«

On Sat, 22 Mar 2008 18:33:03 -0700, Keith Thompson wrote:

> CBFalconer <> writes:
>> Harald van D?k wrote:
>> ... snip ...
>>>
>>> No. jacob gave the source code size because he was directly asked for
>>> the source code size. santosh asked: "What would be the size of
>>> download of the full sources for lcc-win32?" jacob could have answered
>>> with "There is no download of the full sources for lcc-win32, you
>>> can't ask for the size of something that doesn't exist!!", but that
>>> would have been unhelpful and annoying. Instead, he answered santosh's
>>> question. I don't see why you're reading any more into it.
>>
>> Because, in an earlier message, Jacob specifically brought up the
>> subject of the source code and its size. It doesn't really matter,
>> since I assume it to be meaningless, because not available.
>
> It seem to me that source code size is a perfectly valid (though
> imprecise) metric for the complexity and maintainability of a piece of
> softwre, whether the source code is available or not.

Rather depends on the coding style of the developer, no?

Given the same problem, one developer may produce a small program, say
20-30 lines, while another produces 200-300 lines. Which is more
maintainable? No way to tell without seeing; the 200+ line program may
be excessively bloated, or the 20+ line program may be unmaintainably
terse. Or both, or neither.

Kelsey Bjarnason wrote:
> On Sat, 22 Mar 2008 18:33:03 -0700, Keith Thompson wrote:
>
>> CBFalconer <> writes:
>>> Harald van D?k wrote:
>>> ... snip ...
>>>> No. jacob gave the source code size because he was directly asked for
>>>> the source code size. santosh asked: "What would be the size of
>>>> download of the full sources for lcc-win32?" jacob could have answered
>>>> with "There is no download of the full sources for lcc-win32, you
>>>> can't ask for the size of something that doesn't exist!!", but that
>>>> would have been unhelpful and annoying. Instead, he answered santosh's
>>>> question. I don't see why you're reading any more into it.
>>> Because, in an earlier message, Jacob specifically brought up the
>>> subject of the source code and its size. It doesn't really matter,
>>> since I assume it to be meaningless, because not available.
>> It seem to me that source code size is a perfectly valid (though
>> imprecise) metric for the complexity and maintainability of a piece of
>> softwre, whether the source code is available or not.
>
> Rather depends on the coding style of the developer, no?
>
> Given the same problem, one developer may produce a small program, say
> 20-30 lines, while another produces 200-300 lines. Which is more
> maintainable? No way to tell without seeing; the 200+ line program may
> be excessively bloated, or the 20+ line program may be unmaintainably
> terse. Or both, or neither.
>

For a single program yes. But for the source code of several programs,
each of them of a certain complexity, that is no longer the case.

The code is terse, yes, I do not like bloated code, but it is much more
maintainable than 10 times more code *of course*

Have you looked into gcc's code?

Have you ever tried to debug it?

Go ahead and you will find why lcc-win is much more easy!



--
jacob navia
jacob at jacob point remcomp point fr
logiciels/informatique
http://www.cs.virginia.edu/~lcc-win32

Kelsey Bjarnason <> writes:

> On Sat, 22 Mar 2008 18:33:03 -0700, Keith Thompson wrote:
>
>> CBFalconer <> writes:
>>> Harald van D?k wrote:
>>> ... snip ...
>>>>
>>>> No. jacob gave the source code size because he was directly asked for
>>>> the source code size. santosh asked: "What would be the size of
>>>> download of the full sources for lcc-win32?" jacob could have answered
>>>> with "There is no download of the full sources for lcc-win32, you
>>>> can't ask for the size of something that doesn't exist!!", but that
>>>> would have been unhelpful and annoying. Instead, he answered santosh's
>>>> question. I don't see why you're reading any more into it.
>>>
>>> Because, in an earlier message, Jacob specifically brought up the
>>> subject of the source code and its size. It doesn't really matter,
>>> since I assume it to be meaningless, because not available.
>>
>> It seem to me that source code size is a perfectly valid (though
>> imprecise) metric for the complexity and maintainability of a piece of
>> softwre, whether the source code is available or not.
>
> Rather depends on the coding style of the developer, no?
>
> Given the same problem, one developer may produce a small program, say
> 20-30 lines, while another produces 200-300 lines. Which is more
> maintainable? No way to tell without seeing; the 200+ line program may
> be excessively bloated, or the 20+ line program may be unmaintainably
> terse. Or both, or neither.

That is rather stating the obvious.

But size of the C source is probably quite relevant, generally, to how
difficult or large the job of maintaining it is.

As Keith said "imprecise" - but to ridicule it as being applicable is
being ridiculous.

Kelsey Bjarnason said:

> On Sat, 22 Mar 2008 18:33:03 -0700, Keith Thompson wrote:
>
<snip>
>>
>> It seem to me that source code size is a perfectly valid (though
>> imprecise) metric for the complexity and maintainability of a piece of
>> softwre, whether the source code is available or not.
>
> Rather depends on the coding style of the developer, no?
>
> Given the same problem, one developer may produce a small program, say
> 20-30 lines, while another produces 200-300 lines. Which is more
> maintainable? No way to tell without seeing; the 200+ line program may
> be excessively bloated, or the 20+ line program may be unmaintainably
> terse. Or both, or neither.

Whilst that is true, such differences will in most cases be swamped in
larger systems.

I have found, *in general*, that the complexity and maintenance cost of a
system are roughly in proportion to its size. This is hardly surprising.
Nor is it surprising that there are occasional exceptions.

--
Richard Heathfield <http://www.cpax.org.uk>
Email: -http://www. +rjh@
Google users: <http://www.cpax.org.uk/prg/writings/googly.php>
"Usenet is a strange place" - dmr 29 July 1999

jacob navia wrote:

> Kelsey Bjarnason wrote:
>> On Sat, 22 Mar 2008 18:33:03 -0700, Keith Thompson wrote:
>>
>>> CBFalconer <> writes:
>>>> Harald van D?k wrote:
>>>> ... snip ...
>>>>> No. jacob gave the source code size because he was directly asked
>>>>> for the source code size. santosh asked: "What would be the size
>>>>> of download of the full sources for lcc-win32?" jacob could have
>>>>> answered with "There is no download of the full sources for
>>>>> lcc-win32, you can't ask for the size of something that doesn't
>>>>> exist!!", but that would have been unhelpful and annoying.
>>>>> Instead, he answered santosh's question. I don't see why you're
>>>>> reading any more into it.
>>>> Because, in an earlier message, Jacob specifically brought up the
>>>> subject of the source code and its size. It doesn't really matter,
>>>> since I assume it to be meaningless, because not available.
>>> It seem to me that source code size is a perfectly valid (though
>>> imprecise) metric for the complexity and maintainability of a piece
>>> of softwre, whether the source code is available or not.
>>
>> Rather depends on the coding style of the developer, no?
>>
>> Given the same problem, one developer may produce a small program,
>> say
>> 20-30 lines, while another produces 200-300 lines. Which is more
>> maintainable? No way to tell without seeing; the 200+ line program
>> may be excessively bloated, or the 20+ line program may be
>> unmaintainably
>> terse. Or both, or neither.
>>
>
> For a single program yes. But for the source code of several programs,
> each of them of a certain complexity, that is no longer the case.
>
> The code is terse, yes, I do not like bloated code, but it is much
> more maintainable than 10 times more code *of course*
>
> Have you looked into gcc's code?
>
> Have you ever tried to debug it?
>
> Go ahead and you will find why lcc-win is much more easy!

But comparing gcc with lcc-win isn't really fair is it? Quite apart from
the fact that we have no access to lcc-win's source to *make* a
comparison at all, gcc is bigger and more complex source-wise, probably
because it does a whole lot more. It's backend has been ported to
dozens of chips and it does a lot more optimisations than lcc-win. It's
probably also been designed to be more extendable than lcc-win.

Kelsey Bjarnason <> writes:
> On Sat, 22 Mar 2008 18:33:03 -0700, Keith Thompson wrote:
[...]
>> It seem to me that source code size is a perfectly valid (though
>> imprecise) metric for the complexity and maintainability of a piece of
>> softwre, whether the source code is available or not.
>
> Rather depends on the coding style of the developer, no?
>
> Given the same problem, one developer may produce a small program, say
> 20-30 lines, while another produces 200-300 lines. Which is more
> maintainable? No way to tell without seeing; the 200+ line program may
> be excessively bloated, or the 20+ line program may be unmaintainably
> terse. Or both, or neither.

Of course. The word "imprecise" was intended to cover all that.

--
Keith Thompson (The_Other_Keith) <kst->
Nokia
"We must do something. This is something. Therefore, we must do this."
-- Antony Jay and Jonathan Lynn, "Yes Minister"

I guess the point of having a fair idea about what you are trying to
acheive before setting out to write a singlew line of code, would be
prudent to assume in the first instance?

[snips]

On Sun, 23 Mar 2008 22:03:39 +0100, jacob navia wrote:

> For a single program yes. But for the source code of several programs,
> each of them of a certain complexity, that is no longer the case.

Really. You never looked at Jeff Relf's code, then, saw all the screwed-
up macros and whatnot, which made virtually everything he did
unmaintainable?

Oh, right, there's a magic force which prevents those macros being used
in other programs, I forgot. Silly me.