Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Do we need a new maintainer?
23 views
Skip to first unread message
Peter C. Chapin
Sep 30, 2011, 6:54:40 PM9/30/11
to
My activity level in the Open Watcom project has been low to
non-existent for quite some time now. This is not due to a lack of
interest, but rather due to the many other pressures on my time. I'm
sure most of you know exactly what I'm talking about.
non-existent for quite some time now. This is not due to a lack of
interest, but rather due to the many other pressures on my time. I'm
sure most of you know exactly what I'm talking about.
A low level of activity from a single contributor is not necessarily bad
or a problem except that officially I'm the Open Watcom maintainer. It's
been over a year since Open Watcom 1.9 was released and yet there is no
2.0 release in the works. What's more I don't see a 2.0 release
happening any time soon.
I have this belief that I will have more time for Open Watcom in the
future. Right now that means at least a year in the future. This
concerns me. I'd hate to block progress in the project by being (for all
practical purposes) unavailable. What's worse is that I'd hate to block
someone else with more time resources from doing a better and more
active job as maintainer. So here is what I'm saying:
1. I'm willing to continue as project maintainer but I want to be honest
with the community: I'm not likely to be able to do much of anything
(such as release 2.0) for 12 months or more. If everyone is okay with
that, we are good.
2. I'm willing to step down as maintainer if there is someone else with
more resources who is interested in doing the job. I'd be willing to
interact with that person to help streamline any transition.
Actually there is a larger issue at play here. In general activity in
Open Watcom is much less these days than it was in the past (it's not
just me). We've lost a couple of our major contributors and even those
who are still around have been busy with other things. I'm not sure what
this means for the long term health of Open Watcom, but it is worrisome.
I'd welcome any thoughts people might have about this.
Peter
Vesa Jääskeläinen
Oct 1, 2011, 1:56:43 AM10/1/11
to
On 1.10.2011 1:54, Peter C. Chapin wrote:
> Actually there is a larger issue at play here. In general activity in
> Open Watcom is much less these days than it was in the past (it's not
> just me). We've lost a couple of our major contributors and even those
> who are still around have been busy with other things. I'm not sure what
> this means for the long term health of Open Watcom, but it is worrisome.
> I'd welcome any thoughts people might have about this.
Hi All,
> Actually there is a larger issue at play here. In general activity in
> Open Watcom is much less these days than it was in the past (it's not
> just me). We've lost a couple of our major contributors and even those
> who are still around have been busy with other things. I'm not sure what
> this means for the long term health of Open Watcom, but it is worrisome.
> I'd welcome any thoughts people might have about this.
Sorry for the long post ;)
I think one way forward would be to define what Open Watcom currently
is. Then think should there be a reason re-classify and/or split it to
components.
Lots of the code shows its age and do not really stay on the leading
edge anymore. One question is - can some of that be replaced with
already existing components from external sources.
Some problems that I see is in example C library (especially for the
Linux environment as that is a bit limited). Could that in example be
combined with Bionic or other easily licensable C libraries. Those who
are not aware Bionic is licensed with BSD. Same goes for C++ library. Or
could this work be combine with some other compiler project? Strength
that I see in current is that it is not GPL (nor LPGL) and it is easy to
control where it is going.
Another issue is IDE (and other GUI tools). They are showing their age.
Would it be feasible to think about replacing those with more recent
components?. Personally I use Eclipse/CDT for GCC/RVCT based
development, it should not to be too hard to integrate IDE functionality
to Eclipse based solution. Problem with Eclipse/CDT is its debuggers.
They are not the leading edge... Eclipse's license however is rather
nice for extending it.
Then we have the installer. That should be rather easy to replace with
other solutions. Why not pick some project out there and then improve
that project with the needs?
When the problem has been sliced smaller sized problems then it might be
easier to handle. If those sliced problems (or components if you wish)
would have own maintainers (internal or external to project) then it
would be easier to share responsibility of maintenance.
As you said there the real problem is the available resources. Without
the business being in background then only way to get resources is by
giving those (possibly new) resources something that they are looking
for. For some its fame, for some its about learning something new, for
some its feeling of being part of something greater, and for some it is
that they need to extend it for their reasons. Now if you think this
from out side angle and looking at Open Watcom project -- it is not
really easy to get into it.
If you hide surrounding complexity then it is easier to understand by
others. By hiding I actually mean that lets say installer would be in
its own repo with only dependency being to rest of OW tree would be the
compiler itself (which should be the release compiler). If the eyes see
too much they get scared.
If we think about the learning something new -- only component I see in
OW is compilers that would have general interest. In universities there
are compiler courses. They are of course a bit slim in way.. but if the
project could provide good learning material on the topic then some of
those readers could be tempted to contribute. If there is no "getting
into dirty details" guidance then it is rather hard to get resources to
understand how things work.
Last but not least is the roadmap - what is planned and when. If you
tell that there are these _new_ features that would be nice to get then
it is easier to interest. Bug fixing gets boring for the long term. You
need to have some way to keep interest for the resource.
I personally see Open Watcom as a tool suite. From there I have just
been using compilers. Nothing more.
Anyway... my thoughts on the topic.
And no -- I don't have excessive free time to increase my activity level.
Thanks,
Vesa Jääskeläinen
Leif Ekblad
Oct 1, 2011, 3:40:00 AM10/1/11
to
I have no problem with the next release happening in about 6 months, but 12
months or more seems a little bit too long.
Just like you, I don't have much spare time, and most of it goes into
updates to RDOS, and eventually finalizing the device-driver support in OW,
so I cannot take on more in the OW project (or anything else).
Regards,
Leif Ekblad
months or more seems a little bit too long.
Just like you, I don't have much spare time, and most of it goes into
updates to RDOS, and eventually finalizing the device-driver support in OW,
so I cannot take on more in the OW project (or anything else).
Regards,
Leif Ekblad
Leif Ekblad
Oct 1, 2011, 3:56:30 AM10/1/11
to
Vesa Jääskeläinen:
> Lots of the code shows its age and do not really stay on the leading
> edge anymore. One question is - can some of that be replaced with
> already existing components from external sources.
Also possible.
> Some problems that I see is in example C library (especially for the
> Linux environment as that is a bit limited). Could that in example be
> combined with Bionic or other easily licensable C libraries. Those who
> are not aware Bionic is licensed with BSD. Same goes for C++ library. Or
> could this work be combine with some other compiler project? Strength
> that I see in current is that it is not GPL (nor LPGL) and it is easy to
> control where it is going.
Not. If you replace the C library with something else, you'd break most
of it, and besides, who would be interested in maintaining a GCC clone?
> Another issue is IDE (and other GUI tools). They are showing their age.
> Would it be feasible to think about replacing those with more recent
> components?. Personally I use Eclipse/CDT for GCC/RVCT based
> development, it should not to be too hard to integrate IDE functionality
> to Eclipse based solution. Problem with Eclipse/CDT is its debuggers.
> They are not the leading edge... Eclipse's license however is rather
> nice for extending it.
That might be a good idea, but OTOH the IDE is ok and functional.
> Then we have the installer. That should be rather easy to replace with
> other solutions. Why not pick some project out there and then improve
> that project with the needs?
Absolutely
> When the problem has been sliced smaller sized problems then it might be
> easier to handle. If those sliced problems (or components if you wish)
> would have own maintainers (internal or external to project) then it
> would be easier to share responsibility of maintenance.
It is also more complex to install if components come from many different
sources. The current solution, with everything integrated, is easy to
install.
> As you said there the real problem is the available resources. Without
> the business being in background then only way to get resources is by
> giving those (possibly new) resources something that they are looking
> for. For some its fame, for some its about learning something new, for
> some its feeling of being part of something greater, and for some it is
> that they need to extend it for their reasons. Now if you think this
> from out side angle and looking at Open Watcom project -- it is not
> really easy to get into it.
Right. But then are other projects of similar complexity any easier?
I don't know, but from my earlier experience with GCC, GCC is
not easier, and also have a horrible review process that takes almost
forever.
> Last but not least is the roadmap - what is planned and when. If you
> tell that there are these _new_ features that would be nice to get then
> it is easier to interest. Bug fixing gets boring for the long term. You
> need to have some way to keep interest for the resource.
>
> I personally see Open Watcom as a tool suite. From there I have just
> been using compilers. Nothing more.
Personally, for me, OW is the primary development platform.
I use it both commersially and privately. I anticipate to stay with OW
for main application and device-driver development for the forseable
future. Even if nobody else contributes to the project.
I suspect it is the "64-bit hype" that is the main problem for more
general interest in OW. A fully functional 64-bit environment for
some major operating system(s) would increase interest.
Another problem is general compability and adherence to C/C++
standards.
Regards,
Leif Ekblad
> I think one way forward would be to define what Open Watcom currently
> is. Then think should there be a reason re-classify and/or split it to
> components.
Possible.
> is. Then think should there be a reason re-classify and/or split it to
> components.
> Lots of the code shows its age and do not really stay on the leading
> edge anymore. One question is - can some of that be replaced with
> already existing components from external sources.
> Some problems that I see is in example C library (especially for the
> Linux environment as that is a bit limited). Could that in example be
> combined with Bionic or other easily licensable C libraries. Those who
> are not aware Bionic is licensed with BSD. Same goes for C++ library. Or
> could this work be combine with some other compiler project? Strength
> that I see in current is that it is not GPL (nor LPGL) and it is easy to
> control where it is going.
of it, and besides, who would be interested in maintaining a GCC clone?
> Another issue is IDE (and other GUI tools). They are showing their age.
> Would it be feasible to think about replacing those with more recent
> components?. Personally I use Eclipse/CDT for GCC/RVCT based
> development, it should not to be too hard to integrate IDE functionality
> to Eclipse based solution. Problem with Eclipse/CDT is its debuggers.
> They are not the leading edge... Eclipse's license however is rather
> nice for extending it.
> Then we have the installer. That should be rather easy to replace with
> other solutions. Why not pick some project out there and then improve
> that project with the needs?
> When the problem has been sliced smaller sized problems then it might be
> easier to handle. If those sliced problems (or components if you wish)
> would have own maintainers (internal or external to project) then it
> would be easier to share responsibility of maintenance.
sources. The current solution, with everything integrated, is easy to
install.
> As you said there the real problem is the available resources. Without
> the business being in background then only way to get resources is by
> giving those (possibly new) resources something that they are looking
> for. For some its fame, for some its about learning something new, for
> some its feeling of being part of something greater, and for some it is
> that they need to extend it for their reasons. Now if you think this
> from out side angle and looking at Open Watcom project -- it is not
> really easy to get into it.
I don't know, but from my earlier experience with GCC, GCC is
not easier, and also have a horrible review process that takes almost
forever.
> Last but not least is the roadmap - what is planned and when. If you
> tell that there are these _new_ features that would be nice to get then
> it is easier to interest. Bug fixing gets boring for the long term. You
> need to have some way to keep interest for the resource.
>
> I personally see Open Watcom as a tool suite. From there I have just
> been using compilers. Nothing more.
I use it both commersially and privately. I anticipate to stay with OW
for main application and device-driver development for the forseable
future. Even if nobody else contributes to the project.
I suspect it is the "64-bit hype" that is the main problem for more
general interest in OW. A fully functional 64-bit environment for
some major operating system(s) would increase interest.
Another problem is general compability and adherence to C/C++
standards.
Regards,
Leif Ekblad
Peter C. Chapin
Oct 1, 2011, 8:48:43 AM10/1/11
to
On 2011-10-01 03:40, Leif Ekblad wrote:
> I have no problem with the next release happening in about 6 months, but 12
> months or more seems a little bit too long.
Six months sounds reasonable. Perhaps we should plan on that. That would
> I have no problem with the next release happening in about 6 months, but 12
> months or more seems a little bit too long.
take us to the first of March 2012. I could perhaps make that happen.
Peter
Peter C. Chapin
Oct 1, 2011, 8:57:08 AM10/1/11
to
On 2011-10-01 01:56, Vesa Jääskeläinen wrote:
> Sorry for the long post ;)
Thanks for all your comments. You bring up many points that might be
> Sorry for the long post ;)
best served in separate threads. However my general reaction is that if
we had to the time to split Open Watcom up into semi-independent
subprojects we'd probably have the time to make forward progress on the
system as it is currently organized. In other words the split you are
talking about isn't something a project starved for time could realistic
attempt to do. The easiest path by far is the path that entails the
fewest changes.
More specifically, one of Open Watcom's strengths is that it is highly
independent of other projects. I originally got interested in Open
Watcom because I saw it as a viable alternative on Windows to Visual
C/C++ and gcc. If Open Watcom borrows too much from other projects that
source of appeal (to me) is greatly reduced.
On the other hand I can appreciate that a project with minimal human
resources needs to take what it can "off the shelf."
Peter
Lynn McGuire
Oct 3, 2011, 11:53:41 AM10/3/11
to
is independent of all other projects. OW is hopelessly
behind the compiler development curve but that does not
mean that it needs to be discarded. Instead,
occasional pruning will work to fix the bugs that popup
and some forward progress will be made by people trying
to add specific features for their needs.
I keep hoping that some high energy person who does
not need a day job will come in here and clean the
place up but that is really a thankless and herculean
task.
Is Michal still working in the background ? What about
Steve ? They both know OW forwards and backwards.
Lynn
Steven Levine
Oct 3, 2011, 12:28:35 PM10/3/11
to
On Mon, 3 Oct 2011 15:53:41 UTC, Lynn McGuire <l...@winsim.com> wrote:
Hi,
> I keep hoping that some high energy person who does
> not need a day job will come in here and clean the
> place up but that is really a thankless and herculean
> task.
It would be nice, but regardless of the state of the economy, folks
with skills tend to stay busy.
> Is Michal still working in the background ?
Yes.
> What about
> Steve ?
I've not seen any postings from Stephen Howe for quite a while.
Hopefully, he's OK.
FWIW, if you are curious as to what is going on
p4 changes -m20
is handy. If want to see just a specific user, it's
p4 changes -m20 -u username
To see the client names, it's
p4x clients
This can also be done from the PerForce web gui.
Steven
--
---------------------------------------------------------------------
Steven Levine <ste...@earthlink.bogus.net>
eCS/Warp/DIY etc. www.scoug.com www.ecomstation.com
---------------------------------------------------------------------
Hi,
> I keep hoping that some high energy person who does
> not need a day job will come in here and clean the
> place up but that is really a thankless and herculean
> task.
with skills tend to stay busy.
> Is Michal still working in the background ?
> What about
> Steve ?
I've not seen any postings from Stephen Howe for quite a while.
Hopefully, he's OK.
FWIW, if you are curious as to what is going on
p4 changes -m20
is handy. If want to see just a specific user, it's
p4 changes -m20 -u username
To see the client names, it's
p4x clients
This can also be done from the PerForce web gui.
Steven
--
---------------------------------------------------------------------
Steven Levine <ste...@earthlink.bogus.net>
eCS/Warp/DIY etc. www.scoug.com www.ecomstation.com
---------------------------------------------------------------------
E. S. Fabian
Oct 3, 2011, 1:04:36 PM10/3/11
to
From: Steven Levine
same name) in another NG (the JPsoft Forum) with some frequency. IIRC he
indicated some time ago that he is very busy...
--
Steve Fabian
| Lynn McGuire wrote:
|| What about Steve ?
|
| I've not seen any postings from Stephen Howe for quite a while.
| Hopefully, he's OK.
He must be OK, I see postings from him (really him and not someone with the
|| What about Steve ?
|
| I've not seen any postings from Stephen Howe for quite a while.
| Hopefully, he's OK.
same name) in another NG (the JPsoft Forum) with some frequency. IIRC he
indicated some time ago that he is very busy...
--
Steve Fabian
Nomen Nescio
Oct 4, 2011, 6:47:51 AM10/4/11
to
"Steven Levine" <ste...@nomail.earthlink.net> wrote:
> On Mon, 3 Oct 2011 15:53:41 UTC, Lynn McGuire <l...@winsim.com> wrote:
>
> Hi,
>
> > I keep hoping that some high energy person who does
> > not need a day job will come in here and clean the
> > place up but that is really a thankless and herculean
> > task.
>
> It would be nice, but regardless of the state of the economy, folks
> with skills tend to stay busy.
If only that were even close to being true. Reality is, the cheap stay busy
> On Mon, 3 Oct 2011 15:53:41 UTC, Lynn McGuire <l...@winsim.com> wrote:
>
> Hi,
>
> > I keep hoping that some high energy person who does
> > not need a day job will come in here and clean the
> > place up but that is really a thankless and herculean
> > task.
>
> It would be nice, but regardless of the state of the economy, folks
> with skills tend to stay busy.
and those with skills look for jobs. Got outsourcing?
0 new messages