/dev/chat

Jenny D

06:04

@FaheemMitha I should maybe have said "for supposed professional sysadmins" instead - the point is that home and dev setups are off-topic there, and I'd say that using a Raspberry Pi for a webserver isn't in line with "reasonable business information technology management practices"... Doesn't mean the questions are necessarily bad, and I'd much rather put them somewhere they can be answered than having to close them.

Faheem Mitha

06:48

@JennyD SF has a lot of arbitrary exclusions.

Jenny D

@FaheemMitha Yes, it's a lot harder to draw a clear line between professional and non- than between "is it a unix or not".

Faheem Mitha

@JennyD But somehow they manage to draw that line.

Jenny D

@FaheemMitha We do our best... But we also get a lot of complaints about all of us being elitist jerks (usually with a lot more four-letter words)

Jenny D

07:31

Although, to be fair, some of us are.

java-devel

13:27

Hi!

Can this question be reworded?

I edited this question

terdon

@java-devel What question?

java-devel

That I linked above

1

Q: E: Package 'gstreamer0.10-plugins-bad' has no installation candidate

I turned on the "Video Capture" option in VirtualBox, I recorded the video in webm format and I shutted down the virtual machine and try to play this video I got the error An error occured. Could not demultiplex stream in Totem video player. mpv video player can play this video. This implies that...

linux

terdon

13:52

@java-devel Ah, you hadn't actually linked anything above. OK, so what do you want to do to this question?

You want it reopened?

@java-devel is your main question how to get totem to play your video?

Also, please [edit] and clarify what OS you are using. Is the host Debian or the client? or both?

Anyway, I reopened it since it is indeed clearer now.

java-devel

14:58

@terdon Done.

terdon

Thanks

Faheem Mitha

15:23

In my usual off-topic mode, this is an interesting question, because Academia doesn't generally show much support for or understanding of free software. But here they are telling this student to "suck it up" (excuse the Americanism).

20

Q: Supervisor does not allow use of commercial software for student project, against student's interests

My supervisor does not allow me to use commercial software for my project. I believe that they want to reuse, modify and build upon what I do and thus are interested only in tools/libraries that are widely-accepted in the scientific community (which I do not intend to pursue a career in). The c...

ethics

Kusalananda

15:54

@FaheemMitha That question doesn't even mention what the software is. It's just playing on emotions and opinions.

Faheem Mitha

@Kusalananda Presumably it's some proprietary software. Though the poster uses the term "commercial".

terdon

@FaheemMitha Not quite. I wouldn't use closed source in academia either. You have no way of knowing what the damn thing does or if it actually works if it hasn't been published.

Faheem Mitha

@terdon You're missing my point. Mostly Academia seems fine with proprietary software.

(I don't know how much you follow it, but I used to follow it quite regularly.)

But anyway, off-topic.

Still looking for someone to repro my weird Plasma bug...

I wonder how annoying it would be if I started writing "shew" instead of "show". Would anyone notice?

terdon

16:35

@FaheemMitha I mean I understand the OP's professor's position. Academia is fine with proprietary software yes, but not in your area of expertise. For example, wet lab biologists will use proprietary stuff, but a bioinformatician working in the field will not unless he or she understands the software well enough to judge it.

Kusalananda

@FaheemMitha You're making sweepings assumptions about "Academia". I don't recognize this from my experience. Yes, we use commercial software (Slack, for communication, for example), but in research they use well established applications with support from either us or the makers of the software (which may well be proprietary, as long as it's "industry standard", but more often than not it's open source).

For all that we know, the commercial software that the OP wants to use may be Excel. The text doesn't tell.

Faheem Mitha

16:55

@terdon To be clear, by Academia I meant Academia SE the site. Sorry, I realise that was unclear.

@Kusalananda Again, by Academia I meant the site Academia SE.

That was an misleading bit of abbreviation.

As far as Academia the "institution" goes, that's probably too broad to say anything about.

Though my experience of universities is that nobody gives a fig about free software. But my own experience is just a couple of data points.

Kusalananda

17:11

@FaheemMitha Academic software developed using government funding, in Sweden, is required to be open source.

Faheem Mitha

@Kusalananda That's true of the NHS, too, I think. But my point still holds. Nobody cares. Maybe they are required to do such and such as a condition of their funding. That doesn't fall under my definition of "caring", at least.

It might be interesting to study how these funding requirements came about. I doubt govt people, by and large, care about free software either.

terdon

@FaheemMitha That depends entirely on the field. Bioinformaticians tend to be pro-open source, for example, while wet lab biologists don't care and don't even know what it means.

Very few people who are not computer-savvy care about FOSS and that is completely understandable.

Faheem Mitha

@terdon Well, yes. Fair point. I should have said that "in my experience".

I guess I meant that someone who is required to free his code as a condition of a grant isn't by a conventional measure, at least, pro-free software.

I haven't been expressing myself very well here. But fortunately this isn't the Oxford Union.

terdon

Heh

In my experience, every single researcher I've ever met or heard of who produces code does so under an open source license.

You can't get published if you don't, after all.

What would you say? I ran my program and got great results?

How can we know it's true?

Faheem Mitha

@terdon Well, the code itself might be free, but it could depend on proprietary software.

That's the rub.

For example, I can write a Matlab script, and put it under a free license.

terdon

17:25

@FaheemMitha Oh sure, but your code needs to be open.

Faheem Mitha

But you can't run it unless you have Matlab.

terdon

Yes.

I mean yes, you're right, you can't run it.

Faheem Mitha

Though I think that the script itself probably counts as free software.

This point actually came up recently, but I didn't verify it then either.

Of course I could go the lazy route and post on the Open Source SE.

terdon

In any case, the vast majority of scientists produce open code. You need to be able to have other people review it and you need to be able to share it easily.

Faheem Mitha

@terdon Open in the sense above - yes, of course.

Though I'd be surprised if it wasn't a dupe.

The Open Source SE will probably spank me, but here goes...

0

Q: Can software that depends on proprietary software be free?

[It's hard to believe this hasn't been asked here already, but I can't find it.] The question is very simple. Suppose I write a Matlab script and place it under the GPL. The question is whether the result can be considered free software. On the one hand, the script itself is clearly free. On the...

free-software-definition

@terdon Oh, and in my (again, limited) experience, the review of scientific code is like the Loch Ness monster. There are rumors of its existence, but it has never been verified.

user70680

17:45

This could be answered with some "java trap"

user70680

from that time where JRE/JDK was not totally free - gnu.org/philosophy/java-trap.html

terdon

@FaheemMitha Oh it exists. They don't really review the implementation, but they most certainly review the algorithm.

Faheem Mitha

@terdon Isn't the algorithm normally separate from the code?

By which I mean to say, one would not normally define an algorithm in code.

But perhaps I'm misunderstanding you.

terdon

Yes. That's why I clarified that they review the algorithm and not its implementation. Usually.

Faheem Mitha

@terdon Ah, ok. But that's not code review, then.

Michael Homer

19:55

Lots of venues have, usually separate, artefact review processes that precisely do review the implementation code.

And even in paper review, if you provided access to the software I will always attempt to give it a try.

Faheem Mitha

Hi Michael. How are you doing?

Transcript for