Theme: AI

OVERSING UPDATE FOR JANUARY 2016 – HAPPY NEW YEAR. 🙂

I love our product Oversing. I wish I had ten more guys and six more months to add the few nits I think would make it easier for newbies to work with, and add a lot more reports. But I love our product. Damn. Really.

It’s so interesting – it’s an essay on the evolution of technology – that no one started with task and project management and added email to it, and instead, started with email and tried to incorporate tasks into it. I guess we could say that Lotus sort of tried.

Here is a list of things I told board and investors I want to add next (this) year to make it easier to use for newbies.

IMPERFECT FEATURES

There are a few issues I don’t like but we can fix later.

(a) Technically, an appointment is just a subtask for any task. But because of my own legacy thinking from our TIMESYS application, we implemented forecasting/appointments as a pseudo-task rather than as a subtask of a task. For a variety of reasons we wanted appointments to be configured as a general business rule independent of the program/project. So we separated data and workflow. This wasn’t necessary. We could have maintained a universal appointment workflow and program/project specific task workflows. This decision has added unnecessary complexity and we might want to change that in the future.

(b) We need more granular ability to share workflow choices between task types. Right now they are ‘cloned’ not ‘shared’ and this is a mistake. It was just the easiest solution. What we need instead is the ability to point a workflow choice (a workflow button) at a single configuration and share it with any of the workflow entity types. And we need a ‘report’ that shows the contents of these choices so that users can edit complex sets with greater ease.

(c) I want the Board View to hide subtasks of Stories / Problems / Opportunities / Challenges (the ‘Story’ type). (I am losing this battle with Kirill for some reason.) My vision is that the user works with stories as discreet units: as sets that must be completed. And that he does not work ‘across’ stories. in other words, a story is a micro-project or micro-job component of a deliverable (finished state).

(d) the decision to use the backlog for both sprint/iteration backlog and project backlog instead of creating a separate backlog for sprints/iteration(work in progress) has been a mistake and has created unnecessary complexity in the application and unnecessary confusion for the user. I want to change that. (Another battle I have lost. ) In our earliest draft this made more sense than it does today.

(e)I thought that the decision to use a [Scheduled] “state” rather than just a flag was a mistake that I made to save memory in the gantt chart. But this is merely a corollary to the backlog vs todo-backlog in that it’s ‘on the way out’ just as the todo-backlong is ‘on the way in’. I should have figured this out earlier but believe it or not I am as great a victim of priors as anyone else. 🙂

(f) There are still problems with cascading states between parent and child containers. And there are still problems with cascading dependencies on the gantt chart. This is because we do checking only on the backend, rather than both front and back, and reverse it if the back has changed. (a cost cutting measure I took at the time on purpose but maybe shouldn’t have).

(g) Customer/Client needs to be renamed Company/Contact, and the relationship between contact and user better integrated as a one user to many contact relationship. This again is a hangover from our first draft when we did not anticipate including customers clients and contacts as users.

(h) My attempt to encroach on Slack’s functionality probably added confusion rather than clarity. At the top of our activity stream we included the ability to make comments without opening a page and entering the comment in context. I increasingly feel that this should be reduced to a command line for advanced users at the bottom of the page.

MISSING FEATURES THAT MATTER

There are some features that we are missing that I believe hurt us, but I had to cut.

(a) The subtasks tab’s single-depth list of tasks should be added to the workspace. The outline invites abuse of the agile method and I want to answer that, as well as give people a very simple list view to work from without any of the monetary and scheduling values displayed.

(b) Likewise a view of ‘trello’ style columns rather than the full board would help new users. Trello’s visual list-architecture is halfway between agile boards and lists. And for newer users it’s really useful, since it removes discretion from the person doing the work, whereas the full agile board view is more of a funnel approach consisting of a lot of discretion on behalf of the person doing the work. I find myself creating lots more deliverables and relying upon the Tree (outline) view simply because ordering work rather than leaving the sequence of work to the person doing it is not really possible on the board view. Now, the board view is ‘ordered’ separately from the outline view. But without single columns it’s not intuitive that the board view consists of ordered lists just as the outline view is an ordered and hierarchical list.

(c) Next, the “Views” feature that retains the same program selections but changes panels and filters. I should have pressed for this earlier. Sharing of views, and better sharing of favorites, workspaces, appears to be necessary to get to the level of simplicity I feel we need for users to incrementally learn the product. IE: we want to share Workspaces (layouts and selections), Views (layouts), and Favorites (individual items). By doing so we allow the users to create custom experiences for different roles in the company.

(c) We had to cut the content management feature (a content tab, and the integration with WordPress) and I think this single feature assists in positioning us in the market quite differently by leapfrogging Atlassian. All the “document” components and workflow are there (Folder,Document, Section, Article, Card, Citation, Definition) . But we could not even start working on the api for displaying them in web sites.

(d) For simplicity purposes we did not use amazon’s search functionality (as did Slack) for our Activity Stream and Search functionality. I can ‘sense’ this when using the activity stream. It’s just work. We can do it pretty easily with a little time.

(e) Most importantly perhaps, is the fact that the gantt chart functionality is weak without visibility of appointments (scheduled resources) and this must be done for the rather amazing resourcing functionality to be easily used by managers of many projects – common in big companies. This is the only one that gives me consternation because I know from experience that it will take some time to implement.

WHY “DATE STUFF” IS HARDER FOR US TO PROGRAM IN OVERSING

Well, you know, we have one large architectural difference in that ‘scheduling’ can be done a)any time during a week, b) any time during the day c) at a specific time during a day. And that one can forecast time loosely (the equivalent of a PM sprint budgets), or schedule it (reserve someone’s time). So nothing on the market that does all this and we have to write or rewrite all these libraries by hand. (and the truth is everything we find on the market leaks like a sieve anyway).

CLOSING

Anyway, I thought I would share. (I love sharing. lol)

We are just knocking out bugs, almost all of which are usability issues or tweaks. And trying to get the reports done. (Honestly I have postponed a lot of very minor language or clarity issues because they won’t inhibit use enough to delay release.)

I was wrong about reports. There are just four categories of reports – although many variations within each – and we can let people start using it as soon as the primary report in each category is completed.

We did it. We created a new generation “erp” at the high end of the current market, and with the ambitions of reaching the top of the market, with just four of us really, over two and a half years.

Pretty crazy. But over the next two or three years we are going to make something very amazing that will change the workplace forever by eliminating politics through transparency and truthfulness. This will advance healthy organizations, improve the health of marginal organizations, and illustrate the failings of dysfunctional organizations.

***When human capital is the only remaining marginal difference, then the culture of one’s organization is the only remaining competitive advantage.***

OVERSING UPDATE.

Well. My second update today.

We purchased a well known gannt chart control from the best firm and it leaks like hell.

At this point we may have to ( as usual) write something ourselves that doesn’t leak.

And maybe that’s ok.

OVERSING UPDATE

We are down to 111 open issues that we must close before we open the beta program, and of them 56 are reports we need to complete (50%). Leaving 55 issues that are sources of confusion for users because they do not work as expected.

There are two controls: timecards and forecasts that are not right (they need full workflows, rather than the simple states).

The workspace is not displaying the management query when no programs, projects, or deliverables are selected (meaning

Saving the workspace is still a non-standard operation.

The Calendar Control on the workspace isn’t quite right.

The Progress Bars on the schedule panel aren’t updating

Drag and Drop isn’t working on Table View.

The subtasks of Stories are still showing on the board view.

The tasks that need scheduling this week aren’t correctly displaying on the panel.

Smart Search is missing some variables.

Permissions on who can configure workflows.

There is a usability change we want to make, which is to move the task-types into a panel, and out from under the task filter panel so that we can show counts.

We want to break the ribbon for selecting states into stages so the user more clearly sees the relationship between stages(which cross programs and projects) to stages which are specific to programs and projects.)

NITS

The rest are largely nits, most of which fall into usability.

BIG STUFF

Web site content.

Documentation.

I”M NOT SURE

That costs are working correctly.