On Sat, 11 Jun 2011, David M. Syzdek wrote:
> Arthur,
>
> I would beg to differ. Crippling developers with older machines, presumably
> less memory/less powerful CPU, would be a mistake. As modern IDEs progress
> and add features to assist developers, these features often come at a cost
> of higher CPU and memory requirements. For instance, I use Xcode 4 for most
> of my development even for open source projects due to the awesome, as in
> full of awe, debugging and integration capabilities it has due to LLVM (for
> example, the analyzer functionality:
> http://iphonedevelopertips.com/xcode/static-code-analysis-clang-and-xcode-3-2.html).
> It has a much greater memory foot print than Xcode 3.0 which had a greater
> impact than using just bash/vi/gcc. Before anyone shouts foul for comparing
> an OS X IDE to Linux development, I routinely use Xcode4 to analyze my code
> for linux/freebsd daemons.
>
> I do fully support using older hardware as test servers/workstations for
> development so the developers are able to see the impact of performance.
>
> --David M. Syzdek
I understand that people want the latest toys, but your response doesn't
really address anything I said. Nor does it counter the concept that maybe,
just maybe, even our IDEs would be be better written if developers were more
conscious of system resource impact.
Testing for performance on older hardware isn't the same thing as
actually living with it. It gives you a whole new level of incentives to be
less sloppy. I think I can guarantee firefox wouldn't have had the endless
years of constant memory leaks if they had to live with it themselves. But
no one really made it a priority because multi-gigabyte machines made a
bloated browser monstrosity a lot less noticeable.
And any way you cut it, there really isn't a good excuse for why a 1GHz
machine with twice the RAM of a 500MHz machine *isn't* actually twice as
fast for real world use...
--Arthur Corliss
Live Free or Die
---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.
Received on Sun Jun 12 09:12:50 2011
This archive was generated by hypermail 2.1.8 : Sun Jun 12 2011 - 09:12:50 AKDT