[gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Dhruba Bandopadhyay]

Hello.

This is to the devs.

With the release of firebird 0.6 many changes are being pondered.

- migration of bin & cvs packages and executable to new name
- conflict with firebird-db
- src_uri from mozilla.org change
- firebird-bin deprecation in favour of firebird-cvs
- official maintainers of both bin and cvs (cvs bug has been assigned
but bin bug has not)

Since I've not seen much activity or talk by devs on the above issues I
thought I'd ask what their intentions were and what might take shape in
the near future regarding mozilla et al.

I'd be glad to hear your thoughts as I'm sure would many others.

With regards.


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Russell Tennant]

> Hello.
>
> This is to the devs.
>
> With the release of firebird 0.6 many changes are being pondered.
>
> - migration of bin & cvs packages and executable to new name
> - conflict with firebird-db
> - src_uri from mozilla.org change
> - firebird-bin deprecation in favour of firebird-cvs
> - official maintainers of both bin and cvs (cvs bug has been assigned
> but bin bug has not)
>
> Since I've not seen much activity or talk by devs on the above issues I
> thought I'd ask what their intentions were and what might take shape in
> the near future regarding mozilla et al.
>
> I'd be glad to hear your thoughts as I'm sure would many others.
>
> With regards.

Why not call it MozillaFirebird like they do at mozilla.org. This should
avoid a name collision.




--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Chris PeBenito]

[-- Attachment #1: Type: text/plain, Size: 1247 bytes --]

I'm not handling the firebird browser stuff, so I can't address all of
your concerns, but technically speaking, there would be no name
collision in portage.  Firebird the database is dev-db/firebird. 
Firebird the browser would be net-www/firebird, net-www/firebird-bin,
etc.  I don't know which one would be chosen if you type "emerge
firebird".  To be sure to get the right package, you'd have to specify
the category on the emerge commandline too ("emerge
net-www/firebird").   Naming it MozillaFirebird would be a violation of
Gentoo's development policy because of the capitalization
(mozillafirebird or mozilla-firebird would be ok).

On Sun, 2003-05-18 at 23:54, Russell Tennant wrote:
> > - conflict with firebird-db

> Why not call it MozillaFirebird like they do at mozilla.org. This should
> avoid a name collision.

-- 
Chris PeBenito
<pebenito@gentoo.org>
Developer, Gentoo Linux
Hardened Gentoo Project
 
"Engineering does not require science. Science helps
a lot, but people built perfectly good brick walls
long before they knew why cement works."-Alan Cox

Public Key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE6AF9243
Key fingerprint = B0E6 877A 883F A57A 8E6A  CB00 BC8E E42D E6AF 9243


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Spundun Bhatt]

On Sun, 2003-05-18 at 18:14, Dhruba Bandopadhyay wrote:

> - firebird-bin deprecation in favour of firebird-cvs
Is there any reason now to not have a standard build-from-source-tarball
ebuild? I thought it was not supplied previously because of some
apprehensions about mozilla not behaving well when compiled with gtk2.
But I have been using phoenix-cvs from about 2 weeks now and its very
stable.... so may be mozilla(-firebird) and gtk2 are getting along well
now.
Spundun
p.s. - I am not really a dev.
also I think mozilla-firebird is a descent replacement for now (from
Chris' suggestions)


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Christian Gut]

[-- Attachment #1: Type: text/plain, Size: 811 bytes --]

On Mon, 2003-05-19 at 03:14, Dhruba Bandopadhyay wrote:

> With the release of firebird 0.6 many changes are being pondered.
> 
> - migration of bin & cvs packages and executable to new name
> - conflict with firebird-db
> - src_uri from mozilla.org change
> - firebird-bin deprecation in favour of firebird-cvs
> - official maintainers of both bin and cvs (cvs bug has been assigned
> but bin bug has not)
> 
> Since I've not seen much activity or talk by devs on the above issues I
> thought I'd ask what their intentions were and what might take shape in
> the near future regarding mozilla et al.

http://bugs.gentoo.org/show_bug.cgi?id=21145
-src and -bin ebuilds in there..
one uses MozillaFirebird as name allready... would recommend this, too
-- 
Christian Gut <cycloon@is-root.org>

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Alastair 'liquidx' Tse]

[-- Attachment #1: Type: text/plain, Size: 1587 bytes --]

On Mon, 2003-05-19 at 06:38, Chris PeBenito wrote:
> etc.  I don't know which one would be chosen if you type "emerge
> firebird". 

That is one of the reasons why net-www/firebird would be a bad choice.
Currently, it is ambigious which one would be chosen.

> net-www/firebird").   Naming it MozillaFirebird would be a violation of
> Gentoo's development policy because of the capitalization
> (mozillafirebird or mozilla-firebird would be ok).

Indeed, capitialisation is frowned upon for a package name in portage. I
support it being called mozilla-firebird or even mozilla-browser as it
will be renamed to after 1.4[1]. However, I'm not the maintainer of
phoenix or mozilla, so it is ultimately up to them.

I have no comment about -bin and -cvs. I don't personally use phoenix so
I'm not sure why we have those compared with the regular source release.

Cheers,

Alastair

[1] http://www.mozilla.org/roadmap/branding.html
"When referring to Thunderbird or Firebird before or during the 1.4
release cycle, make sure to use the project name with Mozilla pre-pended
as "Mozilla Thunderbird" or "Mozilla Firebird" instead of Mozilla alone
or Firebird/Thunderbird alone. ... Use the names "Mozilla Browser" and
"Mozilla Mail" to describe the Firebird and Thunderbird projects after
the 1.4 release."

-- 
Alastair 'liquidx' Tse
 >> Gentoo Developer
 >> http://www.liquidx.net/ | http://cvs.gentoo.org/~liquidx/
 >> GPG Key : http://cvs.gentoo.org/~liquidx/liquidx_gentoo_org.asc
 >> FingerPrint : 579A 9B0E 43E8 0E40 EE93  BB1C 38CE 1C7B 3907 14F6


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Georgi Georgiev]

On Mon, May 19, 2003 at 10:56:48AM +0100, Alastair 'liquidx' Tse wrote:
> On Mon, 2003-05-19 at 06:38, Chris PeBenito wrote:
> > etc.  I don't know which one would be chosen if you type "emerge
> > firebird". 
> 
> That is one of the reasons why net-www/firebird would be a bad choice.
> Currently, it is ambigious which one would be chosen.

Well, bad choice or not - there is plenty of ambiguity in the portage tree
already.

chutz@tank / $ cd /usr/portage
chutz@tank portage $ for i in *-*/*; do echo *-*/`basename $i`; done | grep \  | wc -l
     88

(most of these are from app-emacs and app-xemacs, but there are a bunch of
others, too)

This might even teach users to always use "emerge -p" before emerging anything
and also people might get in the habbit of using the complete package name. A
nice enhancement of portage would be to refuse operation on any ambiguous
packages. (this is off-topic, but I think it's worth mentioning)

Personally I'd vote for a "mozilla-browser" if there were a poll.

-- 
 /^^^^^^^^^^^^^^^^^^^^^^^^^^^\/^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\
/ Georgi Georgiev    (-<     / MMM-MM!! So THIS is BIO-         \
\ chutz@chubaka.net  /\   .o)\ NEBULATION!                      /
/ +81(90)6266-1163  V_/_ |(/)/                                  \
\___________________________/\__________________________________/

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Alastair 'liquidx' Tse]

[-- Attachment #1: Type: text/plain, Size: 673 bytes --]

On Mon, 2003-05-19 at 12:01, Georgi Georgiev wrote:
> Well, bad choice or not - there is plenty of ambiguity in the portage tree
> already.

Yes, I think most of the devs are aware of the duplicate names in
portage. We will be trying to reduce the number of these when the
(stable) portage manages package moving/renaming cleanly. 

But right now, it would definately help to not introduce more :)

Cheers,
-- 
Alastair 'liquidx' Tse
 >> Gentoo Developer
 >> http://www.liquidx.net/ | http://cvs.gentoo.org/~liquidx/
 >> GPG Key : http://cvs.gentoo.org/~liquidx/liquidx_gentoo_org.asc
 >> FingerPrint : 579A 9B0E 43E8 0E40 EE93  BB1C 38CE 1C7B 3907 14F6


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Martin Schlemmer]

[-- Attachment #1: Type: text/plain, Size: 465 bytes --]

On Mon, 2003-05-19 at 06:54, Russell Tennant wrote:

> Why not call it MozillaFirebird like they do at mozilla.org. This should
> avoid a name collision.
> 

Another option is also to call it 'Firebird', as portage is currently
still case sensitive.  There was some talk about case a while back, but
I cannot remember what the conclusion was.


-- 

Martin Schlemmer
Gentoo Linux Developer, Desktop/System Team Developer
Cape Town, South Africa



[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Christian Axelsson]

[-- Attachment #1: Type: text/plain, Size: 406 bytes --]

On Mon, 2003-05-19 at 03:14, Dhruba Bandopadhyay wrote:

> - conflict with firebird-db

How about when typing 'emerge firebird' the user get a message
that looks something like:

	Multiple matches against "firebird". Possible options are:
	dev-db/firebird
	net-www/firebird

I think this would be a nice solution to nameconflicts.

My 2 cents

-- 
Christian Axelsson
smiler@lanil.mine.nu

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Dhruba Bandopadhyay]

On Sun, 2003-05-25 at 13:43, Christian Axelsson wrote:
> On Mon, 2003-05-19 at 03:14, Dhruba Bandopadhyay wrote:
> 
> > - conflict with firebird-db
> 
> How about when typing 'emerge firebird' the user get a message
> that looks something like:

No.  I'm told emerge processes are supposed to be non-interactive and I
support that thought.  I'm hoping there's a way to name them
differently.  Incidentally, is it the naming and SRC_URI that is causing
the delay?


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Christian Axelsson]

[-- Attachment #1: Type: text/plain, Size: 449 bytes --]

On Sun, 2003-05-25 at 14:46, Dhruba Bandopadhyay wrote:

> No.  I'm told emerge processes are supposed to be non-interactive and I
> support that thought.

Actually I think dev-db/firebird and net-www/firebird are different
enought, maybe removing the shortcut 'emerge firebird' and enforce the
user to specify the path is the best (TM) way to go? Like how it's done
with the rsync package.

-- 
Christian Axelsson
smiler@lanil.mine.nu

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Georgi Georgiev]

On 25/05/2003 at 14:56:32(+0200), Christian Axelsson used 0.9Kbytes, just to say:
> On Sun, 2003-05-25 at 14:46, Dhruba Bandopadhyay wrote:
> 
> > No.  I'm told emerge processes are supposed to be non-interactive and I
> > support that thought.
> 
> Actually I think dev-db/firebird and net-www/firebird are different
> enought, maybe removing the shortcut 'emerge firebird' and enforce the
> user to specify the path is the best (TM) way to go? Like how it's done
> with the rsync package.

I agree. An error message does not break non-intereactivity.

-- 
 /^^^^^^^^^^^^^^^^^^^^^^^^^^^\/^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\
/ Georgi Georgiev    (-<     / A man wrapped up in himself      \
\ chutz@chubaka.net  /\   .o)\ makes a very small package.      /
/ +81(90)6266-1163  V_/_ |(/)/                                  \
\___________________________/\__________________________________/

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Zack Gilburd]

[-- Attachment #1: signed data --]
[-- Type: text/plain, Size: 826 bytes --]

On Sunday 25 May 2003 06:16, Georgi Georgiev wrote:
> On 25/05/2003 at 14:56:32(+0200), Christian Axelsson used 0.9Kbytes, just to 
say:
> > On Sun, 2003-05-25 at 14:46, Dhruba Bandopadhyay wrote:
> > > No.  I'm told emerge processes are supposed to be non-interactive and I
> > > support that thought.
> >
> > Actually I think dev-db/firebird and net-www/firebird are different
> > enought, maybe removing the shortcut 'emerge firebird' and enforce the
> > user to specify the path is the best (TM) way to go? Like how it's done
> > with the rsync package.
>
> I agree. An error message does not break non-intereactivity.

How about keeping the names unique.  The file that would go in /usr/bin for 
either package would be MozillaFirebird, so why not call them 
MozillaFirebird-bin and MozillaFirebird-cvs?

[-- Attachment #2: signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Jon Portnoy]

On Sun, May 25, 2003 at 01:46:19PM +0100, Dhruba Bandopadhyay wrote:
> On Sun, 2003-05-25 at 13:43, Christian Axelsson wrote:
> > On Mon, 2003-05-19 at 03:14, Dhruba Bandopadhyay wrote:
> > 
> > > - conflict with firebird-db
> > 
> > How about when typing 'emerge firebird' the user get a message
> > that looks something like:
> 
> No.  I'm told emerge processes are supposed to be non-interactive and I
> support that thought.  I'm hoping there's a way to name them
> differently.  Incidentally, is it the naming and SRC_URI that is causing
> the delay?
> 

More like the fact that our phoenix maintainer (phoen][x, incidentally) 
left.

-- 
Jon Portnoy
avenj/irc.freenode.net

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Matthew Kennedy]

Christian Axelsson <smiler@lanil.mine.nu> writes:

> On Mon, 2003-05-19 at 03:14, Dhruba Bandopadhyay wrote:
>
>> - conflict with firebird-db
>
> How about when typing 'emerge firebird' the user get a message
> that looks something like:
>
> 	Multiple matches against "firebird". Possible options are:
> 	dev-db/firebird
> 	net-www/firebird
>
> I think this would be a nice solution to nameconflicts.

What naming conflict? 

There is 

   emerge dev-db/firebird
or 
   emerge net-www/firebird

This is not the first time two packages have a name conflict,
requiring the category name to resolve.

Matt

-- 
Matthew Kennedy
Gentoo Linux Developer

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Kumba]

Matthew Kennedy wrote:
> What naming conflict? 
> 
> There is 
> 
>    emerge dev-db/firebird
> or 
>    emerge net-www/firebird
> 
> This is not the first time two packages have a name conflict,
> requiring the category name to resolve.
> 
> Matt


	Not all users are going to immediately know the different categories 
offhand.  I'd have to opt for the ability that if portage detected a 
naming conflict, it'd show both matches + categories so the user would 
know which one he/she wanted.

	OTOH, calling it net-www/mozillafirebird isn't a bad idea.  I don't 
think we need caps in there, caps seems to be reserved for dev-perl 
stuff only.

--Kumba



--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Georgi Georgiev]

On 28/05/2003 at 04:47:01(-0400), Kumba used 0.7Kbytes just to say:
> 	Not all users are going to immediately know the different categories 
> offhand.  I'd have to opt for the ability that if portage detected a 
> naming conflict, it'd show both matches + categories so the user would 
> know which one he/she wanted.

And abort execution, right?

> 	OTOH, calling it net-www/mozillafirebird isn't a bad idea.  I don't 
> think we need caps in there, caps seems to be reserved for dev-perl 
> stuff only.

Since the caps are used for splitting the two words, an extra dash would be
needed instead of the caps. Concatenating the two makes it hard to read and personally, it would be hard for me to type.

It would have been nice if the official distribution had a name gentoo could
adopt, but looking at ftp://ftp.mozilla.org/pub/firebird/nightly/latest-trunk/
one can see:

Firebird-mac.dmg.gz
MozillaFirebird-i686-pc-linux-gnu.tar.gz
MozillaFirebird-win32.zip

So we're back where we started. They have both "firebird" and "mozilla
firebird".

-- 
 /^^^^^^^^^^^^^^^^^^^^^^^^^^^\/^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\
/ Georgi Georgiev    (-<     / I put up my thumb... and it      \
\ chutz@chubaka.net  /\   .o)\ blotted out the planet Earth. -- /
/ +81(90)6266-1163  V_/_ |(/)/ Neil Armstrong                   \
\___________________________/\__________________________________/

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Troy Dack]

On Wed, 2003-05-28 at 18:22, Matthew Kennedy wrote:
> Christian Axelsson <smiler@lanil.mine.nu> writes:
> 
> > On Mon, 2003-05-19 at 03:14, Dhruba Bandopadhyay wrote:
> >
> >> - conflict with firebird-db
> >
> > How about when typing 'emerge firebird' the user get a message
> > that looks something like:
> >
> > 	Multiple matches against "firebird". Possible options are:
> > 	dev-db/firebird
> > 	net-www/firebird
> >
> > I think this would be a nice solution to nameconflicts.

I think this would be a nice solution too.

> What naming conflict? 
> 
> There is 
> 
>    emerge dev-db/firebird
> or 
>    emerge net-www/firebird
> 
> This is not the first time two packages have a name conflict,
> requiring the category name to resolve.

The best/most often forgotten when trying to emerge a single
package/most annoying example of this is:

	emerge rsync
or
	emerge net-misc/rsync

and just for reference: http://bugs.gentoo.org/show_bug.cgi?id=2365

-- 
Troy Dack					http://linux.tkdack.com
<troy@tkdack.com>				http://webportage.sf.net

Public Key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x4D90BE3C
Key fingerprint = 1F3D 6C15 16AA 09D5 0C96  92E5 FD89 16F9 4D90 BE3C


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Kumba]

Georgi Georgiev wrote:

> On 28/05/2003 at 04:47:01(-0400), Kumba used 0.7Kbytes just to say:
> 
>>	Not all users are going to immediately know the different categories 
>>offhand.  I'd have to opt for the ability that if portage detected a 
>>naming conflict, it'd show both matches + categories so the user would 
>>know which one he/she wanted.
> 
> 
> And abort execution, right?

	Yes, abort execution so the user could modify their parameter.


> 
> 
>>	OTOH, calling it net-www/mozillafirebird isn't a bad idea.  I don't 
>>think we need caps in there, caps seems to be reserved for dev-perl 
>>stuff only.
> 
> 
> Since the caps are used for splitting the two words, an extra dash would be
> needed instead of the caps. Concatenating the two makes it hard to read and personally, it would be hard for me to type.
> 
> It would have been nice if the official distribution had a name gentoo could
> adopt, but looking at ftp://ftp.mozilla.org/pub/firebird/nightly/latest-trunk/
> one can see:
> 
> Firebird-mac.dmg.gz
> MozillaFirebird-i686-pc-linux-gnu.tar.gz
> MozillaFirebird-win32.zip
> 
> So we're back where we started. They have both "firebird" and "mozilla
> firebird".


	I wish they'd just kept it "phoenix", but I heard the BIOS company 
wanted to sue (must be desperate for cash like SCO).  I also wonder if 
the Mozilla people did do their homework before choosing "firebird" so 
they atleast knew of another opensource project out there with the same 
name.

	Eitherway, yeah, net-www/mozilla-firebird sounds pretty good.  bit 
lengthy, but it's not the only one.  I've a feeling we'ed need a 
net-mail/mozilla-thunderbird in the future as well, just to keep 
consistency.

--Kumba




--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Michael Kohl]

[-- Attachment #1: Type: text/plain, Size: 1033 bytes --]

On Wed, 28 May 2003 05:20:44 -0400
Kumba <kumba@gentoo.org> wrote:

> Eitherway, yeah, net-www/mozilla-firebird sounds pretty good.  bit 
> lengthy, but it's not the only one.  I've a feeling we'ed need a 
> net-mail/mozilla-thunderbird in the future as well, just to keep 
> consistency.

"4.  Use the names "Mozilla Browser" and "Mozilla Mail" to describe the
Firebird and Thunderbird projects after the 1.4 release." [1]

How about using this names (mozilla-browser, mozilla-mail)? This would
also spare us the trouble of renaming everything again after 1.4 is
released.

And to keep consistent with the above, the current Mozilla ebuild could
be renamed to mozilla-application-suite (or mozilla-suite):
"1.  When referring to a SeaMonkey-based release use the phrase "Mozilla
Application Suite" as the name of the app suite." [1]

Just my 2 cents,
Michael

[1] http://www.mozilla.org/roadmap/branding.html)+


-- 
www.cargal.org 
GnuPG-key-ID: 0x90CA09E3
Jabber-ID: citizen428 [at] cargal [dot] org
Registered Linux User #278726

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Jon Portnoy]

On Wed, May 28, 2003 at 06:06:37PM +0900, Georgi Georgiev wrote:
> On 28/05/2003 at 04:47:01(-0400), Kumba used 0.7Kbytes just to say:
> > 	Not all users are going to immediately know the different categories 
> > offhand.  I'd have to opt for the ability that if portage detected a 
> > naming conflict, it'd show both matches + categories so the user would 
> > know which one he/she wanted.
> 
> And abort execution, right?

Policy is to avoid naming conflicts by using different names whenever 
possible.

> 
> > 	OTOH, calling it net-www/mozillafirebird isn't a bad idea.  I don't 
> > think we need caps in there, caps seems to be reserved for dev-perl 
> > stuff only.
> 
> Since the caps are used for splitting the two words, an extra dash would be
> needed instead of the caps. Concatenating the two makes it hard to read and personally, it would be hard for me to type.

Caps are against policy.

-- 
Jon Portnoy
avenj/irc.freenode.net

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[George Shapovalov]

On Wednesday 28 May 2003 07:06, Jon Portnoy wrote:
> Policy is to avoid naming conflicts by using different names whenever
> possible.
Um, I don't think we have any policy on this matter. I was bringing this issue 
about a month ago and found that some devs believe we cannot have duplicate 
names (or avoid them whereever possible) others think oterwise...
There was a thread initiated by not very happy user whose ebuild was renamed 
with the explanation that portage cannot reliably deal with this situation, 
however later on he found two identically named packages.

The thread was on -dev and I think general consesus was that it would be nice 
to allow duplicate names, we only need to make portage stop and ask for 
clarification if duplicates are found during emerge (emerge search is fine 
already). I think there is even a bug about this..

I remember doing a search for duplicates and finding there were multiple. 
Therefore it is not just a simple issue of enforcinf different names on all 
new packages..  Unfortunately that conversation did not finish up at that 
time, however this might have made it into udder project.. Nonetheless IMHO 
it would be nice to settle on certain policy so that we can at least avoid 
increasing inconsistency..

George



--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Jon Portnoy]

On Wed, May 28, 2003 at 11:13:58AM -0700, George Shapovalov wrote:
> On Wednesday 28 May 2003 07:06, Jon Portnoy wrote:
> > Policy is to avoid naming conflicts by using different names whenever
> > possible.
> Um, I don't think we have any policy on this matter. I was bringing this issue 
> about a month ago and found that some devs believe we cannot have duplicate 
> names (or avoid them whereever possible) others think oterwise...
> There was a thread initiated by not very happy user whose ebuild was renamed 
> with the explanation that portage cannot reliably deal with this situation, 
> however later on he found two identically named packages.
> 
> The thread was on -dev and I think general consesus was that it would be nice 
> to allow duplicate names, we only need to make portage stop and ask for 
> clarification if duplicates are found during emerge (emerge search is fine 
> already). I think there is even a bug about this..
> 
> I remember doing a search for duplicates and finding there were multiple. 
> Therefore it is not just a simple issue of enforcinf different names on all 
> new packages..  Unfortunately that conversation did not finish up at that 
> time, however this might have made it into udder project.. Nonetheless IMHO 
> it would be nice to settle on certain policy so that we can at least avoid 
> increasing inconsistency..
> 

Policy is to avoid naming conflicts. :)

There are indeed ebuilds that conflict. Those are bugs that there is 
currently no clean fix for. Committing _more_ conflicting ebuilds is 
definitely not okay.


-- 
Jon Portnoy
avenj/irc.freenode.net

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Jon Portnoy]

On Wed, May 28, 2003 at 11:13:58AM -0700, George Shapovalov wrote:
> On Wednesday 28 May 2003 07:06, Jon Portnoy wrote:
> > Policy is to avoid naming conflicts by using different names whenever
> > possible.
> Um, I don't think we have any policy on this matter. I was bringing this issue 
> about a month ago and found that some devs believe we cannot have duplicate 
> names (or avoid them whereever possible) others think oterwise...
> There was a thread initiated by not very happy user whose ebuild was renamed 
> with the explanation that portage cannot reliably deal with this situation, 
> however later on he found two identically named packages.


To clarify myself, Portage does need functionality to deal with 
duplicates.

In the meantime, we shouldn't ever commit ebuilds with conflicting 
names. That's been stated policy for some time, but the policy guide 
hasn't been updated to reflect it, unfortunately.


-- 
Jon Portnoy
avenj/irc.freenode.net

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Matthew Kennedy]

Jon Portnoy <avenj@gentoo.org> writes:

> Policy is to avoid naming conflicts. :)
>
> There are indeed ebuilds that conflict. Those are bugs that there is 
> currently no clean fix for. Committing _more_ conflicting ebuilds is 
> definitely not okay.
>

Jon (and others), 

I'm not so sure.  I have app-emacs and app-xemacs and there are
*many* ebuilds which "conflict" in their naming, and this is
intentional.

For example, the "ilisp" package appears in both app-emacs and
app-xemacs.  So if I'm to avoid this naming "conflict" because
somemone typing "emerge ilisp" might get "app-xemacs/ilisp" instead
of "app-emacs/ilisp", then naturally I should prefix the ebuild names
with "emacs" and "xemacs", right?  So I'd end up with

   emerge emacs-ilisp
or
   emerge xemacs-ilisp

But wait!  Thats what categories are for -- ie. defining some
namespace. Thus there's no need, and this more or less backs up my
argument that the category name + package name forms a "fully resolved
and unique" symbol for emerge to act upon.

So the user should do:

   emerge app-emacs/ilisp
or
   emerge app-xemacs/ilisp

What of the case where the user is surprised by the "short form" case:

   emerge ilisp

...well this is where emerge should abort and report back to the user
something like this: "ERROR: Short form 'ilisp' matches
app-emacs/ilisp and app-xemacs/ilisp -- please use a fully qualified
name instead".  (Currently emerge emerges the first one it sees.)

I really think ambiguities should be resolved by using the category
name -- otherwise I can't see why we sort packages into categories
at all.

Matt   

-- 
Matthew Kennedy
Gentoo Linux Developer

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Dave Nellans]

[-- Attachment #1: Type: text/plain, Size: 2059 bytes --]

this is something we discussed about a month ago and never developed an
official policy for then did we? george can you comment on this? 
perhaps its time we do

dave

On Wed, 2003-05-28 at 15:04, Matthew Kennedy wrote:
> Jon Portnoy <avenj@gentoo.org> writes:
> 
> > Policy is to avoid naming conflicts. :)
> >
> > There are indeed ebuilds that conflict. Those are bugs that there is 
> > currently no clean fix for. Committing _more_ conflicting ebuilds is 
> > definitely not okay.
> >
> 
> Jon (and others), 
> 
> I'm not so sure.  I have app-emacs and app-xemacs and there are
> *many* ebuilds which "conflict" in their naming, and this is
> intentional.
> 
> For example, the "ilisp" package appears in both app-emacs and
> app-xemacs.  So if I'm to avoid this naming "conflict" because
> somemone typing "emerge ilisp" might get "app-xemacs/ilisp" instead
> of "app-emacs/ilisp", then naturally I should prefix the ebuild names
> with "emacs" and "xemacs", right?  So I'd end up with
> 
>    emerge emacs-ilisp
> or
>    emerge xemacs-ilisp
> 
> But wait!  Thats what categories are for -- ie. defining some
> namespace. Thus there's no need, and this more or less backs up my
> argument that the category name + package name forms a "fully resolved
> and unique" symbol for emerge to act upon.
> 
> So the user should do:
> 
>    emerge app-emacs/ilisp
> or
>    emerge app-xemacs/ilisp
> 
> What of the case where the user is surprised by the "short form" case:
> 
>    emerge ilisp
> 
> ...well this is where emerge should abort and report back to the user
> something like this: "ERROR: Short form 'ilisp' matches
> app-emacs/ilisp and app-xemacs/ilisp -- please use a fully qualified
> name instead".  (Currently emerge emerges the first one it sees.)
> 
> I really think ambiguities should be resolved by using the category
> name -- otherwise I can't see why we sort packages into categories
> at all.
> 
> Matt   
-- 
Dave Nellans
http://lucy.wox.org/~dnellans/
dnellans@cs.utah.edu

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[George Shapovalov]

On Wednesday 28 May 2003 14:14, Dave Nellans wrote:
> this is something we discussed about a month ago and never developed an
> official policy for then did we? george can you comment on this?
> perhaps its time we do
Yep, that's a nice wrap up of my previous comment :). We did discuss it but it 
never got settled, probably because there was much talk about starting udder 
system, so this particular one got lost..

BTW just did some digging: the bug in question is #19402. The text 
unfortunately is not very clear :(, but I managed to find my way around gmane 
finally :), here is the thread:
http://news.gmane.org/onethread.php?group=gmane.linux.gentoo.devel&root=%3C1050450162.18807.6.camel%40malfus%3E

Also, just got a direct message from a user, but I am sure he meant to send it 
to the list. I'll ask him to forward his responce to a proper place, William, 
if you read this this is your hint ;).
Short wrap-up (to avoid unnecessary spamming): his responce goes in line with 
what mkkennedy just said and many others mentioned in the cited thread.

George



--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[William F Pearson III]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

<snip>
| The thread was on -dev and I think general consesus was that it would
be nice
| to allow duplicate names, we only need to make portage stop and ask for
| clarification if duplicates are found during emerge (emerge search is
fine
| already). I think there is even a bug about this..
<snip>
| Nonetheless IMHO
| it would be nice to settle on certain policy so that we can at least
avoid
| increasing inconsistency..
|
| George
|

Oops, Sorry George, this lists operates differently than my LUG's
mailing list, and I'm sure that's not the last time I'll make that
mistake. :) Here's what I had to say:

I'm an end user, but I subscribe to gentoo-dev to stay current on the
direction of Gentoo, and give my input when I feel it could be helpful.
I think the package names should be intuitive, so if i know a program is
named "foo" i can "emerge foo" and install that package. Of course there
is always the possiblity that there will be "app-editors/foo" and
"net-www"/foo, but as an end user I might not be aware of one or the
other of these packages. The best recourse, from my perspective, would a
message from portage that looks like the same output from "emerge -s
foo" so if I typed:
#emerge foo
I would get:
Portage has returned more than one app named foo, which one would you
like to install?[1,2]
1    app-editors/foo
~        version: 0.0.1
~        homepage: foo.com
~        Description: foo editor
2    net-www/foo
~        version 0.1.0
~        homepage: foo.org
~        Description: a FooBar implementation
:
Of course this is easy for me to say because I doubt I'll be the one
coding it, and this feature would probably slow down portage a little
for any emerge, because it would have to check for duplicates first for
all emerges. I'll shut up now, but I couldn't resist giving my 2 cents.
- --
- -
William F. Pearson, III
william@techwill.net
www.techwill.net
gpg key http://www.techwill.net/files/wfp3.gpg
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQE+1UyVpfPDiNqvYvURAtkZAJ471dKvZ2d6JzN0GXB8Mzsi58tw9QCeO3lR
Aku1gI0obB9ag5lrcT9xRbA=
=pDZq
-----END PGP SIGNATURE-----


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Aron Griffis]

Matthew Kennedy wrote:[Wed May 28 2003, 05:04:34PM EDT]
> But wait!  Thats what categories are for -- ie. defining some
> namespace. Thus there's no need, and this more or less backs up my
> argument that the category name + package name forms a "fully resolved
> and unique" symbol for emerge to act upon.

Actually, that's a misconception of the purpose of categories in
portage.  It's never been to provide namespaces; that is a side-effect.
Categories make it easier for users to find what packages are available
for a given type of task.  That's their purpose. 

There was a discussion a while ago (um, a couple years ago) about
removing categories completely and having a flat list of packages.  The
argument against this was that the categories provide an easy way for
users (and developers) to see what is available in a given area.  It was
not intended for categories to provide namespaces.

Aron

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Aron Griffis]

Jon Portnoy wrote:[Wed May 28 2003, 03:21:30PM EDT]
> In the meantime, we shouldn't ever commit ebuilds with conflicting 
> names. That's been stated policy for some time, but the policy guide 
> hasn't been updated to reflect it, unfortunately.

This is my understanding also.  The app-xemacs category unfortunately
deviates from this policy.

Aron

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[John Robinson]

> There was a discussion a while ago (um, a couple years ago) about
> removing categories completely and having a flat list of packages.  The
> argument against this was that the categories provide an easy way for
> users (and developers) to see what is available in a given area.  It was
> not intended for categories to provide namespaces.

I wasn't around a few years ago, but I've watched this conversation go
back and forth from the previous one to the current. The following seems
(to me) to be a good summary of the two leading points and
counterpoints:


Q: Shouldn't we list the two packages for an interactive choice when two
or more are available?

A: Interactive choice is against the design of emerge. We need to be
able to use it more easily in scripts than that would allow. The best we
can do is to generate an error with that information and let the user
pick and re-run emerge.


Q: Why don't we let categories define namespace?

A: Because we've never used categories to define namespace.


Is there a reason why emerge couldn't have a command line switch to
cause it to deviate from its current behavior and give the user an
interactive choice? I gather that the error-choice option is currently
in the bugs db waiting for implementation, but a lot of users have
seemed to like the idea of the interactive option, which I think a
command line switch could integrate well. Is there a reason why this
should be avoided?

Also, although it isn't currently being done this way, is there a reason
why (given the above changes) the categories shouldn't be used to define
namespace? Unless I'm wrong, that would only involve a policy update, if
a way of avoiding merging the wrong package in the case of a collision
were in place. It seems like the intuitive way to go; most people seem
to assume it works that way until they're told otherwise. Additionally,
a lot of the discussion on this list of late (namely the Firebird issue)
would've been avoided if that system were in place, and I imagine the
same will be true of the future. Unless there's something I'm unaware
of, this system would allow emerge to continue to function just as it
currently does (besides the error on finding duplicate packages) while
relieving a lot of naming headaches.


- John Robinson

-- 

Love justice; desire mercy.


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Matt Tucker]

-- John Robinson <strider@aravir.net> spake thusly:

> Q: Shouldn't we list the two packages for an interactive choice when
> two or more are available?
> 
> A: Interactive choice is against the design of emerge. We need to be
> able to use it more easily in scripts than that would allow. The best
> we can do is to generate an error with that information and let the
> user pick and re-run emerge.

<snip>

> Is there a reason why emerge couldn't have a command line switch to
> cause it to deviate from its current behavior and give the user an
> interactive choice? I gather that the error-choice option is currently
> in the bugs db waiting for implementation, but a lot of users have
> seemed to like the idea of the interactive option, which I think a
> command line switch could integrate well. Is there a reason why this
> should be avoided?

Another option would be to allow emerge to be interactive when run from
a shell and to throw errors or make assumptions at other times. This
would be similar to the way ls shows colored output when run from a
shell, but not when piped to less or run from a script.

If a switch were used, I would suggest defaulting to interactive
behavior with a switch to disable it.


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[John Robinson]

> Another option would be to allow emerge to be interactive when run from
> a shell and to throw errors or make assumptions at other times. This
> would be similar to the way ls shows colored output when run from a
> shell, but not when piped to less or run from a script.

If I'm not mistaken, LS outputs color input (if it's told to) all the
time, and it is filtered by other programs. Maybe a shell programmer can
comment, 'cause I'm not much of one, but is there a way to ascertain the
context in which a script is run? I didn't think there was one.

> If a switch were used, I would suggest defaulting to interactive
> behavior with a switch to disable it.

While I agree in theory, in practice this would break any script that
assumes non-interactive behavior as the default. Given that, I think the
only option is the reverse.

- John Robinson

-- 

Love justice; desire mercy.


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Tavis Ormandy]

On Thu, May 29, 2003 at 09:31:01AM -0400, John Robinson wrote:
> 
> If I'm not mistaken, LS outputs color input (if it's told to) all the
> time, and it is filtered by other programs. Maybe a shell programmer can
> comment, 'cause I'm not much of one, but is there a way to ascertain the
> context in which a script is run? I didn't think there was one.
> 

you can detect a terminal on a file descriptor with test -t (help test
if you use bash)

ls uses --color=auto, to show colour only if stdout is a tty,
--color=always to show colour no matter what.

-- 
-------------------------------------
taviso@sdf.lonestar.org | finger me for my gpg key.
-------------------------------------------------------

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[John Robinson]

> you can detect a terminal on a file descriptor with test -t (help test
> if you use bash)
> 
> ls uses --color=auto, to show colour only if stdout is a tty,
> --color=always to show colour no matter what.

Cool. Good to know.

In that case, the switch to auto-detection seems like a good one to me,
but now I wonder why it hasn't already been implemented. I don't know
why emerge was designed with its current philosophies, or which ones
affect this part of its operation, so (with my changed opinion) I think
I'm gonna have to check out and let somebody who does know speculate on
why it would be a good or bad idea. =)

Cheers,

John Robinson

-- 

Love justice; desire mercy.


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Matthew Kennedy]

Aron Griffis <agriffis@gentoo.org> writes:

> Jon Portnoy wrote:[Wed May 28 2003, 03:21:30PM EDT]
>> In the meantime, we shouldn't ever commit ebuilds with conflicting 
>> names. That's been stated policy for some time, but the policy guide 
>> hasn't been updated to reflect it, unfortunately.
>
> This is my understanding also.  The app-xemacs category unfortunately
> deviates from this policy.

Suggestions welcome.

-- 
Matthew Kennedy
Gentoo Linux Developer

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Matthew Kennedy]

Aron Griffis <agriffis@gentoo.org> writes:

> Matthew Kennedy wrote:[Wed May 28 2003, 05:04:34PM EDT]
>> But wait!  Thats what categories are for -- ie. defining some
>> namespace. Thus there's no need, and this more or less backs up my
>> argument that the category name + package name forms a "fully resolved
>> and unique" symbol for emerge to act upon.
>
> Actually, that's a misconception of the purpose of categories in
> portage.  It's never been to provide namespaces; that is a side-effect.
> Categories make it easier for users to find what packages are available
> for a given type of task.  That's their purpose. 

Aron, 

This is true.  However I seriously doubt the use-case where a user
goes into /usr/portage/* and browses for packages that way.  I've
never done that for the time I've been a developer and also the time
before that when I was a user.  Of course this doesn't mean other
don't... I'd be surprised though.

I think it most people will more likely use the emerge search option
to find what they need.

That aside, using the category name to resolve ambiguous short-form
package names seems to work perfectly *today*, and perhaps we should
adopt its use officially in the documentation as it is a solution to
cases where we have ambiguous short-form package names.  

As far as I can tell,

   emerge app-editors/emacs
   emerge app-editors/emacs-21.3-r1

do the same thing as

   emerge emacs
   emerge emacs-21.3-r1


I think we've uncovered two angels to this thread: 1. how to resolve
an ambiguous short-form package names (I propose the above usage of
category names 'cause it works great already) and 2. what to do when
the user does try to emerge an ambiguously named short-form package
name.

In the second case you could have:

  1. emerge the first foo portage finds (undesirable, but what happens
     currently)

  2. prompt interactively (this would be the first time portage 
     prompts for user interaction afaik)

  3. fail with an error message indication a long-form etc. must be
     given to resolve one or more packages.

I don't really care either way between 2 and 3.

Matt

-- 
Matthew Kennedy
Gentoo Linux Developer

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Georgi Georgiev]

On 30/05/2003 at 02:03:52(-0500), Matthew Kennedy used 2.2Kbytes just to say:
> 2. what to do when the user does try to emerge an ambiguously named
> short-form package name.
> 
> In the second case you could have:
> 
>   1. emerge the first foo portage finds (undesirable, but what happens
>      currently)
> 
>   2. prompt interactively (this would be the first time portage 
>      prompts for user interaction afaik)
> 
>   3. fail with an error message indication a long-form etc. must be
>      given to resolve one or more packages.
> 
> I don't really care either way between 2 and 3.

Shouldn't we define what "ambiguous" is? After all let's not forget that the
worst case is:

emerge rsync vs. emerge net-misc/rsync

Currently, only option 1 works with "emerge rsync". Maybe a warning should be
given every time a user tries to "emerge rsync" so that they get used to the
alias sync.

-- 
 /^^^^^^^^^^^^^^^^^^^^^^^^^^^\/^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\
/ Georgi Georgiev    (-<     / Chance is perhaps the work of    \
\ chutz@chubaka.net  /\   .o)\ God when He did not want to      /
/ +81(90)6266-1163  V_/_ |(/)/ sign. -- Anatole France          \
\___________________________/\__________________________________/

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Christian Axelsson]

> In the second case you could have:
>
>   1. emerge the first foo portage finds (undesirable, but what happens
>      currently)
>
>   2. prompt interactively (this would be the first time portage
>      prompts for user interaction afaik)
>
>   3. fail with an error message indication a long-form etc. must be
>      given to resolve one or more packages.
>

I _really_ vote for #3 in this case. It's nothing wrong to give an error,
afterall, we have separate namespaces, why not take advantage of that?
I think that the 'emerge foo' instead of 'emerge category/foo' is just
pure lazyness and nothing we should hangon to. No, let's print a list with
the avaible options instead.

My €2

-- 
Christian Axelsson
smiler@lanil.mine.nu



--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Grant Goodyear]

[-- Attachment #1: Type: text/plain, Size: 631 bytes --]

On Fri, 2003-05-30 at 03:03, Matthew Kennedy wrote:
> This is true.  However I seriously doubt the use-case where a user
> goes into /usr/portage/* and browses for packages that way.  I've
> never done that for the time I've been a developer and also the time
> before that when I was a user.  Of course this doesn't mean other
> don't... I'd be surprised though.

For what it's worth, I frequently use the categories as a way to search
for things.  When I decided that I wanted to look at cd recording guis,
the first thing I did was see what we have in app-cdr.

-g2boojum-
-- 
Grant Goodyear <g2boojum@gentoo.org>

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Mozilla: Firebird-bin & Firebird-cvs

[Matthew Kennedy]

Matthew Kennedy <mkennedy@gentoo.org> writes:

> Jon Portnoy <avenj@gentoo.org> writes:

http://cvs.gentoo.org/~mkennedy/emerge-2.0.48-r1-shortname.patch



-- 
Matthew Kennedy
Gentoo Linux Developer

--
gentoo-dev@gentoo.org mailing list