Forum OpenACS Q&A: Who wants to built .WRK?
I'm picturing a system that pulls together the best from OpenACS
and .LRN, configured for intranet and extranet out of the box, with
all the relevant applications added and portletized, with some
groups setup for employees, partners, customers, with default portal
layouts, with a setup wizard that guides you through creating users,
uploading your company logo, etc. And administrator and end-user
We don't need all of the above right away, but I'll definitely be
moving towards making this happen in a very short timeframe.
Are there other people out there who would be interested in sharing
the burden on this work? Then I think we should team up and flesh
out the work that needs to be done, so we don't waste effort by
I feel like we've got to move fast if we want to make this a strong
intranet platform. The market for closed source intranet
collaboration software already seems settled, the open source market
may not be just yet.
I'll put together a plan for what I think needs to be done. And I'm going to put up a server where I play with putting it together.
Ben, it would be really useful if you could give us a short description of what you think would be necessary. After all, you're the main architect behind .LRN, on which we'll be building.
We should also set up a forum somewhere. Either a bboard on openacs.org, or something more involved on the new openacs.org. Of course, this is the kind of thing that .WRK would be ideally suited for :)
Expect more from me next week.
From a technical standpoint most of us know that this means that we need to add another step to the installation process that loads parameters into the database and/or writes to .info files, etc.
I would be interested in being involved.
I agree, and I don't.
My model is the Danish company Groupcare (www.groupcare.biz). They're not specialized in any vertical, yet they seem to have a pretty good business. Of course, they went after this market much ealier and have really strong marketing here. They also happen to be business school people, not engineers.
I think if we can have what they have, plus a bunch of things they don't, like Wimpy, we'll be in good shape.
I also think that we basically have all the components that we need, we just need to package and configure it, clean up the user experience, polish a couple of the important packages that are still lacking, and write user documentation.
Then we can build verticals and add functionality on that basis.
Having clients/verticals in mind can help focus technical development by finding out what people really want before building it. Designing for specific contrasting verticals can help flesh out true generalized functionality. It can also lead to partnerships with actual clients/users that can help .WRK adoption down the road.
We agree on what needs to be done. I'm just suggesting an additional technical step (adding pre-packaged configuration functionality) and essentially a formalized process of tracking user scenarios (usually a design step) that allows us to reuse the information to build the pre-packaged configurations.
I want a generalized .WRK. One that can easily be used by several verticals without having to be refactored. We already have a lot going for us with .LRN and the other things you mentioned.
I think we probably should setup a project subsite for .WRK over at dev.openacs.org. There was some concern about splitting development discussions into seperate forums instances, so I am not sure exactly how we want to set it up.
Chances are this type of project will improve packages outside of the immediate .WRK project including the core.
This is great ! Count me in !
The products called eRooms (don;t know if anyone familiar with it) at www.eroom.net.
I've played with it a bit and too be honest its a prime candidate fro taking on. It seamingly only really implements glorifed bulletin boards and chat.. with a bit of file management.. but, as its licensing policy is on a 'per number of rooms' basis, as you can see it gets expensive as an organisation grows! An open source solution to rival it has numerous USPs...
So... thats at least partly what I'd like to see in a dotWRK..
There's also inclusion.net, formerly Planet Intra, also known as Intra.Net, which looks pretty good. I particularly like their home page for talking about deployment phases, verticals, what value the customer will get from using this software, and a short ROI story.
There's an old demo of an old system up at http://yourcompany.intra.net/
that's another thing (besides CMS) that needs to be streamlined in the future.
-Berklee is using volano (http://www.volano.com), which is proprietary.
-Malte will improve his jabber solution (http://www.sussdorff-roy.com/jabber/jabber_inst_main_doc.html) and release it around the end of the year.
Anybody else doing anything on chat solutions?
Re: Chat, that's definitely something to discuss, but I'm more concerned about webmail. I haven't followed its development. What's it up to now?
What I meant to say is that they may lack some functionality, but what they have is very clean, easy to use, it seems finished, polished. They also have this downloadable component thing, that lets you drag and drop documents and stuff.
Just to clarify.
A triumph of presentation over substance! ;0)
Isn't that like saying 'I have the great new microwave, really quiet, nice looking, dead reliable.... mind you it only cooks chicken.'?
No, seriously you're right. Not disagreeing that its well presented and resonable smooth (despite crashes, but heh, windoze..) but I fidn their marketing/claims to somewhat overstate the case.
The feedback I've had is that its restrictive, too narrow and too expensive, given that this (v.large) client had been given another impression....
But then also shows a good way to postion any .WRK. As functionally general and feature rich platform and allow the 'presentation' to be a bespoke element (as often this is what a customer means by configurable).
So... is someone organising some effort on this (v.keen), firstly does anyone own dotWRK? if not should someone register it? OF perhaps, or we can if you like. Probly only covers the UK though.
I don't think its a million miles off dotLRN is some senses, so perhaps it'd be quite easy to 'dive-in' and experiment with some development on it?
I'm available to help, if the product we develop works for the company I'm at.
already been done)?
I'd like to set up a project workspace on dev.openacs.org, so if one of the dev.openacs.org hackers is reading this and would create a space for us, that'd be hunky-dory. Thanks! :)
Here's a partial to-do list of things that I've thought of. When we get the project space up, it'll be easier to keep track of everything and develop this collaboratively. It goes without saying that we're building this on dotLRN.
1) Configure for intranet use. This includes at a minimum changing text at the level of changing "class" to "project". The extended version would be to factor out the class community into a separate package and have the dotlrn core package not know anything about that community type in particular. Ben said he estimated this at 2 solid hacker-weeks worth of work for someone who knows the stuff intimately. So definitely not trivial.
2) Simple CMS: Edit-this-page, ability to attach a directory of static pages to a community, something is needed. Again, we can do something quick for now, then expand later.
What I'm looking for, though, is really something that's compelling out of the box, so I'm going to add these things, too:
3) Graphic design: Make it look good. Don't shoot me! I think this would really, really make a difference. The quick solution would be to slap on some templates, for example using those from Sloanspace as a starting point, if they're willing to share. Deeper version would be to find a solution that would work for all of OpenACS, with skins, widgets, site-wide consistent design, etc.
4) User interface cleanup: Group admin UI is the classic, but let's walk through the whole UI and clean it up, streamline it.
5) And then let's configure it out of the box with an Employee group, departments groups, the ability to invite people to join your project group (extranet?), perhaps there's a default "company" community that spans the whole company, with subcommunities for each department ... I haven't thought this through all that much, but you get the picture.
There's lots more we can add, obviously, but I think this are the critical things that I think would make it compelling for more people to buy into this platform.
I'm researching a number of competing intranet systems these days, and I'll come up with my thoughts for what this thing should look like.
I think there'll be plenty of room for different people pushing different visions, all variations on a theme, and for this thing developing in really exciting ways over time.
My focus right now is on doing the least amount of work possible that we need to do to start attracting people's attention (and money) for real.
6) Write documentation/online help text for both the administrator user ("Congratulations, you just installed dotWRK, now here's how you get started using it ...") and end-user documentation (for the employees of the company using this intranet).
I heard from several people who use it that a Wiki has proven to be an extremely useful tool in their intranets.
Imagine a Wiki that's integrated with OpenACS, like the versioning of the content repository, search, linking to arbitrary objects - wouldn't that be cool? Sorry, just a note to all those hackers out there that have too much time
On items listed/suggested above (by Lars) -- are you sure that, say, any sort of project-tracking could be done via existing class functionality of dotLRN? I somehow doubt it, especially if one were to consider nesting of projects, sub-projects, sub-sub-projects, ..., (n*sub)-projects... This is something that was lacking in ACS-classic project management -- only 2 levels (a parent and a sub) were possible...
I'll try to get the time to do a few mockups, maybe even over the weekend.
Anyway, in a prototype for our organization's management to see, I had mapped our Org Chart into communities with the following structure:
- dotMECH (entire org)
For Terms, I used Fiscal Year boundaries (Oct->Sep in our case). Department was a singleton to represent our entire org. Subject was each customer (internal or external) for whom a project was being done.
Each Class was a discrete project, changing
- Professor -> Manager (not necessarily Proj Mgr). This is the Supervisor/Manager ultimately resp.
- Course Administrator -> Customer Service Rep
- Course Assistant -> Proj Mgr
- Teaching Assist -> Tech Lead(s)
- Student -> Team members I had looked around the (Oracle) SQL to see where to change labeling to reflect non-LRN use, but stopped to start looking at some stuff with ticket-tracker, workflow and lars-blogger to see how to fit them better into a "unified" project planning, tasking, and reporting environment (This was for another presentation to my mgmt).
Then ideally would be to "portlet" them to make all that available to the dotWRK environment.
I'm working on a Wiki package. I am writing it from scratch, but I've wondered if building on ETP2 wouldn't be a better way.
I'm trying to make the Wiki package general so that with some modifications you can have different applications. Especifically, I want a Cookbook package with linking, ratings, moderation. I don't know if it'll end up being Wiki->Cookbook or two separate packages. We'll see.
The mock-up is at http://www.collaboraid.biz/extranet/dotwrk/mock-up-home .
More info at http://www.collaboraid.biz/extranet/dotwrk/mock-ups.
Yet more info at the .WRK project home at http://dev.openacs.org/projects/dotwrk/.
Yes, shaping up much as I'd expect. A portal-style view of corporate intranet is pretty much the way to go I think... however...
I think there's a few areas which are critically important to an intranet-in-a-box that are going to have to be included fairly early on, the danger being that if not all we end up with is a giant, super-fly message board.
One of the things complicating an intranet-in-a-box is that any company of any significant size is a collection of distinct, though related, functions. i.e. Marketing, Sales, Deliver, Support HR and so on...
I personally feel that in many respects the non-tech functions are the primary driver, as its often these folks who drive new investment in stechnologies like .WRK. Its also interesting to note thats its also these areas of a business that are not well served by existing proprietary systems. (got a partner in marketing).
I'd like to see things like these given high priority early on.
- Account Management
- Contact/Sales databases
- Campaign Management (bulk mailing, auto responses etc).
- Good story on how the intranet functions with existing web investment. Maybe this is XML/SOAP interfaces, maybe its a 'move your website to .WRK.. and so on..
- Integration is key. Someone's already mentione Lotus Notes, and its products like these that dotWRK *must* offer interfaces to. Companies will already have investment in these areas, and products that don't/can't integrate are likely to be un-selected.
- HR functions... Timesheets, Holiday booking, expense claims....
- Project SubSites. Easy to create, subsited sections for individual projects/managers/teams...
- Communication functionality. At it simplest CHAT.... figuring that most companies interested in intranets are probably also geographically separated.
Sounds a lot doesn't it but most of the above already has a basis/start-point in existing ACS functionality. We've already built a really simple contact/sales db app which might make a start for some of the things above.
I totally agree that these other areas are super-important. At first I thought you'd be suggesting that I created areas for them on the intranet, like the Planet.Intra demo site linked from the dotWRK home page.
But then I realized that you're actually asking for functionality and integration. That's cool, but would obviously require some funding to get implemented. I've added it to the mock-up, so it can properly serve as a product vision.
I've added the items you mention but I'm not sure about the structure that's there right now.
- Our company is very project based. We've got a bazillion projects going on at the same time, and the ability to track what is happening with each project and what has been done already is extremely important. For this reason, the projects have to be *very* flexible.
- It was a fairly easy modification to ACS 3.4.10 to make more than 2 levels of projects. Our Intranet has unlimited levels of projects.
- We should make it really easy to add different types of data to the calendar. We should also make it easy for companies to decide how their calendars are shown (group calendars, personal calendars, or maybe just one shared calendar in the company).
- We need a very easy to use "collaborative to-do list" aka ticket tracker. These should be linked to projects, and shown on the calendar. The project page should show 1) what needs to be done on the project, 2) what has been done, as far as hours, closed tickets, etc and shown in one place, 3) what the current state of the project is (who is responsible for it, when if any the deadline for the project is, etc..
- If we made a migration path from the old ACS 3.4.x Intranet to the new dotWRK, I'm sure we'd have a lot of early adopters that could test this code out It would also pull in more developrs, like me for example. Wishful thinking perhaps...
- In our company, we are not geographically dispersed, so let's make sure that it's easy for a small 25 person company in one location to use this too. I'm sure I'll have more thoughts later. :)
I agree with your recommendations for functionality.
It also presents an interesting dimension that's building. People seem to be describing dotWRK not really as a corporate intranet, but increasingly as a project management tool..... and further to that the focus (and I'm only going on the few postings dotted abot so far) seems to be a little biased towards software companies!
Folks, I know this sounds a bit trite, but there are companies out there that do stuff other than develop software ;), perhaps we should consider what the difference in needs is between tech companies using intranets and more 'traditional' companies.
I'm worried that the requirements being built up are being largely by developers. Do we have anyone in the community *not* involved with tech as their main business. CAn we persuade them to get involved in this informal 'requirements capture'?
My partner is a marketing executive, and even a short discussion with her has radically changed my idea of what a company might want from an intranet (and she's in a tech company).
Features of our Intranet:
<li>Profiles of employees.
<li>Books, magazine, studies database. Where can I find a book? They are often spread all over the company.
<li>Projects. It is supposed to show who is working at which project. You can upload .ppt, .doc etc. further explaining your work. There is no project planing or collaboration yet, but it surely would be helpful.
<li>Processes. Explains the process of every job position. They are mainly displayed as graphics defining the workflow of i.e. a call center agent. Some workflows are deeply branchen (if/else etc.). I have never looked at the workflow module, but it would be great, if it could define branched workflows and display them graphically.
<li>Forms. Content Repository for different forms.
Simon, could you expand on that, please? What did you learn?
Well I won't go inot it all here, but I'll try to summarise the main points:
- 1) Technology is largely irrelevant i.e. they don't care whats its made from, just what it can do.
- 2) That the most likely 'sponsors' for intranets-in-a-box are not likely to be the tech/internal folks, but much more likely to be salespeople, marketing and account management.
- 3) Integration with existing technology figures really high. i.e. going in with a 'this is a great toolkit... but it works on its own' is a sure fire non-starter.
- 4) Don't build a super-functional clever-wizzo system. Academice may be prepared to experiment and try new approaches, businesses much less so. Somehting that does simple things really well is much more preferable.
- 5) The kinds of people who will make extensive use (sales, marketing, management) are also the kinds of people who aren't usually prepared to adopt complex, rigid, involved or prescriptive systems. The more flexibly the platform can be used the better. I seem to remember aD found this same thing with Seimens.. i,e, you can build clever knowledge-databases, informaiton analysis and so on... but salesmen end up doing it all with informal bulletin boards ;)
- 6) Make it work on their PCs with little effort. i.e.
- 7) Make it solve existing business problems, not provide clever functionality that enhances existing operations. i.e. don;t give them something they never had before, replace something they never liked.
As you can see this kind of thinking suggests a different kind of platform.
One comment that was made was 'if it worked with Lotus Notes, thats more important than if it was all that good'......
It also seemed to me that integration was far more salent than operation...
1. "Log" package - posts "description", "done by", "date" with options to activate/display fields in "duration", "confirmed by", "authorized by", "comments/notes" "WBS/outline reference" "job code" "status" etc. where fields displayed can be filtered/sorted by content, and ability to add or change values and headings based on permissions.
2. accounting package with simple ledger functionality (account, item ref, original date, to/from, for/exp-code, credit/debit, project/client ref, their ref, statement date, papers filed, tax ref etc.), and again extra fields (other1, other2) included (so a DB/OpenACS expert doesn't have to get involved everytime a project is a little different) --saves time for everyone.
3. time ledger package that integrates with calendar (adds and removes alerts to it as schedule changes). again, extra fields
4. research package/info-log that tracks progress of unknowns to facts. This could be a table consisting of fields: Info type - (about subject from other than subject, from itself, from subject about other), entry date, citation/source, comments etc. Subject may be Person, Entity, Witness, Topic, Product/project feasibility...
All four of these are a variation on a cluster of similar functions and procedures that perhaps can be rolled into one package and mounted/configured in different ways --more than one instance working in the same "realm" ....perhaps integrating across package instances hmm.. that may be too much too fast! =)
Many times, extra functionality is not as important as an appropriate place to put the info, authority to edit/change it, and ability to customize how it is reported.
The packages should include a few extra blank long fields with generic field names--where the displayed fieldnames can be changed.
Also, the ability to export/import to/from a spreadsheet (as tab delimited text or CSV) would be useful in case some field data needs changed based on functions or batch editing, for charting etc.
Perhaps some fields could contain custom calculations... starting with (filtered) pure tcl represented 1 line equations that reference other fields in the same table set...
think customizable, flexible, integrable
be describing dotWRK not really as a corporate intranet, but increasingly as a
project management tool..... and further to that the focus (and I'm only going
on the few postings dotted abot so far) seems to be a little biased towards
software companies! </I>
Hear hear! A lot of the work I've had to do customizing our ACS based Intranet
has been to make it useful for non-software companies. For example,
changing the ticket tracker to be a real collaborative to-do list (very
Project management is going to be extremely important for us. I'm not really
sure what everybody is agreeing on with the idea to build dotWRK, but
depending on the form it takes, I might be on board. Can we agree what our
goals are for it?
become an admin? :)
- What is the difference between departments and groups? They seem to be the same thing.
- The idea of projects and project management seems to be missing from this particular mockup. Our small company of 25 people has about 1000 projects going on at once, and each of these projects has to keep track of a vast array of information: who is in charge of it, what has been done for it (project reports), what needs to be done for it (a modified ticket tracker)... It would be nice if it were possible to add arbitrary fields of information to track throughout each projects... I'll put up some mockups of some ideas.
- so if I wanted to add in a collaborative to-do list (simplified and refined ticket tracker), I could do that as a portlet, and it would appear under Company? Would it also show up as a link under each group/department, and/or under the projects, if projects are going to be a part of it?
- It's nice having the control panel where it is.
- Other ideas for portlets: room reservations, company glossary, hour logging
- Just playing devil's advocate: what is the advantage to me as a developer in writing an Intranet on dotLRN instead of just OACS? I'll put up some mockups to hopefully aid in this discussion.
I've only put up the things that were already readily available as dotLRN applets, which means that a number of applications are still missing. There's a bit of work remaining in dotLRN'ifying all the other applications out there, which I haven't tackeled yet--but as soon as we get a client for this, we'd love to . That includes project management, hour logging, room reservations, form generator, company glossary, knowledge management, wiki, task list, workflow, etc.
Departments are groups, too, yes. The idea is that department groups are centered around departments, whereas the other groups will typically be centered around a project. I just figured that Departments were special yet common enough to warrant their own little special treatment.
If you built a collaborative to-do list and wrapped it to become a dotLRN applet, yet, it would show up in all the groups where you create an instance of it.
Regard dotLRN vs. OACS as the basis for an intranet, I think the dotLRN groups and portlets infrastructure is just subsites and OACS groupware functionality done right. If you want groupware functionality in your intranet, I don't see any downside to building on the dotLRN codebase.
I'll be on vacation for a week from now, but I'll be looking forward to your mock-ups.
Just a small comment about your very rough mock up for what your dummy intranet might look like. Calenders, inside companies, are often more useful if they are a group calender - so you can see what the whole organisation is doing at a glance. The choice to personalise it, or hide it all together, would be up to the user. Just a quick thought.
also, you would probably want to be able to have degrees of visibility. For example as Director I'd probably want certain things private, but visible to co-directors and my PA perhaps (if I had one ;) but not visible to other employees.
So you'd want to be able to create closed-user-groups.
As well... and this is something I've recently been asked by a customer, GSM adn PDA integration is key... SMS notifications of diary changes entries for example, and being able to sync the system.
I guess you;ve already used the SMS Broker, so you'll know it won;t affect the application, other than you should put in the option to put an SMS on the queue at these key points.
If you say PDA/Mobile integration to a customer who want an intranet their eyes light up ;)
discretion of the person setting up the Intranet.
I'll get my mockups up today or tomorrow, so that you all can see how our
company is dealing with these issues.
dotWRK. They are a combination of the way our Intranet is currently, and
where it will be soon. I edited out any information that should be kept within
our company, but I'd like to only send it out by request instead of putting it out
on the internet at large.
I think the tickets and calendar may be of particular interest.
If you'd like to see these mockups, please email me, with the following in the
body of your email:
"I will not distribute these mockups without permission. I am also not working
in the baking industry and do not currently have a client in the baking
Sorry for the legalese-ishness. I just don't want to give away too much
information on our intranet to potential competitors.
I posted a thread on a possible scheduling app for dotWRK: http://openacs.org/bboard/q-and-a-fetch-msg.tcl?msg_id= 0006gy&topic_id=11&topic=OpenACS
Danielle, Simon, Jade, excellent feedback, thanks. Hopefully someone in the community will soon find the resources or clients to push ahead on this for real!
Has any-one ported the room reservation system to 4.5/6 and prepared to let the cat out of the bag?
I came across an interesting discussion on the display of project management information. Specifically, the presentation of gantt charts and alternatives are discussed. I thought it might be of interest:
A significant leap in improving the way project information is displayed (and entered) would, I think, make dotWRK standout well above any available competition.
This thread seems to have died a couple of months
Did it move to another location?
I think the dotWRK concept is wonderful.
In fact, I'm drooling over it.
Be glad to help in any way that
I can. About all I can do is write
I think we're at the point where we've proven that the concept is the right on, we just need one or two or more organizations who need it and want to fund the development work needed to really make it happen.
We're working on that part, but we're definitely interested in finding more.
I think the navigation bar on the left could be taken out. Perhaps it could be replaced with a bar at the top or something? It just seems to take a lot of unnecessary space.
Eventually, I'll get around to posting what we've done on our Intranet. I've just updated the interface, so the mockups I've sent out are old. I think it's improved quite a bit..
Are other people in the community pitching in with this work? Are there any items that need some volunteer help?
So let's continue to share, discuss, and so on, and if you have something you're working on, by all means, feel free to share what you've done.
But in my opinion, dotWRK won't become a reality until someone releases some code. :)
It will happen, but it needs some attention and developer time. I am using ACS 3.4 for our Intranet, with a heavily customized interface. We've been using it for about two years, and I've learned a lot in the process, and hope to use that to help in the .WRK effort.
However, I'd like .WRK to be a pretty flexible entity.
One of the major problems with the ACS 3.4 version of the Intranet package was that it was written so specifically for ArsDigita's needs.
One concern I've had is that if one of us goes off and writes this package, it may do the same thing all over again. That certainly is what my own tendency would be.
Another problem I'm going to face, that many of the rest of you won't, is that I have to migrate all of my old data from ACS 3.4.
We've heavily customized the ticket tracker, to where it is not for tracking bugs, for is a collaborative to-do list. I've thought about porting it to OpenACS/dotWRK. I'm not really happy with any of the ticket trackers right now. None of them are general enough to be used for all to-do items, and I kind of like the interface we've developed here. However, it is a fair amount of work.
The whole project is fairly large though, especially because it's difficult for those involved to sit down and plan out how it might work. Collaborative programming works, but it's easier if you have something concrete to start from. Or at least that's my (limited) experience.
Althoug I agree with Jade that it is most likely that things will speed up more and more with a growing code-base, I also think that the project needs what in the oacs community seems to be called a 'gatekeeper' - otherwise I see a risk that people start independent dotWrk-a-like things (I know I would), which distracts valuable mantime from dotWrk.
Is there already somebody (actively) assuming that role? If time is an issue - I can 'donate' time to the project, not a whole lot (say 8h/week) but on a fairly structural basis for the coming 3-4 months at least ('cause it has a strong relation with part of my daytime job).
IRL pulling large projects is a big chunk of what I do, so besides coding I can bring something to the table here in the area of planning/coordinating/pushing.
To summarize: I've time to offer to spend on dotWrk-a-like things. I can either spend it on a 'fork' or on 'the real thing'.
Correct me if I'm wrong.
You're absolutely right that we need someone to lead this project. I suspect that I'd be the current lead or something, but as you've all noted, I haven't had the time to follow through on this.
So your offer is most welcome.
I think a good place to start would be to catalog the current state of affairs:
- the interested people and what they're willing and able to contribute
- the tasks, what needs to be done, not just in terms of code, but also in terms of design, infrastructure service modules, enhancements, information architecture, etc.
I'm still very determined to present a proposal for a new information architecture/user interface scheme around the time of the conference here in Copenhagen, but people shouldn't be waiting for me before taking action.
In fact, we should have visible project leads for all of our subprojects -- openacs.org, documentation, dotWRK, CMS, dotLRN, etc. -- but that's another story.
Somewhere in the coming 2-3 days I'll craft a mail and send it to that list in order to probe for interest and capabilities.
Parallel to that action I'll see if I can come up with a proposal for a 'plan of attack'.
How can I publish any results in that area on the project page?