llvm dev - May 2017 - Permissions for llvm-mirror - Setting up Libc++ Appveyor builders

> On Apr 20, 2017, at 7:03 PM, Eric Fiselier <eric at efcs.ca> wrote:
> 
> 
> 
> On Thu, Apr 20, 2017 at 4:55 PM, Mehdi Amini <mehdi.amini at apple.com
<mailto:mehdi.amini at apple.com>> wrote:
> 
>> On Apr 20, 2017, at 12:30 PM, Eric Fiselier <eric at efcs.ca
<mailto:eric at efcs.ca>> wrote:
>> 
>> 
>> 
>> On Thu, Apr 20, 2017 at 11:06 AM, Mehdi Amini <mehdi.amini at
apple.com <mailto:mehdi.amini at apple.com>> wrote:
>> 
>>> On Apr 20, 2017, at 12:39 AM, Eric Fiselier <eric at efcs.ca
<mailto:eric at efcs.ca>> wrote:
>>> 
>>> 
>>> 
>>> On Wed, Apr 19, 2017 at 11:19 PM, Mehdi Amini <mehdi.amini at
apple.com <mailto:mehdi.amini at apple.com>> wrote:
>>> What would be the status of these buildbots? Is it for your private
usage?
>>> 
>>> I intend for them to be public Windows buildbots for libc++.
>> 
>> 
>> I’m not sure it’d be OK to send blame email to contributors based on
this though.
>> 
>> I don't see why not (at least once the bot is stable). Can you
elaborate?
> 
> I don’t think there is a precedent for having a bot that does not checkout
from SVN and email using the canonical SVN revision number.
> 
> I understand share your concern about the emails using nonsensical git
hashes as opposed to SVN revision numbers.
> Apart from that I have no idea why the version control used by the CI
matters. It should have to effect on the build or the results.
I don’t know, but other people may have other concern that I don’t necessarily
anticipate..
>  
> 
> So it does not seems like a given to me  (not that I’m against it) and
would require a discussion on the mailing-list first IMO.
> 
> We might as well continue having it now since it's been started.
I'll re-raise the issue in a month or so when the bot is
> actually stable enough to consider sending emails.
I think it deserve a separate thread to deal with this, so that the thread title
help making sure no-one miss the discussion.

— 
Mehdi

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20170420/df6b12c9/attachment.html>

Hi,
Fragmentation of the bots is not ideal.  While I totally understand Eric's
reasoning for doing so, and am delighted to see  another Windows bot
helping to ensure quality code, I think this is a bit of a slippery slope.
We already have the main Buildbots, Green Dragon, Chapuni's Bots, a myriad
of private bots and now this new bot.  I don't think it is reasonable to
ask people to have to visit several different places to check if their
commit broke something.  Anecdotally, I find many people will only look at
lab.llvm.org:8011.  They will check Green Dragon if they get an email and
then only if the email does not get lost in the noise.  Some folks are
regular users of bb.pgr.jp but the actual number or regular visitors I
don't know.

Perhaps it would be a more beneficial discussion to talk about putting
together some type of portal which displays results from all the CI systems
in one place?  I for one would love an interface which would be able to
show any given commit and how is has performed across the board.  Something
like this could possibly allow for anyone to stand up a bot or CI
infrastructure of their choosing and then have that system integrated in a
common reporting platform so it is useful and accessible to everyone.

Thoughts?

-Mike

On Thu, Apr 20, 2017 at 7:22 PM, Mehdi Amini via llvm-dev <
llvm-dev at lists.llvm.org> wrote:
>
> On Apr 20, 2017, at 7:03 PM, Eric Fiselier <eric at efcs.ca> wrote:
>
>
>
> On Thu, Apr 20, 2017 at 4:55 PM, Mehdi Amini <mehdi.amini at
apple.com>
> wrote:
>
>>
>> On Apr 20, 2017, at 12:30 PM, Eric Fiselier <eric at efcs.ca>
wrote:
>>
>>
>>
>> On Thu, Apr 20, 2017 at 11:06 AM, Mehdi Amini <mehdi.amini at
apple.com>
>> wrote:
>>
>>>
>>> On Apr 20, 2017, at 12:39 AM, Eric Fiselier <eric at efcs.ca>
wrote:
>>>
>>>
>>>
>>> On Wed, Apr 19, 2017 at 11:19 PM, Mehdi Amini <mehdi.amini at
apple.com>
>>> wrote:
>>>
>>>> What would be the status of these buildbots? Is it for your
private
>>>> usage?
>>>>
>>>
>>> I intend for them to be public Windows buildbots for libc++.
>>>
>>>
>>>
>>> I’m not sure it’d be OK to send blame email to contributors based
on
>>> this though.
>>>
>>
>> I don't see why not (at least once the bot is stable). Can you
elaborate?
>>
>>
>> I don’t think there is a precedent for having a bot that does not
>> checkout from SVN and email using the canonical SVN revision number.
>>
>
> I understand share your concern about the emails using nonsensical git
> hashes as opposed to SVN revision numbers.
> Apart from that I have no idea why the version control used by the CI
> matters. It should have to effect on the build or the results.
>
>
> I don’t know, but other people may have other concern that I don’t
> necessarily anticipate..
>
>
>
>>
>> So it does not seems like a given to me  (not that I’m against it) and
>> would require a discussion on the mailing-list first IMO.
>>
>
> We might as well continue having it now since it's been started.
I'll
> re-raise the issue in a month or so when the bot is
> actually stable enough to consider sending emails.
>
>
> I think it deserve a separate thread to deal with this, so that the thread
> title help making sure no-one miss the discussion.
>
> —
> Mehdi
>
>
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20170421/cdd2bbb4/attachment-0001.html>

I'm definitely in favor of efforts to defragment our many CI systems, but I
also don't want to stop people from doing more testing. :)

If AppVeyor is providing free testing resource to OSS projects, then great,
let's try using it and see how it goes.

On Fri, Apr 21, 2017 at 6:56 AM, Mike Edwards via llvm-dev <
llvm-dev at lists.llvm.org> wrote:
> Hi,
> Fragmentation of the bots is not ideal.  While I totally understand
Eric's
> reasoning for doing so, and am delighted to see  another Windows bot
> helping to ensure quality code, I think this is a bit of a slippery slope.
> We already have the main Buildbots, Green Dragon, Chapuni's Bots, a
myriad
> of private bots and now this new bot.  I don't think it is reasonable
to
> ask people to have to visit several different places to check if their
> commit broke something.  Anecdotally, I find many people will only look at
> lab.llvm.org:8011.  They will check Green Dragon if they get an email and
> then only if the email does not get lost in the noise.  Some folks are
> regular users of bb.pgr.jp but the actual number or regular visitors I
> don't know.
>
> Perhaps it would be a more beneficial discussion to talk about putting
> together some type of portal which displays results from all the CI systems
> in one place?  I for one would love an interface which would be able to
> show any given commit and how is has performed across the board.  Something
> like this could possibly allow for anyone to stand up a bot or CI
> infrastructure of their choosing and then have that system integrated in a
> common reporting platform so it is useful and accessible to everyone.
>
> Thoughts?
>
> -Mike
>
> On Thu, Apr 20, 2017 at 7:22 PM, Mehdi Amini via llvm-dev <
> llvm-dev at lists.llvm.org> wrote:
>
>>
>> On Apr 20, 2017, at 7:03 PM, Eric Fiselier <eric at efcs.ca>
wrote:
>>
>>
>>
>> On Thu, Apr 20, 2017 at 4:55 PM, Mehdi Amini <mehdi.amini at
apple.com>
>> wrote:
>>
>>>
>>> On Apr 20, 2017, at 12:30 PM, Eric Fiselier <eric at efcs.ca>
wrote:
>>>
>>>
>>>
>>> On Thu, Apr 20, 2017 at 11:06 AM, Mehdi Amini <mehdi.amini at
apple.com>
>>> wrote:
>>>
>>>>
>>>> On Apr 20, 2017, at 12:39 AM, Eric Fiselier <eric at
efcs.ca> wrote:
>>>>
>>>>
>>>>
>>>> On Wed, Apr 19, 2017 at 11:19 PM, Mehdi Amini <mehdi.amini
at apple.com>
>>>> wrote:
>>>>
>>>>> What would be the status of these buildbots? Is it for your
private
>>>>> usage?
>>>>>
>>>>
>>>> I intend for them to be public Windows buildbots for libc++.
>>>>
>>>>
>>>>
>>>> I’m not sure it’d be OK to send blame email to contributors
based on
>>>> this though.
>>>>
>>>
>>> I don't see why not (at least once the bot is stable). Can you
elaborate?
>>>
>>>
>>> I don’t think there is a precedent for having a bot that does not
>>> checkout from SVN and email using the canonical SVN revision
number.
>>>
>>
>> I understand share your concern about the emails using nonsensical git
>> hashes as opposed to SVN revision numbers.
>> Apart from that I have no idea why the version control used by the CI
>> matters. It should have to effect on the build or the results.
>>
>>
>> I don’t know, but other people may have other concern that I don’t
>> necessarily anticipate..
>>
>>
>>
>>>
>>> So it does not seems like a given to me  (not that I’m against it)
and
>>> would require a discussion on the mailing-list first IMO.
>>>
>>
>> We might as well continue having it now since it's been started.
I'll
>> re-raise the issue in a month or so when the bot is
>> actually stable enough to consider sending emails.
>>
>>
>> I think it deserve a separate thread to deal with this, so that the
>> thread title help making sure no-one miss the discussion.
>>
>> —
>> Mehdi
>>
>>
>> _______________________________________________
>> LLVM Developers mailing list
>> llvm-dev at lists.llvm.org
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>>
>>
>
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20170502/3a9420f1/attachment-0001.html>