Mongo seems to have won out in popularity over Couch. I'm just googling around to see what happened to the whole "couchapp" craze, but there site won't load. No so many new question on S.O. with that tag.
 
^ simplicity and ease of use wins again...it's so much slower than the alternatives but I really want the easy replication/sharding and geolocation and stuff...too many luxuries to choose any other DB for now
 
Slow only matters if it means "not fast enough" ;)
Whereas "easy" is always better.
 
@goldilocks "fast/easy but falls over in a light breeze" is also not useful in the long run
 
1:44 PM
@slm have a look at the comments here
0
A: /dev/disk/by-uuid/ not working on one machine

AlvaroIs udev working properly on your server? udev is the device mapper in charge of the creation of those symlinks, so it can be that it is not working, or that it is not configured properly. Is it running? (i.e. /etc/init.d/udev status) It may be that it is not configured properly, so check if yo...

 
@jenny thanks for mount tip, I was too blind to see it
 
@pinkunicorn You're welcome! It's really hard to catch one's own typos...
 
@JennyD Sure, using the wrong tool for the job is using the wrong tool. However, again citing the wikipedia example, I'm not sure what we're referring to. Obviously, PHP does not blow over so easily.
 
+ what did you really mean about erlang? I only know that couchDB is built using it
 
Is it one of them new fangled things like Haskell? ;P
 
1:46 PM
@pinkunicorn It is a language used for building services that do not go down - basically using breakable parts to build an unbreakable system
 
slm
@Anthon Is this guy for real?
 
@JennyD Implies this is a unique trait of erlang which seems a little credulous. I like to argue BTW...
 
slm
@Anthon - Martin likes to throw bounties on Q's that are weakly asked as a way to get others to do all the work (IMO).
 
I love to argue: better still I like to be shown that I am wrong and learn why
 
@slm I don't know but Stewarts answer, without looking at the history of the answer ...
 
1:49 PM
@goldilocks I noticed. I don't think I ever said it's a unique trait - but unlike e.g. PHP, Java, etc, this language was constructed for that specific purpose, which naturally affects what choices the developers make when they implement the language.
@goldilocks For things built in erlang, see e.g. Facebook chat.
 
@slm did you mean to write to Alvaro, because you wrote to me "You have 200 rep" . I know I have less than you, but it is not that bad....
 
slm
@Anthon - yes I typed @ A and your name was 1st, I fixed it
 
But after having worked with a system built in a language that has "keeping traffic flowing regardless of what elements of the system crash" as the top priority, hearing that I don't understand crashable code was... amusing.
 
so small parts like a simple async message relay that fail nicely?
 
@pinkunicorn could you maybe expand a little?
 
1:52 PM
@JennyD Great, but the idea that there aren't robust long running, highly stable systems using PHP or Java is also pretty credulous. Of course there are It's one thing for developers of something to claim something, it's another for it to have real significance.
 
@goldilocks I don't believe I said that either.
 
trying to get a feel for the concept from you, based on what you said about "using breakable parts to build an unbreakable system"
 
I did say that PHP is not a suitable language to do it. Doesn't mean it can't be done, and I will be impressed when it is. There's an entire hotel in Sweden built from ice - obviously it can be done, but most people won't even try, they're satisfied making snowmen.
 
@JennyD I was responding to the idea that the developers of erlang have important priorities that the developers of PHP and Java don't, and that this really adds up to anything much. One thing I do remember about it is it is an incredible memory pig -- way, way, way worse than any language I've seen in fact.
 
slm
@goldilocks I don't think she said that either, maybe she was implying it but I didn't get that. All languages can be used correctly and incorrectly.
 
1:55 PM
@slm Yeah, that's the point I'm driving at.
 
@goldilocks I am saying that they have different priorities. Do you not agree that this affects the languages?
 
slm
Languages are like the 50 hammers at home depot, you can use a rubber mallet all the time to put nails in...or use another more appropriate one...
 
with PHP it's easy to get the templating right - that's what PHP was from the ground up. Using it for your app logic, regular jobs, routing and so on is harder and easy to do wrong
 
@JennyD Absolutely languages are different. However, high level abstract claims about "breakability" are just that: high level abstract claims. I don't doubt there's logic to make that more concrete, but I'm questioning the premise that there is some significant "breakability" issue with other languages such as PHP and Java.
 
(hence wikipedia, wordpress using PHP: focus on content)
 
1:57 PM
@pinkunicorn And a language that prioritises message passing among nodes, has built-in easy to use methods to sync nodes up when one has been offline, etc , but does not prioritise templating is an entirely different beast.
 
agree totally
 
slm
you can usually tell when you're using a lang. in a inappropriate way b/c you have to stick more to conventions and best practices, these are sort of like bandaids over the lang. to protect you.
 
Thus, prioritising keeping the service online is built into the language itself, in a way that it is not in PHP or Java or C or most others.
 
@JennyD Inline templating is a pretty zany thing, I don't really like it and it's probably the major reason PHP gets derided -- it makes the whole thing seem very hackish. Vis. redundancy an nodes, it sounds like a nice way to specialize a language, but I'd presume it is easy enough to accomplish in Java?
And most everything else, actually.
 
really really depends what a 'node' is...a machine, an apache instance, a db replica
(or I've misunderstood something)
 
2:06 PM
@pinkunicorn A machine in a distributed system en.wikipedia.org/wiki/Node_(networking)
Which are common place. Not that I've ever worked on one, but redundancy would seem to be pretty important. I believe google is the master of this and they use predominantly C++.
 
^ not so sure it must be a machine...but arguaing about semantics is boring
 
@goldilocks You don't have to speculate; it is on record. He resigned to found the GNU project and the FSF. He didn't want to have to worry about MIT making an intellectual property claim on his work.
 
@pinkunicorn I guess you could consider two applications working together the same way.
 
@pinkunicorn An erlang instance within the distributed system - it is usually on another server, though of course you can have several nodes on one server if you like it
 
I could have a DB server and a nginx and a hand-written c daemon running on a machine, and each one could be up or down
 
2:10 PM
@FaheemMitha Ah so it was a long time ago then and ever since. Really bugs me this thing about believing he had a chair because I have some kind of obviously false memory about having looked it up and read about. Or else someone remarking in an article that "Stallman's office is down the hall from mine".
@pinkunicorn Yeah this is my point, the idea is integral to "distributed systems", not erlang. Although erlang may be tailor made for such, apparently.
 
@goldilocks Yes. A language which is built to run on multiple connected node and has that built in is fundamentally different from a language which has the distributed part tacked on.
 
slm
@Anthon - you've seen this Q&A about using type vs. which?
 
@JennyD See, I'm not sure if I agree with that. If you consider a language that's implemented in C, or even implemented in itself (ala Java, I believe -- not sure which of these two erlang is), then all features are essentially pieces tacked together. You could say, e.g., C is not designed to do anything except implement an operating system, and that's literally true. However, that does not mean that's there's any awkwardness in doing just about anything with it....
 
slm
54
Q: Why not use "which"? What to use then?

Stephane ChazelasWhen looking for the path to an executable or checking what would happen would you enter a command name in a Unix shell, there's a plethora of different utilities (which, type, command, whence, where, whereis, whatis, hash...). We often hear that which should be avoided. Why? What should we use ...

 
@goldilocks I thought so, too, before I started using more than one type of language.
 
2:22 PM
@JennyD ...e.g., on the one had OOP in C is awkward and debatable impossible, but on the other there are oodles of OOP languages that are written in C.
-> on the one hand
 
@goldilocks I think you and I do not even agree on what constitutes a language.
You really don't see any difference between a language where syntax and usage encourages you to work in one way (e.g. by message passing between nodes rather than by calling functions), and one where that way needs a metric shedload of extra code?
 
@JennyD Seems pretty simple, turing completeness and all that. I'm not questioning the value of erlang or your use of it, just the idea that all kinds of other languages aren't well suited to distributed systems. A topic I know almost nothing about, lol, except I am sure there are piles of them using Java components.
 
@goldilocks So if you know nothing about it, how about maybe, you know, reading about it a bit before dissing it?
And if you say that a language implemented in C is no different than C, then we do not agree at all on anything at all. Which, yeah, kinda obvious, so never mind.
 
@JennyD Okay, so the exact same sort of claim can be made for PHP as a perfectly fine technology. Like I said, I don't really like inline templating, but lots of people consider it a boon to productivity and they are probably right.
 
@goldilocks No. The next question is whether the design and implementation is good, which is a separate question from whether it's suited for the domain.
Is the syntax understandable? Are built-in functions consistently named? Are error messages useful?
 
2:32 PM
@JennyD The design is not a separate question from applicability to a domain. Implementation is a fine grained topic, but in the end it is hard to claim that something used on millions upon millions of servers has serious implementation issues without it simply amount to something like "I don't like MS Windows". I don't, but that doesn't mean there's actually anything in particular wrong with it.
 
PHP is a great templating language, from many viewpoints the best, abused horribly because eventually the developers need to do more with it. I would argue the application of the language is usually the fail point, but that's no surprise since PHP has wrappers for just about anything out of the box
 
@pinkunicorn I note how you do not answer the questions I pose as a bare minimum of what a language should achieve in order to not be a huge stinking pile.
In any case, I'm heading home now. Cya.
 
@no was out for moment
 
@JennyD Based on what you've been saying about specialization and erlang, how does it make sense to say "erlang is great" but "PHP sucks"? Going back to erlangs completely outrageous memory requirements, someone could claim "C++ is great, erlang sucks". Etc.
 
@jenny I consider them indirect/rhetorical. Be more direct, we'll all benefit since you seem to know what you're talking about
 
2:37 PM
@slm what about that question? WRT the guy not finding his yum after mounting some partition on /usr?
 
@pinkunicorn I'm kind of reminded of what I use to like to call "python ego" or "python envy". Like what's up with all the defensiveness? ;)
Some brands seem to attract this more than others.
 
no idea...php isn't something that you can use as a programmer to show off, that's for sure. I've acheived a lot with it but it hasn't been elegant, and that does have a lot to do with the language itself. As a businessman, it stands out as a great choice, even if it doesn't have the geek cred of python, perl or c++
 
Yeah. And an arrogant C programmer can also say, "You're all pathetic because you need these crutches." It's kind of Chevy vs. Ford for the most part.
IMO
 
Not 'need'...you can use a glass to drink water or you can drink from the tap...if you have nothing to wash your glass, it might be better to drinnk from the tap
 
Even if you have a glass it would be better to drink from the tap because it's faster and involves fewer moving parts, lol.
 
3:05 PM
Fortran is where its at. I refuse to read the backlog of this lengthy discussion but feel compelled to add an opinion.
 
slm
3:17 PM
@Anthon Yeah I've been mulling that one over in my mind. I'm assuming that it's either a path issue, or when he mounts it he might be missing mounting options.
 
@goldilocks Universities don't give chairs to programmers, even possibly genius-level programmers.
An obvious source of info about Stallman if you want to look is the Williams book "Free as in Freedom".
Free as in Freedom: Richard Stallman's Crusade for Free Software (ISBN 0-596-00287-4) is a free book licensed under the GNU Free Documentation License about the life of Richard Stallman, written by Sam Williams and published by O'Reilly Media on March 1, 2002. Williams conducted several interviews with Stallman during the writing of the book, as well as with classmates, colleagues of Stallman, and his mother. The book has received positive reviews. Structure The book is divided into a preface, thirteen chapters, an epilogue, three appendices and an index. A copy of the GNU Free Docume...
I think Stallman could possibly be called a researcher; certainly some of his sotware work was researchy, but not the kind of academic research work universities like. You have to be published in peer-reviewed journals.
 
slm
I'd buy that argument
 
@slm Hmm?
What argument?
@casey :-)
 
slm
He's called a researcher
@Anthon - I left the yum OP a comment asking for more details.
I believe there is additional info that's missing from this Q that's stalling any of us from being able to help you further. For example this system is not an actual physical host with CentOS on it, correct? It's a virtual machine (I'm assuming it's using XEN), can you confirm this? Also a simple ls -l /usr/bin/yum should reveal if yum is installed or not, it many not be working because of other reasons but please do confirm this too. — slm 41 secs ago
Does anyone have any ideas on this overscanning issue with GNOME desktop?
2
Q: Gnome 3 desktop display gone rogue?

Question OverflowI am having trouble logging out and shutting down normally using the Gnome 3 desktop interface. This is what I was shown on random occasions when I turn on the computer: The arrows point to where things are not normal. Right clicking the dropdown arrow on the system top bar produces the half-h...

 
3:37 PM
@slm I believe his problem is the screen...
 
slm
The monitor?
 
yeah...
 
slm
that was my thought too, you can scale the display to accommodate the DE
 
I have to press "Auto" each time I turn on the monitor
 
slm
OK, I'll see if he's done that
 
3:39 PM
@slm is he is mounting something on his /usr directory. I wanted him to use which (appropriate or not) to see that yum is in /usr/bin. No wonder that is inaccessible after /usr has whatever is on the partition he parks on top of that.
 
slm
@anthon I have no idea what he's doing, the devices he's mounting are typically used w/ Xen VMs
 
3:54 PM
@FaheemMitha No I thought he had a physics chair, but it seems he never actually completed his PhD.
I see him as an innovator but not really a significant researcher, computing wise.
Not like say, Richie et. al.
Hmm, but "While working (starting in 1975) as a research assistant [7] at MIT under Gerry Sussman, Stallman published a paper (with Sussman) in 1977 on an AI truth maintenance system, called dependency-directed backtracking.[13] This paper was an early work on the problem of intelligent backtracking in constraint satisfaction problems. As of 2003, the technique Stallman and Sussman introduced is still the most general and powerful form of intelligent backtracking." So that's that.
 
02:00 - 16:0016:00 - 00:00

« first day (1245 days earlier)      last day (3700 days later) »