iamb-dev-internal — Discussion of the gory details involved in a rewrite for the next major version

Re: [Iamb-dev-internal] query language

[Matt C. <su...@qx...>]

On Fri, 2 Mar 2001, Rich Bowen wrote:

> select * from streed where timestamp > $time

we're talking to a socket here. the query language i'm looking for is for
the API module to talk to the socket. when can make the API to the socket
do anything you want on the user end. kernel end however i'm not sure we
want to go that direction. i was thinking something very simple like

STREE 1 
^ get stree for forumID 1

ADD 1 5[6]
^ ADD ArticleID 6 as a child of article 5 in forumID 1

DEL 1 6
^ delete article 6 from forumID 1

etc etc. you get the direction i'm going i gather. 

and DBD::RAM will not be doing the heavy lifting in the final
version. most of the data storage will be internal data structures in ram
that are written back to sql via dbi periodically.  DBD::RAM is a
convienent means to the ends of a test database without having to set
anything up :)

m.

Re: [Iamb-dev-internal] query language

[Rich B. <rb...@rc...>]

Matt Cashner wrote:
> 
> i'm working on the guts of B<streed> and it suddenly occurs to me that i
> have no language defined for speaking to the server. now, it could be that
> you all leave it up to me and i build the daemon and a module to speak to
> the daemon and no one really cares as long as it works. but i figure you
> all actually want some input. so, what kind of commands should the server
> accept on a socket and what sort of spooge should it emit?
> 
> and as a word of preparation, i'm using the uber cool DBD::RAM for RDBMS
> emulation in testing, so when i release a pre-alpha of this, you'll want
> to make sure you have DBD::RAM installed :)

Personally, I think something as closely approaching SQL as possible
would be very cool.

select * from streed where timestamp > $time

that sort of thing.

Not sure how easy that is, but one would think that with DBD::RAM doing
most of the heavy lifting there, it would not be too much of a hardship.

-- 
Rich Bowen <rb...@rc...>
Come see me at Apachecon! -- http://www.apachecon.com/

[Iamb-dev-internal] query language

[Matt C. <su...@qx...>]

i'm working on the guts of B<streed> and it suddenly occurs to me that i
have no language defined for speaking to the server. now, it could be that
you all leave it up to me and i build the daemon and a module to speak to
the daemon and no one really cares as long as it works. but i figure you
all actually want some input. so, what kind of commands should the server
accept on a socket and what sort of spooge should it emit?

and as a word of preparation, i'm using the uber cool DBD::RAM for RDBMS
emulation in testing, so when i release a pre-alpha of this, you'll want
to make sure you have DBD::RAM installed :)

m.

Re: [Iamb-dev-internal] Streed

[Rich B. <rb...@rc...>]

Matt Cashner wrote:
> 
> i have just committed to cvs (dev.rcbowen.com) a functional socket
> layer called streed. this is the beginnings of the stree daemon i've
> been babbling about for so long. as the commit note says, the stree
> layer is absent and i need to define a query language which i will rfc
> first. anyway, dinner guests calleth.

In case an official edict is necessary, it should be noted that one is
implied in the above note. This branch of the code is no longer using
SourceForge for CVS services. Any code in CVS on SourceForge, or which
is in the future added to CVS on sourceforge, in the -internals tree, or
the -dev tree, or whatever it is called, is no longer particularly
relevant.

The stable branch is still out there, but if I happen to have a
hankering to start working on it again, I will likely move it to
cvs.rcbowen.com

-- 
Rich Bowen <rb...@rc...>
Come see me at Apachecon! -- http://www.apachecon.com/

[Iamb-dev-internal] Streed

[Matt C. <su...@qx...>]

i have just committed to cvs (dev.rcbowen.com) a functional socket 
layer called streed. this is the beginnings of the stree daemon i've
been babbling about for so long. as the commit note says, the stree
layer is absent and i need to define a query language which i will rfc
first. anyway, dinner guests calleth.

m.

Re: [Iamb-dev-internal] ssh

[Rich B. <rb...@rc...>]

Matt Cashner wrote:
> 
> can we get openssh out here? ala rpm -e ssh  and install openssl and
> openssh.  plain ol non-open_source ssh does strange things and i cant
> figure out how to do key-based auth on a user end. i'm doing what the docs
> say and its not working. it may be that the key structures are
> different. i'm not sure. anyway...

I've tried uninstalling SSH and anything related to it, via RPM, and via
trying to ferret out files and remove them myself. Neither method has
had the desired effect. 

I have installed OpenSSH and OpenSSL probably a dozen times, just hoping
that I had done something wrong all the times before and I would get it
right the next time.

I think that by this time I have screwed things up so thoroughally that
I need to just dump the whole server and start over. Which I will do, as
soon as I figure out what data on here needs to be backed up, and which
I can just ditch.

-- 
  Rich Bowen --  Director of Web Application Development
http://www.cre8tivegroup.com/  --  ri...@cr...
  Come see me at Apachecon! -- http://www.apachecon.com/

Re: [Iamb-dev-internal] grrr

[Rich B. <rb...@rc...>]

Matt Cashner wrote:
> 
> On Wed, 7 Feb 2001, Rich Bowen wrote:
> 
> > make things easier). The only problem at the moment is that I've got a
> > strange version of sshd, and I don't know how to upload your keys so you
> > don't have to put in your password every time.
> 
> um, i think we can do that :)  ~/.ssh/authorized_keys

Yeah. Tried that. Doesn't work. I think it has something to with with
just having the wrong/different version. It stores authorized keys
somewhere else, apparently. Or I'm doing something else wrong.
 
> > And for some reason
> > (because it's RedHat?) although I've rebuilt sshd about 12 times, I
> > still get the old version.
> 
> mayhaps redhat came with ssh and it is installed in a different place than
> where the Makefile is putting it. might need to uninstall an rpm or two
> first.

I'll try that. I think I did already, but I don't remember for sure.
 
> > And scp doesn't work, which is related, but
> > different.
> 
> symlink /usr/local/bin/scp into /usr/bin

Yeah. Did that too. Not joy yet.

-- 
  Rich Bowen --  Director of Web Application Development
http://www.cre8tivegroup.com/  --  ri...@cr...
  Come see me at Apachecon! -- http://www.apachecon.com/

[Iamb-dev-internal] recommendation

[Matt C. <su...@qx...>]

this is a bit premature but as we move forward in code, i'd like us to do
the html ui of iamb using lots of css kept in a standard.css file
someplace (tho that name is not important). that way folks can redefine
styles and color without having to hack templates or (god
forbid) code. thoughts?

m.

Re: [Iamb-dev-internal] grrr

[Matt C. <su...@qx...>]

On Wed, 7 Feb 2001, Rich Bowen wrote:

> CVS on cvs.rcbowen.com is at /home/cvs, access via ssh. You each have

the code has been imported into module iamb-dev.  might want to fix
syncmail tho. it mailed pf...@mo... the notification message :) but
that and the keys issue aside we're ready cvs wise on dev.rcbowen to move
ahead.

m.

[Iamb-dev-internal] ssh

[Matt C. <su...@qx...>]

can we get openssh out here? ala rpm -e ssh  and install openssl and
openssh.  plain ol non-open_source ssh does strange things and i cant
figure out how to do key-based auth on a user end. i'm doing what the docs
say and its not working. it may be that the key structures are
different. i'm not sure. anyway...

m.

Re: [Iamb-dev-internal] grrr

[Matt C. <su...@qx...>]

On Wed, 7 Feb 2001, Rich Bowen wrote:

> make things easier). The only problem at the moment is that I've got a
> strange version of sshd, and I don't know how to upload your keys so you
> don't have to put in your password every time. 

um, i think we can do that :)  ~/.ssh/authorized_keys

> And for some reason
> (because it's RedHat?) although I've rebuilt sshd about 12 times, I
> still get the old version. 

mayhaps redhat came with ssh and it is installed in a different place than
where the Makefile is putting it. might need to uninstall an rpm or two
first.

> And scp doesn't work, which is related, but
> different. 

symlink /usr/local/bin/scp into /usr/bin

m.> 

Re: [Iamb-dev-internal] grrr

[Rich B. <rb...@rc...>]

Matt Cashner wrote:
> 
> can anyone get a valid checkout of the new stree.t and Stree.pm ?  because
> of sourceforge's bastard connection my ssh/cvs sessions keep timing out
> and i think the commit borked. i am so annoyed.

I got:

rbowen@rhiannon:/tmp$ cvs -d
rb...@cv...:/cvsroot/iamb co iamb-dev
cvs server: Updating iamb-dev
cvs server: Updating iamb-dev/Community
cvs server: Updating iamb-dev/IAMB
cvs server: Updating iamb-dev/RFC
U iamb-dev/RFC/RFC_01.pod
U iamb-dev/RFC/RFC_02.pod
U iamb-dev/RFC/RFC_03.pod
cvs server: Updating iamb-dev/pm
cvs server: Updating iamb-dev/pm/IAMB
U iamb-dev/pm/IAMB/Stree.pm
cvs server: Updating iamb-dev/pm/t
U iamb-dev/pm/t/stree.t
cvs server: Updating iamb-dev/scripts
cvs server: Updating iamb-dev/setup
cvs server: Updating iamb-dev/templates


CVS on cvs.rcbowen.com is at /home/cvs, access via ssh. You each have
accounts (krietz and sungo, but I can change these names if it would
make things easier). The only problem at the moment is that I've got a
strange version of sshd, and I don't know how to upload your keys so you
don't have to put in your password every time. And for some reason
(because it's RedHat?) although I've rebuilt sshd about 12 times, I
still get the old version. And scp doesn't work, which is related, but
different. 

I really need to rebuild that box some day, but I'll probably wait for a
2.4 distribution.

I think that there are still mailing lists on there still. I'll need to
check.

The only issue would be backups, which happen just once a week, if I
remember. Not very reliable. I hope to get a tape drive eventually,
which will make this more regular, but backing up to a CDR on a Windows
box across the network is hard to automate.

-- 
  Rich Bowen --  Director of Web Application Development
http://www.cre8tivegroup.com/  --  ri...@cr...
  Come see me at Apachecon! -- http://www.apachecon.com/

Re: [Iamb-dev-internal] grrr

[Ken R. <kr...@qx...>]

> > Methinks that exposure will happen if and when we get this thing out
> > there and known. Until then, the likelihood of a quality coder happening
> > along is so small that it isn't worth the aggravation (to Matt at least,
> > since I am, thankfully, out of that loop).
> 
> I certainly hope you're not out of that loop. At some point, I would
> hope that you'll be directly committing code changes. I don't want to
> have to send all diffs through Matt from here on out. Particularly once
> we have parts of the development carved up amoung us.

Hmm. Not very clear there. I was referring to the aggravation Matt
was having with SourceForge. That is not something I would really
enjoy doing. But as grand poobah of the dev branch, Matt does have
to wrestle that gorilla.

And, of course, I do plan on coding, and committing directly, and
all that.

-- Ken

Re: [Iamb-dev-internal] grrr

[Ken R. <kr...@qx...>]

> > there and known. Until then, the likelihood of a quality coder happening
> > along is so small that it isn't worth the aggravation (to Matt at least,
> > since I am, thankfully, out of that loop).
> 
> which loop? i'm confused.

Running off tests.... :-)

The one that administrates the code base.

-- Ken

Re: [Iamb-dev-internal] grrr

[Matt C. <su...@qx...>]

On Tue, 6 Feb 2001, Matt Cashner wrote:

> can anyone get a valid checkout of the new stree.t and Stree.pm ?  because
> of sourceforge's bastard connection my ssh/cvs sessions keep timing out
> and i think the commit borked. i am so annoyed.

nm. blew away the tree and did a fresh co and i have a seemingly
functional tree. sf--

m.

Re: [Iamb-dev-internal] grrr

[Matt C. <su...@qx...>]

can anyone get a valid checkout of the new stree.t and Stree.pm ?  because
of sourceforge's bastard connection my ssh/cvs sessions keep timing out
and i think the commit borked. i am so annoyed.

m.

Re: [Iamb-dev-internal] grrr

[Matt C. <su...@qx...>]

On Tue, 6 Feb 2001, Rich Bowen wrote:

> I certainly hope you're not out of that loop. At some point, I would
> hope that you'll be directly committing code changes. I don't want to
> have to send all diffs through Matt from here on out. Particularly once
> we have parts of the development carved up amoung us.

no, there will be none of that. everything is in cvs and will be developed
in cvs from here on out. the sending patches to me thing was so we could
get the module acceptable to ken before blessing it and deeming it good.

m.

Re: [Iamb-dev-internal] grrr

[Rich B. <rb...@rc...>]

Ken Rietz wrote:
> 
> > I've been moderately disillusioned with SourceForge also. The other
> > major benefit that it was supposed to give us - exposure to a larger
> > group of potential developers - has also not happened, and I don't know
> > if it will or not. But I'm equally not sure if having it somewhere else
> > is a particular detriment to that.
> 
> Methinks that exposure will happen if and when we get this thing out
> there and known. Until then, the likelihood of a quality coder happening
> along is so small that it isn't worth the aggravation (to Matt at least,
> since I am, thankfully, out of that loop).

I certainly hope you're not out of that loop. At some point, I would
hope that you'll be directly committing code changes. I don't want to
have to send all diffs through Matt from here on out. Particularly once
we have parts of the development carved up amoung us.

--
  Rich Bowen --  Director of Web Application Development
http://www.cre8tivegroup.com/  --  ri...@cr...
  Come see me at Apachecon! -- http://www.apachecon.com/

Re: [Iamb-dev-internal] I wanna write code, durn it!

[Rich B. <rb...@rc...>]

Ken Rietz wrote:
> 
> > I wasn't thinking so much of converting databases now. What I'm more
> > interested in is a method that I can in some sense feed an entire thread
> > to, and get a stree. I'm not quite sure if it's even an important
> > question, but it will certainly help in the testing phase to be able to
> > get a set of article objects and create a stree out of them.
> 
> Since I seem to be the local stree surgeon, I'll tackle it, if you
> can tell me the structure of the thread you want to feed it. It won't
> happen until tomorrow, though, since I still have to write a test
> with 3 versions tonight.

I don't know, really. I was thinking that I would just

select * from table where threadID = 12;

Then somehow plant the parent one, then each of it's kids. It does not
even really need to be in the API (I don't think) since it's really just
for testing. It could just be a supplementary script

-- 
  Rich Bowen --  Director of Web Application Development
http://www.cre8tivegroup.com/  --  ri...@cr...
  Come see me at Apachecon! -- http://www.apachecon.com/

Re: [Iamb-dev-internal] I wanna write code, durn it!

[Rich B. <rb...@rc...>]

Ken Rietz wrote:
> 
> > i was just going to commit source. i will wait for your patch.
> 
> Viola, or whatever that French thingy is.

I think it's an accordian.

-- 
  Rich Bowen --  Director of Web Application Development
http://www.cre8tivegroup.com/  --  ri...@cr...
  Come see me at Apachecon! -- http://www.apachecon.com/

Re: [Iamb-dev-internal] grrr

[Matt C. <su...@qx...>]

On Tue, 6 Feb 2001, Ken Rietz wrote:

> there and known. Until then, the likelihood of a quality coder happening
> along is so small that it isn't worth the aggravation (to Matt at least,
> since I am, thankfully, out of that loop). 

which loop? i'm confused.

Re: [Iamb-dev-internal] grrr

[Matt C. <su...@qx...>]

tonight i have effectively been working on a 2400 baud modem connection to
sourceforge.  cvs changes takes 5 minutes to proces becuse of the ssh
stuff going on over such a slow link.  i could have had this done hours
ago if it werent for this issue.  as we speak, i am waiting for the cvs
add for the test suit and moduel to finish so i can commit them. which
will take another 5 minutes.  

Re: [Iamb-dev-internal] grrr

[Ken R. <kr...@qx...>]

> I've been moderately disillusioned with SourceForge also. The other
> major benefit that it was supposed to give us - exposure to a larger
> group of potential developers - has also not happened, and I don't know
> if it will or not. But I'm equally not sure if having it somewhere else
> is a particular detriment to that.

Methinks that exposure will happen if and when we get this thing out
there and known. Until then, the likelihood of a quality coder happening
along is so small that it isn't worth the aggravation (to Matt at least,
since I am, thankfully, out of that loop). Returning to dev.rcbowen.com
won't stop developers once they know about IAMB, and will also provide
us with a greater measure of control over who gets to do what once
somebody notices us. Sounds like an overall win.

-- Ken

Re: [Iamb-dev-internal] grrr

[Rich B. <rb...@rc...>]

Matt Cashner wrote:
> 
> i would like to suggest that we move the source tree off of
> sourceforge. i'd like it someplace that a) is not so damn slow  b) is not
> so damn broken  c) is in our control.  frankly, i'm just tired of dealing
> with the issues of sourceforge when have the ability to host these
> services jsut as well if not better than sourceforge.

The various things that we moved to SourceForge to gain simply have not
materialized, and so I would not be averse to moving things back to
dev.rcbowen.com. Basically all that we've ever used on SourceForge was
CVS and the mailing lists. Moving back to dev.rcbowen.com would give us
this, plus RT. The people that appear to be the ones that will be active
developers already have acconts on this machine, so ssh access to CVS
there should not be a problem. And we could set RT up again.

I've been moderately disillusioned with SourceForge also. The other
major benefit that it was supposed to give us - exposure to a larger
group of potential developers - has also not happened, and I don't know
if it will or not. But I'm equally not sure if having it somewhere else
is a particular detriment to that.

-- 
  Rich Bowen --  Director of Web Application Development
http://www.cre8tivegroup.com/  --  ri...@cr...
  Come see me at Apachecon! -- http://www.apachecon.com/

Re: [Iamb-dev-internal] I wanna write code, durn it!

[Matt C. <su...@qx...>]

On Tue, 6 Feb 2001, Ken Rietz wrote:

> 
> > i was just going to commit source. i will wait for your patch.
> 
> Viola, or whatever that French thingy is.

thanks, applied.  

will be in cvs momentarily.

m.