And some say FOSS sucks
-
An article I stumbled upon claims FOSS has better code quality than industry average.
I, personally, am not sure what to make of it, but I'm sure some regulars here will have a strong opinion.
-
It says for a certain LOC range their automated tool found lower (but the other range mentioned FOSS lost badly). So it doesn't really say what you are using to bait people.
-
That's impossible. The FOSS people never do what I tell them.
-
"The 2012 Coverity Scan Open Source Report details the analysis of the Scan services's most active open source projects, totaling over 68 million lines of open source software code. In addition, the report details the results of over 380 million lines of proprietary software code from a sample of anonymous Coverity users."
I suspect their sample size for FOSS products may be too damn small. They don't mention how many products cover all those lines.
In any case, they want me to register to see their report; fuck that, I'll just smugly ignore them instead.
-
Wow, that's basically just one big ad. How many times are they going to mention the name of their company (complete with link!) in the "article" (7 pages of ads, with some image captions).
I don't even have to look at their methodology to see how retarded this is. For one, they're treating every single defect as equal. This is fucking moronic.
Two, quality is more than just a measure of defects-per-line. How the hell you could compare FOSS to proprietary when there are so many areas where FOSS doesn't even compete is beyond me. It's like saying "Hey, my FOSS car drives as good as your proprietary car! Of course it lacks reverse, any forward gear over first and brakes, but those aren't 'defects', they're just completely fucking missing features!"
Let's compare the following FOSS vs. proprietary programs:
- A not-shit image-editing program.
- A video program that doesn't have a case of HIV.
- A filesystem with modern features like usable block-level snapshots. (ZFS doesn't count--it was created as a proprietary software product by a proprietary company. It wasn't developed by the FOSS community. Just like Java doesn't count--it was open sourced after a decade of proprietary development. FOSStards trying to claim credit for that is like me trying to claim credit for bringing a chick to orgasm after the vibrator warmed her up for 20 minutes and all I did is pant and groan for 8 seconds before collapsing and falling asleep.)
- A desktop environment that isn't an ineptly-assembled piece of shit.
Once FOSS creates these things, then we can actually look at shit like defects-per-line. Until then, I don't give a fuck if Go has fewer bugs per-line than Photoshop, because Go is fucking useless.
Seriously, how fucking dumb do you have to be to not get this? Hello?? Do you even understand the concept of trying to compare things that are actually comparable??
-
@morbiuswilters said:
3) A filesystem with modern features like usable block-level snapshots.
Btrfs
4) A desktop environment that isn't an ineptly-assembled piece of shit.
KDE
1) A not-shit image-editing program.
2) A video program that doesn't have a case of HIV.Fuck that. I browse the web from my VT220.
...bringing a chick to orgasm after the vibrator warmed her up for 20 minutes and all I did is pant and groan for 8 seconds before collapsing and falling asleep...
I see that you know my wife!
-
KDE is terrible.
-
@lucas said:
KDE is terrible.
Have you actually used it, or are you just saying that because you heard other people who didn't use KDE say that?
-
@morbiuswilters said:
Two, quality is more than just a measure of defects-per-line. How the hell you could compare FOSS to proprietary when there are so many areas where FOSS doesn't even compete is beyond me. It's like saying "Hey, my FOSS car drives as good as your proprietary car! Of course it lacks reverse, any forward gear over first and brakes, but those aren't 'defects', they're just completely fucking missing features!"
Yes, the title should read "PR Report Finds Open Source Software Sucks Less In Coverity Scan than Industry Average". But after all, everybody blatantly lies in the article titles. And I am not really going into discussion about the usefulness of the coverity scan metric.
But I would suspect that the "Industry Average" is skewed. Only companies that actually try to suck less licence Coverity and run these types of tests, whereas the open source projects are just grabbed from the repo or scanned by the one single person who cares.
PS: Googling found this as first result: http://scan.coverity.com/ (It works! This is the default web page for this server.)
-
@spamcourt said:
Btrfs
Not stable. Isn't even close to being ready for production use. Will probably limp along for another 5 years before finally being killed.
@spamcourt said:
KDE
HA HA HA.
-
@Ben L. said:
Have you actually used it, or are you just saying that because you heard other people who didn't use KDE say that?
Yes.
It is like most open source DEs everything is there but the polish is severely lacking. Also every single bloody thing in the UI has slightly out of proportion dimensions, the default which I find infuriating. Transparency is stuck all over the place because it is flashy while decent UI stuff is overlooked.
Also while MATE and XFCE work fairly well on my ageing Dell Notebook, the newer KDE and GNOME work like utter dog shit thanks to the compositing, however similar effects work fine in Windows Vista and 7.
Open source is really really good at creating things like libraries, APIs, tools etc and utterly sucks at creating anything compelling that rivals commercial software (with the exception of the odd success story).
-
@Ben L. said:
@lucas said:
Does it matter? He's right either way.KDE is terrible.
Have you actually used it, or are you just saying that because you heard other people who didn't use KDE say that?
-
@lucas said:
...and utterly sucks at creating anything compelling that rivals commercial software (with the exception of the odd success story).
Yeah, and take a look at Firefox if you want to see how long they can actually maintain that success.
-
@veggen said:
An article I stumbled upon claims FOSS has better code quality than industry average.
Who gives a shit about code quality when they have terrible product quality?
-
@blakeyrat said:
@veggen said:
An article I stumbled upon claims FOSS has better code quality than industry average.
Who gives a shit about code quality when they have terrible product quality?
A guy who is ugly, fat, has questionable hygiene and a tiny wang is going to play up his encyclopedic knowledge of Hobbits as much as possible.
-
@morbiuswilters said:
@blakeyrat said:@veggen said:
A guy who is ugly, fat, has questionable hygiene and a tiny wang is going to play up his encyclopedia knowledge of Hobbits as much as possible.An article I stumbled upon claims FOSS has better code quality than industry average.
Who gives a shit about code quality when they have terrible product quality?The thing is even by the questionable measurements they used it was FOSS with .44 to other stuff with .98 but only in half to one mil LOC things. In 1+ mil FOSS had .75 and other stuff had .66; so using your example the dude has encyclopedic knowledge of Hobbits but knows less about everything else from Tolkien's work than some that watched the movies only. Not only are the bragging about something that doesn't matter (at least to most people) but even there if you look closely they are worse.
-
@lucas said:
Open source is really really good at creating things like libraries, APIs, tools etc and utterly sucks at creating anything compelling that rivals commercial software (with the exception of the odd success story).
Another way of putting this would be that open source is really bad at designing UIs, for the simple reason that open source UIs are being designed by programmers while commercial products actually have workers who's job description is "UI Designer", acceptance testing, etc. Especially when the UI is a GUI, open source is gaggingly bad at it. If this is your primary definition of product quality, open source will never win because UI design experts don't contribute their free time to open source the way programmers do.
-
@locallunatic said:
So it doesn't really say what you are using to bait people.
Quite possibly, but it still works.
-
@jello said:
If this is your primary definition of product quality, open source will never win because UI design experts don't contribute their free time to open source the way programmers do.
UI is an important part of the whole. You can make the most ass-kickingest software ever, but the if the UI is unusable then it's basically a shit product.
Of course, UI isn't FOSS's only problem. Not by a long shot. There's the ridiculous fragmentation and forking, resulting in 20 mediocre versions of Foo instead of one good one. For example, how many FOSS languages are there? Now how many have mature IDEs, debuggers, libraries, support, etc.? Even big languages like Python tend to fall short on that. (Then add in the fact that the Python devs insist on releasing a completely new, not-backwards-compatible version of the language every few years, but they can't even be arsed to fix the fucking GIL.)
There's the attitude of so many FOSS projects which is "You will use it the way we tell you and if you complain we will just reply with some snotty bullshit like 'You have source, just change it yourself.'"
Then there are the fucking crazy ideological kicks FOSS projects will go on, completely fucking over users to achieve some kind of OCD "purity", whether with regard to some arbitrary ideals chosen by the project, or just unyielding adherence to a spec, no matter how broken it is, and no matter how many existing users it fucks over.
The real problem of FOSS is that there's little incentive to improve, hardly any sense of competition. It's just a bunch of anti-social nerds pissing away their free time, which means they have no reason to give a shit about their users. In fact, they tend to develop this bizarre, put-upon attitude of "You aren't appreciating enough of what I am doing for free!", where the costs of FOSS aren't in dollars and cents, but in not being able to criticize anything these prima donnas do. Of course, nobody made them donate their time for free, but they start assuming people owe them something since they're doing some ill-defined "good" for the world.
-
@veggen said:
@locallunatic said:
So it doesn't really say what you are using to bait people.
Quite possibly, but it still works.Agreed. I enjoy a good FOSS bashing. Thanks!
-
@morbiuswilters said:
Let's compare the following FOSS vs. proprietary programs:
- A not-shit image-editing program.
- A video program that doesn't have a case of HIV.
- A filesystem with modern features like usable block-level snapshots. (ZFS doesn't count--it was created as a proprietary software product by a proprietary company. It wasn't developed by the FOSS community. Just like Java doesn't count--it was open sourced after a decade of proprietary development. FOSStards trying to claim credit for that is like me trying to claim credit for bringing a chick to orgasm after the vibrator warmed her up for 20 minutes and all I did is pant and groan for 8 seconds before collapsing and falling asleep.)
- A desktop environment that isn't an ineptly-assembled piece of shit.
1, and 4 involve GUIs, so I would agree.
2. usually involves needing to reverse engineer a proprietary codec (and possibly also encryption scheme) somewhere where said reverse engineering is legal, which severely limits participation.
For 3, what about LVM? It had block level snapshots before ZFS. Strictly speaking it isn't a filesystem, but there are a number of features that people get excited about in ZFS that LVM had first.
-
@jello said:
2. usually involves needing to reverse engineer a proprietary codec (and possibly also encryption scheme) somewhere where said reverse engineering is legal, which severely limits participation.
My Linux system has tons of (probably illegal) codecs on it. When I said "video program" I meant "video editing program", like Final Cut Pro.
@jello said:
For 3, what about LVM? It had block level snapshots before ZFS. Strictly speaking it isn't a filesystem, but there are a number of features that people get excited about in ZFS that LVM had first.
And other systems had them before LVM. But, anyway, LVM snaphots are not "usable". For one, take 3 snapshots and watch your write performance drop 90%. For two, there's no way to export a binary diff between snapshots for backing up off-site (or even onto another on-site server.) So either you have to mount the snapshot and tar up all of the data on it for each backup (ha!) or you just have to keep your only backup on the same machine you're trying to back up (also ha!) This is why people spend so much on high-end SANs and NASes: they at least have these features built-in.
-
@jello said:
@lucas said:
Open source is really really good at creating things like libraries, APIs, tools etc and utterly sucks at creating anything compelling that rivals commercial software (with the exception of the odd success story).
Another way of putting this would be that open source is really bad at designing UIs, for the simple reason that open source UIs are being designed by programmers while commercial products actually have workers who's job description is "UI Designer", acceptance testing, etc. Especially when the UI is a GUI, open source is gaggingly bad at it. If this is your primary definition of product quality, open source will never win because UI design experts don't contribute their free time to open source the way programmers do.
What other definition could you possibly use for a product that is intended to be used directly (as opposed to a library or plugin etc.). A product is only as good (to the actual users) as its interface allows it to be. If the users can't manage to do what they want with it, or if it's just really annoying to do, then the product has failed.
-
@jello said:
Another way of putting this would be that open source is really bad at designing UIs
I was quite disappointed that this underline was not a link to an horrible open source UI example. You're such a tease, like those slot machines that slow down when you got two 7s and you only need a 3rd one to win the jackpot.
-
@Ronald said:
You're such a tease, like those slot machines that slow down when you got two 7s and you only need a 3rd one to win the jackpot.
Or those hookers who slow down when you're only 7s from your own jackpot. And then they die.
-
@morbiuswilters said:
@Ronald said:
You're such a tease, like those slot machines that slow down when you got two 7s and you only need a 3rd one to win the jackpot.
Or those hookers who slow down when you're only 7s from your own jackpot. And then they die.
Hookers are so mainstream, you should try both ends of the street spectrum instead (runaway kids and hobo amputees).
-
@morbiuswilters said:
@spamcourt said:
Btrfs
Not stable. Isn't even close to being ready for production use.
SUSE and (urp) Oracle disagree: https://btrfs.wiki.kernel.org/index.php/Getting_started#Enterprise_distributions_with_btrfs_support
I've been using btrfs on my machines for nearly three years now. In all that time it hasn't eaten my data, and functioning cow snapshots have made hot backups super-easy.For two, there's no way [for LVM to] to export a binary diff between snapshots for backing up off-site
Btrfs does this: https://btrfs.wiki.kernel.org/index.php/Incremental_Backup
And then you lie and tell the paramedics she died right after you finished.
Oh, shit, that's because they can't really tell the TOD, amirong?
-
@spamcourt said:
I've been using btrfs on my machines for nearly three years now.
Production machines? What kind of load? I tried btrfs not long ago on a freakin' workstation and it was a disaster. I was getting all kinds of warnings and shit in dmesg. I did not feel confident storing my data on it, so I went with ZFS. (And it's not like installing ZFS isn't a fucking hassle, but it feels much more stable and safe.)
@spamcourt said:
Btrfs does this: https://btrfs.wiki.kernel.org/index.php/Incremental_Backup
When I checked a few months ago, btrfs didn't have a "send". I've been waiting for them to implement this since 2008. Still, I feel more comfortable using ZFS, which has had send functionality for quite some time.
@spamcourt said:
Oh, shit, that's because they can't really tell the TOD, amirong?
Not to the nearest 7 seconds, which is all the time I need.
-
@morbiuswilters said:
Production machines? What kind of load?
Nah. I use it on a well used but mostly read-only file server, my C++ dev machine at work, and my Gentoo-Linux-running laptop. (That laptop has suffered innumerable sudden power losses (related to my dumb ass forgetting to plug it in) and several kernel panics (related to gentoo-hardened, not to btrfs), but btrfs has never eaten my data.)
I tried btrfs not long ago on a freakin' workstation and it was a disaster. I was getting all kinds of warnings and shit in dmesg.
Were you using Ubuntu? I never see shit in my dmesg on my Gentoo boxes, but my Ubuntu machine often screams about slowpath_inside_fastpath stuff. Frankly, I think the Ubuntu kids aren't very good at choosing kernels.
Still, I feel more comfortable using ZFS...
Fair enough.
-
@morbiuswilters said:
@jello said:
Almost there, but not quite, is [url="http://www.videolan.org/vlmc/"]VLMC[/url], from the VideoLAN people.2. usually involves needing to reverse engineer a proprietary codec (and possibly also encryption scheme) somewhere where said reverse engineering is legal, which severely limits participation.
My Linux system has tons of (probably illegal) codecs on it. When I said "video program" I meant "video editing program", like Final Cut Pro.
-
@DescentJS said:
What other definition could you possibly use for a product that is intended to be used directly (as opposed to a library or plugin etc.). A product is only as good (to the actual users) as its interface allows it to be. If the users can't manage to do what they want with it, or if it's just really annoying to do, then the product has failed.
QFT
-
The only reasonable response is that the difference of quality between two proprietary product (even by the same firm) and the difference of quality between two open products (even by the same people) is higher than the average difference between FOSS and proprietary. So condemning a product because he is proprietary or open is exactly as stupid as condemning someone because he is black or white. Also, I suppose that's for the troll, but a lot of FOSS code is done by firms.
-
@DescentJS said:
What other definition could you possibly use for a product that is intended to be used directly (as opposed to a library or plugin etc.). A product is only as good (to the actual users) as its interface allows it to be. If the users can't manage to do what they want with it, or if it's just really annoying to do, then the product has failed.
If you have a program that's super duper easy to use, and you can put in all of your data, etc, but it ends up with the wrong answer, is that "better software" than the program which is a bit clunky and annoying to use, but gives you perfect output? What if the clunky software finishes its task in a minute but the other takes a day? Despite what you (or blakeyrat) think, that's not me excusing a bad interface, but exposing your lack of imagination.
-
@boomzilla said:
@DescentJS said:
What other definition could you possibly use for a product that is intended to be used directly (as opposed to a library or plugin etc.). A product is only as good (to the actual users) as its interface allows it to be. If the users can't manage to do what they want with it, or if it's just really annoying to do, then the product has failed.
If you have a program that's super duper easy to use, and you can put in all of your data, etc, but it ends up with the wrong answer, is that "better software" than the program which is a bit clunky and annoying to use, but gives you perfect output? What if the clunky software finishes its task in a minute but the other takes a day? Despite what you (or blakeyrat) think, that's not me excusing a bad interface, but exposing your lack of imagination.
Plus a lot of commerical, propretiary software is a lot shitter than FOSS and then you lost because you can't fix it yourself. Of course stuff like MS Office, Adobe Photoshop etc. are better than the FOSS versions but then the budgets behind them are gigantic. If you look at less common proprietary stuff that don't have a 1 Billion $ budget, there is just a lot of crap around.
-
"If the users can't manage to do what they want with it"
I would imagine this would include pulling out correct data.
-
@beginner_ said:
Plus a lot of commerical, propretiary software is a lot shitter than FOSS and then you lost because you can't fix it yourself. Of course stuff like MS Office, Adobe Photoshop etc. are better than the FOSS versions but then the budgets behind them are gigantic. If you look at less common proprietary stuff that don't have a 1 Billion $ budget, there is just a lot of crap around.
Yes, I bring this up all the time whenever stuff like MS Office is trotted out to prove that proprietary is always better than FOSS. I think this study supports my thesis in that the Coverity scans were seeing stuff that may never be released. And even for products that are released, the source code isn't, which is what TFA is about. It's quite possible, of course, to have code full of WTFs that ends up being a reasonable end result.
-
@boomzilla said:
Yes, I bring this up all the time whenever stuff like MS Office is trotted out to prove that proprietary is always better than FOSS.
As a side issue, one may wonder why it takes these gigantic budgets to create good big software. Maybe that's just how it works? Foobar2000 is closed-source, 100% fr€€, and good software, but it's also small software. It's a media player with maybe 3 or 4 people behind it.
Take a bigger project, like Paint.Net, with the same order of magnitude of resources behind it, and suddenly the cracks starts showing.
So my hypothesis is that every equal piece of added functionality or polish takes twice as much resource, as software progresses and develops.
-
@dhromed said:
So my hypothesis is that every equal piece of added functionality or polish takes twice as much resource, as software progresses and develops.
Yes, I think there's definitely merit in that. The overall complexity and scope have a major impact. And that's before you even take the people aspect into account.
-
@Ben L. said:
@lucas said:
KDE is terrible.
Have you actually used it, or are you just saying that because you heard other people who didn't use KDE say that?
No, he's right, in a Churchillian way. It's the most terrible desktop out there, except for all the others.
-
@drurowin said:
Almost there, but not quite, is VLMC, from the VideoLAN people.
The only reason VLC's so good is because VideoLan is in some third-world country with no patent laws. Imagine the awesome technology we'd have if a country like the US abolished patents entirely.
-
VLC is shit. The ONLY thing it can do competently is play videos, and even that it frequently fucks up. (I particularly love when it opens a movie off-screen because it saved its maximized window size as its normal window size.) The stream/convert feature, you know, *the thing it was designed to do*, is a complete mess.
I really feel like I'm in a parallel universe when people can say "VLC is so good". I mean, what the hell is BAD to you!?
-
I prefer Media Player Classic to VLC. It's GPL, but Windows-only.
Funny, the UI is actually based on the old Windows Media Player. I'd use WMP but it's slower than ass, crashes regularly, and has corrupted my media library for the last time.
-
@blakeyrat said:
VLC is shit. The ONLY thing it can do competently is play videos, and even that it frequently fucks up. (I particularly love when it opens a movie off-screen because it saved its maximized window size as its normal window size.) The stream/convert feature, you know, the thing it was designed to do, is a complete mess.
I've never tried to stream anything with it, but I've never had it open up off screen. Generally, it just starts playing and I watch.
@blakeyrat said:
I really feel like I'm in a parallel universe when people can say "VLC is so good". I mean, what the hell is BAD to you!?
The worst is MS MediaPlayer, which has controls that you can't figure out, and in whose media library you can never find what you're looking for. At least, it used to be. I'll allow that perhaps things have improved since I entered my time pod, but given the UI disaster that is Windows 8, I'm not giving benefit of the doubt.
-
@boomzilla said:
The worst is MS MediaPlayer, which has controls that you can't figure out, and in whose media library you can never find what you're looking for. At least, it used to be. I'll allow that perhaps things have improved since I entered my time pod, but given the UI disaster that is Windows 8, I'm not giving benefit of the doubt.
Media Player in Windows 8 is exactly the same as it was in Windows 7: godawful.For actual music management and music playback, I use the Zune desktop software, because it's not totally awful and it syncs up with my Zune and my phone (yes I have a Zune, the 32GB HD was cheaper than an 8GB iPod Touch. Also I hate iTunes.) and use VLC or Media Player Classic for playing videos. Media Player is promptly forgotten about and only used when I accidentally click on it.
-
@beginner_ said:
@boomzilla said:
@DescentJS said:
What other definition could you possibly use for a product that is intended to be used directly (as opposed to a library or plugin etc.). A product is only as good (to the actual users) as its interface allows it to be. If the users can't manage to do what they want with it, or if it's just really annoying to do, then the product has failed.
If you have a program that's super duper easy to use, and you can put in all of your data, etc, but it ends up with the wrong answer, is that "better software" than the program which is a bit clunky and annoying to use, but gives you perfect output? What if the clunky software finishes its task in a minute but the other takes a day? Despite what you (or blakeyrat) think, that's not me excusing a bad interface, but exposing your lack of imagination.
Plus a lot of commerical, propretiary software is a lot shitter than FOSS and then you lost because you can't fix it yourself. Of course stuff like MS Office, Adobe Photoshop etc. are better than the FOSS versions but then the budgets behind them are gigantic. If you look at less common proprietary stuff that don't have a 1 Billion $ budget, there is just a lot of crap around.
Exactly. Sturgeon's Law applies here, folks.
-
Yes, Windows Media Player is also awful.
Software quality isn't a teeter-totter, people. VLC being bad doesn't mean WMP is good.
-
@boomzilla said:
Yes, I bring this up all the time whenever stuff like MS Office is trotted out to prove that proprietary is always better than FOSS.
Speaking of Office, what the hell did they do with 2013? They offer 3 shades of white as "colour schemes", require a Windows Live account for activation, and somehow managed to break the caret completely (by default it does this distracting smooth motion, but even if you disable that it lags behind when typing, just not as much). AND THE RIBBON SCREAMS AT YOU.
-
I like the smooth typing thing.
-
@blakeyrat said:
Software quality isn't a teeter-totter, people. VLC being bad doesn't mean WMP is good.
Yes. As someone who likes VLC, I was just answering your question. At least we can agree on something, though.
-
@ender said:
They offer 3 shades of white as "colour schemes"
The default white is awful (someone should be fired for that), however I find the light grey setting to be fine. The dark grey setting is just weird.@ender said:require a Windows Live account for activation
Less issues with activation - that being said I was under the impression that was only with the subscription versions needed the live account. Retail boxes have a product key as normal, and my Dreamspark copy of Visio had an option for a key, which I used because I had one.@ender said:nd somehow managed to break the caret completely (by default it does this distracting smooth motion, but even if you disable that it lags behind when typing, just not as much)
I like the smooth motion. There's something nice about it I can't really describe beyond liking it.@ender said:AND THE RIBBON SCREAMS AT YOU.
Maybe it's just me with my tiny high resolution screen at 100% DPI, or maybe it's because I'm not easily distracted by what's outside of the document I'm working on, but I really do not notice the all caps at all. They're just there. That being said, if you find it irksome you can force it to whatever case you like simply by going into the Customize Ribbon screen, and renaming each label by putting a space on the start or end of their name to get the desired effect, or give them entirely different names.
