osfree.org

#1028 RE: [osFree] Project Definition
Expand Messages

John P Baker
Jul 23, 2004

Michael



It is good to hear from you. As stated, I will be making a project definition outline available on the web over the weekend.

Once that is done, I will be seeking to dole out specific sections to interested parties.



John P Baker

Software Engineer



From: michael [mailto:michael@...]
Sent: Friday, July 23, 2004 22:56
To: osFree@yahoogroups.com
Subject: Re: [osFree] Project Definition



Hi,

my name is Michael and I've reading this discussion since mounths/years.
I'm interested in documentation.
My knowledge based on Warp 4 and Windows NT/XP.

Regards,

Michael

>
> If you have an interest in contributing to this effort, please drop me a
> note, indicating your specific area of expertise.
>

#1029 Project Definition
Expand Messages

John P Baker
Jul 23, 2004

Ok. I have put up my basic project definition web page.



It is a very rough outline at this point. But you should be able to see the direction in which I am heading.



The URL is: http://home.earthlink.net/~jbaker314/



John P Baker

Software Engineer

#1030 Re: [osFree] Project Definition
Expand Messages

Lynn H. Maxson
Jul 23, 2004
I've reviewed your website and the project outline contained
therein. I'm not too sure how to respond. Under
"Documentation" I see nothing related to a design document. I
have already done a complete data (interface) analysis of the
OS/2 API, which I could submit. However, it's in PL/I and not C.

Normally projects occur with a predecessor/successor
sequence that indicates the order of execution as well as
depicting parallel processes. I realise that you offer it as a
suggested form subject to change. If we are agreed on the
design and its documentation as a first stage, then we could
set aside, i.e. hide from view, following stages. That would
allow us to deal with the sub-projects of the first stage as
they arise.

#1031 Re: [osFree] Project Definition
Expand Messages

Davey Brain
Jul 24, 2004
John P Baker wrote:

<snip>

> There are several items of information for which I need to identify
> appropriate sources. Your help will be appreciated.
>
> For a filetype of “AVI”, where can I find detailed information for the
> internal structure thereof? Are there multiple implementations
> (versions) of the internal structure of this filetype? If so, I need to
> identify appropriate sources for each implementation (version).
>
> For a filetype of “BMP”, where can I find detailed information for the
> internal structure thereof? I am aware that there exist multiple
> implementations (versions) of the internal structure of this filetype.

I only have a few mins before work, but I googled on "format information
BMP" and found this right off:

Format Index: <http://www.daubnet.com/formats/index.html>
AVI: <http://www.daubnet.com/formats/AVI.html>
BMP: <http://www.daubnet.com/formats/BMP.html>

I used to know a very good format def site but that was a long time &
many hard drives ago...I'll attempt to find it when I'm off work.

> If there available an “INF” viewer that runs on Windows?

At Hobbes for Win16:

<http://hobbes.nmsu.edu/cgi-bin/h-viewer ... /iview.zip>

Viewing the file: /pub/windows/iview.zip
./ Go to /pub/windows Current Directory
iview.zip
IBM IPF for Windows Version 2.1 (1998/Jun/02).
Tool to view OS/2 INF help files under Windows.
2003/03/13 Compressed archive, 187.62Kb
Copyright IBM Corp 1981, 1992, Copyright SingaLab Pte. Ltd. 1993, 1994

> If you have an interest in contributing to this effort, please drop me a
> note, indicating your specific area of expertise.

I can help in GUI design & documentation plus thorough alpha/beta
testing and can help in any website docs & design (9 years experience in
web & graphics design). Using OS/2 since IBM/MS OS/2 V1.1. I know some C
but little C++...willing to learn more.

> If your are looking for a hand-out, and have no interest in making a
> contribution, please don’t waste your time nor mine.

I hope not

> Please understand that I do not intend to write the first line of code
> until a reasonably comprehensive project definition is completed and the
> pre-requisite documentation is in a useable form. You may take that to
> mean that no code development will begin until next year.

Gives me time to learn more programming skills.

> John P Baker
>
> Software Engineer

WarpDavey
--
Davey Brain
"Hail to the speaker, hail to the knower;
joy to he who has understood, delight to they who have listened." --Odin
Uptime is since the last storm
This OS/2 uptime is 15 d 20 h 06 m 37 seconds

#1032 RE: [osFree] Project Definition
Expand Messages

John P Baker
Jul 24, 2004

What you refer to is exactly the kind of input that I am looking for. The items listed so far are just a small example of what will have to be included in the project definition.



Predecessor/successor information will have to be provided for each and every item.



Every API, of which I included “DosClose” and “DosOpen” as examples, will have to be identified, documented, and inserted as appropriate.



Every command will have to be identified, documented, and inserted as appropriate.



Some other considerations:


Do we need to continue to support 16-bit APIs and the associated “thunking”?
Do we need to support an Windows interface layer or just native “protected mode” OS/2 sessions?
What hardware should we support? Do we establish a minimum of a “Pentium 4”?



Every suggestion for improvement to the design is welcome.



As far as hiding elements that we will get to down the road, I agree. However, I have go to learn a bit more about how to perform selective exports from Microsoft Project before that happens. Actually, I would like to be able to have multiple web pages where you would be able to click on a major item and open a new web page with all of the sub-items for that major item. All in good time.



John P Baker

Software Engineer

From: Lynn H. Maxson [mailto:lmaxson@...]
Sent: Saturday, July 24, 2004 02:30
To: osFree@yahoogroups.com
Subject: Re: [osFree] Project Definition



I've reviewed your website and the project outline contained
therein. I'm not too sure how to respond. Under
"Documentation" I see nothing related to a design document. I
have already done a complete data (interface) analysis of the
OS/2 API, which I could submit. However, it's in PL/I and not C.

Normally projects occur with a predecessor/successor
sequence that indicates the order of execution as well as
depicting parallel processes. I realise that you offer it as a
suggested form subject to change. If we are agreed on the
design and its documentation as a first stage, then we could
set aside, i.e. hide from view, following stages. That would
allow us to deal with the sub-projects of the first stage as
they arise.

#1033 RE: [osFree] Project Definition
Expand Messages

Lynn H. Maxson
Jul 24, 2004
John,

The project control itself has a documentation dynamic that
increases the people resources required as activities and their
relationships increase and change. Having been associated
with project control since its introduction in the 50's as PERT I
question its viability as a "management" tool in a purely
volunteer effort.

I don't question that some things, activities, precede, follow,
or overlap others. That understanding and agreement among
participants can lead to a more orderly process. I do question
the ability to set either dates or duration in any meaningful
manner. At this point you do not have the ability to control or
allocate volunteer resources to maintain critical path
schedules as the future need may arise.

You offer some hint at understanding the hypermedia
capabilities of using a website and html. You will not find the
skill in doing so evenly distributed, either easily learned or
applied. When they discover that the people cost in writing
documentation, its development and maintenance, easily
exceeds that of writing source code (programming) and that
testing which requires the synchronization of both again easily
doubles the people cost, we will have a better grasp on the
need for maintenance contracts.<g>

I don't say these things to detract from what you have started
here. I have always found the OS/2 replacement project
doable, but I see no need to introduce elements (activities)
before their "time". Such an "unfinished" set can seem
overwhelming and discouraging however factual.

Now I'm working on a software tool called the "Developer's
Assistant" to drastically reduce the people cost in numbers
and in effort. It uses the same database, as opposed to a file,
system for both source text and code. It not only allows full
"literate programming" a la Donald Knuth but applies it
globally to all documentation, including user guides, general
information, and reference manuals.

As long as you use the same tools as M$, IBM, and others you
will incur the same or greater people costs. IBM decided with
respect to OS/2 that it would no longer sustain the same level
and transition its support to eventually remove it from the
marketplace. The OS/2 community has neither the numbers
nor the financial wherewithal to assume what IBM found
necessary to sustain OS/2.

In short that means we cannot use the same tools and
methods in creating an OS/2 replacement...or compete with
whatever new incompatibilities M$ chooses to introduce. We
cannot work harder. Therefore we must work smarter.

You have chosen a project management style which to the
best of my knowledge no other open source project has
attempted. It's one that even closed source finds difficult to
manage and maintain. It requires both an additional skill set
as well as people resource.

I would probably suggest that we consider only development
a project control scheme for a sub-project like "Design an
OS/2 Replacement" with general activities like "Gather
Requirements", "Document APIs", etc.. Then we could break
this general activities as we gain familiarity into the more
detailed activities of the "information units", e.g. each API.

In my examination of the APIs I found particularly disturbing
the generic use of the variable name "rc" for "return code".
As the same values have different meanings depending upon
API involved it appears to me that each different set of return
codes have a unique name. I also don't happen to have C's
penchant for multiple fancy names for the same data types,
of which two dominate, a full word binary value or an
address. As a PL/I person not given to confusing things I'm
inclined to define the first as "dcl xxx fixed bin (31);" and the
second as "dcl yyy ptr;".

Just a few thoughts for your consideration.

#1034 RE: [osFree] Project Definition
Expand Messages

John P Baker
Jul 24, 2004

You raise many valid points.



I will start by saying that I didn’t get started until the late 1960s.



However, I have found over the years that the successful projects have been those which have been well-defined “before” coding has begun. This is not to say that significant changes to the project definition have not been made during the coding process. Far from it. On the other hand, the large scale goals defined in the project plan have “almost always” remained intact.



By the way, I am sitting here looking at one of my several editions of “The Art of Computer Programming” (Donald Knuth).



I can not agree with you in respect to the people cost issue. In my view, IBM’s failure with OS/2 was more due to IBM’s lack of a clear vision for OS/2.



You are correct that working smarter is critical to the success of this project. That is why I believe that the production of a project definition is critical.



I learned years ago about what we called “Black Box Programming”. Each component had a well-defined and well-documented external interface. The exact style of the implementation of the internals was not considered critical. In a way, you could say that “Black Box Programming” was one of the predecessors to “Object-Oriented Programming”.



As for the OS/2 APIs, I believe that we need to identify and document each and every one. Does this mean that we need to implement every OS/2 API? No. I think that the OS/2 API set can be improved upon. I have a particular interest in security and auditability. I am leaning towards a much more object-oriented API set.



As for a development language, that is down the road. Certainly, what you are doing is of great interest. But am I prepared at this point to commit to a specific third generation language or fourth generation programming tool? No, it is too early in the process.



I do not take your comments as a detraction at all. Rather, this is exactly the kind of give and take that is crucial to a successful project.



John P Baker

Software Engineer

From: Lynn H. Maxson [mailto:lmaxson@...]
Sent: Saturday, July 24, 2004 10:42
To: osFree@yahoogroups.com
Subject: RE: [osFree] Project Definition



John,

The project control itself has a documentation dynamic that
increases the people resources required as activities and their
relationships increase and change. Having been associated
with project control since its introduction in the 50's as PERT I
question its viability as a "management" tool in a purely
volunteer effort.

I don't question that some things, activities, precede, follow,
or overlap others. That understanding and agreement among
participants can lead to a more orderly process. I do question
the ability to set either dates or duration in any meaningful
manner. At this point you do not have the ability to control or
allocate volunteer resources to maintain critical path
schedules as the future need may arise.

You offer some hint at understanding the hypermedia
capabilities of using a website and html. You will not find the
skill in doing so evenly distributed, either easily learned or
applied. When they discover that the people cost in writing
documentation, its development and maintenance, easily
exceeds that of writing source code (programming) and that
testing which requires the synchronization of both again easily
doubles the people cost, we will have a better grasp on the
need for maintenance contracts.<g>

I don't say these things to detract from what you have started
here. I have always found the OS/2 replacement project
doable, but I see no need to introduce elements (activities)
before their "time". Such an "unfinished" set can seem
overwhelming and discouraging however factual.

Now I'm working on a software tool called the "Developer's
Assistant" to drastically reduce the people cost in numbers
and in effort. It uses the same database, as opposed to a file,
system for both source text and code. It not only allows full
"literate programming" a la Donald Knuth but applies it
globally to all documentation, including user guides, general
information, and reference manuals.

As long as you use the same tools as M$, IBM, and others you
will incur the same or greater people costs. IBM decided with
respect to OS/2 that it would no longer sustain the same level
and transition its support to eventually remove it from the
marketplace. The OS/2 community has neither the numbers
nor the financial wherewithal to assume what IBM found
necessary to sustain OS/2.

In short that means we cannot use the same tools and
methods in creating an OS/2 replacement...or compete with
whatever new incompatibilities M$ chooses to introduce. We
cannot work harder. Therefore we must work smarter.

You have chosen a project management style which to the
best of my knowledge no other open source project has
attempted. It's one that even closed source finds difficult to
manage and maintain. It requires both an additional skill set
as well as people resource.

I would probably suggest that we consider only development
a project control scheme for a sub-project like "Design an
OS/2 Replacement" with general activities like "Gather
Requirements", "Document APIs", etc.. Then we could break
this general activities as we gain familiarity into the more
detailed activities of the "information units", e.g. each API.

In my examination of the APIs I found particularly disturbing
the generic use of the variable name "rc" for "return code".
As the same values have different meanings depending upon
API involved it appears to me that each different set of return
codes have a unique name. I also don't happen to have C's
penchant for multiple fancy names for the same data types,
of which two dominate, a full word binary value or an
address. As a PL/I person not given to confusing things I'm
inclined to define the first as "dcl xxx fixed bin (31);" and the
second as "dcl yyy ptr;".

Just a few thoughts for your consideration.

#1035 RE: [osFree] Project Definition
Expand Messages

Lynn H. Maxson
Jul 24, 2004
John,

I don't think differ widely on any point, except possibly why
IBM withdrew OS/2 marketing and then support. To the best
of my knowledge IBM had an 18-month cycle on fixpaks. In
order to properly release a tested version every six months,
the schedule they maintained, they had three teams working
separately and in parallel. That basically was and is their
approach on all "strategic" software.

Furthermore all studies published thus far and reports given
me by closed source vendors indicates that open source
projects are less efficient and require more people resources
to achieve the same or equivalent output in a given interval.
So far my observations on other OS/2 projects such as those
on Netlabs seems to confirm this. You could look at Innotek,
who in fact gets paid to write the software which we get for
free, to appreciate the difference.

Project management implies control. In fact the two words
become interchangeable in this context. In a volunteer effort
in which people assign themselves to activities and not
allocated per se, the larger the participation the less
"predictable" control you can exercise. I don't disagree with
what you want or your reasons for wanting it. I just have no
example in the open source arena for its successful
application. Not even Linux has a proven example.

The issue does not lie in ordering activities, of saying some
precede or succeed or parallel others. This ordering makes up
a project plan. I do question that you can apply meaningful, in
terms of predictable, durations, particularly when the
participants for the most part have no experience on which to
base the estimates.

You can argue that you can make the changes as participants
gain in such experience. However, it is the rate at which you
would have to make those changes and the number of such
changes that will drive up your people resource cost. That
doesn't include the global changes that will occur when
certain when you do not meet (exceed) scheduled durations
which inter-relate with other activities, which now you must
also adjust.

I think you want to keep the maintenance of the schedule, the
people resource involved, to a minimum to allow for its
application elsewhere. For that reason I would suggest
de-emphasizing durations, start dates, and end dates entirely
for activities. I realise that this does away with the core of
project control (or management), but it does allow us to
detail activities down to the necessary level and ordering
them logically. It allow us to see what needs to get done
without infusing it with scheduling details that have no real
accuracy except after the fact. That commits minimal people
resources to the tracking and management effort without
losing the overall vision.

As an advocate of structured analysis (dataflows) and design
(structure charts) I agree with your views on black box
programming, the detail IPO model. You have two sets of
specification, a data one for the I(nput) and the O(utput), and
one for the P(rocess) which connects them. I would suggest
that we use PL/I as their formal specified form instead of C or
any C derivative. We need an "exact", "unambiguous"
specification for both data and process (the rules which bind
them).

Unless we know that no OS/2 application uses a particular API
I suggest we fully document all others whose use we know or
don't know. While we can look to create a different set of
APIs for writing new applications for the replacement product,
I don't see giving up the investment we have currently made
in existing products. We want to give the OS/2 replacement a
continuing future without having to forego our investment in
its past.

It appears to me then that we need to settle on a
specification language with neither the limits nor confusion of
C. I propose PL/E or it rename as SL/I. I think the advantages
of logic programming, specifically related to enforcing data
rules in software and not programmers programming, will
quickly become evident. If it wasn't easier to learn and
implement than C or any C-derivative, I wouldn't offer it.

Once people realise that software can perform analysis,
design, construction, and testing based on an unordered set of
specifications, then this effort which seems almost
overwhelming at the moment comes well within our comfort
zone.

It's really logic programming and not a particular language
that makes this possible. I see no reason why we cannot use
the development of a design document to make participants
more savvy on logic programming. In that manner bring more
of the participants into the programming fold because it just
got a whole lot easier.

#1036 Re: [osFree] Project Definition
Expand Messages

Tom Lee Mullins
Jul 24, 2004

> I have begun the process of creating a project definition within Microsoft
> Project Professional for the creation of an OS/2 follow-on.
>
> I see it as irony. Like my use of a Win-pc to post (at work)

to an OS/2 or eCS grouop.

>
> And before anyone returns a snide comment, yes, I know that it is a
> "Microsoft" product, but it is what I have, so we will just have to live
> with it.
>

Like those who create ReactOS - http://www.reactos.com - who compile
their source code using a Win compiler.

>
>
> I am going to setup a very vanilla web-site over the weekend where I can
> make a web version of the project definition available.
>

I look forward to seeing the site.

>
>
> As additional items to be added are identified, the project definition and
> the web version thereof will be updated.
>
>
>
> Please note that this is an outline which will address both code and
> documentation. Many entries will occur multiple times. For example, an
> API
> definition will occur under a "Code" grouping as well as under a
> "Documentation" grouping.
>
>
>
> There are several items of information for which I need to identify
> appropriate sources. Your help will be appreciated.
>
>
>
> For a filetype of "AVI", where can I find detailed information for the
> internal structure thereof? Are there multiple implementations (versions)
> of the internal structure of this filetype? If so, I need to identify
> appropriate sources for each implementation (version).
>
>
>
> For a filetype of "BMP", where can I find detailed information for the
> internal structure thereof? I am aware that there exist multiple
> implementations (versions) of the internal structure of this filetype.
>
>
>
> If there available an "INF" viewer that runs on Windows?
>

Have you checked http://hobbes.nmsu.edu to see if there is one?

>
>
> If you have an interest in contributing to this effort, please drop me a
> note, indicating your specific area of expertise.
>

I have no expertise but I am interested. I will be taking a
programming course.

>
>
> If your are looking for a hand-out, and have no interest in making a
> contribution, please don't waste your time nor mine.
>

Perhaps in the future after I have learned more on programming.

>
>
> Please understand that I do not intend to write the first line of code
> until
> a reasonably comprehensive project definition is completed and the
> pre-requisite documentation is in a useable form. You may take that to
> mean
> that no code development will begin until next year.
>
>
>
> John P Baker
>
> Software Engineer
>
>

BigWarpGuy
* * * * * * * * * * * * * * * * * * * * * * * * *
* * * * eComStation * * * * * * * * * * * * * * *
* * the os for the internet generation * * * * * *
* * * * * * * * * * * * * * * * * * * * * * * * *
* * * http://www.ecomstation.com * * * * * * * * *
* * * * * * * * * * * * * * * * * * * * * * * * *
* * * * http://www.os2world.com * * * * * * * * * *
* * * * * * * * * * * * * * * * * * * * * * * * *
* * * * * best site for OS/2 Warp Users * * * * * *
* * * * * * * * * * * * * * * * * * * * * * * * *

#1037 Security Computing Platform
Expand Messages

John P Baker
Jul 25, 2004

I have updated the projection definition to reflect a “Research” phase, consisting of a “Gathering Information” phase and of a “Propose target capabilities” phase.



Please contribute ideas for new elements to be incorporated.



John P Baker

Software Engineer