/dev/chat - 2014-10-22 (page 1 of 2)

« first day (1474 days earlier) ← previous day next day → last day (3472 days later) »

12:02 AM

finally got this file transfer working again. I was happy go lucky with rsync until the login node killed my ssh session for oversubscribing CPU time on an interactive node :/ Have to go through some proprietary managed BS to get my files with very little real documentation on setting it up. fun...

@FaheemMitha Sorry, I wasn't here, what's up?

@terdon hey. thanks for getting back to me.

@terdon see

6

A: Biological data being used by an unpublished research paper is considered proprietary

The copyright is probably on the full database release flatfile and the formatted entries ... you will find similar conditions for UniProt/SwissProt so it is not so unusual. The restrictions on scripts are common to prevent server performance hits from a large number of requests. You can simply...

This method worked for me for a while. but the service that i was using shut down, so, the method suggested there, namely "http://srs.ebi.ac.uk/srsbin/cgi-bin/wgetz?[IMGTLIGM-ID:a00673|a01650]+-view+Fas‌taSeqs+-ascii"

does not work any longer.

My first thought was to switch to another server, but these things don't seem to have standardized interfaces.

The IMGT db is still there - ftp.ebi.ac.uk/pub/databases/imgt I was debating downloading the entire db and parsing it locally. Maybe that would be less prone to breakage. I dunno.

@FaheemMitha Wow!

OK, what exactly are these data?

Im that case, there a systematic way of finding out where these IMGT files live on other servers, in case this goes down?

@terdon Wow?

@FaheemMitha I can't believe they won't allow automatic retrieval of their sequences.

That's ridiculous!

12:07 AM

As you can see, this was all necessitated by a crazy French lady.

@terdon I'm glad you agree. It's nuts.

Did I not mention this to you earlier?

No. Wow. I've never seen anything like that before, I'll ask my housemate if he knows this guy. What a bitch!

@terdon Not a guy. A woman. Professor Marie something.

Yeah, I always get confused with those names. Marie-Paul could go either way I think. Like Jean-Marie. Perhaps it's the first of the two that decides gender, I don't know.

@terdon anyway, the answer I linked is from a DD. Which you could perhaps guess. All the other answers are completely useless. I posted on debian-science and Peter Rice replied on SE (and to me personally).

Anyway, what's so special about these data?

12:10 AM

@terdon Hmm, good point. I'm not 100% it is a woman.

What's a DD?

@FaheemMitha It is, I checked.

@terdon I need it for my research project.

@terdon Debian Developer.

@FaheemMitha Yes, but why not use the normal, public databases? What's special about hers?

@terdon The imgt one? I think that is the only one that has the info the way I want it. I'm certainly no bioinformatics expert, so I could be wrong. But, it has been working for me for a while, so I'd rather not change.

So, with this service shut down, it switching to another one easy? And what happens if that shuts down?

And what is SRS anyway? Don't know what that means.

@FaheemMitha I have no idea. If you choose to use some kind of quirky data, you're on your own. I have no idea what this thing is though it does look legit. But as to whether it will go down, your guess is as good as mine.

12:13 AM

Peter also suggested that I could download the whole db, and then parse it locally.

@terdon Ok.

Yes, that makes sense.

What does?

@FaheemMitha EBI is the European Center for Bioinformatics, it's one of the main databases. SRS is something or other, but the main point there is EBI.

@terdon yay

@FaheemMitha The thing is, of all you need are some sequences, it's trivial to get them from uniprot or ncbi or ebi or wherever that lets you do what you want with them.

@Braiam :)

12:16 AM

@terdon I'm not exactly sure how one would do that. So this IMGT thing means nothing to you?

@FaheemMitha No but then, I don't work in immunology. It might be huge there. What data do you need?

Just to be clear, this EBI thing is not going away any time soon?

If all you want is the fasta sequence of a gene, use the big databases.

@terdon big?

@FaheemMitha ncbi, ebi, ensembl, uniprot, unigene, depends on what data you need.

Can you edit your question and give a link to the actual data? I don't see this "6 sequence FASTA format" I'm supposed to click on. Where is it?

12:18 AM

@terdon Ok. I think maybe only ebi has it, not sure.

@terdon Edit the biology.sx q?

One sec, let me look.

@FaheemMitha You mean academia right?

@FaheemMitha Not a chance, they share data.

@terdon right

@terdon Those wretched people changed the interface. Sorry about that.

I'll have to figure out what they did. Sigh.

@FaheemMitha Can you just tell me what the data are?

@terdon Ok, I guess more poking around is necessary. I'll write to Peter.

Fasta sequences apparently but of what?

12:21 AM

@terdon One sec. I'll try to find a link

@FaheemMitha I'm sure I can help you if you only tell me what the data are! Like with words? :P

@terdon Does the top of this page mean anything to you?

imgt.org/ligmdb/documentation#h1_0

apt users, this works right? sudo apt-get install --reinstall grep?

@terdon I haven't thought about this for a bit. Let me try to load it up in my head.

@FaheemMitha nope

12:23 AM

@Braiam Works for what?

@Braiam Yes, I ran a similar thing a few days ago.

@terdon ok

@FaheemMitha unix.stackexchange.com/review/suggested-edits/66015

someone through that the order influence things...

as long as the packages are in the last position apt-get doesn't care

@terdon description of the data sets direct from my paper draft

\item RSS data sets : Human and mouse 12RSS (Recombination Signal Sequences) data sets
downloaded from \url{http://www.itb.cnr.it/rss/stats/HS12RSS.fasta}
and \url{http://www.itb.cnr.it/rss/stats/MM12RSS.fasta} (see
\cite{rsssite}). These data sets consist of 118 and 125 sequences
respectively, all of length 28. We remove duplicates from these
sets, obtaining two sets of 111 and 120 sequences respectively.

\item Gene data sets: Two collections of genes, one human and the other mouse, from
IMGT's LIGM-DB nucleotide sequences (see \cite{ligmdb}). The genes

@terdon edited your answer ..

12:28 AM

@FaheemMitha OK, usually, such data are reported in a paper and are in the public domain. I have never seen anything similar to this before. Still, the link there seems to work and since it is linking back to their servers, I don't see what the problem is. Just make your scripts polite. Make them wait before requesting additional sequences so the server is not spammed.

@terdon if that still doesn't mean anything to you, that is a problem, because my own understanding of these things is only the sketchiest. here is the imgt-db thing on ebi - ftp.ebi.ac.uk/pub/databases/imgt

@terdon What link is that? the erb one?

@FaheemMitha No, the one from your latex there.

I was previously using this srs service to return fasta files the way i liked it. they shut it down.

@AvinashRaj Yes, I just saw. Umm, why?

the cnr.it one?

12:29 AM

What's the point of using a lookbehind there?

@FaheemMitha Yes, itb.cnr.it/rss/stats/HS12RSS.fasta

op says one or more occurances..

@terdon Ok, those people are not making any difficulties. the french people are.

i didn't add any lookbehind there.

@AvinashRaj Hang on, I'd missed the one or more requirement.

(?:blah blah) = Non-capturing group.

i added an example.

i think it explains you clearly

12:31 AM

@terdon on srs.ebi.ac.uk they mention IMGTLIGM, but only the ebi link is given

@AvinashRaj Indeed it does, cool, thanks!

come on @terdon. Why thanks... :-)

@FaheemMitha And they give an ftp.ebi.ac.uk address for it! Cool, use that. That should be very stable.

@terdon I already have one. I was wondering where else one could find this dataset if i needed to. Peter mentioned ftp.ebi.ac.uk/pub/databases/imgt in his answer.

@FaheemMitha Ah, well, if the EBI has it you should be fine for ever really. Short of nuclear war I'd say.

12:36 AM

@terdon ok

@FaheemMitha Yes, that's listed there on the page you gave me, in the IMGTLIGM entry.

@terdon right. but no other sources are given.

Well, the EBI is really about as safe as it gets.

European Bioinformatics Institute

The European Bioinformatics Institute (EBI) is a centre for research and services in bioinformatics, and is part of European Molecular Biology Laboratory (EMBL). EMBL-EBI is listed in the Registry of Research Data Repositories re3data.org. == About the EMBL-EBI == The roots of the EMBL-EBI lie in the EMBL Nucleotide Sequence Data Library (now known as EMBL-Bank), which was established in 1980 at the EMBL laboratories in Heidelberg, Germany and was the world's first nucleotide sequence database. The original goal was to establish a central computer database of DNA sequences, to supplement sequences...

Publicly funded, not for profit, etc etc.

@terdon ok. so, any tips on how to go about parsing the that db fasta file for what i want?

i mean ftp.ebi.ac.uk/pub/databases/imgt/LIGM-DB/imgt.fasta.Z specifically.

@FaheemMitha Depends. What is it that you want?

12:47 AM

Hmm, maybe I could just look it up by accesion number, in which case it would be easy

@terdon I need to pull out some of those sequences.

@FaheemMitha Ah, sudo apt-get install exonerate

@terdon if you have a moment, can you (a) download that file and (b) tell me if you recognize the identifier used?

@terdon hmm?

That is a great tool that among many other things includes a nifty utility to extract a sequence from a multifasta file

@terdon Ah, ok.

@FaheemMitha I'm doing so.

12:49 AM

@terdon Thanks.

@FaheemMitha I explain in more detail here:

3

A: Getting matched fasta file

The simple case you show is trivial. Your sequences are never more than a single line so you can simply use grep to search for each of your IDs and the line after them: grep -Fwf list.txt -A 1 seq.fasta | grep -v '^--$' > out.fasta The grep -v '^--$' simply filters out the lines with -- that...

Oh, and you can also use my script, I link to it at the end of that answer.

@terdon That's very helpful. thanks.

:)

can this put multiple sequences in a single fasta file?

that's the way i have it set up currently.

OK, those are EBI accession numbers, is that your dataset?

12:52 AM

@terdon I think so. Those are an international standard? One sec, let me paste the numbers I am using.

This is for mice.

@FaheemMitha Of course. Both my script and the, much faster, fastaindex/fastafetch approach. Just use fastafetch -f seq.fasta -i seq.idx -Fq <(sort -u list.txt ) > out.fa

@FaheemMitha Yes, one of the standards.

Which means that you don't need to interact with this lady in any way. Cite here paper if you use her data but retrieve the data directly from any other of the myriad of databases that will it. Give me a couple of IDs and I'll give you a few ways of getting the sequence.

paste.lisp.org/display/144122

@terdon ^^. Are those EBI accession numbers?

Yup, looks like.

Man, you python people and your horrible, hard to parse list syntax.

here is one random one -> >AJ132683 Mus musculus IgVk he24 gene

@terdon Would you prefer Common Lisp?

Actually, that is YAML format.

@terdon does that mean I don't actually need to use this db?

@FaheemMitha Yes, that's what it means.

12:56 AM

@terdon mm?

@FaheemMitha I'd prefer one value per line so I can use it directly :)

what python?

though downloading this fasta file and parsing through it seems like a reasonable and robust way to do things.

@Braiam maybe it wasn't but it looked like it.

@terdon You mean my paste? It is easy enough to munge. Were you planning to do something to it?

12:57 AM

looks more like json than python

@FaheemMitha Yes, but there you might run into legal stuff since you're redistributing the data. Retrieving it from public servers might be safer.

I can send you a one per line list if you really want

@Braiam It's YAML

@terdon Well, I can tell users to download the big data file, then parse it.

YAML is approximately a json superset, i think. to a first approximation, anyway.

I'm using a python yaml thingy with it.

Any idea how widely this exonerate thing is available? And in this simple case, does it really buy much?

@FaheemMitha It's in the repos, I explain all that in my answer. Just apt-get install. Anyway, is this something you want your users to do or something you want your program to do? This will get all sequences in a couple of minutes:

while read i; do wget -O - ebi.ac.uk/ena/data/view$i"&display=fasta" >> accs.fa; done < accs

accs is the list of accessions, one per line.

Else, run this to add the urls to the file of accessions:

sed -i -r 's#(.*)#ebi.ac.uk/ena/data/view\1&display=fasta#' accs

Then, get all the sequences in 13 seconds on my shitty connection:

@terdon i was actually about to ask you this. :-)

$ time wget -i accs -O - > accs.fa 2>/dev/null

real	0m13.144s
user	0m0.032s
sys	0m0.020s

1:05 AM

@terdon Are you still in FR?

Yup, not for long now though.

@terdon Where to next?

@terdon I don't understand this bit.

@terdon This fetches the fasta files. That I get.

@FaheemMitha So does the other. It's just that the bash loop takes a list of accessions as input while the second takes a list of URLs:

$ head -2 accs
ebi.ac.uk/ena/data/view/AE000665AE000665display=fasta
ebi.ac.uk/ena/data/view/AF018146AF018146display=fasta

@terdon Oh, which is better?

The latter, by far.

The bash one will start a separate wget for each sequence while the second will only launch one.

1:09 AM

@terdon Ok. I'll play around with it. Thanks. So this can be done for other db servers across the world? Is there a standardized interface?

@FaheemMitha Of course not! Standards? In biology? You must be kidding.

@terdon heh

Nah, they often provide the classic approaches such as restful services and the like. This is one of the best IMO, it's simply http://server.url.com/SEQ_ACCESSION&display=fasta

@terdon ok. just for the sake of comparison, can you give me a way to do this for one other server?

Oh, and yes, ignore the example above, the URLs are wrong.

1:11 AM

preferably one of the well known ones.

@FaheemMitha I don't know of one that has the EBI accessions available that way off the top of my head.

@terdon ignore what?

@terdon Ok.

2 mins ago, by terdon

$ head -2 accs
http://www.ebi.ac.uk/ena/data/view/AE000665AE000665display=fasta
http://www.ebi.ac.uk/ena/data/view/AF018146AF018146display=fasta

@terdon Ok. Are those example scripts earlier correct?

@terdon like this one?

@FaheemMitha Yes, only the sed was wrong. The whole procedure is 1) start with a list of accession numbers, one per line (call it accs.txt) 2) run this sed to make each line a URL:

sed -i -r 's#(.*)#ebi.ac.uk/ena/data/view\1\&display=fasta#' accs.txt

1:14 AM

@terdon I see

Then use accs.txt?

3) feed it to wget

wget -i accs.txt -O - > accs.fa 2>/dev/null

Took 24 seconds on my machine.

no, hang on, accs.txt is the list of acc nos.

you lost me. what does the complete thing look like? afk for a few min.

@Seth you can't close as duplicated a question of "how to solve X issue" to another that asks "you need this information to solve X issue"

oh, i see, the accs.txt is modified in place. clever

@FaheemMitha The sed was sed -i so it changed the original file, it is now a list of URLs

1:16 AM

@terdon right. i'll give that a try. ok, really afk now.

the later won't solve the X issue by itself, but will help to get on

Good luck.

i guess that is an alternative to dloading a big data file, and then pulling out the stuff i need

someone is seeing that chat doesn't scroll automagically?

@Braiam Nope, just did for me.

1:17 AM

@Braiam You can when all that information is missing. Sorta. It isn't perfect workflow, but a lot quicker and easier than hunting down that link every time. Either way, we should have our own local copy, even if you just copy paste it (attribution necessary, ofc).

@Seth err no, that's unproductive

you will close as duplicated, then people will think if they follow the answer the issue is solved, when is not

true enough. See my meta comment.

@terdon grrr... F5!

@terdon i just realised it must be 2 am in France. Don't mean to keep you.

I do have more questions, but they can wait.

@FaheemMitha terdon goes to sleep at 6am, so don't worry :P

1:25 AM

No, 3.25 am.

@Braiam He does?

Hi @Seth

@FaheemMitha there's enough terdon for at least 2 hours :P

Hi @Faheem

uggh, SE is being slow right now.

@FaheemMitha check his chat user page chat.stackexchange.com/users/47392/terdon

Chromium has been crashing a lot recently. wheezy. Anyone else?

I just tried using hjkl for navigation while editing a post >.>

1:28 AM

well, the invidual tabs crash, at the drop of a hat. the whole thing keeps running though.

user image

UGH! SE team use iMacs...

@Braiam I see a lot of macbooks, but no imacs

^

and if I was going to complain, I'd rather complain about the windows infrastructure powering SE than the use of a unix derivative OS on Apple hardware

though I am biased, since (full disclosure) I do own a macbook pro in addition to my many self built machines that run linux

@casey I reckon macs are powerful little beast... up to a price

1:41 AM

everyone loves those osx systems, we're in the minority

I consider them the Audi of the laptops

same here

they are good, but way too expensive and difficult to upgrade

they're for ppl that want to be hardcore Unix types but aren't really

I have 1 for work. It's Unix enough

but it just irks me

I don't like the "you can't do whatever you want with your hardware" philosophy either

1:45 AM

I'd never buy a mac.

@FaheemMitha Depends on the day but I do often, yes.

@terdon ok. up for a couple more questions?

Try me

@Braiam with the macbook and imac / mac pro, its not so much that philosophy as with the iphone/ipad. I can install linux on the macbook fairly easily with (almost) no trace of OS X left behind (need to keep the iSight driver and the lcd color profiles). Or it'll happily (and supported) multi-boot into linux or windows.

The philosophy of the laptop and desktops is more "you can't run our software unless you buy the hardware" but they don't seem to care what you do with the hardware later

@casey Will it? I thought it was quite easily done but officially illegal.

1:48 AM

@terdon bootcamp supports it out of the box

@casey Huh, OK. I must be thinking of running OSX on non macs, That's not allowed is it?

@terdon ok, so I'm a bit puzzled about this sequence thing. This woman behaves like these sequences are her personal property. Is that because her lab produced this data?

thats where the drama is. apple considers the hackintoshes as contraband

user image

though I don't see the point of the hackintoshes. OSX isn't the choice I would make.

1:50 AM

But it seems you can get it directly from this ebi place without doing anything special. Have they simply added her data in with all the other stuff?

@FaheemMitha More because you would have been using her servers. There is no way she can expect to exert control over them if you're retrieving from public databases. The more I think about this, the more convinced I am that there is a misunderstanding here.

@terdon I was very explicit. I wrote her more than one message about it. Mostly because I couldn't believe it.

If you cared, I could send you copies. But I don't suppose you do.

@FaheemMitha No, she probably submitted it herself, which is why I find her stance puzzling. I don't think she understood you only wanted the sequences alone and not all the other bells and whistles offered by her institute. They seqs were almost certainly made available along with the NAR paper she mentioned.

but, just for my information, it is possible her group came up with these sequences?

@FaheemMitha I'll take your word for it :)

@FaheemMitha I'm sure they did. Not come up with them, sequence them.

1:53 AM

@terdon that's what i meant. the point is, they produced this data in the lab. but now it is part of all the other genomic data publicly available?

@FaheemMitha Yes, if you're getting it from EBI

@terdon as I recall, i sent her a copy of my paper, as well as an explanation of what i was trying to do. I optimistically thought she might be interested, but mostly she seemed uncomprehending.

@FaheemMitha That last word you used there is why I think there's a misunderstanding. She probably had no idea what you weer talking about.

@terdon oh, and about your sed thingy above. will it fetch all the sequences as a single file, in a single call to the server?

@Braiam me neither

1:56 AM

@terdon I guess i could go back and check my email, but I think I was very clear. I always try to be, and living in India, I get plentry of practice.

Not to sound racist, but by and large, if Indians can misunderstand something in some crazy way, they will.

@FaheemMitha The wget thing? I guess so, based on the speed.

cnet.com/news/…

@FaheemMitha Well, remember she's not a native English speaker, not a mathematician and probably computer illiterate.

@terdon Ok. I'll play with it. Maybe get back to you with questions.

big reason

> Apple represents a design aesthetic to which open-source developers aspire. While many open-source developers probably relate to the utilitarian interface of Windows, I suspect that more wish that they could churn out products that looked more like OS X.

1:58 AM

@terdon One would hope someone who has set up a big db online would have basic computer skills.

@slm Yuck.

Apple SUX.

I think free software devs could use training in UI. Design aesthetics? What's that?

Oh, and help writing good documentation, too.

There are some programs with pretty nice user interfaces out there, but also plenty that are not.

@FaheemMitha No, she hires people with computer skills. For all you know, you interacted with some yahoo from the accounting department.

@terdon You mean I was not exchanging emails with this professor at all?

I sure sounded like I was.

She did respond to an earlier query I had about the db. Not very helpfully.

I mean, a quasi-technical biology type query.

Though I really struggle talking to biologists. It always feels like we are talking different languages.

@FaheemMitha Well, having read through her CV, I expect her to be very knowledgeable about the biology and largely clueless about the IT side.

Well, time to try to go back to sleep, I think.

@FaheemMitha You are, which is why I come back to the whole misunderstanding thing.

2:03 AM

@terdon Plausible.

Hell, I still don't know what you want to do with her data and I speak both languages :)

Though I recently met a chemist who is writing a CL implementation based on LLVM from scratch, so one cannot generalize.

@terdon Hey, I sent you my paper. :-)

@FaheemMitha And you're chatting with a biologist who's a mod on U&L, I am not generalizing :)

@FaheemMitha Is it that one?

@terdon I thought you were a bioinformatician.

@terdon hmm?

@FaheemMitha speak her language, propose to be co-PI on a grant proposal :)

2:05 AM

The paper? yes, I think I've only sent you one.

@casey lol.

@FaheemMitha I'm a biologist by training. Well, my undergraduate degree was in pure Biology, I have no official programming or computers training.

@terdon I see. I didn't know that. Well, you do pretty well, but it explains some of the more puzzling gaps...

I'm a math person with no official programming or computer training either. Never taken a computer class in my life.

@FaheemMitha Heh, I'm sure :)

0

Q: Can't Connect To Site Cloned with Kali Linux

I'm using Kali's Social Engineering Toolkit's credentials harvester. I have it clone www.facebook.com and I can connect to my cloned site using the machine's internal IP address as long as I connect using the same machine that's running the harvester. When I try to connect using any other compute...

I bet his followup Q will be "How can I route all web traffic in my dorm through my local machine?"

@casey Is that what I think it is? Is he asking for help to harvest credentials from facebook?

2:09 AM

@terdon Given that hes effectively running a MitM program (I can only assume) and he wants to connect from non-localhost, I'm guessing that is exactly what he wants to do

@FaheemMitha anyway, if all you want to do is include the simple sequences with your academic paper, that seems to come under research and educational purposes by definition. I think you're getting worked up over nothing.

@casey That's what it sounds like. Delete with extreme prejudice?

@terdon given no justification of a legitimate use, thats my vote

@terdon Worked up? Me?

yea, just cite her paper and use her data

she doesn't need to be in the loop

@FaheemMitha Worried. Just use the data.

2:11 AM

she probably cares more about her h-index anyway

How the hell does one clone facebook? Or am I misunderstanding?

@terdon If you mean my earlier questions, I was just wondering how things worked with the whole - where do sequences come from - deal.

@casey I wasn't planning to keep her in the loop.

reject pls unix.stackexchange.com/review/suggested-edits/66026

@FaheemMitha Ah, well, usually they come from people like her, who produce the sequences and submit them to the public databases. As she has done. Then, the rest of us use those sequences and cite the paper that reported them. That's all you need to do. Unless, that is, you are using it for some kind of profit making scheme in which case, things are different.

@Braiam you don't think a description is useful here?

@terdon Right. Agreed.

@FaheemMitha its likely just proxying facebook so if you don't notice the odd url and lack of an ssl cert, it looks exactly like facebook, because it is. Its just that the proxy is going to record your login details

2:14 AM

@FaheemMitha describing what "upstart" is, is like saying to a friend "here you can buy hamburgers" when you enter a MacDonnals...

@Braiam Done and edited.

@casey eww. bad man.

so phishing, then? he's got a nerve posting on a public forum.

@FaheemMitha hey, the guy that managed "the (bad side of the) silk road" also did the same

@FaheemMitha I think most of the kali-linux "pen testers" think we are clueless that they are a bunch of noobs that don't have a clue themselves.

kali is like a magnet to that type

@Braiam No idea who that is.

2:17 AM

@Braiam there was a good side?

179

Q: Did the Stack Exchange staff members assist in the apprehension of Ross Ulbricht?

The recent arrest of someone behind Silk Road is in the news today, he was arrested in a San Francisco public library. A couple of background articles here and here. To what extent, if any, did staff and/or admins have in assisting law enforcement agencies in arresting this programmer? Did the...

discussion legal stackexchange-team

@casey I'm aware, that right now there's a thing called "the silk road" to buy/sell things with bitcoins

wee! the active tab is broken :D

Apparently drug lords program too. Good to know.

@terdon i'm off. Thanks very much for the help.

@FaheemMitha You're welcome.

1 hour later…

3:38 AM

you guys never heard of silk road?

You run a binary and nothing happens, how do you debug what's doing?

So the answer for this would be to check if we have execute permission on the binary right?

@Ramesh define "nothing"

I'd open a debugger and take a look, but thats just me.

Well, the question doesn't say much.

@casey Which debugger?

which question?

is necessary to have lan when there's networking?

3:49 AM

@casey Not a site question. Was referring to some questions here.

it won't be a permissions thing. Try it, the result is something (a permission denied error) rather than nothing

If a program did nothing that I expected something out of I have a few options. I can strace it. I can run it in a debugger (e.g. gdb).

I can see if it has more verbose output options or debug output options

@Ramesh strace binary?

@Braiam But it will just say me the sys calls involved right?

But I think the question is somewhat unclear.

@Ramesh they are moronic questions, thinking that they are smart ones

better waste your time solving Google's... those are interesting

4:06 AM

you're 5 mm tall, in the bottom of a blender and the blade is being turned on in 5 seconds. what do you do?

Puzzling

ok, maybe not a good puzzling question.

@casey I guess you could lie down?

iirc the answer is "jump out". the basis being that your strength / weight ratio increases as you diminish in size, so like a small jumping spider you just.. jump out

interesting.

In short, if were you shrunk to 1/10 your present height, your muscles would be only 1/100 as powerful—but you'd weigh a mere 1/1,000 as much. All else being equal, small creatures are "stronger" in lifting their bodies against gravity. Were you shrunk to nickel size, you'd be strong enough to leap like Superman, right out of the blender. Think of the feats performed by fleas in a flea circus.

quoted here.

@casey I believe that standard procedure is to splatter yourself all over the walls of the blender.

4:13 AM

that question was lifted from that google movie

the interns?

i think

It seems it was indeed an actual question.

imdb.com/title/tt2234155

i'm sure it was, that whole movie was a hiring ad for google i think

Use a programming language to describe a chicken.

4:27 AM

@Ramesh, I'm not sure that's a good idea. Toolchain compilation is a moving target, and any reference that is not actively maintained becomes unusable rather quickly. — Simon Richter 1 min ago

It seems a valid point. What should I tell?

@Ramesh strace?

4:44 AM

Oh, I see @Braiam said that already.

4 hours later…

8:37 AM

@terdon This doesn't work. Does it really work for you? I tried it one just one line.

The links at the bottom do work, but they return a whole web page -> ebi.ac.uk/Tools/dbfetch

2 hours later…

11:04 AM

@FaheemMitha Yes, it worked for me. How does it fail for you? What links work? The bottom of what? What was the "one line"?

1 hour later…

12:06 PM

BTS, slow for anyone?

forget it, is the ocsp server

1 hour later…

1:23 PM

Bytes transferred: 399,931,115,013 (32%).

Progress!

14 hours down, ~28 to go

1

Q: Generate MIPS architecture assembly code on a X86 machine

I need to generate MIPS specific code on my machine when I run my C program. When I simply run, gcc -O2 -S -c hello.c On my system, I get the hello.s which seems to generate some assembly code but it doesn't seem to be MIPS specific code. The contents of hello.s file is as below. .file ...

compiling cross-compilation mips

@Ramesh A couple nitpicks for you

what your answer describes is not a cross compile toolchain

what your answer describes is installing a MIPS OS image into a virtual machine and running a native toolchain within the virtualized environment

a cross compile toolchain does not need a VM, it runs on your native host but produces non-native objects. You would need a VM to run the code, but that VM doesn't need to have any development tools because all of the compiling happens on your native host

setting up gcc to cross compile involves compiling gcc (if no pre-build packages for your target exist). e.g. GCC configured on your debian x86_64 box with something like ./configure --target=mips64-elf --host=x86_64-pc-linux-gnu --prefix=/opt/gcc-mips

which when built will produce a compiler that runs on x86_64 but produces output suitable for MIPS. This is a cross compiler

1:39 PM

@terdon Sorry, I guess that was a terrible error report.

Off to dinner in a few, I'll reply later.

@Ramesh and my last nitpick, compilers do not convert code to binary, they compile code to a binary format :)

@casey Sounds more like correcting mis-statements than nitpicks.

@casey This one does sound like a nitpick. :-)

@FaheemMitha i suppose

I'd edit it myself but it would be too major, so I'd rather he did it.

I'd keep up to that paragraph he defines a cross compiler and I would strike the entirety of the VM and qemu related text

@casey Nothing wrong with editing it yourself.

well, my issue is really that I have no desire to setup a cross compile toolchain (for MIPS or anything else) so I'm not in a position to test and verify how to really get it setup and usable.

1:54 PM

@casey Fair enough.

I don't think I've never set one up either. Nor have I ever actually seen one in use. Hey, maybe cross-compilers are an urban legend!

nah, I used to have an ARM cross-compile environment

@casey Oh. No longer, I suppose.

« first day (1474 days earlier) ← previous day next day → last day (3472 days later) »

Transcript for

Oct '1422

/dev/chat

General discussion for unix.stackexchange.com. If you have a q...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile