Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.advocacy > #240400 > unrolled thread

This years Coverity scan

Back to article view | Back to comp.os.linux.advocacy

Contents

  This years Coverity scan Peter Köhlmann <peter-koehlmann@t-online.de> - 2014-04-26 23:42 +0200
    Re: This years Coverity scan Snit <usenet@gallopinginsanity.com> - 2014-04-26 17:24 -0700
    Re: This years Coverity scan GreyCloud <mist@cumulus.com> - 2014-04-26 19:32 -0600
      Re: This years Coverity scan Snit <usenet@gallopinginsanity.com> - 2014-04-26 19:06 -0700
        Errors per line of code ? !  Jeff-Relf.Me <@.> - 2014-04-26 19:23 -0700
          Re: Errors per line of code ? ! Steve Carroll <fretwizzen@gmail.com> - 2014-04-26 20:25 -0700
            @snit should publish his monthly rate.  Jeff-Relf.Me <@.> - 2014-04-26 21:34 -0700
              Re: @snit should publish his monthly rate.  Snit <usenet@gallopinginsanity.com> - 2014-04-26 21:40 -0700
                @snit should publish his monthly rate.  Jeff-Relf.Me <@.> - 2014-04-26 22:27 -0700
                  Re: @snit should publish his monthly rate.  Snit <usenet@gallopinginsanity.com> - 2014-04-26 23:16 -0700
                    @snit is bat·shit crazy.  Jeff-Relf.Me <@.> - 2014-04-27 00:51 -0700
                      Re: @snit is bat·shit crazy.  Snit <usenet@gallopinginsanity.com> - 2014-04-27 09:20 -0700
              Re: @snit should publish his monthly rate. Steve Carroll <fretwizzer@gmail.com> - 2014-04-27 11:09 -0700
    Re: This years Coverity scan Silver Slimer <slvrslmr@lv.c> - 2014-04-27 11:00 -0400
      Re: This years Coverity scan 	flatfish+++ <phlatphish@yahoo.com> - 2014-04-27 11:38 -0400
        Re: This years Coverity scan Silver Slimer <slvrslmr@lv.c> - 2014-04-27 12:41 -0400
          Re: This years Coverity scan GreyCloud <mist@cumulus.com> - 2014-04-27 11:20 -0600
            Re: This years Coverity scan Silver Slimer <slvrslmr@lv.c> - 2014-04-27 15:13 -0400
      Re: This years Coverity scan Snit <usenet@gallopinginsanity.com> - 2014-04-27 09:31 -0700
    Re: This years Coverity scan DFS <nospam@dfs.com> - 2014-04-28 19:30 -0400
      Re: This years Coverity scan Peter Köhlmann <peter-koehlmann@t-online.de> - 2014-04-29 01:42 +0200
        Re: This years Coverity scan Chris Ahlstrom <OFeem1987@teleworm.us> - 2014-04-28 20:09 -0400
          Re: This years Coverity scan DFS <nospam@dfs.com> - 2014-04-29 19:50 -0400
        Re: This years Coverity scan William Poaster <wp@induh-vidual.net> - 2014-04-29 09:18 +0100
        Re: This years Coverity scan chrisv <chrisv@nospam.invalid> - 2014-04-29 07:00 -0500
          Re: This years Coverity scan 	flatfish+++ <phlatphish@yahoo.com> - 2014-04-29 08:13 -0400
            Re: This years Coverity scan Hadron<hadronquark@gmail.com> - 2014-04-29 14:32 +0200

Page 1 of 2  [1] 2  Next page →

#240400 — This years Coverity scan

Coverity scan has again published their report about error density.

This time OSS code was in *all* categories better than CSS code (last year 
it was better in all categories with code below 1.000.000 lines).
Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS

The result this year was 0.59 per kloc for OSS, 0.72 for CSS

The linux code they scanned ( 8.5 million lines of code) resulted in 0.61 
per kloc

So far for the repeated claims of "hobby ware" made by some dimbulbs

The code quality of OSS software was *better* acrosss the board compared to 
CSS code

[toc] | [next] | [standalone]

#240416

On 4/26/14, 2:42 PM, in article ljh98o$tra$1@dont-email.me, "Peter Köhlmann"
<peter-koehlmann@t-online.de> wrote:

> Coverity scan has again published their report about error density.
> 
> This time OSS code was in *all* categories better than CSS code (last year
> it was better in all categories with code below 1.000.000 lines).
> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS
> 
> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
> 
> The linux code they scanned ( 8.5 million lines of code) resulted in 0.61
> per kloc
> 
> So far for the repeated claims of "hobby ware" made by some dimbulbs
> 
> The code quality of OSS software was *better* acrosss the board compared to
> CSS code
> 
Hey, so why not use your Voldermort software and show how Windows or OS X
has 1/10th the errors KDE has.

I look forward to it!


-- 
* Mint MATE: Trash, Panel, Menu:    <http://youtu.be/C0y74FIf7uE>
* Mint KDE bugs or Easter eggs?     <http://youtu.be/CU-whJQvtfA>
* Mint KDE working with folders:    <http://youtu.be/7C9nvniOoE0>
* Mint KDE creating files:          <http://youtu.be/N7-fZJaJUv8>
* Mint KDE help:                    <http://youtu.be/3ikizUd3sa8>
* Mint KDE general navigation:      <http://youtu.be/t9y14yZtQuI>
* Mint / LibreOffice:               <http://youtu.be/USU9iqppfto>
* Win 8 / Word 2013:                <http://youtu.be/z8gUL2TCcV4>
* Mavericks / Pages 5.1:            <http://youtu.be/D3BPWANQoIk>

[toc] | [prev] | [next] | [standalone]

#240435

On 4/26/2014 3:42 PM, Peter Köhlmann wrote:
> Coverity scan has again published their report about error density.
> 
> This time OSS code was in *all* categories better than CSS code (last year 
> it was better in all categories with code below 1.000.000 lines).
> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS
> 
> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
> 
> The linux code they scanned ( 8.5 million lines of code) resulted in 0.61 
> per kloc
> 
> So far for the repeated claims of "hobby ware" made by some dimbulbs
> 
> The code quality of OSS software was *better* acrosss the board compared to 
> CSS code
> 
Where is this report?

[toc] | [prev] | [next] | [standalone]

#240437

On 4/26/14, 6:32 PM, in article
rYGdnWdV59u4w8HOnZ2dnUVZ_gqdnZ2d@bresnan.com, "GreyCloud" <mist@cumulus.com>
wrote:

> On 4/26/2014 3:42 PM, Peter Köhlmann wrote:
>> Coverity scan has again published their report about error density.
>> 
>> This time OSS code was in *all* categories better than CSS code (last year
>> it was better in all categories with code below 1.000.000 lines).
>> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS
>> 
>> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
>> 
>> The linux code they scanned ( 8.5 million lines of code) resulted in 0.61
>> per kloc
>> 
>> So far for the repeated claims of "hobby ware" made by some dimbulbs
>> 
>> The code quality of OSS software was *better* acrosss the board compared to
>> CSS code
>> 
> Where is this report?
> 
<http://blog.coverity.com/2014/04/15/hot-presses-coverity-scan-report-2013>


-- 
* Mint MATE: Trash, Panel, Menu:    <http://youtu.be/C0y74FIf7uE>
* Mint KDE bugs or Easter eggs?     <http://youtu.be/CU-whJQvtfA>
* Mint KDE working with folders:    <http://youtu.be/7C9nvniOoE0>
* Mint KDE creating files:          <http://youtu.be/N7-fZJaJUv8>
* Mint KDE help:                    <http://youtu.be/3ikizUd3sa8>
* Mint KDE general navigation:      <http://youtu.be/t9y14yZtQuI>
* Mint / LibreOffice:               <http://youtu.be/USU9iqppfto>
* Win 8 / Word 2013:                <http://youtu.be/z8gUL2TCcV4>
* Mavericks / Pages 5.1:            <http://youtu.be/D3BPWANQoIk>

[toc] | [prev] | [next] | [standalone]

#240440 — Errors per line of code ? ! 

<PRE Style='White-Space: PRE !important; Font-Size: 18px !important;Font-Family: monospace !important;'> 
Errors per line of code ? !

My C/C++, Visual Studio 2013 code uses 150 columns.
ScreenShot: http://Jeff-Relf.Me/SourceCode.PNG
-----
  http://Jeff-Relf.Me/DirectX.CPP  http://Jeff-Relf.Me/Glyph.JPG
  http://Jeff-Relf.Me/X.CPP  http://Jeff-Relf.Me/X.PNG
  http://Jeff-Relf.Me/X.HTM  http://Jeff-Relf.Me/X.EXE

[toc] | [prev] | [next] | [standalone]

#240444 — Re: Errors per line of code ? !

On Sunday, April 27, 2014 2:23:45 AM UTC, Jeff-Relf.Me wrote:
>  
> Errors per line of code ? !
> 
> My C/C++, Visual Studio 2013 code uses 150 columns.
> ScreenShot: http://Jeff-Relf.Me/SourceCode.PNG
> -----
>   http://Jeff-Relf.Me/DirectX.CPP  http://Jeff-Relf.Me/Glyph.JPG
>   http://Jeff-Relf.Me/X.CPP  http://Jeff-Relf.Me/X.PNG
>   http://Jeff-Relf.Me/X.HTM  http://Jeff-Relf.Me/X.EXE

How many errors per line of code do you see here from Michael Glasser of Prescott Computer Guy (http://prescottcomputerguy.com)

http://prescottcomputerguy.com/web_design.html

Even the image he uses for his "techniques" has at least one error in the code coloring. Quick, Michael Glasser of Prescott, Arizona, fix it! You know you will. You read my every post.

[toc] | [prev] | [next] | [standalone]

#240449 — @snit should publish his monthly rate. 

<PRE Style='White-Space: PRE !important; Font-Size: 18px !important;Font-Family: monospace !important;'> 
Steve⋅Carroll,  You asked me: 
> How many errors per line of code do you see here 
> from Michael Glasser of Prescott Computer Guy. 
> http://prescottcomputerguy.com

He should use "Quick Sort"; "Bubble Sort" is bad news.
https://en.wikipedia.org/wiki/Quicksort

He should publish his monthly rate,
-not- his "per minute" rate... it looks bad.
Do his customers dump him after a few minutes ? !

I've held my current job for 21 years.

I wouldn't use any of the video players he mentioned.
I use Zoom Player, but MPC-HC is also good.

[toc] | [prev] | [next] | [standalone]

#240450 — Re: @snit should publish his monthly rate.

On 4/26/14, 9:34 PM, in article Jeff-Relf.Me@Apr.26{21.34.Seattle.2014},
"Jeff-Relf.Me" <@.> wrote:

> ?
> Steve・Carroll,  You asked me:
>> How many errors per line of code do you see here
>> from Michael Glasser of Prescott Computer Guy.
>> http://prescottcomputerguy.com
> 
> He should use "Quick Sort"; "Bubble Sort" is bad news.
> https://en.wikipedia.org/wiki/Quicksort

I have used both.

> He should publish his monthly rate,
> -not- his "per minute" rate... it looks bad.
> Do his customers dump him after a few minutes ? !

How I charge is irrelevant to COLA.
 
> I've held my current job for 21 years.

Mazel Tov! 
 
> I wouldn't use any of the video players he mentioned.

I did not mention any.

> I use Zoom Player, but MPC-HC is also good.

Mazel Tov! 



-- 
* Mint MATE: Trash, Panel, Menu:    <http://youtu.be/C0y74FIf7uE>
* Mint KDE bugs or Easter eggs?     <http://youtu.be/CU-whJQvtfA>
* Mint KDE working with folders:    <http://youtu.be/7C9nvniOoE0>
* Mint KDE creating files:          <http://youtu.be/N7-fZJaJUv8>
* Mint KDE help:                    <http://youtu.be/3ikizUd3sa8>
* Mint KDE general navigation:      <http://youtu.be/t9y14yZtQuI>
* Mint / LibreOffice:               <http://youtu.be/USU9iqppfto>
* Win 8 / Word 2013:                <http://youtu.be/z8gUL2TCcV4>
* Mavericks / Pages 5.1:            <http://youtu.be/D3BPWANQoIk>

[toc] | [prev] | [next] | [standalone]

#240452 — @snit should publish his monthly rate. 

<PRE Style='White-Space: PRE !important; Font-Size: 18px !important;Font-Family: monospace !important;'> 
@snit,  You replied to me, saying:
> > I wouldn't use any of the video players he mentioned.
> 
> I did not mention any.

How is it that you don't know 
what's on your own website ? !

You mention them here:
http://prescottcomputerguy.com/tools-media.html

[toc] | [prev] | [next] | [standalone]

#240453 — Re: @snit should publish his monthly rate.

On 4/26/14, 10:27 PM, in article Jeff-Relf.Me@Apr.26{22.27.Seattle.2014},
"Jeff-Relf.Me" <@.> wrote:

>  
> @snit,  You replied to me, saying:
>>> I wouldn't use any of the video players he mentioned.
>> 
>> I did not mention any.
> 
> How is it that you don't know
> what's on your own website ? !

I never said I didn't!

> You mention them here:
> http://prescottcomputerguy.com/tools-media.html

Irrelevant to COLA.



-- 
* Mint MATE: Trash, Panel, Menu:    <http://youtu.be/C0y74FIf7uE>
* Mint KDE bugs or Easter eggs?     <http://youtu.be/CU-whJQvtfA>
* Mint KDE working with folders:    <http://youtu.be/7C9nvniOoE0>
* Mint KDE creating files:          <http://youtu.be/N7-fZJaJUv8>
* Mint KDE help:                    <http://youtu.be/3ikizUd3sa8>
* Mint KDE general navigation:      <http://youtu.be/t9y14yZtQuI>
* Mint / LibreOffice:               <http://youtu.be/USU9iqppfto>
* Win 8 / Word 2013:                <http://youtu.be/z8gUL2TCcV4>
* Mavericks / Pages 5.1:            <http://youtu.be/D3BPWANQoIk>

[toc] | [prev] | [next] | [standalone]

#240454 — @snit is bat·shit crazy. 

<PRE Style='White-Space: PRE !important; Font-Size: 18px !important;Font-Family: monospace !important;'> 
You are bat·shit crazy, @snit.

I can't think of anyone here, on Usenet,
that is as insane as you.

Mental patients, in the loony bin,
make more sense than you.

[toc] | [prev] | [next] | [standalone]

#240471 — Re: @snit is bat·shit crazy. 

On 4/27/14, 12:51 AM, in article Jeff-Relf.Me@Apr.27{0.51.Seattle.2014},
"Jeff-Relf.Me" <@.> wrote:

>  
> You are bat·shit crazy, @snit.
> 
> I can't think of anyone here, on Usenet,
> that is as insane as you.
> 
> Mental patients, in the loony bin,
> make more sense than you.

Quote what you did not understand and I will try to rephrase it using
smaller words or at a lower level for you. Then maybe it will make sense to
you. 


-- 
* Mint MATE: Trash, Panel, Menu:    <http://youtu.be/C0y74FIf7uE>
* Mint KDE bugs or Easter eggs?     <http://youtu.be/CU-whJQvtfA>
* Mint KDE working with folders:    <http://youtu.be/7C9nvniOoE0>
* Mint KDE creating files:          <http://youtu.be/N7-fZJaJUv8>
* Mint KDE help:                    <http://youtu.be/3ikizUd3sa8>
* Mint KDE general navigation:      <http://youtu.be/t9y14yZtQuI>
* Mint / LibreOffice:               <http://youtu.be/USU9iqppfto>
* Win 8 / Word 2013:                <http://youtu.be/z8gUL2TCcV4>
* Mavericks / Pages 5.1:            <http://youtu.be/D3BPWANQoIk>

[toc] | [prev] | [next] | [standalone]

#240487 — Re: @snit should publish his monthly rate.

On Saturday, April 26, 2014 10:34:29 PM UTC-6, Jeff-Relf.Me wrote:
>  
> Steve⋅Carroll,  You asked me: 

Nah... I didn't ask you anything, the person forging my name may have but not I. These forgeries are, most likely, penned by Snit playing the victim again. My gmail address is fretwizzer, not fretwizzen (like the forgeries often use, though, there have been others). Here's a post from me dating back to Apr. of 2010, the forgeries using fretwizzen began sometime in Oct. of 2012.

https://groups.google.com/forum/#!original/comp.sys.mac.advocacy/GPOCFLSb6Ks/68O0KHFwgEgJ

As you said, Snit is "bat·shit crazy" and prone to  creating weird 'victim' scenarios to drum up sympathy to cover the scent of his constant trolling.

[toc] | [prev] | [next] | [standalone]

#240463

On 2014-04-26 5:42 PM, Peter Köhlmann wrote:
> Coverity scan has again published their report about error density.
>
> This time OSS code was in *all* categories better than CSS code (last year
> it was better in all categories with code below 1.000.000 lines).
> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS
>
> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
>
> The linux code they scanned ( 8.5 million lines of code) resulted in 0.61
> per kloc
>
> So far for the repeated claims of "hobby ware" made by some dimbulbs
>
> The code quality of OSS software was *better* acrosss the board compared to
> CSS code

That's fine. However, explain why updates basically break a system, why 
LibreOffice is devoid of a strong number of features standard to MS 
Office since the 90's and why open-source drivers perform so terribly in 
comparison to the closed-source ones. The quality of the code doesn't 
matter as much as the end result. If I have to fear updating my system 
because it might not work after a reboot, I'm not better off with 
GNU/Linux. If I can't rely on my office suite to simply rotate an image, 
I'm not better off with an open-source suite and if I have to suffer 
through screen-tearing and fractional performance because I 
philosophically chose a free driver over a closed one, I'm not much 
better off in GNU/Linux.

-- 
Silver Slimer
Wikipedia & OpenMedia Supporter

[toc] | [prev] | [next] | [standalone]

#240469

On Sun, 27 Apr 2014 11:00:19 -0400, Silver Slimer wrote:

> On 2014-04-26 5:42 PM, Peter Köhlmann wrote:
>> Coverity scan has again published their report about error density.
>>
>> This time OSS code was in *all* categories better than CSS code (last year
>> it was better in all categories with code below 1.000.000 lines).
>> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS
>>
>> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
>>
>> The linux code they scanned ( 8.5 million lines of code) resulted in 0.61
>> per kloc
>>
>> So far for the repeated claims of "hobby ware" made by some dimbulbs
>>
>> The code quality of OSS software was *better* acrosss the board compared to
>> CSS code
> 
> That's fine. However, explain why updates basically break a system, why 
> LibreOffice is devoid of a strong number of features standard to MS 
> Office since the 90's and why open-source drivers perform so terribly in 
> comparison to the closed-source ones. The quality of the code doesn't 
> matter as much as the end result. If I have to fear updating my system 
> because it might not work after a reboot, I'm not better off with 
> GNU/Linux. If I can't rely on my office suite to simply rotate an image, 
> I'm not better off with an open-source suite and if I have to suffer 
> through screen-tearing and fractional performance because I 
> philosophically chose a free driver over a closed one, I'm not much 
> better off in GNU/Linux.

This is a case of the operation being a success but the paitent
dying.

Consider two 100 story high rise buildings being constructed right
across the street from each other. One Open Source and the other
closed source.

The Open Source building has every nail, screw and girder in place.
The many eyes have looked at and vetted the entire structure and it
is literally perfect.

The closed source building has some problems. Some of the sheet rock
has nail pops in it. The carpet is crooked and the tiles need to be
re-done due to cracks.

Yet on a beautiful Saturday the closed source building has lines of
tourists waiting to go up to the observation deck to take in the
view.
There is not a single person waiting in line outside the Open Source
building.

Why?

Well, the closed source building has elevators installed and the Open
Source building has stairs so the tourists have to walk 100 floors up
and down to the observation deck.

That's the difference between Open Source and closed source software.
And just 10 minutes using GNU/Linux/FOSS on a desktop system is more
than enough time to prove this.

[toc] | [prev] | [next] | [standalone]

#240480

On 2014-04-27 11:38 AM,  flatfish+++ wrote:
> On Sun, 27 Apr 2014 11:00:19 -0400, Silver Slimer wrote:
>
>> On 2014-04-26 5:42 PM, Peter Köhlmann wrote:
>>> Coverity scan has again published their report about error density.
>>>
>>> This time OSS code was in *all* categories better than CSS code (last year
>>> it was better in all categories with code below 1.000.000 lines).
>>> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS
>>>
>>> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
>>>
>>> The linux code they scanned ( 8.5 million lines of code) resulted in 0.61
>>> per kloc
>>>
>>> So far for the repeated claims of "hobby ware" made by some dimbulbs
>>>
>>> The code quality of OSS software was *better* acrosss the board compared to
>>> CSS code
>>
>> That's fine. However, explain why updates basically break a system, why
>> LibreOffice is devoid of a strong number of features standard to MS
>> Office since the 90's and why open-source drivers perform so terribly in
>> comparison to the closed-source ones. The quality of the code doesn't
>> matter as much as the end result. If I have to fear updating my system
>> because it might not work after a reboot, I'm not better off with
>> GNU/Linux. If I can't rely on my office suite to simply rotate an image,
>> I'm not better off with an open-source suite and if I have to suffer
>> through screen-tearing and fractional performance because I
>> philosophically chose a free driver over a closed one, I'm not much
>> better off in GNU/Linux.
>
> This is a case of the operation being a success but the paitent
> dying.
>
> Consider two 100 story high rise buildings being constructed right
> across the street from each other. One Open Source and the other
> closed source.
>
> The Open Source building has every nail, screw and girder in place.
> The many eyes have looked at and vetted the entire structure and it
> is literally perfect.
>
> The closed source building has some problems. Some of the sheet rock
> has nail pops in it. The carpet is crooked and the tiles need to be
> re-done due to cracks.
>
> Yet on a beautiful Saturday the closed source building has lines of
> tourists waiting to go up to the observation deck to take in the
> view.
> There is not a single person waiting in line outside the Open Source
> building.
>
> Why?
>
> Well, the closed source building has elevators installed and the Open
> Source building has stairs so the tourists have to walk 100 floors up
> and down to the observation deck.
>
> That's the difference between Open Source and closed source software.
> And just 10 minutes using GNU/Linux/FOSS on a desktop system is more
> than enough time to prove this.

I love the examples and that's more or less accurate. I would have added 
that the closed-source building has reflective windows covering each 
floor of the building whereas the open-source one doesn't have windows 
at all.

-- 
Silver Slimer
Wikipedia & OpenMedia Supporter

[toc] | [prev] | [next] | [standalone]

#240482

On 4/27/2014 10:41 AM, Silver Slimer wrote:
> On 2014-04-27 11:38 AM,  flatfish+++ wrote:
>> On Sun, 27 Apr 2014 11:00:19 -0400, Silver Slimer wrote:
>>
>>> On 2014-04-26 5:42 PM, Peter Köhlmann wrote:
>>>> Coverity scan has again published their report about error density.
>>>>
>>>> This time OSS code was in *all* categories better than CSS code
>>>> (last year
>>>> it was better in all categories with code below 1.000.000 lines).
>>>> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS
>>>>
>>>> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
>>>>
>>>> The linux code they scanned ( 8.5 million lines of code) resulted in
>>>> 0.61
>>>> per kloc
>>>>
>>>> So far for the repeated claims of "hobby ware" made by some dimbulbs
>>>>
>>>> The code quality of OSS software was *better* acrosss the board
>>>> compared to
>>>> CSS code
>>>
>>> That's fine. However, explain why updates basically break a system, why
>>> LibreOffice is devoid of a strong number of features standard to MS
>>> Office since the 90's and why open-source drivers perform so terribly in
>>> comparison to the closed-source ones. The quality of the code doesn't
>>> matter as much as the end result. If I have to fear updating my system
>>> because it might not work after a reboot, I'm not better off with
>>> GNU/Linux. If I can't rely on my office suite to simply rotate an image,
>>> I'm not better off with an open-source suite and if I have to suffer
>>> through screen-tearing and fractional performance because I
>>> philosophically chose a free driver over a closed one, I'm not much
>>> better off in GNU/Linux.
>>
>> This is a case of the operation being a success but the paitent
>> dying.
>>
>> Consider two 100 story high rise buildings being constructed right
>> across the street from each other. One Open Source and the other
>> closed source.
>>
>> The Open Source building has every nail, screw and girder in place.
>> The many eyes have looked at and vetted the entire structure and it
>> is literally perfect.
>>
>> The closed source building has some problems. Some of the sheet rock
>> has nail pops in it. The carpet is crooked and the tiles need to be
>> re-done due to cracks.
>>
>> Yet on a beautiful Saturday the closed source building has lines of
>> tourists waiting to go up to the observation deck to take in the
>> view.
>> There is not a single person waiting in line outside the Open Source
>> building.
>>
>> Why?
>>
>> Well, the closed source building has elevators installed and the Open
>> Source building has stairs so the tourists have to walk 100 floors up
>> and down to the observation deck.
>>
>> That's the difference between Open Source and closed source software.
>> And just 10 minutes using GNU/Linux/FOSS on a desktop system is more
>> than enough time to prove this.
> 
> I love the examples and that's more or less accurate. I would have added
> that the closed-source building has reflective windows covering each
> floor of the building whereas the open-source one doesn't have windows
> at all.
> 
The misleading part of the coverity stuff is that they only have about
1100 customers using it, and surely not MS doing it for their own code.
I don't see any MS products from them as well, so I don't see how they
can base their claims on finding more bugs than CSS.  Sounds pretty
suspicious to me.

[toc] | [prev] | [next] | [standalone]

#240495

On 2014-04-27 1:20 PM, GreyCloud wrote:

>> I love the examples and that's more or less accurate. I would have added
>> that the closed-source building has reflective windows covering each
>> floor of the building whereas the open-source one doesn't have windows
>> at all.
>>
> The misleading part of the coverity stuff is that they only have about
> 1100 customers using it, and surely not MS doing it for their own code.
> I don't see any MS products from them as well, so I don't see how they
> can base their claims on finding more bugs than CSS.  Sounds pretty
> suspicious to me.

The way I see it, it's probably worth it to throw the open-source people 
a bone every now and then. I would imagine that the people writing 
software for the GNU/Linux probably take some pride in their work 
whereas the people coding for commercial entities do it for a paycheck. 
One is consumed with trying to get software to run as efficiently as 
possible with optimizations in every direction whereas the other is only 
concerned with getting it out there in time for the deadline. Commercial 
coders take into consideration that people running their software will 
do so on hardware which is state-of-the-art for the release year whereas 
open-sourcers, I assume, want it to work properly regardless of what the 
hardware is.

It's a shame that the end result isn't necessarily stable.

-- 
Silver Slimer
Wikipedia & OpenMedia Supporter

[toc] | [prev] | [next] | [standalone]

#240477

On 4/27/14, 8:00 AM, in article ljj624$hbn$1@dont-email.me, "Silver Slimer"
<slvrslmr@lv.c> wrote:

> On 2014-04-26 5:42 PM, Peter Köhlmann wrote:
>> Coverity scan has again published their report about error density.
>> 
>> This time OSS code was in *all* categories better than CSS code (last year
>> it was better in all categories with code below 1.000.000 lines).
>> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS
>> 
>> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
>> 
>> The linux code they scanned ( 8.5 million lines of code) resulted in 0.61
>> per kloc
>> 
>> So far for the repeated claims of "hobby ware" made by some dimbulbs
>> 
>> The code quality of OSS software was *better* acrosss the board compared to
>> CSS code
> 
> That's fine. However, explain why updates basically break a system, why
> LibreOffice is devoid of a strong number of features standard to MS
> Office since the 90's and why open-source drivers perform so terribly in
> comparison to the closed-source ones. The quality of the code doesn't
> matter as much as the end result. If I have to fear updating my system
> because it might not work after a reboot, I'm not better off with
> GNU/Linux. If I can't rely on my office suite to simply rotate an image,
> I'm not better off with an open-source suite and if I have to suffer
> through screen-tearing and fractional performance because I
> philosophically chose a free driver over a closed one, I'm not much
> better off in GNU/Linux.

With the rotation, try masking it first and then using even the rotate 90
degrees option LibreOffice offers. Funny results. I should make a video
showing it.


-- 
* Mint MATE: Trash, Panel, Menu:    <http://youtu.be/C0y74FIf7uE>
* Mint KDE bugs or Easter eggs?     <http://youtu.be/CU-whJQvtfA>
* Mint KDE working with folders:    <http://youtu.be/7C9nvniOoE0>
* Mint KDE creating files:          <http://youtu.be/N7-fZJaJUv8>
* Mint KDE help:                    <http://youtu.be/3ikizUd3sa8>
* Mint KDE general navigation:      <http://youtu.be/t9y14yZtQuI>
* Mint / LibreOffice:               <http://youtu.be/USU9iqppfto>
* Win 8 / Word 2013:                <http://youtu.be/z8gUL2TCcV4>
* Mavericks / Pages 5.1:            <http://youtu.be/D3BPWANQoIk>

[toc] | [prev] | [next] | [standalone]

#240746

On 4/26/2014 5:42 PM, Peter Köhlmann wrote:
> Coverity scan has again published their report about error density.
>
> This time OSS code was in *all* categories better than CSS code (last year
> it was better in all categories with code below 1.000.000 lines).
> Tested was NetBSD, FreeBSD, LibreOffice and Linux among other OSS

Just those 4 projects represented 18% of the total scanned OSS code. 
Those 4 alone could skew the results in favor of OSS.


> The result this year was 0.59 per kloc for OSS, 0.72 for CSS
>
> The linux code they scanned ( 8.5 million lines of code) resulted in 0.61
> per kloc
>
> So far for the repeated claims of "hobby ware" made by some dimbulbs

By definition it's hobbyware, since most of it is created by unpaid 
developers in their spare time, for fun.

And it shows in the usage, where you get ridiculous bugs such as mice 
being identified as keyboards, same app listed 16x, apps not sorted 
alphabetically, Trash file handling inconsistencies, inability to sort 
numbers (KDiskFree), etc.

The most bullshit bugs I've ever seen have been in open source.


> The code quality of OSS software was *better* acrosss the board compared to
> CSS code

Coverity's code is closed source.  So is yours.  And Creepy's.  And 
SMelzzzzz's. And every developer's here on cola.

I downloaded
http://softwareintegrity.coverity.com/rs/coverity/images/2013-Coverity-Scan-Report.pdf

Look at Table 4: it's horseshit.  Their supposed "Average Defect 
Density" increased 300%+ from 2009 to 2010, then suddenly dropped by 44% 
in 2011.

Ridiculous.

[toc] | [prev] | [next] | [standalone]

Page 1 of 2  [1] 2  Next page →

Back to top | Article view | comp.os.linux.advocacy

csiph-web