Hi Joep,
On 04/23/12 08:40 am, Joep Suijs wrote:
I don't particular want to delete these samples. I suggest to remove
samples that have the risk of getting outdated. There are about 10
board files, so each test file delivers up to 10 samples. Once an
issues is resolved or the test file is extended, all the samples are
regenerated.If a sample fails to regenerate, it remains behind and
no-one notices. IMO this is an undesirable situation.
On the other hand, there is a reason for a sample generation to fail.
The two particular samples at hand are special in the way that only
one of the 10 board files has supplied a sample and currently, none
is. Maybe the sample requires something specific in the board file. If
this requirement is removed, you'll probably have half a dozen samples
for this library.
>
We had a discussion on samples before and it turns out that everyone
has its own approach. This is quite inefficient but I don't think this
is likely to change. And since I am the only one that use the
jallib.py sample generation function, I feel it is my responsibility
to keep that part up to date. I did most of the cleanup about a year
ago and AFAIK this is the last remaining issue. IMO it would be wise
to have separate directories at jallib for (at least) the samples Rob
generates, the samples from jallib.py and the rest, which are mainly
manually maintained. If Matt regenerates his samples, it should have
its own directory too. I'm not sure what the impact on the generation
processes would be.
Good points!
But we probably should have two discussion threads:
- structure of the repository
- structure of the distribution package
because these have different requirements (and different 'users').
For the repository the highest priority is probably maintenance.
For the distibution package the most important is probably
categorisation of the samples.
I wonder how use full over 1000 samples in one directory are for an
average user and if either reduction of the number of samples or some
structure would be helpful (e.g. separate directories for samples
generated by method and one for the manual maintained ones).
To start with we could separate the elementary (generated!) blink
samples from the rest, both in the repository and in the distribution
package. So there will be about 600 'real' samples left, and because of
the naming convention of the samples it will not be too difficult for a
user to find what he is looking for.
At the moment I do not see how to split the samples further in
categories in an 'obvious' way for the user. Maybe we should not put
the device as first part of the name but as last part (and the primary
library as first part).
Regards, Rob.
--
R. Hamerling, Netherlands --- http://www.robh.nl
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/jallib?hl=en.
