On Fri, 29 May 2020 at 00:49, Craig Russell <apache....@gmail.com> wrote: > > It looks like the best approach is to do the svn operations in two steps: > > 1. Update the svn foundation/members.txt based on the request (active, > emeritus, deceased) which is already in place. > 2. Perform the moves in svn documents > (emeritus-request-received>emeritus-request-rescinded, > emeritus-request-received->emeritus) which is new code to be written. > > These are two different svn repositories so they need to be in two svn > commits anyway.
Not AFAIK. All the ASF documents are in the same SVN repo, which is rooted at: https://svn.apache.org/repos/private > And if the operations are retried on failure there is nothing to be done to > coordinate the two operations. Meanwhile, I am sure that I do not know how to > make the roster ruby/svn code retriable. A problem arises if the second request fails even after retry. There will be an outstanding change which has to be performed at some later stage. > I plan to start on figuring out how to do svn move from emeritus to > emeritus-request-rescinded. Do you mean emeritus-request => emeritus-request-rescinded ? > Thanks, > Craig > > > On May 26, 2020, at 10:37 PM, Sam Ruby <ru...@intertwingly.net> wrote: > > > > On Wed, May 27, 2020 at 12:38 AM Craig Russell <apache....@gmail.com> wrote: > >> > >> I'm not having any success understanding the memstat.json.rb code: > >> # update members.txt > >> _svn.update members_txt, message: message do |dir, text| > >> > >> Does this call the code in svn.rb? > >> def self.update(path, msg, env, _, options={}) > >> > >> I guess I'm not so good at reading the ruby... > > > > The definition of _svn is here: > > > > https://github.com/apache/whimsy/blob/4e89c0e2a1069fbe7b6f4472722836c12a846e6c/www/roster/models/svn.rb#L29 > > > > It does indeed call the update method in lib/whimsy/asf/svn.rb. > > > > As an aside, the JavaScript module system, like Python, requires each > > module to explicitly import what it needs, this makes it easier to > > identify where the definition of a given method is. Enough so that > > IDEs like VSCode allow you to control-click (or command-click on > > MacOS) on a function call and it will take you to its definition. > > > >> I'm more used to the read/update/write/repeat/commit style of database > >> access, where the memstat.json.rb would > >> - read the members.txt > >> - using the action (emeritus, active, deceased, request_emeritus, > >> rescind_emeritus) edit the members.txt > >> - move the file from emeritus-requests-received to emeritus (emeritus) > >> - remove the file from emeritus-requests-received (rescind) > >> - svn commit the changes > > > > That's indeed how it was done on the original whimsy_vm (effectively > > whimsy_vm1). It routinely would get 'wedged' (that's the term you > > used at the time) and I would have to ssh into that machine, run a > > combination of svn reverts and svn cleanups until the directory was > > clean again, then the operation could be attempted again. Perhaps you > > might remember this happening. > > > > The second iteration creates a fresh checkout of the directory in > > question, runs your block against that directory and commits the > > results. Should the directory get 'wedged', the operation will fail, > > but no matter what happens that temporary directory will be thrown > > away, so the operation can be attempted again. This is the code you > > are now looking at. > > > > The third iteration (present only in the Ruby board agenda tool at the > > moment) takes this a step further and retries operations. Unlike the > > secretary workbench and even to some extent the roster tool, the board > > agenda tool is often used by multiple people concurrently, often with > > changes to the same lines (most notably, the approval lines) on the > > day before the board meeting. If a commit fails, a sleep for a random > > time occurs and the operation is attempted again with a fresh checkout > > and re-invoking the same block. I even have a test tool that > > simulates 9 directors committing at once: > > > > https://github.com/apache/whimsy/blob/master/www/board/agenda/test/stresstest.rb > > > > The fourth iteration (currently being prototyped in > > JavaScript/Node.js) is intended to incorporate the above retry logic > > in production, but will also employ different strategies in > > development and test. > > > > In the fullness of time, all of this should hopefully converge. > > Meanwhile, build on the strategy that you feel most comfortable with, > > and over time it will likely be rewritten as new strategies emerge. > > > >> Craig > > > > - Sam Ruby > > > >>> On May 26, 2020, at 6:09 PM, Sam Ruby <ru...@intertwingly.net> wrote: > >>> > >>> On Tue, May 26, 2020 at 8:44 PM Craig Russell <apache....@gmail.com> > >>> wrote: > >>>> > >>>> Sorry for not being clear. In the whimsy/lib there is a helper file > >>>> svn.rb that has a number of functions including update(path, msg, env, > >>>> _, options={}) that is used by roster memstat.json.rb and it contains > >>>> the commit in the function. > >>>> > >>>> I did not see a move or remove function in svn.rb. Should I be looking > >>>> somewhere else for examples of svn commands? Should I look at adding > >>>> move and remove functions to svn.rb? > >>> > >>> Not currently. > >>> > >>> I'm convinced that the right way forward is for all svn access to be > >>> done via what you are calling helpers. This not only makes the code > >>> cleaner > >>> > >>> Ruby: File.read(File.join(ASF::SVN['foundation_board'], > >>> 'board_agenda_06_17.txt') > >>> JS: Board.read('board_agenda_06_17.txt') > >>> > >>> ... it also enables redirecting all commits to a local repository in > >>> development, and mocking the repository in test. > >>> > >>> In any case, all of the svn operations should be updated to make use > >>> of --password-from-stdin where available, which unfortunately won't be > >>> until whimsy-vm is upgraded to Ubuntu 20.04. > >>> > >>>> Thanks, > >>>> Craig > >>> > >>> - Sam Ruby > >>> > >>>>> On May 26, 2020, at 5:23 PM, Sam Ruby <ru...@intertwingly.net> wrote: > >>>>> > >>>>> On Tue, May 26, 2020 at 8:15 PM Craig Russell <apache....@gmail.com> > >>>>> wrote: > >>>>>> > >>>>>> The cancel button now works (thanks, Sam). On to the server side. > >>>>>> > >>>>>> To implement active->emeritus we need to update members.txt and move > >>>>>> the request file from emeritus-requests-received to emeritus. > >>>>>> > >>>>>> - Is there an svn function to move a file from one directory to > >>>>>> another? > >>>>> > >>>>> http://svnbook.red-bean.com/en/1.6/svn.ref.svn.c.move.html > >>>>> > >>>>>> - Is there a way to have both things done in the same svn commit? > >>>>> > >>>>> If both the source and target have a common parent, do the svn commit > >>>>> from that parent. > >>>>> > >>>>> When the secretary files board minutes, > >>>>> 1) the minutes are moved to the website. Those source and target > >>>>> don't share a common parent, so two commits are required. > >>>>> 2) the agenda is moved to a subdirectory. The source and target share > >>>>> a common parent, so one commit suffices. > >>>>> > >>>>>> To implement request_emeritus_status we need to create an email. Easy.. > >>>>>> > >>>>>> To implement rescind_emeritus_request we need to svn rm the file from > >>>>>> emeritus-requests-received. > >>>>>> > >>>>>> - Is there an svn function to delete a file? > >>>>> > >>>>> http://svnbook.red-bean.com/en/1.6/svn.ref.svn.c.delete.html > >>>>> > >>>>>> Thanks, > >>>>>> Craig > >>>>> > >>>>> - Sam Ruby > >>>>> > >>>>>>> On May 26, 2020, at 3:20 PM, Craig Russell <apache....@gmail.com> > >>>>>>> wrote: > >>>>>>> > >>>>>>> Here are the files. > >>>>>>> <main.js.rb><memstat.js.rb> > >>>>>>> And the console logs: > >>>>>>> [Log] dblclick event.currentTarget: [object HTMLDivElement] (app.js, > >>>>>>> line 3268) > >>>>>>> [Log] dblclick event.currentTarget.dataset: [object DOMStringMap] > >>>>>>> (app.js, line 3269) > >>>>>>> [Log] dblclick event.currentTarget.dataset.edit: memstat (app.js, > >>>>>>> line 3270) > >>>>>>> [Log] canx called from Cancel button (app.js, line 4021) > >>>>>>> [Log] event: [object MouseEvent] parent: [object HTMLButtonElement] > >>>>>>> (app.js, line 4024) > >>>>>>> [Log] parent: [object HTMLButtonElement] memstatElement: [object > >>>>>>> HTMLDivElement] (app.js, line 4025) > >>>>>>> [Log] preventing default (app.js, line 4026) > >>>>>>> [Log] memstatElement.dataset: [object DOMStringMap] (app.js, line > >>>>>>> 4027) > >>>>>>> [Log] memstatElement.dataset.edit: memstat (app.js, line 4028) > >>>>>>> [Log] submit event: [object Object] (app.js, line 3291) > >>>>>>> [Log] submit target: [object HTMLButtonElement] (app.js, line 3292) > >>>>>>> [Log] submit parent: [object HTMLButtonElement] (app.js, line 3293) > >>>>>>> [Log] submit parent.getAttribute(data_cancel): null (app.js, line > >>>>>>> 3294) > >>>>>>> [Log] submit target.dataset_edit: undefined (app.js, line 3295) > >>>>>>> [Log] submit target.dataset: [object DOMStringMap] (app.js, line > >>>>>>> 3296) > >>>>>>> [Log] submit target.dataset.edit: undefined (app.js, line 3297) > >>>>>>> [Log] submit memstatElement.dataset.edit: memstat (app.js, line 3298) > >>>>>>> [Log] submit cancel_submit: null (app.js, line 3300) > >>>>>>> > >>>>>>>> On May 26, 2020, at 2:19 PM, Sam Ruby <ru...@intertwingly.net> wrote: > >>>>>>>> > >>>>>>>> On Tue, May 26, 2020 at 1:32 AM Craig Russell <apache....@gmail.com> > >>>>>>>> wrote: > >>>>>>>>> > >>>>>>>>> But calling preventDefault on the event doesn't appear to do > >>>>>>>>> anything. It still calls the POST behavior, and does not cause the > >>>>>>>>> inline edit menu to disappear. And setting the edit function to nil > >>>>>>>>> doesn't do anything either. > >>>>>>>>> > >>>>>>>>> Maybe there is something else that I need to do? > >>>>>>>> > >>>>>>>> I'll admit that I'm not clear on what you are trying to accomplish, > >>>>>>>> but apparently the problem here is that there is another piece of > >>>>>>>> code > >>>>>>>> that attaches an event handler (and calls preventDefault): > >>>>>>>> > >>>>>>>> https://github.com/apache/whimsy/blob/d3246f107a35f4d989350f4c1ca64366c98ef423/www/roster/views/person/main.js.rb#L341 > >>>>>>>> > >>>>>>>> Perhaps it would be best to add an attribute to the button, and have > >>>>>>>> the submit method remove the buttons and exit early: > >>>>>>>> > >>>>>>>> diff --git a/www/roster/views/person/main.js.rb > >>>>>>>> b/www/roster/views/person/main.js.rb > >>>>>>>> index 1b5ffdb8..f498489f 100644 > >>>>>>>> --- a/www/roster/views/person/main.js.rb > >>>>>>>> +++ b/www/roster/views/person/main.js.rb > >>>>>>>> @@ -355,6 +355,12 @@ class Person < Vue > >>>>>>>> form = jQuery(event.currentTarget).closest('form') > >>>>>>>> target = event.target > >>>>>>>> > >>>>>>>> + # if button is a cancel button, don't submit and remove buttons > >>>>>>>> + if target.getAttribute('data-cancel') > >>>>>>>> + @edit = null > >>>>>>>> + return > >>>>>>>> + end > >>>>>>>> + > >>>>>>>> # serialize form > >>>>>>>> formData = form.serializeArray(); > >>>>>>>> > >>>>>>>> diff --git a/www/roster/views/person/memstat.js.rb > >>>>>>>> b/www/roster/views/person/memstat.js.rb > >>>>>>>> index 39367c40..e8d3d318 100644 > >>>>>>>> --- a/www/roster/views/person/memstat.js.rb > >>>>>>>> +++ b/www/roster/views/person/memstat.js.rb > >>>>>>>> @@ -35,6 +35,8 @@ class PersonMemberStatus < Vue > >>>>>>>> _button.btn.btn_primary 'move to emeritus', > >>>>>>>> name: 'action', value: 'emeritus' > >>>>>>>> end > >>>>>>>> + > >>>>>>>> + _button.btn.btn_secondary 'cancel', data_cancel: true > >>>>>>>> end > >>>>>>>> end > >>>>>>>> end > >>>>>>>> > >>>>>>>> - Sam Ruby > >>>>>>> > >>>>>>> Craig L Russell > >>>>>>> c...@apache.org > >>>>>>> > >>>>>> > >>>>>> Craig L Russell > >>>>>> c...@apache.org > >>>>>> > >>>> > >>>> Craig L Russell > >>>> c...@apache.org > >>>> > >> > >> Craig L Russell > >> c...@apache.org > >> > > Craig L Russell > c...@apache.org >
