kicad-developers team mailing list archive

["Chris Albertson" <albertson.chris@...>]

On 10/26/07, Richard A Burton <richardaburton@...> wrote:
>
>
> How much work would it be for you to do this? And how easy would it be
> for the rest of us to maintain it?
>




I think if an evolutionary approach is taken is would be easy. Keep the
existing files intact. An autotools based system that mimics the existing
system with only very minimal checks is easy to implement and maintain.
You gain a lot with just this minimal level. For example you get quite a
few
usful make targets like "make dist" (for packaging up a release) and
"make check" for regression testing. Many others too. All basically for
"free".

Next, simple tests in the configure script are easy. For example checking
for
a dependent library and stopping if not found. Just going to this simple
level gives you some nice meaningful diagnostics during the configure
process, beter then compile errors later

Going to the next level is harder, this is actually making the code portable
and finding people to test it on all the available platforms but using GNU
autotools just as a check to verify the environment is easy.


-- 
=====
Chris Albertson
Redondo Beach, California
 ------=_Part_9313_11625335.1193420962458 Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

<br><br><div><span class="gmail_quote">On 10/26/07, <b class="gmail_sendername">Richard A Burton</b> &lt;<a href="mailto:richardaburton@...";>richardaburton@...</a>&gt; wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">









<div style="background-color: rgb(255, 255, 255);">


<div style="width: 655px;">
<div style="padding: 0pt 15px 0pt 0pt; width: 490px; float: left;">


<div>
<p><span class="q"></span><br>
How much work would it be for you to do this? And how easy would it be<br>
for the rest of us to maintain it?</p></div></div></div></div></blockquote><div><br><br>&nbsp;</div><br><div><br>I think if an evolutionary approach is taken is would be easy.&nbsp; Keep the<br>existing files intact.&nbsp; An autotools based system that mimics the existing
<br>system with only very minimal checks is easy to implement and maintain.<br>You gain a lot with just this minimal level.&nbsp; For example you get quite a few<br>usful make targets like &quot;make dist&quot; (for packaging up a release) and
<br>&quot;make check&quot; for regression testing.&nbsp; Many others too.&nbsp; All basically for<br>&quot;free&quot;.&nbsp;&nbsp; <br><br>Next, simple tests in the configure script are easy.&nbsp; For example checking for<br>a dependent library and stopping if not found.&nbsp; Just going to this simple
<br>level gives you some nice meaningful diagnostics during the configure<br>process, beter then compile errors later<br><br>Going to the next level is harder, this is actually making the code portable<br>and finding people to test it on all the available platforms but using GNU
<br>autotools just as a check to verify the environment is easy.<br>&nbsp;</div><br></div>-- <br>=====<br>Chris Albertson<br>Redondo Beach, California
 ------=_Part_9313_11625335.1193420962458--