Velocity Reviews - Computer Hardware Reviews

Velocity Reviews > Newsgroups > Computing > NZ Computing > Closed-source software follies

Reply
Thread Tools

Closed-source software follies

 
 
Lawrence D'Oliveiro
Guest
Posts: n/a
 
      01-25-2007
Here's an entertaining story
<http://www.infoworld.com/article/07/01/23/05OPrecord_1.html> about some of
the antics that software companies get up to, such as buying up an absolute
crap piece of software simply because of the prospect of getting income
from its existing customers (who must have been trapped by vendor lock-in,
otherwise they'd have gone if they had any sense).

What's not explicitly stated in the article is that all the products were
closed-source. It's hard to see anything like that happening in an
open-source project--at least, not for long: after the community had had
enough of laughing itself silly over the code, it would either junk it in
favour of a rewrite, or dive in and fix it.
 
Reply With Quote
 
 
 
 
Cadae
Guest
Posts: n/a
 
      01-25-2007
"Lawrence D'Oliveiro" <(E-Mail Removed)_zealand> wrote in message
news:ep9j70$egg$(E-Mail Removed)...
> Here's an entertaining story
> <http://www.infoworld.com/article/07/01/23/05OPrecord_1.html> about some
> of
> the antics that software companies get up to, such as buying up an
> absolute
> crap piece of software simply because of the prospect of getting income
> from its existing customers (who must have been trapped by vendor lock-in,
> otherwise they'd have gone if they had any sense).


Great story. The company clearly wasn't listening to its technical advisors,
and paid the price.

> What's not explicitly stated in the article is that all the products were
> closed-source. It's hard to see anything like that happening in an
> open-source project--at least, not for long: after the community had had
> enough of laughing itself silly over the code, it would either junk it in
> favour of a rewrite, or dive in and fix it.


Quality of code is independent of which license scheme is used to distribute
it, you can't assume closed-source = = bad and open-source = = good.

For example look at this - one of the most important Linux library routines-
malloc:
http://www.linuxhq.com/kernel/v1.0/3/lib/malloc.c If malloc doesn't work
properly, it can bring down an entire system.
Note the use of goto statements, and the length of the functions - this is
an analysis and maintenance nightmare. Does it work ? - maybe. Is it well
engineered and thus reliable ? No.
It'd been in use for years, and no one in the open-source community "dived
in and fixed it".
Compare with the contemporary malloc.c library in Visual Studio - there's a
huge qualitative gap in favour of the Microsoft code.

PC




 
Reply With Quote
 
 
 
 
thingy
Guest
Posts: n/a
 
      01-25-2007
Lawrence D'Oliveiro wrote:
> Here's an entertaining story
> <http://www.infoworld.com/article/07/01/23/05OPrecord_1.html> about some of
> the antics that software companies get up to, such as buying up an absolute
> crap piece of software simply because of the prospect of getting income
> from its existing customers (who must have been trapped by vendor lock-in,
> otherwise they'd have gone if they had any sense).
>
> What's not explicitly stated in the article is that all the products were
> closed-source. It's hard to see anything like that happening in an
> open-source project--at least, not for long: after the community had had
> enough of laughing itself silly over the code, it would either junk it in
> favour of a rewrite, or dive in and fix it.


Not unknown, in fact depressingly frequent...one of the worst.......POS
(point of sale) stuff I have come across, it was locked down so you
could not upgrade the box hardware without a new key....the vendor's
ex-support guy who bought the code from the company that went broke
wanted an huge monthly fee for "support" and another huge fee for a new
key....the retailer was over a barrel....hardly making any money but had
to come up with silly fees.........

This is one of the reasons I am so keen on FOSS....the end user has the
code so can go to any other service company who supports that code and
use them....companies have to compete on costs and quality....so the end
user is not ass raped by greedy slaes ppl.

regards

Thing









 
Reply With Quote
 
Lawrence D'Oliveiro
Guest
Posts: n/a
 
      01-26-2007
In message <45b8717d$(E-Mail Removed)>, Cadae wrote:

> For example look at this - one of the most important Linux library
> routines- malloc:
> http://www.linuxhq.com/kernel/v1.0/3/lib/malloc.c If malloc doesn't work
> properly, it can bring down an entire system.
> Note the use of goto statements, and the length of the functions - this is
> an analysis and maintenance nightmare.


You _did_ notice that what you referenced was not the actual source of the
Linux kernel memory allocator, but a patch dating from 1993, did you not?

The current Linux kernel memory allocator is actually called kmalloc, and
you'll find it defined here <http://lxr.linux.no/source/mm/slab.c>, in a
lot cleaner and more maintainable fashion than the obsolete example you
posted.

 
Reply With Quote
 
Cadae
Guest
Posts: n/a
 
      01-26-2007
"Lawrence D'Oliveiro" <(E-Mail Removed)_zealand> wrote in message
news:epbrqv$h0a$(E-Mail Removed)...
>
> You _did_ notice that what you referenced was not the actual source of the
> Linux kernel memory allocator, but a patch dating from 1993, did you not?


That's the point - essentially the same code with the same poor coding
practice remained in use from 91 with only minor patches through to at least
96.
And that was a critical part of Linux for years. No one "dived in and fixed
it" as you claim happens to open-source code. Even when they did fix parts
of it, as per the patch example, the important coding faults didn't get
fixed. There's a wealth of similarly poor code in the open-source world I
can point to - I've seen absolutely abominable open-source code ... and I've
also seen very good open-source code. The same is true for "closed-source"
code.

> The current Linux kernel memory allocator is actually called kmalloc, and
> you'll find it defined here <http://lxr.linux.no/source/mm/slab.c>, in a
> lot cleaner and more maintainable fashion than the obsolete example you
> posted.


Only after many years had passed did this essential bit of code get its
long-overdue improvements.



PC



 
Reply With Quote
 
Dianthus Mimulus
Guest
Posts: n/a
 
      01-26-2007
On Fri, 26 Jan 2007 21:26:17 +1300, Cadae wrote:

> That's the point - essentially the same code with the same poor coding
> practice remained in use from 91 with only minor patches through to at least
> 96.
> And that was a critical part of Linux for years. No one "dived in and fixed
> it" as you claim happens to open-source code. Even when they did fix parts
> of it, as per the patch example, the important coding faults didn't get
> fixed.


Surely it has already been fixed, has it not?


> There's a wealth of similarly poor code in the open-source world
> I can point to - I've seen absolutely abominable open-source code ...
> and I've also seen very good open-source code.


If you saw "absolutely abominable" code, did you submit a fix?


> The same is true for "closed-source" code.


How would you *know that* if you cannot see the source code?


--
Dianthus Mimulus

Microsoft's business practises exposed in court:
http://www.maxframe.com/DR/Info/full...#_Toc447960918
 
Reply With Quote
 
Lawrence D'Oliveiro
Guest
Posts: n/a
 
      01-26-2007
In message <45b9bb2e$(E-Mail Removed)>, Cadae wrote:

> "Lawrence D'Oliveiro" <(E-Mail Removed)_zealand> wrote in message
> news:epbrqv$h0a$(E-Mail Removed)...
>>
>> You _did_ notice that what you referenced was not the actual source of
>> the Linux kernel memory allocator, but a patch dating from 1993, did you
>> not?

>
> That's the point - essentially the same code with the same poor coding
> practice remained in use from 91 with only minor patches through to at
> least 96.
> And that was a critical part of Linux for years. No one "dived in and
> fixed it" as you claim happens to open-source code.


So where is that code now?

 
Reply With Quote
 
Cadae
Guest
Posts: n/a
 
      01-29-2007
"Dianthus Mimulus" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> Surely it has already been fixed, has it not?


Yes - but it took years before it was fixed. No one dived in and fixed the
bad coding ( as you claim happens ) even tho it was in a super-critical part
of the linux code.

> If you saw "absolutely abominable" code, did you submit a fix?


Way too much bad code around to fix - and why should I fix it ? After all,
open source suffers from tragedy of the commons - i.e. given the choice of
putting effort into something that will pay the family bills (and give me a
sense of achievement) and something that will only give me a sense of
achievement, I'm not irrational enough to turn myself into a pauper or
parasite of the state by wasting my valuable time fixing open source.

It only makes rational sense to put time into open source if I were to
benefit in some secondary manner as a large company or government
organisation can e.g. by getting training or support contracts, or getting a
piece of hardware operational. None of those apply to me. I like programming
and am good enough to make money at it, so it makes much more sense to
produce "closed source" software and earn money directly from it than try to
benefit from secondary, indirect and unreliable sources of income (support,
training etc) that open source provides.

>
> How would you *know that* if you cannot see the source code?


Easy - there's much non-"open source" code around that's viewable. For
example take any copy of Visual Studio and look at the source code in the
libraries. It's not "open source" but it is readable source code. Don't
forget that the official definition of "open source" as used in relation to
Linux is very specific - it applies to the licensing terms, not whether or
not the source code is viewable.

I've been involved in many projects that have given me access to other's
"closed" source code. Some has been brilliant, some has been rotten - you
can't assume as you did that closed source code is bad and that all open
source code is good.


PC


 
Reply With Quote
 
Cadae
Guest
Posts: n/a
 
      01-29-2007
"Lawrence D'Oliveiro" <(E-Mail Removed)_zealand> wrote in message
news:epe203$ekn$(E-Mail Removed)...
>
> So where is that code now?


I don't understand your question.

PC


 
Reply With Quote
 
Lawrence D'Oliveiro
Guest
Posts: n/a
 
      01-30-2007
In message <45bdc1a5$(E-Mail Removed)>, Cadae wrote:

> "Lawrence D'Oliveiro" <(E-Mail Removed)_zealand> wrote in message
> news:epe203$ekn$(E-Mail Removed)...
>>
>>In message <45b9bb2e$(E-Mail Removed)>, Cadae wrote:

>
>>> "Lawrence D'Oliveiro" <(E-Mail Removed)_zealand> wrote in
>>> message news:epbrqv$h0a$(E-Mail Removed)...
>>>>
>>>> You did notice that what you referenced was not the actual source of
>>>> the Linux kernel memory allocator, but a patch dating from 1993, did
>>>> you not?
>>>

>> That's the point - essentially the same code with the same poor coding
>>> practice remained in use from 91 with only minor patches through to at
>>> least 96. And that was a critical *part of Linux for years. No
>>> one "dived in and fixed it" as you claim happens to open-source code.

>
>> So where is that code now?

>
> I don't understand your question.


That code was replaced with much improved code. Someone _did_ "dive in and
fix it", because it was open-source code, and anybody could do that, so
someone did. Which proves my point that it's much harder for crufty code to
persist in open-source projects than in closed-source ones.

For your next example, would you like to choose one that actually backs up
your point, rather than one which reinforces mine?
 
Reply With Quote
 
 
 
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Re: NYC train & bus follies PluckThePig=^||^ Computer Support 0 06-24-2010 04:05 AM
DOS Follies (Was: Docstrings...) D'Arcy J.M. Cain Python 1 03-04-2010 02:50 AM
DLL Call Follies Kythe Perl Misc 6 09-11-2006 12:33 PM
DVD Verdict reviews: ALL I WANT FOR CHRISTMAS, CARTOON NETWORK CHRISTMAS: YULETIDE FOLLIES, and more! DVD Verdict DVD Video 0 12-10-2004 10:10 AM
"titicut follies" Vlvetmorning98 DVD Video 6 02-29-2004 09:13 PM



Advertisments