| |
"Tim Courtney" <tim@zacktron.com> wrote in message news:GCMIxI.BvL@lugnet.com...
> I've been pushing for this inside for quite some time, with little response.
> Even splitting up the work for a big update among some of us who discuss
> LDraw.org issues offline.
>
> I'm still mixed, between the need for an update, and the understanding that • this
> is a volunteer project that no one gets paid for, though.
Just sent another email to a few people about this. I hope there'll be some
progress.
-Tim
|
|
| |
"Tim Courtney" <tim@zacktron.com> writes:
> [...] "pushing for this inside" [...] "among some of us who discuss
> LDraw.org issues offline" [...] "sent another email to a few
> people" [...]
It seems to me that LDraw has become a project of secrecy and inside
work. When did it stop being an "open" project? When did discussions
concerning LDraw cease to be done in the open? I think I missed this.
I realize that whoever work with the project may want some "anonymity"
to avoid being "spammed" with beginner's questions and so on. It is
probably hard to direct people to the newsgroups. Some people tend to
want to email a single person with a question, rather than posting on
a newsgroup. But is there really a need to hide all the work from the
public anyway?
I realize also that discussions on the newsgroup tend to get all noisy
and go nowhere. This is probably frustrating to work with, and I can
see the desire to go underground. It is probably a lot easier to get
things done when working alone, not having to achieve consensus among
a diverse group as the people (left) on lugnet.cad.*.
Fredrik
|
|
| |
"Fredrik Glöckner" <fredrik.glockner@bio.uio.no> wrote in message
news:qrd1yq98itn.fsf@hephaistos.uio.no...
> "Tim Courtney" <tim@zacktron.com> writes:
>
> > [...] "pushing for this inside" [...] "among some of us who discuss
> > LDraw.org issues offline" [...] "sent another email to a few
> > people" [...]
>
>
> It seems to me that LDraw has become a project of secrecy and inside
> work. When did it stop being an "open" project? When did discussions
> concerning LDraw cease to be done in the open? I think I missed this.
>
> I realize that whoever work with the project may want some "anonymity"
> to avoid being "spammed" with beginner's questions and so on. It is
> probably hard to direct people to the newsgroups. Some people tend to
> want to email a single person with a question, rather than posting on
> a newsgroup. But is there really a need to hide all the work from the
> public anyway?
Its not a matter of hiding the work, see below.
> I realize also that discussions on the newsgroup tend to get all noisy
> and go nowhere. This is probably frustrating to work with, and I can
> see the desire to go underground. It is probably a lot easier to get
> things done when working alone, not having to achieve consensus among
> a diverse group as the people (left) on lugnet.cad.*.
This is the case. Newsgroup discussions get noisy - people tend to have long
drawn out arguments, and most of the time those people aren't even directly
involved. We discussed this in NY when just the LDraw people got together
(aside from LEGO). Because of the efforts to draft the parts license, form an
official organization, etc, we have found it more appropriate to move some
discussion off of LUGNET. This is because we want progress - a smaller group
who is more dedicated can stay focused easier and more will get done. This is
from experience.
This is the best explanation I can give. Its not out of a desire to be secret,
elite, or anything of that nature. Its a desire that a lot of us who are doing
the work have to keep things rolling and not get bogged down by newsgroup
discussion. It results from feelings expressed at a couple in person meetings -
NYC and my visit to Steve's where Steve, Larry, and I discussed LDraw issues.
The good news from this is, allthough sometimes it moves slow, we are able to
get a lot more done becuase it is small. We've got a parts license 95% done and
ready to go. We'll tackle the organization stuff so LDraw can do more things
and act as an entity.
The newsgroups are still good for discussing, and interacting with those who do
work on LDraw related stuff. They've helped out a lot over the last two years
towards the development of the site, and will continue to be an invaluable
resource. Some of us just find a need to discuss certain things privately so
decisions get made faster - we're the ones doing the work and we're the ones
whose work the decision will directly effect - the work of the website and the
software being released.
Hope this helps
-Tim
|
|
| |
In lugnet.cad.dev, Fredrik Glöckner writes:
> It seems to me that LDraw has become a project of secrecy and inside
> work. When did it stop being an "open" project? When did discussions
> concerning LDraw cease to be done in the open? I think I missed this.
Thank you Fredrik, for so succinctly expressing what I was feeling. I was
building up a lot of frustration because I felt I had to make a loud noise
in order to be heard.
> I realize also that discussions on the newsgroup tend to get all noisy
> and go nowhere. This is probably frustrating to work with, and I can
> see the desire to go underground. It is probably a lot easier to get
> things done when working alone, not having to achieve consensus among
> a diverse group as the people (left) on lugnet.cad.*.
I think the discussions should still be in the open. No consensus needs to
be reached in .cad per se, as long as those responsible for the
administration agree -- but issues like the long-term planning,
organization, and meeting with Lego using NDAs should not be done with the
LDraw users in the dark.
Cheers,
- jsproat
|
|
| |
HeeHeee!
I have great faith in Tim that one day LDraw will become an
open project, but I don't think it has ever been one in the
past! :)
As soon as the parts library is opened up, I promise
I'll check it all into Sourceforge.org, so one can use industry-
standard tools like CVS to update parts and see what is new,
check in bug fixes (if permissions allow), record versioning
information, etc.
-gyug
In lugnet.cad.dev, Fredrik Glöckner writes:
> ">It seems to me that LDraw has become a project of secrecy and inside
> work. When did it stop being an "open" project? When did discussions
> concerning LDraw cease to be done in the open? I think I missed this.
>
|
|
| |
"Paul Gyugyi" <paul@gyugyi.com> writes:
> I have great faith in Tim that one day LDraw will become an open
> project, but I don't think it has ever been one in the past! :)
This sounds almost like communistic ideology: For the project to
ultimately become open, it must first become more closed for a while!
;-)
> As soon as the parts library is opened up, I promise I'll check it
> all into Sourceforge.org, so one can use industry- standard tools
> like CVS to update parts and see what is new, check in bug fixes
> (if permissions allow), record versioning information, etc.
That's undoubtedly an interesting idea.
Fredrik
|
|
| |
In lugnet.cad.dev, Paul Gyugyi wrote:
> As soon as the parts library is opened up, I promise
> I'll check it all into Sourceforge.org, so one can use industry-
> standard tools like CVS to update parts and see what is new,
> check in bug fixes (if permissions allow), record versioning
> information, etc.
Paul,
I don't know if you saw earlier messages about the Part Tracker, but there
was some discussion about whether or not we could use CVS as a backend to
an automated system. What do you think?
There are a number of features we need that (I think) CVS does not offer
'out of the box'. These are:
1. Unofficial/Official flag. When files are submitted, they are
unofficial. Once they're released, they're official. Or turn this around,
and once files are official (ie, certified), they're released.
2. Certification process. After files are submitted, they need to be
peer-reviewed. After they're peer-reviewed and OK'ed, they can be
auto-included in the next update cycle.
3. File-checking. When files are submitted, they should be auto-checked
for syntactic problems.
A couple of issues with using CVS in development with the Parts Tracker
are: (remember, I've got no experience with CVS; some of these could be
ignorance)
1. NIH. Discount this one if you want. Or call it 'unfamiliarity with CVS
causes project to be more complicated rather than less'.
2. CVS gives us functionality that is nice to have, but not required for
the PT.
STeve
|
|
| |
In lugnet.cad.dev, Steve Bliss writes:
> I don't know if you saw earlier messages about the Part Tracker, but there
> was some discussion about whether or not we could use CVS as a backend to
> an automated system. What do you think?
I'll go look up the discussion. I'm just getting back into the
swing of things. I think CVS would be a great way of distributing
the parts. There's clearly some managment tools needed to help
with organizing the submissions and reviews of new parts. Some
of this could be helped with CVS, but you would probably need
a front-end program. If you go to www.sourceforge.com and
search for "wincvs", you'll find the "The CvsGui project" with
MFC C++ sourcecode for calling CVS and other things that will
help tie CVS into a front-end. A popular front end, WinCVS,
should be at http://www.wincvs.org, but that site seems down today.
> There are a number of features we need that (I think) CVS does not offer
> 'out of the box'. These are:
>
> 1. Unofficial/Official flag. When files are submitted, they are
> unofficial. Once they're released, they're official. Or turn this around,
> and once files are official (ie, certified), they're released.
This could be managed with CVS branches. One branch, the unofficial,
can be updated by anyone. The other, the official, is read-only.
The LDRAW Overlord Government is responsible for approving parts,
and moves them from one branch to the other. When you get files
with CVS, it can insert comments in the files that say which
branch the files came from.
> 2. Certification process. After files are submitted, they need to be
> peer-reviewed. After they're peer-reviewed and OK'ed, they can be
> auto-included in the next update cycle.
CVS lets you find out "what is new" in a repository. For instance,
the command "cvs diff -rFEB2001_RELEASE *.dat" would show me all files
that are different since the febuary 2001 release (I assuming I tagged
the files with the FEB2001_RELEASE label at the time of the release).
I'm afraid CVS doesn't give you any additional help in organizing
peer-review, although I'll see if source-forge's bug tracking
abilities can be of use.
> 3. File-checking. When files are submitted, they should be auto-checked
> for syntactic problems.
You'll need a front-end for this.
> A couple of issues with using CVS in development with the Parts Tracker
> are: (remember, I've got no experience with CVS; some of these could be
> ignorance)
>
> 1. NIH. Discount this one if you want. Or call it 'unfamiliarity with CVS
> causes project to be more complicated rather than less'.
I've been using CVS for a few years now. It is complicated, but only
because the tasks involved are complicated. The big advantage is
being able to find out what is new and what has changed
(and who changed it).
> 2. CVS gives us functionality that is nice to have, but not required for
> the PT.
>
> STeve
|
|
| |
In lugnet.cad.dev, Paul Gyugyi writes:
> In lugnet.cad.dev, Steve Bliss writes:
> > I don't know if you saw earlier messages about the Part Tracker, but there
> > was some discussion about whether or not we could use CVS as a backend to
> > an automated system. What do you think?
>
> I'll go look up the discussion. I'm just getting back into the
> swing of things. I think CVS would be a great way of distributing
> the parts. There's clearly some managment tools needed to help
> with organizing the submissions and reviews of new parts. Some
> of this could be helped with CVS, but you would probably need
> a front-end program. If you go to www.sourceforge.com and
> search for "wincvs", you'll find the "The CvsGui project" with
> MFC C++ sourcecode for calling CVS and other things that will
> help tie CVS into a front-end. A popular front end, WinCVS,
> should be at http://www.wincvs.org, but that site seems down today.
>
> > There are a number of features we need that (I think) CVS does not offer
> > 'out of the box'. These are:
> >
> > 1. Unofficial/Official flag. When files are submitted, they are
> > unofficial. Once they're released, they're official. Or turn this around,
> > and once files are official (ie, certified), they're released.
>
>
> This could be managed with CVS branches. One branch, the unofficial,
> can be updated by anyone. The other, the official, is read-only.
> The LDRAW Overlord Government is responsible for approving parts,
> and moves them from one branch to the other. When you get files
> with CVS, it can insert comments in the files that say which
> branch the files came from.
>
> > 2. Certification process. After files are submitted, they need to be
> > peer-reviewed. After they're peer-reviewed and OK'ed, they can be
> > auto-included in the next update cycle.
>
>
> CVS lets you find out "what is new" in a repository. For instance,
> the command "cvs diff -rFEB2001_RELEASE *.dat" would show me all files
> that are different since the febuary 2001 release (I assuming I tagged
> the files with the FEB2001_RELEASE label at the time of the release).
> I'm afraid CVS doesn't give you any additional help in organizing
> peer-review, although I'll see if source-forge's bug tracking
> abilities can be of use.
>
>
> > 3. File-checking. When files are submitted, they should be auto-checked
> > for syntactic problems.
>
> You'll need a front-end for this.
I think a small script running "l3p -check" called from within the CVS
commitinfo file would work great for this.
>
> > A couple of issues with using CVS in development with the Parts Tracker
> > are: (remember, I've got no experience with CVS; some of these could be
> > ignorance)
> >
> > 1. NIH. Discount this one if you want. Or call it 'unfamiliarity with CVS
> > causes project to be more complicated rather than less'.
>
> I've been using CVS for a few years now. It is complicated, but only
> because the tasks involved are complicated. The big advantage is
> being able to find out what is new and what has changed
> (and who changed it).
>
>
> > 2. CVS gives us functionality that is nice to have, but not required for
> > the PT.
> >
> > STeve
|
|
|
