httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Roy T. Fielding" <>
Subject Re: Where to start the Apache 2.0 code base
Date Fri, 06 Aug 1999 00:05:28 GMT
>"Roy T. Fielding" wrote:
>> I know for a fact that this is
>> the way it must be done if we are to support people doing software
>> research on change analysis.
>a) why on Earth should we care about the ease of doing such research?
>Hmm. I don't quite mean that. I mean: it isn't a goal of the Apache
>project to support research on change analysis. You are proposing a
>_huge_ amount of work for something we don't aim to support.

Did you bother to read the follow-on sentence?  The reason we want to
support it is because the same aspects that make such research possible
also make it EASIER to perform the kind of change analysis that a good
developer does every time they want to make a future change to the same
code.  The reason I want to support it is because it is MY RESEARCH, and
one of the main reasons I got involved in Apache, and hence one of the
reasons Apache is open source.

>b) research done on such an artificial construct strikes me as
>completely worthless. The _real_ changes have happened already. Redoing
>them to make research easier will no doubt give a beautiful Popperesque
>view of the process, bearing as little relationship to real life as
>Popper does.

That depends on what types of change the researcher is looking for.
I'm not suggesting we delete all of the experimental repositories.
In fact, had the existing repositories been based on any identifiable
mark in the other repositories, and included change logs that had any
relevance whatsoever to the changes being made, this wouldn't be necessary
(and don't tell me there is one -- I've already looked and the tags are
bogus and the logs are only useful if the person making the change
happens to remember what they were doing on the day in question).

And don't tell me its a lot of work -- you won't be the one doing it,
I have to do it anyway just to understand what the hell was changed,
and I've said a dozen times over that it can be done in parallel
with the other thing you guys were planning to do anyway.  We've wasted
far more time discussing this than it will take to implement it.


View raw message