DrDr experiment
42 views
Skip to first unread message
Jay McCarthy
Apr 19, 2015, 8:13:06 PM4/19/15
to dev
I'm starting an experiment where DrDr will be running all the time,
rather than only on pushes to the main repository. Hopefully this will
make it more useful in the post-single-repo ecosystem, but there may
be some errors. One big downside is that I can't guarantee each push
will be tested individually, instead it considers all commits that
came in since the last time it polled (which happens just before it
runs an integration) to be the same "push".
Jay
--
Jay McCarthy
http://jeapostrophe.github.io
"Wherefore, be not weary in well-doing,
for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
- D&C 64:33
rather than only on pushes to the main repository. Hopefully this will
make it more useful in the post-single-repo ecosystem, but there may
be some errors. One big downside is that I can't guarantee each push
will be tested individually, instead it considers all commits that
came in since the last time it polled (which happens just before it
runs an integration) to be the same "push".
Jay
--
Jay McCarthy
http://jeapostrophe.github.io
"Wherefore, be not weary in well-doing,
for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
- D&C 64:33
Robby Findler
Apr 19, 2015, 8:51:42 PM4/19/15
to Jay McCarthy, dev
On Sunday, April 19, 2015, Jay McCarthy <jay.mc...@gmail.com> wrote:
I'm starting an experiment where DrDr will be running all the time,
rather than only on pushes to the main repository. Hopefully this will
make it more useful in the post-single-repo ecosystem, but there may
be some errors. One big downside is that I can't guarantee each push
will be tested individually, instead it considers all commits that
came in since the last time it polled (which happens just before it
runs an integration) to be the same "push".
Can you say more how this is a downside, given the post-split world we live in? (I can understand that when you compare to the old world but in the new one, this bad property is worse when you wait for pushes in just one repo right?)
Robby
Jay
--
Jay McCarthy
http://jeapostrophe.github.io
"Wherefore, be not weary in well-doing,
for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
- D&C 64:33
--
You received this message because you are subscribed to the Google Groups "Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to racket-dev+...@googlegroups.com.
To post to this group, send email to racke...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/racket-dev/CAJYbDakBWZGKGMVK0LgXh8T7KjQrc5GxGLuLPpBtkJ7ERwv8vA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.
Robby Findler
Apr 19, 2015, 9:25:41 PM4/19/15
to Jay McCarthy, dev
The vast majority of our code is outside the one repo that DrDr
monitors. So already, for all that code, what you describe is the
case. Indeed, it routinely causes me problems and the changes that you
recently pushed make life much better for this majority.
No?
Robby
On Sun, Apr 19, 2015 at 7:55 PM, Jay McCarthy <jay.mc...@gmail.com> wrote:
> then push it (push +1), then the net collection and push that (push
> +2), but during that period of time, DrDr was working, these two
> changes will be combined into a single DrDr test (as well as all other
> changes in any other package.) Previously to this recent change, every
> push to the main repo was Sacred and tested separately, with all
> changes to packages combined... now there is no protection whatsoever
> for any repo/package/etc.
monitors. So already, for all that code, what you describe is the
case. Indeed, it routinely causes me problems and the changes that you
recently pushed make life much better for this majority.
No?
Robby
On Sun, Apr 19, 2015 at 7:55 PM, Jay McCarthy <jay.mc...@gmail.com> wrote:
> On Sun, Apr 19, 2015 at 8:51 PM, Robby Findler
> <ro...@eecs.northwestern.edu> wrote:
>>
>>
>> On Sunday, April 19, 2015, Jay McCarthy <jay.mc...@gmail.com> wrote:
>>>
>>> I'm starting an experiment where DrDr will be running all the time,
>>> rather than only on pushes to the main repository. Hopefully this will
>>> make it more useful in the post-single-repo ecosystem, but there may
>>> be some errors. One big downside is that I can't guarantee each push
>>> will be tested individually, instead it considers all commits that
>>> came in since the last time it polled (which happens just before it
>>> runs an integration) to be the same "push".
>>>
>>
>> Can you say more how this is a downside, given the post-split world we live
>> in? (I can understand that when you compare to the old world but in the new
>> one, this bad property is worse when you wait for pushes in just one repo
>> right?)
>>
>
> If you make a change to the main repo and change the db collection,
> <ro...@eecs.northwestern.edu> wrote:
>>
>>
>> On Sunday, April 19, 2015, Jay McCarthy <jay.mc...@gmail.com> wrote:
>>>
>>> I'm starting an experiment where DrDr will be running all the time,
>>> rather than only on pushes to the main repository. Hopefully this will
>>> make it more useful in the post-single-repo ecosystem, but there may
>>> be some errors. One big downside is that I can't guarantee each push
>>> will be tested individually, instead it considers all commits that
>>> came in since the last time it polled (which happens just before it
>>> runs an integration) to be the same "push".
>>>
>>
>> Can you say more how this is a downside, given the post-split world we live
>> in? (I can understand that when you compare to the old world but in the new
>> one, this bad property is worse when you wait for pushes in just one repo
>> right?)
>>
>
> then push it (push +1), then the net collection and push that (push
> +2), but during that period of time, DrDr was working, these two
> changes will be combined into a single DrDr test (as well as all other
> changes in any other package.) Previously to this recent change, every
> push to the main repo was Sacred and tested separately, with all
> changes to packages combined... now there is no protection whatsoever
> for any repo/package/etc.
Jay McCarthy
Apr 20, 2015, 5:10:44 AM4/20/15
to Robby Findler, dev
> On Sunday, April 19, 2015, Jay McCarthy <jay.mc...@gmail.com> wrote:
>>
>> I'm starting an experiment where DrDr will be running all the time,
>> rather than only on pushes to the main repository. Hopefully this will
>> make it more useful in the post-single-repo ecosystem, but there may
>> be some errors. One big downside is that I can't guarantee each push
>> will be tested individually, instead it considers all commits that
>> came in since the last time it polled (which happens just before it
>> runs an integration) to be the same "push".
>>
>
> Can you say more how this is a downside, given the post-split world we live
> in? (I can understand that when you compare to the old world but in the new
> one, this bad property is worse when you wait for pushes in just one repo
> right?)
>
>>
>> I'm starting an experiment where DrDr will be running all the time,
>> rather than only on pushes to the main repository. Hopefully this will
>> make it more useful in the post-single-repo ecosystem, but there may
>> be some errors. One big downside is that I can't guarantee each push
>> will be tested individually, instead it considers all commits that
>> came in since the last time it polled (which happens just before it
>> runs an integration) to be the same "push".
>>
>
> Can you say more how this is a downside, given the post-split world we live
> in? (I can understand that when you compare to the old world but in the new
> one, this bad property is worse when you wait for pushes in just one repo
> right?)
>
If you make a change to the main repo and change the db collection,
then push it (push +1), then the net collection and push that (push
+2), but during that period of time, DrDr was working, these two
changes will be combined into a single DrDr test (as well as all other
changes in any other package.) Previously to this recent change, every
push to the main repo was Sacred and tested separately, with all
changes to packages combined... now there is no protection whatsoever
for any repo/package/etc.
then push it (push +1), then the net collection and push that (push
+2), but during that period of time, DrDr was working, these two
changes will be combined into a single DrDr test (as well as all other
changes in any other package.) Previously to this recent change, every
push to the main repo was Sacred and tested separately, with all
changes to packages combined... now there is no protection whatsoever
for any repo/package/etc.
Robby Findler
Apr 20, 2015, 8:42:24 AM4/20/15
to Jay McCarthy, dev
So, just to recap, I think this is a great change that will, for most
people whose code DrDr monitors be an improvement and I'm very glad
that this change was made.
Robby
On Sun, Apr 19, 2015 at 8:25 PM, Robby Findler
people whose code DrDr monitors be an improvement and I'm very glad
that this change was made.
Robby
On Sun, Apr 19, 2015 at 8:25 PM, Robby Findler
Reply all
Reply to author
Forward
0 new messages