increasing the coverage of MetacelloBrowser
Locked
 
|
Hi Dale,
I would like to increase the coverage of the browser. However, each test execution pops this message up: -=-=-=-=-=-=-=-=-=-=-=-= You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. ConfigurationOfMBFooTests -=-=-=-=-=-=-=-=-=-=-=-= I have to press many times to execute the tests. Not really convenient. The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. |
Re: increasing the coverage of MetacelloBrowser
|
|
The last time I ran the tests I didn't see those kind of notifiers popping up all of the time, so this is something new or there is something else wrong ... Don't change the tests ... I'll have to see if I can reproduce the problem and try to figure out what happened ...
Dale On Mar 25, 2011, at 7:21 PM, Alexandre Bergel wrote: > Hi Dale, > > I would like to increase the coverage of the browser. However, each test execution pops this message up: > -=-=-=-=-=-=-=-=-=-=-=-= > You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. > > ConfigurationOfMBFooTests > -=-=-=-=-=-=-=-=-=-=-=-= > > I have to press many times to execute the tests. Not really convenient. > The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. > Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? > > Cheers, > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > |
Re: increasing the coverage of MetacelloBrowser
|
|
I just ran the tests and hit the problem and it looks to me like there is a missing or incorrect teardown ... will look at diffs from the last few test checkins and figure out which test got changed last ... it may have been caused by an intermediate checkin since I was transitioning from work in progress at the office and home (had to do a checkpoint checkin before making sure everything was clean) and then I've been too busy working from home to resume my work ... I was working on increasing the test coverage then ...
I'll try to fix this first thing tomorrow morning ... Dale On Mar 25, 2011, at 9:45 PM, Dale Henrichs wrote: > The last time I ran the tests I didn't see those kind of notifiers popping up all of the time, so this is something new or there is something else wrong ... Don't change the tests ... I'll have to see if I can reproduce the problem and try to figure out what happened ... > > Dale > > On Mar 25, 2011, at 7:21 PM, Alexandre Bergel wrote: > >> Hi Dale, >> >> I would like to increase the coverage of the browser. However, each test execution pops this message up: >> -=-=-=-=-=-=-=-=-=-=-=-= >> You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. >> >> ConfigurationOfMBFooTests >> -=-=-=-=-=-=-=-=-=-=-=-= >> >> I have to press many times to execute the tests. Not really convenient. >> The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. >> Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? >> >> Cheers, >> Alexandre >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> >> > |
|
|
Thanks Dale for your swift reply.
Alexandre Le 26 mars 2011 à 00:54, Dale Henrichs <[hidden email]> a écrit : > I just ran the tests and hit the problem and it looks to me like there is a missing or incorrect teardown ... will look at diffs from the last few test checkins and figure out which test got changed last ... it may have been caused by an intermediate checkin since I was transitioning from work in progress at the office and home (had to do a checkpoint checkin before making sure everything was clean) and then I've been too busy working from home to resume my work ... I was working on increasing the test coverage then ... > > I'll try to fix this first thing tomorrow morning ... > > Dale > On Mar 25, 2011, at 9:45 PM, Dale Henrichs wrote: > >> The last time I ran the tests I didn't see those kind of notifiers popping up all of the time, so this is something new or there is something else wrong ... Don't change the tests ... I'll have to see if I can reproduce the problem and try to figure out what happened ... >> >> Dale >> >> On Mar 25, 2011, at 7:21 PM, Alexandre Bergel wrote: >> >>> Hi Dale, >>> >>> I would like to increase the coverage of the browser. However, each test execution pops this message up: >>> -=-=-=-=-=-=-=-=-=-=-=-= >>> You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. >>> >>> ConfigurationOfMBFooTests >>> -=-=-=-=-=-=-=-=-=-=-=-= >>> >>> I have to press many times to execute the tests. Not really convenient. >>> The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. >>> Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? >>> >>> Cheers, >>> Alexandre >>> -- >>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>> Alexandre Bergel http://www.bergel.eu >>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>> >>> >>> >>> >>> >> > |
Re: increasing the coverage of MetacelloBrowser
|
|
Alexandre,
Hmm ... it doesn't appear to be directly related to the MetacelloBrowser tests ... I was in the middle of test development when I ran the full test suite and saw the notifiers as you had reported. This morning I manually removed ConfigurationOfMBFooTests and MBFooTest packages then poked aroun at the suspicious tests without reproducing the problem ... then I ran the full test suite and no notifiiers ... Soooo, my guess is that TestRunner/SUnit might not run the teardown correctly if you stop a test in the middle with a debugger ... it works more often than not, because until last night I had never seen any bad behavior, but I was in the middle of test development so I the testing configuration and package must have been left laying around ... The fact that the test suite runs cleanly once the bogus packages were removed means that the teardown logic is correct ... Will have to keep an eye on things and try to catch the operation that causes the packages to be left around... Dale On Mar 26, 2011, at 5:14 AM, Alexandre Bergel wrote: > Thanks Dale for your swift reply. > > Alexandre > > > > Le 26 mars 2011 à 00:54, Dale Henrichs <[hidden email]> a écrit : > >> I just ran the tests and hit the problem and it looks to me like there is a missing or incorrect teardown ... will look at diffs from the last few test checkins and figure out which test got changed last ... it may have been caused by an intermediate checkin since I was transitioning from work in progress at the office and home (had to do a checkpoint checkin before making sure everything was clean) and then I've been too busy working from home to resume my work ... I was working on increasing the test coverage then ... >> >> I'll try to fix this first thing tomorrow morning ... >> >> Dale >> On Mar 25, 2011, at 9:45 PM, Dale Henrichs wrote: >> >>> The last time I ran the tests I didn't see those kind of notifiers popping up all of the time, so this is something new or there is something else wrong ... Don't change the tests ... I'll have to see if I can reproduce the problem and try to figure out what happened ... >>> >>> Dale >>> >>> On Mar 25, 2011, at 7:21 PM, Alexandre Bergel wrote: >>> >>>> Hi Dale, >>>> >>>> I would like to increase the coverage of the browser. However, each test execution pops this message up: >>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>> You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. >>>> >>>> ConfigurationOfMBFooTests >>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>> >>>> I have to press many times to execute the tests. Not really convenient. >>>> The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. >>>> Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? >>>> >>>> Cheers, >>>> Alexandre >>>> -- >>>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>>> Alexandre Bergel http://www.bergel.eu >>>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>>> >>>> >>>> >>>> >>>> >>> >> |
|
|
Oh, strange. I just run them, and it works.
I will work on increasing the coverage then... Alexandre On 26 Mar 2011, at 13:12, Dale Henrichs wrote: > Alexandre, > > Hmm ... it doesn't appear to be directly related to the MetacelloBrowser tests ... I was in the middle of test development when I ran the full test suite and saw the notifiers as you had reported. > > This morning I manually removed ConfigurationOfMBFooTests and MBFooTest packages then poked aroun at the suspicious tests without reproducing the problem ... then I ran the full test suite and no notifiiers ... > > Soooo, my guess is that TestRunner/SUnit might not run the teardown correctly if you stop a test in the middle with a debugger ... it works more often than not, because until last night I had never seen any bad behavior, but I was in the middle of test development so I the testing configuration and package must have been left laying around ... > > The fact that the test suite runs cleanly once the bogus packages were removed means that the teardown logic is correct ... > > Will have to keep an eye on things and try to catch the operation that causes the packages to be left around... > > Dale > > On Mar 26, 2011, at 5:14 AM, Alexandre Bergel wrote: > >> Thanks Dale for your swift reply. >> >> Alexandre >> >> >> >> Le 26 mars 2011 à 00:54, Dale Henrichs <[hidden email]> a écrit : >> >>> I just ran the tests and hit the problem and it looks to me like there is a missing or incorrect teardown ... will look at diffs from the last few test checkins and figure out which test got changed last ... it may have been caused by an intermediate checkin since I was transitioning from work in progress at the office and home (had to do a checkpoint checkin before making sure everything was clean) and then I've been too busy working from home to resume my work ... I was working on increasing the test coverage then ... >>> >>> I'll try to fix this first thing tomorrow morning ... >>> >>> Dale >>> On Mar 25, 2011, at 9:45 PM, Dale Henrichs wrote: >>> >>>> The last time I ran the tests I didn't see those kind of notifiers popping up all of the time, so this is something new or there is something else wrong ... Don't change the tests ... I'll have to see if I can reproduce the problem and try to figure out what happened ... >>>> >>>> Dale >>>> >>>> On Mar 25, 2011, at 7:21 PM, Alexandre Bergel wrote: >>>> >>>>> Hi Dale, >>>>> >>>>> I would like to increase the coverage of the browser. However, each test execution pops this message up: >>>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>>> You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. >>>>> >>>>> ConfigurationOfMBFooTests >>>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>>> >>>>> I have to press many times to execute the tests. Not really convenient. >>>>> The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. >>>>> Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? >>>>> >>>>> Cheers, >>>>> Alexandre >>>>> -- >>>>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>>>> Alexandre Bergel http://www.bergel.eu >>>>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>> > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. |
Re: increasing the coverage of MetacelloBrowser
|
|
Excellent ... I have some work in progress that I should check in before you get too far ... rename of a command class and I moved update dev from the MetacelloBrowser class to cmd class...hadn't started writing test for that ... got sidetracked, but you should start eith that...
Dale On Mar 26, 2011, at 11:16 AM, Alexandre Bergel wrote: > Oh, strange. I just run them, and it works. > I will work on increasing the coverage then... > > Alexandre > > > On 26 Mar 2011, at 13:12, Dale Henrichs wrote: > >> Alexandre, >> >> Hmm ... it doesn't appear to be directly related to the MetacelloBrowser tests ... I was in the middle of test development when I ran the full test suite and saw the notifiers as you had reported. >> >> This morning I manually removed ConfigurationOfMBFooTests and MBFooTest packages then poked aroun at the suspicious tests without reproducing the problem ... then I ran the full test suite and no notifiiers ... >> >> Soooo, my guess is that TestRunner/SUnit might not run the teardown correctly if you stop a test in the middle with a debugger ... it works more often than not, because until last night I had never seen any bad behavior, but I was in the middle of test development so I the testing configuration and package must have been left laying around ... >> >> The fact that the test suite runs cleanly once the bogus packages were removed means that the teardown logic is correct ... >> >> Will have to keep an eye on things and try to catch the operation that causes the packages to be left around... >> >> Dale >> >> On Mar 26, 2011, at 5:14 AM, Alexandre Bergel wrote: >> >>> Thanks Dale for your swift reply. >>> >>> Alexandre >>> >>> >>> >>> Le 26 mars 2011 à 00:54, Dale Henrichs <[hidden email]> a écrit : >>> >>>> I just ran the tests and hit the problem and it looks to me like there is a missing or incorrect teardown ... will look at diffs from the last few test checkins and figure out which test got changed last ... it may have been caused by an intermediate checkin since I was transitioning from work in progress at the office and home (had to do a checkpoint checkin before making sure everything was clean) and then I've been too busy working from home to resume my work ... I was working on increasing the test coverage then ... >>>> >>>> I'll try to fix this first thing tomorrow morning ... >>>> >>>> Dale >>>> On Mar 25, 2011, at 9:45 PM, Dale Henrichs wrote: >>>> >>>>> The last time I ran the tests I didn't see those kind of notifiers popping up all of the time, so this is something new or there is something else wrong ... Don't change the tests ... I'll have to see if I can reproduce the problem and try to figure out what happened ... >>>>> >>>>> Dale >>>>> >>>>> On Mar 25, 2011, at 7:21 PM, Alexandre Bergel wrote: >>>>> >>>>>> Hi Dale, >>>>>> >>>>>> I would like to increase the coverage of the browser. However, each test execution pops this message up: >>>>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>>>> You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. >>>>>> >>>>>> ConfigurationOfMBFooTests >>>>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>>>> >>>>>> I have to press many times to execute the tests. Not really convenient. >>>>>> The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. >>>>>> Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? >>>>>> >>>>>> Cheers, >>>>>> Alexandre >>>>>> -- >>>>>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>>>>> Alexandre Bergel http://www.bergel.eu >>>>>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>> >> > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > |
|
|
In reply to this post by abergel
Dale,
For my needs, I would like have a complete traceability of every single incremental commit of the browser. I will produce many 1.58.X versions. I hope this is ok with you. Comment of 1.58.1: - broke addBaseline into small methods. The methods was quite complex before 1.58.2 will test all the small methods ... Cheers, Alexandre On 26 Mar 2011, at 14:16, Alexandre Bergel wrote: > Oh, strange. I just run them, and it works. > I will work on increasing the coverage then... > > Alexandre > > > On 26 Mar 2011, at 13:12, Dale Henrichs wrote: > >> Alexandre, >> >> Hmm ... it doesn't appear to be directly related to the MetacelloBrowser tests ... I was in the middle of test development when I ran the full test suite and saw the notifiers as you had reported. >> >> This morning I manually removed ConfigurationOfMBFooTests and MBFooTest packages then poked aroun at the suspicious tests without reproducing the problem ... then I ran the full test suite and no notifiiers ... >> >> Soooo, my guess is that TestRunner/SUnit might not run the teardown correctly if you stop a test in the middle with a debugger ... it works more often than not, because until last night I had never seen any bad behavior, but I was in the middle of test development so I the testing configuration and package must have been left laying around ... >> >> The fact that the test suite runs cleanly once the bogus packages were removed means that the teardown logic is correct ... >> >> Will have to keep an eye on things and try to catch the operation that causes the packages to be left around... >> >> Dale >> >> On Mar 26, 2011, at 5:14 AM, Alexandre Bergel wrote: >> >>> Thanks Dale for your swift reply. >>> >>> Alexandre >>> >>> >>> >>> Le 26 mars 2011 à 00:54, Dale Henrichs <[hidden email]> a écrit : >>> >>>> I just ran the tests and hit the problem and it looks to me like there is a missing or incorrect teardown ... will look at diffs from the last few test checkins and figure out which test got changed last ... it may have been caused by an intermediate checkin since I was transitioning from work in progress at the office and home (had to do a checkpoint checkin before making sure everything was clean) and then I've been too busy working from home to resume my work ... I was working on increasing the test coverage then ... >>>> >>>> I'll try to fix this first thing tomorrow morning ... >>>> >>>> Dale >>>> On Mar 25, 2011, at 9:45 PM, Dale Henrichs wrote: >>>> >>>>> The last time I ran the tests I didn't see those kind of notifiers popping up all of the time, so this is something new or there is something else wrong ... Don't change the tests ... I'll have to see if I can reproduce the problem and try to figure out what happened ... >>>>> >>>>> Dale >>>>> >>>>> On Mar 25, 2011, at 7:21 PM, Alexandre Bergel wrote: >>>>> >>>>>> Hi Dale, >>>>>> >>>>>> I would like to increase the coverage of the browser. However, each test execution pops this message up: >>>>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>>>> You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. >>>>>> >>>>>> ConfigurationOfMBFooTests >>>>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>>>> >>>>>> I have to press many times to execute the tests. Not really convenient. >>>>>> The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. >>>>>> Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? >>>>>> >>>>>> Cheers, >>>>>> Alexandre >>>>>> -- >>>>>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>>>>> Alexandre Bergel http://www.bergel.eu >>>>>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>> >> > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. |
|
|
Comment of 1.58.4:
- We went from 58% to 60% of coverage. Go go go!!! Dale, I recommend you to try out what I did. I did some refactoring of MetacelloBrowser: some methods were really complex and I cut them down. All the tests are green, but you may want to check the OB version. I will continue tomorrow. Cheers, Alexandre On 26 Mar 2011, at 20:49, Alexandre Bergel wrote: > Dale, > > For my needs, I would like have a complete traceability of every single incremental commit of the browser. > I will produce many 1.58.X versions. > I hope this is ok with you. > > Comment of 1.58.1: > - broke addBaseline into small methods. The methods was quite complex before > > 1.58.2 will test all the small methods ... > > Cheers, > Alexandre > > > On 26 Mar 2011, at 14:16, Alexandre Bergel wrote: > >> Oh, strange. I just run them, and it works. >> I will work on increasing the coverage then... >> >> Alexandre >> >> >> On 26 Mar 2011, at 13:12, Dale Henrichs wrote: >> >>> Alexandre, >>> >>> Hmm ... it doesn't appear to be directly related to the MetacelloBrowser tests ... I was in the middle of test development when I ran the full test suite and saw the notifiers as you had reported. >>> >>> This morning I manually removed ConfigurationOfMBFooTests and MBFooTest packages then poked aroun at the suspicious tests without reproducing the problem ... then I ran the full test suite and no notifiiers ... >>> >>> Soooo, my guess is that TestRunner/SUnit might not run the teardown correctly if you stop a test in the middle with a debugger ... it works more often than not, because until last night I had never seen any bad behavior, but I was in the middle of test development so I the testing configuration and package must have been left laying around ... >>> >>> The fact that the test suite runs cleanly once the bogus packages were removed means that the teardown logic is correct ... >>> >>> Will have to keep an eye on things and try to catch the operation that causes the packages to be left around... >>> >>> Dale >>> >>> On Mar 26, 2011, at 5:14 AM, Alexandre Bergel wrote: >>> >>>> Thanks Dale for your swift reply. >>>> >>>> Alexandre >>>> >>>> >>>> >>>> Le 26 mars 2011 à 00:54, Dale Henrichs <[hidden email]> a écrit : >>>> >>>>> I just ran the tests and hit the problem and it looks to me like there is a missing or incorrect teardown ... will look at diffs from the last few test checkins and figure out which test got changed last ... it may have been caused by an intermediate checkin since I was transitioning from work in progress at the office and home (had to do a checkpoint checkin before making sure everything was clean) and then I've been too busy working from home to resume my work ... I was working on increasing the test coverage then ... >>>>> >>>>> I'll try to fix this first thing tomorrow morning ... >>>>> >>>>> Dale >>>>> On Mar 25, 2011, at 9:45 PM, Dale Henrichs wrote: >>>>> >>>>>> The last time I ran the tests I didn't see those kind of notifiers popping up all of the time, so this is something new or there is something else wrong ... Don't change the tests ... I'll have to see if I can reproduce the problem and try to figure out what happened ... >>>>>> >>>>>> Dale >>>>>> >>>>>> On Mar 25, 2011, at 7:21 PM, Alexandre Bergel wrote: >>>>>> >>>>>>> Hi Dale, >>>>>>> >>>>>>> I would like to increase the coverage of the browser. However, each test execution pops this message up: >>>>>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>>>>> You are about to load new versions of the following packages that have unsaved changes in the image. If you continue, you will lose these changes. >>>>>>> >>>>>>> ConfigurationOfMBFooTests >>>>>>> -=-=-=-=-=-=-=-=-=-=-=-= >>>>>>> >>>>>>> I have to press many times to execute the tests. Not really convenient. >>>>>>> The problems comes from MBConfigurationUpdateDevCmdMetacelloTests>>setUp. >>>>>>> Should I unload/deregister the package 'ConfigurationOfMBFooTests' at the beginning of the setUp? >>>>>>> >>>>>>> Cheers, >>>>>>> Alexandre >>>>>>> -- >>>>>>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>>>>>> Alexandre Bergel http://www.bergel.eu >>>>>>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>> >>> >> >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> >> > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. |
Re: increasing the coverage of MetacelloBrowser
|
|
In reply to this post by abergel
On Mar 26, 2011, at 5:49 PM, Alexandre Bergel wrote: > Dale, > > For my needs, I would like have a complete traceability of every single incremental commit of the browser. Could you please explain your "needs" ... I sent mail a while back explaining my thoughts on this and you did not reply, so I assume that you didn't actually read my mail ... if you did, please reply to my arguments and explain why you think you are getting "complete traceability" by creating a metacello version per mcz file commi .. BTW, I would appreciate it if you would add commit comments when you check in a configuration ... not much information there when the comment is blank... Dale |
|
|
> Could you please explain your "needs" ... I sent mail a while back explaining my thoughts on this and you did not reply, so I assume that you didn't actually read my mail ...
No, I have an important deadline at the beginning of April. I think I replied to your email. > if you did, please reply to my arguments and explain why you think you are getting "complete traceability" by creating a metacello version per mcz file commi .. Easy scenario: I want to be able to retrieve the version of the browser with a given coverage. By having all the versions in the configuration, I can just press 'load' and get back to the state where the coverage was 58% for example. I found very cumbersome to have to load the right configuration in which 1.57 was at 58% covered. Again, I am probably the one who have read the most your post about your process. It indeed works (we have been using it for the browser). However, I still prefer to have an explicit version per increment. And reading your post again will probably not change my mind. I will probably finish my experiment tomorrow or Tuesday. If you feel that the 1.58.X versions are not necessary or confusing, then we can just remove them on Wednesday. Until then, I really need to be able to quickly load a version in which, for example, MBInfo was not covered. And crawling over each version of the configuration in the squeaksource repository is really not optimal. I will probably do a script like "ConfigurationOfMetacelloBrowser from: '1.58.1' to: '1.58.10' do: [ ... ]" > BTW, I would appreciate it if you would add commit comments when you check in a configuration ... not much information there when the comment is blank... Yes. I find the comment mechanism in the browser a bit insufficient. I found redundant to have to specify a comment for the metacello version, another for each of packages that has to be saved and another comment for the configuration. You right, I should pay attention for the configuration, I will. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. |
Re: increasing the coverage of MetacelloBrowser
|
|
On Mar 27, 2011, at 6:59 PM, Alexandre Bergel wrote: >> Could you please explain your "needs" ... I sent mail a while back explaining my thoughts on this and you did not reply, so I assume that you didn't actually read my mail ... > > No, I have an important deadline at the beginning of April. > I think I replied to your email. > >> if you did, please reply to my arguments and explain why you think you are getting "complete traceability" by creating a metacello version per mcz file commi .. > > Easy scenario: I want to be able to retrieve the version of the browser with a given coverage. By having all the versions in the configuration, I can just press 'load' and get back to the state where the coverage was 58% for example. I found very cumbersome to have to load the right configuration in which 1.57 was at 58% covered. > > Again, I am probably the one who have read the most your post about your process. It indeed works (we have been using it for the browser). However, I still prefer to have an explicit version per increment. And reading your post again will probably not change my mind. I will probably finish my experiment tomorrow or Tuesday. If you feel that the 1.58.X versions are not necessary or confusing, then we can just remove them on Wednesday. Until then, I really need to be able to quickly load a version in which, for example, MBInfo was not covered. And crawling over each version of the configuration in the squeaksource repository is really not optimal. > > I will probably do a script like "ConfigurationOfMetacelloBrowser from: '1.58.1' to: '1.58.10' do: [ ... ]" Ah, you are interested in explicit checkpoints for a specific purpose? Then I agree that it is not a bad use of versions ... as long as there is an external reason for creating a version.. I disagree with the general practice of creating a Metacello version for each checkin for each unit of work, for simple historical information. When multiple people are working on a project it is not easy to merge changes across metacello versions with no tools support it is a completely manual operation . No need to remove the versions ... might as well keep them around ... > >> BTW, I would appreciate it if you would add commit comments when you check in a configuration ... not much information there when the comment is blank... > > > Yes. I find the comment mechanism in the browser a bit insufficient. I found redundant to have to specify a comment for the metacello version, another for each of packages that has to be saved and another comment for the configuration. You right, I should pay attention for the configuration, I will. The checkpoint dev command takes a single comment for packages, descriptions and config commit message ... the checkpoint dev supports a particular use case: when checkpointing a unit of work it makes sense that the same comment be used in all three places... If you are doing each thing independently then it can be tedious, but I assume that there are generally useful use cases that can be added to the tool, to make the chores less tedious... Dale |
|
|
> Ah, you are interested in explicit checkpoints for a specific purpose? Then I agree that it is not a bad use of versions ... as long as there is an external reason for creating a version..
My experiment is almost done. I will ping you when it is. > I disagree with the general practice of creating a Metacello version for each checkin for each unit of work, for simple historical information. When multiple people are working on a project it is not easy to merge changes across metacello versions with no tools support it is a completely manual operation . Indeed, merge is problematic. I agree with this. Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. |
«
Return to Metacello
|
1 view|%1 views
| Free forum by Nabble | Edit this page |