Re: [fossil-users] Testing. Was: Two trunks?

[Richard Hipp]

On 4/26/15, j. van den hoff  wrote:
>  some other
> wording would be better I believe. actually the previous description
> "multiple leaves on trunk (or branch XXX)" seems much clearer to me.
>

The alternative-fork-warning branch uses this wording, and it also
shows a list of all the open leaves on the branch in question.
-- 
D. Richard Hipp
d...@sqlite.org
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[bch]

On Apr 26, 2015 1:00 PM, "j. van den hoff" 
wrote:
>
> On Sun, 26 Apr 2015 19:51:44 +0200, Matt Welland 
wrote:
>
>> I like this idea. I will test this branch Monday.
>>
>> +1
>>
>> On Sun, Apr 26, 2015 at 10:16 AM, Jan Nijtmans 
>> wrote:
>>
>>> 2015-04-26 12:54 GMT+02:00 Richard Hipp :
>>> > Yes, but it is not a fork.  And so we shouldn't call it "fossil forks"
>>> > since that would prevent us from creating a "fossil forks" command
>>> > that actually lists real forks.
>>> >
>>> > Perhaps the command should be "fossil warnings" or "fossil concerns"
>>> > and it should report all topological features that are worrisome to
>>> > some users.  (Are there any other graph topology features besides
>>> > multiple leaves on the same branch that people are concerned about?)
>>>
>>> Or, maybe just combine it with "fossil info" and use the more general
>
>
> +1
>
>
>>> term "ambigeous branch" (of which "fork" is a special case)
>
>
> which term many people would not understand, I'm afraid... some other
wording would be better I believe. actually the previous description
"multiple leaves on trunk (or branch XXX)" seems much clearer to me.
>

This touches on something I think should be considered, which is the
"judgment" that the report is expressing. This is why I like our "ticket
system" versus, say, a "bug tracker". The implications are:

1) what are a readers expectations of the output

2) (per 1) what are developers expected to support (managing false
negatives, false positives, etc, etc)

3) the general "tone" of the experience

4) level/placement of responsibility that the tool (fossil) and operator
are assuming

>
>>> 
>>>
>>> Regards,
>>> Jan Nijtmans
>>> ___
>>> fossil-users mailing list
>>> fossil-users@lists.fossil-scm.org
>>> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>>>
>
>
> --
> Using Opera's revolutionary email client: http://www.opera.com/mail/
>
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[j. van den hoff]

On Sun, 26 Apr 2015 19:51:44 +0200, Matt Welland   
wrote:



I like this idea. I will test this branch Monday.

+1

On Sun, Apr 26, 2015 at 10:16 AM, Jan Nijtmans 
wrote:


2015-04-26 12:54 GMT+02:00 Richard Hipp :
> Yes, but it is not a fork.  And so we shouldn't call it "fossil forks"
> since that would prevent us from creating a "fossil forks" command
> that actually lists real forks.
>
> Perhaps the command should be "fossil warnings" or "fossil concerns"
> and it should report all topological features that are worrisome to
> some users.  (Are there any other graph topology features besides
> multiple leaves on the same branch that people are concerned about?)

Or, maybe just combine it with "fossil info" and use the more general


+1


term "ambigeous branch" (of which "fork" is a special case)


which term many people would not understand, I'm afraid... some other  
wording would be better I believe. actually the previous description  
"multiple leaves on trunk (or branch XXX)" seems much clearer to me.






Regards,
Jan Nijtmans
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users




--
Using Opera's revolutionary email client: http://www.opera.com/mail/
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[Matt Welland]

I like this idea. I will test this branch Monday.

+1

On Sun, Apr 26, 2015 at 10:16 AM, Jan Nijtmans 
wrote:

> 2015-04-26 12:54 GMT+02:00 Richard Hipp :
> > Yes, but it is not a fork.  And so we shouldn't call it "fossil forks"
> > since that would prevent us from creating a "fossil forks" command
> > that actually lists real forks.
> >
> > Perhaps the command should be "fossil warnings" or "fossil concerns"
> > and it should report all topological features that are worrisome to
> > some users.  (Are there any other graph topology features besides
> > multiple leaves on the same branch that people are concerned about?)
>
> Or, maybe just combine it with "fossil info" and use the more general
> term "ambigeous branch" (of which "fork" is a special case)
> 
>
> Regards,
> Jan Nijtmans
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[Jan Nijtmans]

2015-04-26 12:54 GMT+02:00 Richard Hipp :
> Yes, but it is not a fork.  And so we shouldn't call it "fossil forks"
> since that would prevent us from creating a "fossil forks" command
> that actually lists real forks.
>
> Perhaps the command should be "fossil warnings" or "fossil concerns"
> and it should report all topological features that are worrisome to
> some users.  (Are there any other graph topology features besides
> multiple leaves on the same branch that people are concerned about?)

Or, maybe just combine it with "fossil info" and use the more general
term "ambigeous branch" (of which "fork" is a special case)


Regards,
Jan Nijtmans
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[Richard Hipp]

On 4/26/15, Jan Nijtmans  wrote:
> Although not matching the definition of fork, it's a potential problem...

Yes, but it is not a fork.  And so we shouldn't call it "fossil forks"
since that would prevent us from creating a "fossil forks" command
that actually lists real forks.

Perhaps the command should be "fossil warnings" or "fossil concerns"
and it should report all topological features that are worrisome to
some users.  (Are there any other graph topology features besides
multiple leaves on the same branch that people are concerned about?)

The "fossil forks" command (if it exists) ought to show bifurcations
in the graph - nodes that have two or more children with the same
branch tag.  I see three possible variants:

 fossil forks --unresolved
 fossil forks --resolved
 fossil forks --all

Perhaps the first of these should be the default, since that seems to
be what people consider to be the most harmful.

-- 
D. Richard Hipp
d...@sqlite.org
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[Jan Nijtmans]

2015-04-25 22:54 GMT+02:00 Richard Hipp :
> On 4/25/15, Jan Nijtmans  wrote:
>> 2015-04-25 18:38 GMT+02:00 Andy Bradford:
>> So, let's start testing:
>>  
>
> The "fossil forks" command applied to SQLite reports these four
> check-ins as forks:
>
> https://www.sqlite.org/src/timeline?c=763d2bc74b&unhide
> https://www.sqlite.org/src/timeline?c=7d445e593a&unhide

Those are two tips, both originating from trunk, which both
has the same branch-name. Although not exactly matching
the definition of fork, they should have been closed
before unhiding. So, yes, they indicate a potential problem,
I would like "fossil forks" to report this. Remedy: both of
those leaves should be closed.

> https://www.sqlite.org/src/timeline?c=cbea02d938&unhide
> https://www.sqlite.org/src/timeline?c=bf6ca21b36&unhide

The same here: there are two tips with branch-name
"branch-3.7.16" which forked off trunk. Although not
matching the definition of fork, it's a potential problem
which I would like to be reported by "fossil forks".
Remedy: close at least the second one (but the
first one probably as well)

Regards,
Jan Nijtmans
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[Andy Bradford]

Thus said Matt Welland on Sat, 25 Apr 2015 19:07:11 -0700:

> bf6ca looks very much like a potential accidentally orphaned commit.

Actually, [bf6ca21b] also  looks like a potential leaf of  a fork. Sure,
it was merged, but it remained an  open leaf, the other side of which is
[cbea02d938]:

$ fossil test-shortest-path -R sqlite.fossil cbea02d938 bf6ca21b
   1: 45602 cbea02d93865 2013-04-12 11:52:43 is a child of
   2: 45603 86f26f915298 2013-04-11 19:04:20 is a child of
   3: 45348 527231bc6728 2013-03-29 13:44:34 is a parent of
   4: 45553 bf6ca21b36ac 2013-04-10 03:22:59

So cbea0 and bf6ca both share  a common parent [527231b], and both share
the same branch tag (branch=branch-3.7.16).

https://www.sqlite.org/src/info/527231b
https://www.sqlite.org/src/timeline?n=200&y=ci&v=0&c=cbea02d938&unhide=

So these two at least look like legitimate forks.


The other two, however, do not appear to be forks.

[7d445e59] is  still an open leaf.  Closing it would likely  cause it to
stop showing  up as a  fork, but it  probably shouldn't be  considered a
fork anyway because it has a new  branch=mistake tag that was added by a
control artifact [31197713]  (shouldn't this have an  extra non-digit in
this?).

So perhaps  the ``fossil  forks'' is not  taking into  consideration any
tags that override the original branch tag on the artifact?

[763d2bc7] is  also a  checkin that was  moved to a  new branch  with an
overriding control artifact.


This  brings up  an interesting  question... ``fossil  forks'' currently
shows the leaves of any forks. Should it be extended to show the parents
instead? e.g. fossil forks --parents

Andy
-- 
TAI64 timestamp: 4000553c5d8a


___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[Matt Welland]

bf6ca looks very much like a potential accidentally orphaned commit. Did
mistachkin commit his changes thinking they were on the branch
branch-3.7.16 and part of the merge that led to sessions? Maybe, maybe not.
The commit should be inspected and either merged, closed or named to a
different branch.

The other fork candidates I'd agree seem to be false positives.

On Sat, Apr 25, 2015 at 5:57 PM, Andy Bradford 
wrote:

> Thus said Richard Hipp on Sat, 25 Apr 2015 16:54:51 -0400:
>
> > https://www.sqlite.org/src/timeline?c=763d2bc74b&unhide
> > https://www.sqlite.org/src/timeline?c=7d445e593a&unhide
> > https://www.sqlite.org/src/timeline?c=cbea02d938&unhide
> > https://www.sqlite.org/src/timeline?c=bf6ca21b36&unhide
> >
> > None of those look like forks to me
>
> This seems  to be  a legitimate bug.  They don't look  like forks  to me
> either.
>
> Andy
> --
> TAI64 timestamp: 4000553c3807
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Testing. Was: Two trunks?

[Andy Bradford]

Thus said Richard Hipp on Sat, 25 Apr 2015 16:54:51 -0400:

> https://www.sqlite.org/src/timeline?c=763d2bc74b&unhide
> https://www.sqlite.org/src/timeline?c=7d445e593a&unhide
> https://www.sqlite.org/src/timeline?c=cbea02d938&unhide
> https://www.sqlite.org/src/timeline?c=bf6ca21b36&unhide
> 
> None of those look like forks to me

This seems  to be  a legitimate bug.  They don't look  like forks  to me
either.

Andy
--
TAI64 timestamp: 4000553c3807
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

[fossil-users] Testing. Was: Two trunks?

[Richard Hipp]

On 4/25/15, Jan Nijtmans  wrote:
> 2015-04-25 18:38 GMT+02:00 Andy Bradford:
> So, let's start testing:
>  

The "fossil forks" command applied to SQLite reports these four
check-ins as forks:

https://www.sqlite.org/src/timeline?c=763d2bc74b&unhide
https://www.sqlite.org/src/timeline?c=7d445e593a&unhide
https://www.sqlite.org/src/timeline?c=cbea02d938&unhide
https://www.sqlite.org/src/timeline?c=bf6ca21b36&unhide

None of those look like forks to me
-- 
D. Richard Hipp
d...@sqlite.org
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users