Hi
I've tried to sum up the problems with the wine-users mailing list that are urging many people to call for a forum.winehq.org.
I hope it's useful.
Problems ========
1) Cannot post without configured mail client 2) Browsing old topics and replying/posting new ones happen in two different places (web archive vs. mail client) 3) Current archive web interface unwieldy (many clicks needed to browse, always sorted by month) 4) Missing Googlish search function 5) "Subscribe" wording and web pages suggest your inbox will fill up when doing so 6) Subscribing will cause your inbox to fill up; it's not obvious how to change it; neither how to receive mail on specific topics you DO want once you've changed it 7) Missing "forward all postings re this topic to my inbox" feature 8) Very difficult to post under pseudonym 9) Missing 'sticky' feature (?)
Solution 1: A forum, fx. phpBB ===============================
Pros: ) Easy to set up. We might even be able to sneak it in before AJ returns :-). (Sorry big guy, couldn't resist pulling one on your expense =)..) ) Might attract away from wine-users a high quantity of dumb questions that people ask when they can't be bothered to search the archives. ) Adds a structural approach by categorizing users' various problems, which encourages searching instead of asking.
Cons: ) Dillutes knowledge; some ends up in forum and some on wine-users, unless we nuke wine-users. ) Need to hack the code to get visual integration with WineHQ (menus etc). ) To get single-sign-on, we need either a.) reverse proxy or b.) something LDAPish and some amount of phpBB hacking
Solution 2: Point people at Gmane or Google Groups in a prominent place ========================================================================
Pros: ) We don't have to do anything :-).
Cons: ) If it's not directly on winehq.org, it's probably not official enough that people are going to use it. ) Not everyone has a Gmail account (required to post through Groups) ) Not everyone is adept to using Gmane
Solution 3: WebForum-on-top-of-List ====================================
Pros: ) We concentrate people, allowing web and email users to communicate. ) We concentrate (archived) knowledge in one place.
Cons: ) The reason that forums are often filled with spam might be that it's too easy to post to them. We'd be duplicating that. ) We need to code it all ourselves!
I think solution 1 is good, but I prefer the extra icing on the cake that comes with solution 3.
Now, I'm not high on crystal meth or anything, but I'll just assume for a moment that we all want better accessibility for the newcomers (even with the rise in volume that they bring), and that all agree to go with solution 3 :-). Thus I'll try to outline a battle plan:
Battle-plan for solution 3 ==========================
We'd like non-Wine developers to be able to join in; so let's make sure that this system is generically usable. That means: * Standard user authentication, fx. using LDAP. * Componentize a bit, fx. keep the web interface itself separate from WineHQ menu structure and logo.
We'd need to: * Prioritize features * Setup a mailing list * Setup a simple project web page * Setup a winehq-like staging area for development (follows HEAD/TIP/origin of web repository so everyone can see what the others are doing) * Find developers/volunteers! * Get started on the coding
When prioritizing features, we should take a look around and see if there are software components available that could be reused, fx.: * PHP LDAP auth bindings * Existing, good web interfaces for the wine archives (mailman?) * ?...
Any holes or oversights in that plan?
If solution 3 ever takes off, I will gladly contribute with coding. Realistically, though, i cannot implement something like this all by myself, especially not in any kind of a timely fashion.
Hello, I have noted below problems that we can take care of, and those that aren't our problem, never were, never will be, impossible to be our problem, (you get the picture), because they are PEBKAC errors :)
For all of those that don't know: PEBKAC = Problem Exists Between Keyboard and Chair.
Molle Bestefich wrote:
Hi
I've tried to sum up the problems with the wine-users mailing list that are urging many people to call for a forum.winehq.org.
I hope it's useful.
Problems
- Cannot post without configured mail client
An unconfigued client causes hell for EVERYONE they email, not just the list members. If that's a problem for people, well, paper and pens still exist. this is PEBKAC.
- Browsing old topics and replying/posting new ones happen in two
different places (web archive vs. mail client)
There should be a "server" email address (like majordomo) that you san send email to for requesting old archived posts. It should be flexible. This would allow for this problem to be fixed, and the archives to be access via email (for us that remember email-ftp)
- Current archive web interface unwieldy (many clicks needed to
browse, always sorted by month)
This should be sorted in a 2-pass manner, Each level of posting (the original post is level one, immedate replies are level 2, and all the immediate replies to the first replies would be the seveal level 3s) would be sorted chronologically, and then by level. That's pretty much how I see mailing lists do that. If you don't understand, email me and i'll try to explain better.
Missing Googlish search function
"Subscribe" wording and web pages suggest your inbox will fill up
when doing so
Well, that's not a real problem because it is truthful for some poor people that don't use Yahoo!/GMail/Hotmail/Excite/etc. (those services offer large email boxes)
- Subscribing will cause your inbox to fill up; it's not obvious how
to change it; neither how to receive mail on specific topics you DO want once you've changed it
Solution: Teach users how to properly use their mailbox, or simply get a bigger mailbox. If it helps, I'll give people GMail invites just to counteract this problem (and having to use a stupid web-based email client as well :) this one is partially our problem, partially PEBKAC, partially the email providers problem.
- Missing "forward all postings re this topic to my inbox" feature
- Very difficult to post under pseudonym
Not true, just don't include your real name anywhere in your email setup. I don't. Purely PEBKAC.
- Missing 'sticky' feature (?)
That's not our problem. it's impossible for that to be our problem, because on a mailing list, sticky stuff is client-side only, no matter what your client is, which also implies that it's impossible for us to do jack about it. Purely PEBKAC.
Solution 1: A forum, fx. phpBB
Pros: ) Easy to set up. We might even be able to sneak it in before AJ returns :-). (Sorry big guy, couldn't resist pulling one on your expense =)..) ) Might attract away from wine-users a high quantity of dumb questions that people ask when they can't be bothered to search the archives. ) Adds a structural approach by categorizing users' various problems, which encourages searching instead of asking.
Cons: ) Dillutes knowledge; some ends up in forum and some on wine-users, unless we nuke wine-users. ) Need to hack the code to get visual integration with WineHQ (menus etc). ) To get single-sign-on, we need either a.) reverse proxy or b.) something LDAPish and some amount of phpBB hacking
Solution 2: Point people at Gmane or Google Groups in a prominent place
Pros: ) We don't have to do anything :-).
Cons: ) If it's not directly on winehq.org, it's probably not official enough that people are going to use it. ) Not everyone has a Gmail account (required to post through Groups) ) Not everyone is adept to using Gmane
Solution 3: WebForum-on-top-of-List
Pros: ) We concentrate people, allowing web and email users to communicate. ) We concentrate (archived) knowledge in one place.
Cons: ) The reason that forums are often filled with spam might be that it's too easy to post to them. We'd be duplicating that. ) We need to code it all ourselves!
I think solution 1 is good, but I prefer the extra icing on the cake that comes with solution 3.
Now, I'm not high on crystal meth or anything, but I'll just assume for a moment that we all want better accessibility for the newcomers (even with the rise in volume that they bring), and that all agree to go with solution 3 :-). Thus I'll try to outline a battle plan:
Battle-plan for solution 3
We'd like non-Wine developers to be able to join in; so let's make sure that this system is generically usable. That means:
- Standard user authentication, fx. using LDAP.
- Componentize a bit, fx. keep the web interface itself separate from
WineHQ menu structure and logo.
We'd need to:
- Prioritize features
- Setup a mailing list
- Setup a simple project web page
- Setup a winehq-like staging area for development (follows
HEAD/TIP/origin of web repository so everyone can see what the others are doing)
- Find developers/volunteers!
- Get started on the coding
When prioritizing features, we should take a look around and see if there are software components available that could be reused, fx.:
- PHP LDAP auth bindings
- Existing, good web interfaces for the wine archives (mailman?)
- ?...
Any holes or oversights in that plan?
If solution 3 ever takes off, I will gladly contribute with coding. Realistically, though, i cannot implement something like this all by myself, especially not in any kind of a timely fashion.
On 4/22/06, Segin segin2005@gmail.com wrote:
Hello, I have noted below problems that we can take care of, and those that aren't our problem, never were, never will be, impossible to be our problem, (you get the picture), because they are PEBKAC errors :)
For all of those that don't know: PEBKAC = Problem Exists Between Keyboard and Chair.
I find this a little crude. You're acknowledging the fact that they are problems, but dismissing them and not really offering a solution. It seems to me that a lot of the problems with the mailing list is that the interface is complex, especially for joe user. The least we can do is to make it easy for them, and a forum seems to solve many of the complexity issues. Take for example 1) Cannot post without configured mail client. If there was no need for a mail client, there would be no problem.
Molle Bestefich wrote:
Hi
I've tried to sum up the problems with the wine-users mailing list that are urging many people to call for a forum.winehq.org.
I hope it's useful.
Problems
- Cannot post without configured mail client
An unconfigued client causes hell for EVERYONE they email, not just the list members. If that's a problem for people, well, paper and pens still exist. this is PEBKAC.
- Browsing old topics and replying/posting new ones happen in two
different places (web archive vs. mail client)
There should be a "server" email address (like majordomo) that you san send email to for requesting old archived posts. It should be flexible. This would allow for this problem to be fixed, and the archives to be access via email (for us that remember email-ftp)
- Current archive web interface unwieldy (many clicks needed to
browse, always sorted by month)
This should be sorted in a 2-pass manner, Each level of posting (the original post is level one, immedate replies are level 2, and all the immediate replies to the first replies would be the seveal level 3s) would be sorted chronologically, and then by level. That's pretty much how I see mailing lists do that. If you don't understand, email me and i'll try to explain better.
Missing Googlish search function
"Subscribe" wording and web pages suggest your inbox will fill up
when doing so
Well, that's not a real problem because it is truthful for some poor people that don't use Yahoo!/GMail/Hotmail/Excite/etc. (those services offer large email boxes)
- Subscribing will cause your inbox to fill up; it's not obvious how
to change it; neither how to receive mail on specific topics you DO want once you've changed it
Solution: Teach users how to properly use their mailbox, or simply get a bigger mailbox. If it helps, I'll give people GMail invites just to counteract this problem (and having to use a stupid web-based email client as well :) this one is partially our problem, partially PEBKAC, partially the email providers problem.
- Missing "forward all postings re this topic to my inbox" feature
- Very difficult to post under pseudonym
Not true, just don't include your real name anywhere in your email setup. I don't. Purely PEBKAC.
- Missing 'sticky' feature (?)
That's not our problem. it's impossible for that to be our problem, because on a mailing list, sticky stuff is client-side only, no matter what your client is, which also implies that it's impossible for us to do jack about it. Purely PEBKAC.
Solution 1: A forum, fx. phpBB
Pros: ) Easy to set up. We might even be able to sneak it in before AJ returns :-). (Sorry big guy, couldn't resist pulling one on your expense =)..) ) Might attract away from wine-users a high quantity of dumb questions that people ask when they can't be bothered to search the archives. ) Adds a structural approach by categorizing users' various problems, which encourages searching instead of asking.
Cons: ) Dillutes knowledge; some ends up in forum and some on wine-users, unless we nuke wine-users. ) Need to hack the code to get visual integration with WineHQ (menus
etc).
) To get single-sign-on, we need either a.) reverse proxy or b.) something LDAPish and some amount of phpBB hacking
Solution 2: Point people at Gmane or Google Groups in a prominent place
Pros: ) We don't have to do anything :-).
Cons: ) If it's not directly on winehq.org, it's probably not official enough that people are going to use it. ) Not everyone has a Gmail account (required to post through Groups) ) Not everyone is adept to using Gmane
Solution 3: WebForum-on-top-of-List
Pros: ) We concentrate people, allowing web and email users to communicate. ) We concentrate (archived) knowledge in one place.
Cons: ) The reason that forums are often filled with spam might be that it's too easy to post to them. We'd be duplicating that. ) We need to code it all ourselves!
I think solution 1 is good, but I prefer the extra icing on the cake that comes with solution 3.
Now, I'm not high on crystal meth or anything, but I'll just assume for a moment that we all want better accessibility for the newcomers (even with the rise in volume that they bring), and that all agree to go with solution 3 :-). Thus I'll try to outline a battle plan:
Battle-plan for solution 3
We'd like non-Wine developers to be able to join in; so let's make sure that this system is generically usable. That means:
- Standard user authentication, fx. using LDAP.
- Componentize a bit, fx. keep the web interface itself separate from
WineHQ menu structure and logo.
We'd need to:
- Prioritize features
- Setup a mailing list
- Setup a simple project web page
- Setup a winehq-like staging area for development (follows
HEAD/TIP/origin of web repository so everyone can see what the others are doing)
- Find developers/volunteers!
- Get started on the coding
When prioritizing features, we should take a look around and see if there are software components available that could be reused, fx.:
- PHP LDAP auth bindings
- Existing, good web interfaces for the wine archives (mailman?)
- ?...
Any holes or oversights in that plan?
If solution 3 ever takes off, I will gladly contribute with coding. Realistically, though, i cannot implement something like this all by myself, especially not in any kind of a timely fashion.
James D. Trotter
Why aren't I offering suggestions? I did offer one. The rest are beyond our control. To attempt to fix those would be as intelligent as sticking your hand into a boiling pot of water.
Let's look at that last one, the sticky thing. This usually means to have a email highlighted in a way that it gets the attention of the user. Now, think about it. There is nothing that we can do to address this ourselves. If you say we can, you are an idiot. That is something that must be configured in whatever email client the user is using.
Now let's look at the email client config thing... Well, we could try to fix headers of incoming email, but that's not really a solution. In fact, there is no real solution that we can provide ourselves, and besides, a properly configued email client (to produce proper headers and such) is just proper netiuqitte. There is, again, nothing we can do legally about this (well, we can find a way to break into the users system and configure his email client for him, but were not here to hold hands or commit crimes)
Well, it's pretty obvious you didn't read my comments, think about them, read the problems i commented against, and thought about those, and finally thought all of it together. You are just out to troll, and your reply is just flamebait. Your time would probably best be spent on USENET, as there are plenty of trolls there.
And before anyone says I am trolling, you're almost right, except that I explained why the problems i mention as unfixable by us are just that. Can we configure the users email client for him? No. Therefore we can't fix those problems.
James Trotter wrote:
On 4/22/06, *Segin* <segin2005@gmail.com mailto:segin2005@gmail.com> wrote:
Hello, I have noted below problems that we can take care of, and those that aren't our problem, never were, never will be, impossible to be our problem, (you get the picture), because they are PEBKAC errors :) For all of those that don't know: PEBKAC = Problem Exists Between Keyboard and Chair.I find this a little crude. You're acknowledging the fact that they are problems, but dismissing them and not really offering a solution. It seems to me that a lot of the problems with the mailing list is that the interface is complex, especially for joe user. The least we can do is to make it easy for them, and a forum seems to solve many of the complexity issues. Take for example 1) Cannot post without configured mail client. If there was no need for a mail client, there would be no problem.
Molle Bestefich wrote: >Hi > >I've tried to sum up the problems with the wine-users mailing list >that are urging many people to call for a forum.winehq.org <http://forum.winehq.org> . > >I hope it's useful. > > >Problems >======== > > 1) Cannot post without configured mail client > > An unconfigued client causes hell for EVERYONE they email, not just the list members. If that's a problem for people, well, paper and pens still exist. this is PEBKAC. > 2) Browsing old topics and replying/posting new ones happen in two >different places (web archive vs. mail client) > > There should be a "server" email address (like majordomo) that you san send email to for requesting old archived posts. It should be flexible. This would allow for this problem to be fixed, and the archives to be access via email (for us that remember email-ftp) > 3) Current archive web interface unwieldy (many clicks needed to >browse, always sorted by month) > This should be sorted in a 2-pass manner, Each level of posting (the original post is level one, immedate replies are level 2, and all the immediate replies to the first replies would be the seveal level 3s) would be sorted chronologically, and then by level. That's pretty much how I see mailing lists do that. If you don't understand, email me and i'll try to explain better. > 4) Missing Googlish search function > > > 5) "Subscribe" wording and web pages suggest your inbox will fill up >when doing so > > Well, that's not a real problem because it is truthful for some poor people that don't use Yahoo!/GMail/Hotmail/Excite/etc. (those services offer large email boxes) > 6) Subscribing will cause your inbox to fill up; it's not obvious how >to change it; neither how to receive mail on specific topics you DO >want once you've changed it > > Solution: Teach users how to properly use their mailbox, or simply get a bigger mailbox. If it helps, I'll give people GMail invites just to counteract this problem (and having to use a stupid web-based email client as well :) this one is partially our problem, partially PEBKAC, partially the email providers problem. > 7) Missing "forward all postings re this topic to my inbox" feature > 8) Very difficult to post under pseudonym > > Not true, just don't include your real name anywhere in your email setup. I don't. Purely PEBKAC. > 9) Missing 'sticky' feature (?) > > That's not our problem. it's impossible for that to be our problem, because on a mailing list, sticky stuff is client-side only, no matter what your client is, which also implies that it's impossible for us to do jack about it. Purely PEBKAC. > >Solution 1: A forum, fx. phpBB >=============================== > > Pros: > ) Easy to set up. > We might even be able to sneak it in before AJ returns :-). > (Sorry big guy, couldn't resist pulling one on your expense =)..) > ) Might attract away from wine-users a high quantity of dumb >questions that people ask when they can't be bothered to search the >archives. > ) Adds a structural approach by categorizing users' various >problems, which encourages searching instead of asking. > > Cons: > ) Dillutes knowledge; some ends up in forum and some on wine-users, >unless we nuke wine-users. > ) Need to hack the code to get visual integration with WineHQ (menus etc). > ) To get single-sign-on, we need either a.) reverse proxy or b.) >something LDAPish and some amount of phpBB hacking > > >Solution 2: Point people at Gmane or Google Groups in a prominent place >======================================================================== > > Pros: > ) We don't have to do anything :-). > > Cons: > ) If it's not directly on winehq.org <http://winehq.org>, it's probably not official >enough that people are going to use it. > ) Not everyone has a Gmail account (required to post through Groups) > ) Not everyone is adept to using Gmane > > >Solution 3: WebForum-on-top-of-List >==================================== > > Pros: > ) We concentrate people, allowing web and email users to communicate. > ) We concentrate (archived) knowledge in one place. > > Cons: > ) The reason that forums are often filled with spam might be that >it's too easy to post to them. We'd be duplicating that. > ) We need to code it all ourselves! > > >I think solution 1 is good, but I prefer the extra icing on the cake >that comes with solution 3. > > > >Now, I'm not high on crystal meth or anything, but I'll just assume >for a moment that we all want better accessibility for the newcomers >(even with the rise in volume that they bring), and that all agree to >go with solution 3 :-). Thus I'll try to outline a battle plan: > >Battle-plan for solution 3 >========================== > >We'd like non-Wine developers to be able to join in; so let's make >sure that this system is generically usable. >That means: > * Standard user authentication, fx. using LDAP. > * Componentize a bit, fx. keep the web interface itself separate from >WineHQ menu structure and logo. > >We'd need to: > * Prioritize features > * Setup a mailing list > * Setup a simple project web page > * Setup a winehq-like staging area for development (follows >HEAD/TIP/origin of web repository so everyone can see what the others >are doing) > * Find developers/volunteers! > * Get started on the coding > >When prioritizing features, we should take a look around and see if >there are software components available that could be reused, fx.: > * PHP LDAP auth bindings > * Existing, good web interfaces for the wine archives (mailman?) > * ?... > >Any holes or oversights in that plan? > >If solution 3 ever takes off, I will gladly contribute with coding. >Realistically, though, i cannot implement something like this all by >myself, especially not in any kind of a timely fashion. > > > > >James D. Trotter
On 4/22/06, Segin < segin2005@gmail.com> wrote:
Why aren't I offering suggestions? I did offer one. The rest are beyond our control. To attempt to fix those would be as intelligent as sticking your hand into a boiling pot of water.
Let's look at that last one, the sticky thing. This usually means to have a
email highlighted in a way that it gets the attention of the user. Now, think about it. There is nothing that we can do to address this ourselves. If you say we can, you are an idiot. That is something that must be configured in whatever email client the user is using.
Now let's look at the email client config thing... Well, we could try to fix headers of incoming email, but that's not really a solution. In fact, there is no real solution that we can provide ourselves, and besides, a properly configued email client (to produce proper headers and such) is just proper netiuqitte. There is, again, nothing we can do legally about this (well, we can find a way to break into the users system and configure his email client for him, but were not here to hold hands or commit crimes)
Well, it's pretty obvious you didn't read my comments, think about them, read the problems i commented against, and thought about those, and finally thought all of it together. You are just out to troll, and your reply is just flamebait. Your time would probably best be spent on USENET, as there are plenty of trolls there.
And before anyone says I am trolling, you're almost right, except that I explained why the problems i mention as unfixable by us are just that. Can we configure the users email client for him? No. Therefore we can't fix those problems.
I wasn't trolling, and I'm very sorry if it seemed that way. I didn't mean to upset you either.
I suppose you were offering a suggestion, but if I understood it correctly, it seemed to be: "Let the user fix it." In this case, I believe we can solve it otherwise and spare the user the trouble dealing with such complex matters such as configuring mail clients and dealing with mailing-list interfaces. That's what Mike Hearn's original (re-)proposal suggested, that we set up a forum using phpBB or the like. There wouldn't be need for any mail client whatsoever and the interface is more familiar to the ordinary user. Yeah, call me an idiot, but we could even have stickies!
James Trotter wrote:
On 4/22/06, Segin < segin2005@gmail.com> wrote:
Hello, I have noted below problems that we can take care of, and those that aren't our problem, never were, never will be, impossible to be our problem, (you get the picture), because they are PEBKAC errors :)
For all of those that don't know: PEBKAC = Problem Exists Between Keyboard and Chair.
I find this a little crude. You're acknowledging the fact that they are problems, but dismissing them and not really offering a solution. It seems to me that a lot of the problems with the mailing list is that the interface is complex, especially for joe user. The least we can do is to make it easy for them, and a forum seems to solve many of the complexity issues. Take for example 1) Cannot post without configured mail client. If there was no need for a mail client, there would be no problem.
Molle Bestefich wrote:
Hi
I've tried to sum up the problems with the wine-users mailing list that are urging many people to call for a forum.winehq.org .
I hope it's useful.
Problems
- Cannot post without configured mail client
An unconfigued client causes hell for EVERYONE they email, not just the list members. If that's a problem for people, well, paper and pens still exist. this is PEBKAC.
- Browsing old topics and replying/posting new ones happen in two
different places (web archive vs. mail client)
There should be a "server" email address (like majordomo) that you san send email to for requesting old archived posts. It should be flexible. This would allow for this problem to be fixed, and the archives to be access via email (for us that remember email-ftp)
- Current archive web interface unwieldy (many clicks needed to
browse, always sorted by month)
This should be sorted in a 2-pass manner, Each level of posting (the original post is level one, immedate replies are level 2, and all the immediate replies to the first replies would be the seveal level 3s) would be sorted chronologically, and then by level. That's pretty much how I see mailing lists do that. If you don't understand, email me and i'll try to explain better.
Missing Googlish search function
"Subscribe" wording and web pages suggest your inbox will fill up
when doing so
Well, that's not a real problem because it is truthful for some poor people that don't use Yahoo!/GMail/Hotmail/Excite/etc. (those services offer large email boxes)
- Subscribing will cause your inbox to fill up; it's not obvious how
to change it; neither how to receive mail on specific topics you DO want once you've changed it
Solution: Teach users how to properly use their mailbox, or simply get a
bigger mailbox. If it helps, I'll give people GMail invites just to counteract this problem (and having to use a stupid web-based email client as well :) this one is partially our problem, partially PEBKAC, partially the email providers problem.
- Missing "forward all postings re this topic to my inbox" feature
- Very difficult to post under pseudonym
Not true, just don't include your real name anywhere in your email setup. I don't. Purely PEBKAC.
- Missing 'sticky' feature (?)
That's not our problem. it's impossible for that to be our problem, because on a mailing list, sticky stuff is client-side only, no matter what your client is, which also implies that it's impossible for us to do jack about it. Purely PEBKAC.
Solution 1: A forum, fx. phpBB
Pros: ) Easy to set up. We might even be able to sneak it in before AJ returns :-). (Sorry big guy, couldn't resist pulling one on your expense =)..) ) Might attract away from wine-users a high quantity of dumb questions that people ask when they can't be bothered to search the archives. ) Adds a structural approach by categorizing users' various problems, which encourages searching instead of asking.
Cons: ) Dillutes knowledge; some ends up in forum and some on wine-users, unless we nuke wine-users. ) Need to hack the code to get visual integration with WineHQ (menus
etc).
) To get single-sign-on, we need either a.) reverse proxy or b.) something LDAPish and some amount of phpBB hacking
Solution 2: Point people at Gmane or Google Groups in a prominent
place
========================================================================
Pros: ) We don't have to do anything :-).
Cons: ) If it's not directly on winehq.org, it's probably not official enough that people are going to use it. ) Not everyone has a Gmail account (required to post through Groups) ) Not everyone is adept to using Gmane
Solution 3: WebForum-on-top-of-List
Pros: ) We concentrate people, allowing web and email users to
communicate.
) We concentrate (archived) knowledge in one place.
Cons: ) The reason that forums are often filled with spam might be that it's too easy to post to them. We'd be duplicating that. ) We need to code it all ourselves!
I think solution 1 is good, but I prefer the extra icing on the cake that comes with solution 3.
Now, I'm not high on crystal meth or anything, but I'll just assume for a moment that we all want better accessibility for the newcomers (even with the rise in volume that they bring), and that all agree to go with solution 3 :-). Thus I'll try to outline a battle plan:
Battle-plan for solution 3
We'd like non-Wine developers to be able to join in; so let's make sure that this system is generically usable. That means:
- Standard user authentication, fx. using LDAP.
- Componentize a bit, fx. keep the web interface itself separate from
WineHQ menu structure and logo.
We'd need to:
- Prioritize features
- Setup a mailing list
- Setup a simple project web page
- Setup a winehq-like staging area for development (follows
HEAD/TIP/origin of web repository so everyone can see what the others are doing)
- Find developers/volunteers!
- Get started on the coding
When prioritizing features, we should take a look around and see if there are software components available that could be reused, fx.:
- PHP LDAP auth bindings
- Existing, good web interfaces for the wine archives (mailman?)
- ?...
Any holes or oversights in that plan?
If solution 3 ever takes off, I will gladly contribute with coding. Realistically, though, i cannot implement something like this all by myself, especially not in any kind of a timely fashion.
James D. Trotter
James D. Trotter
Actually, it seems most users like to send messages via their email client, but like in another email i sent, they also resent their email clients for lack of pretty graphics, colors, and the mailing list restriction on annoying HTML markup (namely none allowed) which if you look at it, is really pointless (would you like to get a email with a black background and pink text that's bold and some weird, hard-to-read font? I don't think so)
Also, apology accepted :)
James Trotter wrote:
On 4/22/06, *Segin* < segin2005@gmail.com mailto:segin2005@gmail.com> wrote:
Why aren't I offering suggestions? I did offer one. The rest are beyond our control. To attempt to fix those would be as intelligent as sticking your hand into a boiling pot of water. Let's look at that last one, the sticky thing. This usually means to have a email highlighted in a way that it gets the attention of the user. Now, think about it. There is nothing that we can do to address this ourselves. If you say we can, you are an idiot. That is something that must be configured in whatever email client the user is using. Now let's look at the email client config thing... Well, we could try to fix headers of incoming email, but that's not really a solution. In fact, there is no real solution that we can provide ourselves, and besides, a properly configued email client (to produce proper headers and such) is just proper netiuqitte. There is, again, nothing we can do legally about this (well, we can find a way to break into the users system and configure his email client for him, but were not here to hold hands or commit crimes) Well, it's pretty obvious you didn't read my comments, think about them, read the problems i commented against, and thought about those, and finally thought all of it together. You are just out to troll, and your reply is just flamebait. Your time would probably best be spent on USENET, as there are plenty of trolls there. And before anyone says I am trolling, you're almost right, except that I explained why the problems i mention as unfixable by us are just that. Can we configure the users email client for him? No. Therefore we can't fix those problems.I wasn't trolling, and I'm very sorry if it seemed that way. I didn't mean to upset you either.
I suppose you were offering a suggestion, but if I understood it correctly, it seemed to be: "Let the user fix it." In this case, I believe we can solve it otherwise and spare the user the trouble dealing with such complex matters such as configuring mail clients and dealing with mailing-list interfaces. That's what Mike Hearn's original (re-)proposal suggested, that we set up a forum using phpBB or the like. There wouldn't be need for any mail client whatsoever and the interface is more familiar to the ordinary user. Yeah, call me an idiot, but we could even have stickies!
James Trotter wrote:On 4/22/06, *Segin* < segin2005@gmail.com <mailto:segin2005@gmail.com>> wrote: Hello, I have noted below problems that we can take care of, and those that aren't our problem, never were, never will be, impossible to be our problem, (you get the picture), because they are PEBKAC errors :) For all of those that don't know: PEBKAC = Problem Exists Between Keyboard and Chair. I find this a little crude. You're acknowledging the fact that they are problems, but dismissing them and not really offering a solution. It seems to me that a lot of the problems with the mailing list is that the interface is complex, especially for joe user. The least we can do is to make it easy for them, and a forum seems to solve many of the complexity issues. Take for example 1) Cannot post without configured mail client. If there was no need for a mail client, there would be no problem. Molle Bestefich wrote:Hi
I've tried to sum up the problems with the wine-users mailing
listthat are urging many people to call for a forum.winehq.org
<http://forum.winehq.org> .I hope it's useful.
Problems
- Cannot post without configured mail client
An unconfigued client causes hell for EVERYONE they email, not just the list members. If that's a problem for people, well, paper and pens still exist. this is PEBKAC.
- Browsing old topics and replying/posting new ones happen
in twodifferent places (web archive vs. mail client)
There should be a "server" email address (like majordomo) that you san send email to for requesting old archived posts. It should be flexible. This would allow for this problem to be fixed, and the archives to be access via email (for us that remember email-ftp)
- Current archive web interface unwieldy (many clicks needed to
browse, always sorted by month)
This should be sorted in a 2-pass manner, Each level of posting (the original post is level one, immedate replies are level 2, and all the immediate replies to the first replies would be the seveal level 3s) would be sorted chronologically, and then by level. That's pretty much how I see mailing lists do that. If you don't understand, email me and i'll try to explain better.
Missing Googlish search function
"Subscribe" wording and web pages suggest your inbox will
fill upwhen doing so
Well, that's not a real problem because it is truthful for some poor people that don't use Yahoo!/GMail/Hotmail/Excite/etc. (those services offer large email boxes)
- Subscribing will cause your inbox to fill up; it's not
obvious howto change it; neither how to receive mail on specific topics
you DOwant once you've changed it
Solution: Teach users how to properly use their mailbox, or simply get a bigger mailbox. If it helps, I'll give people GMail invites just to counteract this problem (and having to use a stupid web-based email client as well :) this one is partially our problem, partially PEBKAC, partially the email providers problem.
- Missing "forward all postings re this topic to my inbox"
feature
- Very difficult to post under pseudonym
Not true, just don't include your real name anywhere in your email setup. I don't. Purely PEBKAC.
- Missing 'sticky' feature (?)
That's not our problem. it's impossible for that to be our problem, because on a mailing list, sticky stuff is client-side only, no matter what your client is, which also implies that it's impossible for us to do jack about it. Purely PEBKAC.Solution 1: A forum, fx. phpBB
Pros: ) Easy to set up. We might even be able to sneak it in before AJ returns :-). (Sorry big guy, couldn't resist pulling one on your
expense =)..)) Might attract away from wine-users a high quantity of dumb questions that people ask when they can't be bothered to
search thearchives. ) Adds a structural approach by categorizing users' various problems, which encourages searching instead of asking.
Cons: ) Dillutes knowledge; some ends up in forum and some on
wine-users,unless we nuke wine-users. ) Need to hack the code to get visual integration with
WineHQ (menus etc).) To get single-sign-on, we need either a.) reverse proxy
or b.)something LDAPish and some amount of phpBB hacking
Solution 2: Point people at Gmane or Google Groups in a
prominent place========================================================================
Pros: ) We don't have to do anything :-).
Cons: ) If it's not directly on winehq.org http://winehq.org,
it's probably not officialenough that people are going to use it. ) Not everyone has a Gmail account (required to post
through Groups)) Not everyone is adept to using Gmane
Solution 3: WebForum-on-top-of-List
Pros: ) We concentrate people, allowing web and email users to
communicate.) We concentrate (archived) knowledge in one place.
Cons: ) The reason that forums are often filled with spam might
be thatit's too easy to post to them. We'd be duplicating that. ) We need to code it all ourselves!
I think solution 1 is good, but I prefer the extra icing on
the cakethat comes with solution 3.
Now, I'm not high on crystal meth or anything, but I'll just
assumefor a moment that we all want better accessibility for the
newcomers(even with the rise in volume that they bring), and that all
agree togo with solution 3 :-). Thus I'll try to outline a battle plan:
Battle-plan for solution 3
We'd like non-Wine developers to be able to join in; so let's
makesure that this system is generically usable. That means:
- Standard user authentication, fx. using LDAP.
- Componentize a bit, fx. keep the web interface itself
separate fromWineHQ menu structure and logo.
We'd need to:
- Prioritize features
- Setup a mailing list
- Setup a simple project web page
- Setup a winehq-like staging area for development (follows
HEAD/TIP/origin of web repository so everyone can see what
the othersare doing)
- Find developers/volunteers!
- Get started on the coding
When prioritizing features, we should take a look around and
see ifthere are software components available that could be reused,
fx.:
- PHP LDAP auth bindings
- Existing, good web interfaces for the wine archives (mailman?)
- ?...
Any holes or oversights in that plan?
If solution 3 ever takes off, I will gladly contribute with
coding.Realistically, though, i cannot implement something like this
all bymyself, especially not in any kind of a timely fashion.
James D. TrotterJames D. Trotter
Saturday, April 22, 2006, 4:14:28 PM, Segin wrote:
Actually, it seems most users like to send messages via their emailclient, but like in another email i sent, they also resent their emailclients for lack of pretty graphics, colors, and the mailing listrestriction on annoying HTML markup (namely none allowed) which if youlook at it, is really pointless (would you like to get a email with ablack background and pink text that's bold and some weird, hard-to-readfont? I don't think so)
BTW that didn't stop you from using HTML format which (as you pointed out) is not good for mailing lists.
Oh yeah and that nick of yours...
Vitaliy.
Vitaliy Margolen wrote:
Saturday, April 22, 2006, 4:14:28 PM, Segin wrote:
Actually, it seems most users like to send messages via their emailclient, but like in another email i sent, they also resent their emailclients for lack of pretty graphics, colors, and the mailing listrestriction on annoying HTML markup (namely none allowed) which if youlook at it, is really pointless (would you like to get a email with ablack background and pink text that's bold and some weird, hard-to-readfont? I don't think so)
BTW that didn't stop you from using HTML format which (as you pointed out) is not good for mailing lists.
I thought I set it to plaintext for Wine-devel... maybe it's doing both plain and HTML At any rate, I never used "stupid" HTML (god awful colors/fonts/font sizes)
Oh yeah and that nick of yours...
what about it? :P
Vitaliy.
Segin.
On 4/23/06, Segin segin2005@gmail.com wrote:
I thought I set it to plaintext for Wine-devel... maybe it's doing both plain and HTML At any rate, I never used "stupid" HTML (god awful colors/fonts/font sizes)
Gmail now seems to be activating html automatically in reply of someone sending you an html email. I specifically requested plain text too.
Jesse
Segin wrote:
Why aren't I offering suggestions? I did offer one. The rest are beyond our control. To attempt to fix those would be as intelligent as sticking your hand into a boiling pot of water.
Let's look at that last one, the sticky thing. This usually means to have a email highlighted in a way that it gets the attention of the user. Now, think about it. There is nothing that we can do to address this ourselves. If you say we can, you are an idiot. That is something that must be configured in whatever email client the user is using.
Now let's look at the email client config thing... Well, we could try to fix headers of incoming email, but that's not really a solution. In fact, there is no real solution that we can provide ourselves, and besides, a properly configued email client (to produce proper headers and such) is just proper netiuqitte. There is, again, nothing we can do legally about this (well, we can find a way to break into the users system and configure his email client for him, but were not here to hold hands or commit crimes)
Well, it's pretty obvious you didn't read my comments, think about them, read the problems i commented against, and thought about those, and finally thought all of it together. You are just out to troll, and your reply is just flamebait. Your time would probably best be spent on USENET, as there are plenty of trolls there.
And before anyone says I am trolling, you're almost right, except that I explained why the problems i mention as unfixable by us are just that. Can we configure the users email client for him? No. Therefore we can't fix those problems.
Segin... chill! lol Flamebait aside, The fact of the matter is that while these things are beyond our control, some people just don't use email enough to configure a client, or be bothered with setting up a webmail account.. As for the mailbox filling up I think the complaint is not the mailbox running out of space, its that they dont want to have to delete several hundred (literally) messages when they come back from a weekend, or several thousand when they come back from holiday, and dont tell me to put them on vacation mode in the ml settings as that isnt convenient. People want 1 or 2 clicks or as close to it as possible, or they wont use it. Unfortunately the mailing lists just dont give them that.
Tom
Lots of forums support RSS, and you can "subscribe" to a topic - two ways for forum posts to come to you.
There are lots of people who prefer forums. Everything seems so much easier to me with a proper modern forum. There is so much that's just impossible on a mailing list.
Mailing lists are searchable, but harder than a forum to search. Forums let you search by any of text, title, author, date, category, etc.
You can't cancel or edit posts. Forum moderators can delete spam posts. On a mailing list, once spam gets though to my mailbox it's too late for a moderator to delete it.
Mailing lists can be categorized, but then you have to subscribe to each category individually - hence all user discussion being lumped together into "wine-users" instead of being subcategorized.
Forums puts quoted text in a nice colored box so you can tell unmistakably at a glance. You quote people with greater-than symbols - so I can't as quickly distinguish message text from quoted text and it's ugly. Your probably used to it, but we're not.
Forum admins can give people badges and/or titles to make it clear who's a developer. And you can see anyone's posting history. On a mailing list you can only sometimes tell by email address or signature.
Forum admins can lock (prevent replies to) a topic that's turned into a really bad flamewar.
Sterling Christensen wrote:
Lots of forums support RSS, and you can "subscribe" to a topic
- two ways for forum posts to come to you.
Those are definitely nice features.
There are lots of people who prefer forums. Everything seems so much easier to me with a proper modern forum. There is so much that's just impossible on a mailing list.
It's not impossible per definition, but it would need coding.
If given a handful of coders, I think the majority of forum niceness can be accomplished with a new web interface for the mailing lists.
Mailing lists are searchable, but harder than a forum to search. Forums let you search by any of text, title, author, date, category,
Ok, that sounds nice.
The 'category' thing sounds tricky to implement on top of a mailing list (as I would like it)...
You can't cancel or edit posts. Forum moderators can delete spam posts.
That's just a horrible feature. As with any kind of censorship it will do nothing but make people suspect you for deleting legitimate posts.
Spam should be removed (or maybe categorized under 'spam') by a spam filter. And if one does slip through, it will disappear in the noise in no time. Aren't most forums like about 30.000 pages long? Would you ever notice anything like a single spam mail?
Silly feature, if you ask me.
On a mailing list, once spam gets though to my mailbox it's too late for a moderator to delete it.
Here, have a cookie to go with your spam :-).
Mailing lists can be categorized, but then you have to subscribe to each category individually - hence all user discussion being lumped together into "wine-users" instead of being subcategorized.
Yes. Hmm.
Still advocating the "forum web interface on top of mailing lists" approach, I wonder how that could be done. Perhaps all uncategorized (mailing list originated) postings could be fed through a Hessian matrix and automatically categorized? :-). Or web interface users could be allowed to alter the category of a posting (initially dropped in the "uncategorized" category perhaps). Or both. Or something else. Hmm.
Forums puts quoted text in a nice colored box so you can tell unmistakably at a glance. You quote people with greater-than symbols - so I can't as quickly distinguish message text from quoted text and it's ugly. Your probably used to it, but we're not.
Easily fixed in a web interface: Take all consecutive lines prepended with '>' and put them in a <div> with a border on it.
Or a more flexible solution: Strip all non-[a-zA-Z] characters from each posting and all earlier postings in the same topic. Run a diff against the earlier postings, and any lines that are present counts as "quoted". Put those in boxes..
Forum admins can give people badges and/or titles to make it clear who's a developer.
I'm not even sure the Wine developers like badges. Well, perhaps they do, but then it should probably be honorary ones.
And you can see anyone's posting history.
How does that differ from searching for all postings from a particular e-mail address?
Forum admins can lock (prevent replies to) a topic that's turned into a really bad flamewar.
Sounds like a controversial feature to me..
* Sterling Christensen sterling.christensen@gmail.com [22/04/06, 16:10:06]:
Lots of forums support RSS, and you can "subscribe" to a topic - two ways for forum posts to come to you.
Which will fill my mailbox just like a mailing list does, just that it only says "Someone replied to your forum post. Go <long URL> to see the post" instead of giving me the post right away.
There are lots of people who prefer forums. Everything seems so much easier to me with a proper modern forum. There is so much that's just impossible on a mailing list.
This doesn't really seem like a point that is helping. Assume I reply with "There are lots of people who prefer mailing lists."
...
You can't cancel or edit posts. Forum moderators can delete spam posts. On a mailing list, once spam gets though to my mailbox it's too late for a moderator to delete it.
I've seen deleted forum entries only by people who started a flame out of sheer stupidity and then, when they realized that they were wrong deleted the post, leaving only the replies. I would hope that this sort of thing wouldn't happen for a wine forum anyway.
That being said, forums seem to get many more people who are too lazy to look for the information. If I need to sign up to a mailing list to ask a question, I can as well read the FAQ first.
As for the spam, I do have a moderator for my mailbox called spamassassin...
Mailing lists can be categorized, but then you have to subscribe to each category individually - hence all user discussion being lumped together into "wine-users" instead of being subcategorized.
That's why most email client have an option to use threads (google mail calls it conversations(beta)). If you are not interested in a thread, you can just delete it without reading.
Forums puts quoted text in a nice colored box so you can tell unmistakably at a glance. You quote people with greater-than symbols - so I can't as quickly distinguish message text from quoted text and it's ugly. Your probably used to it, but we're not.
Forums make it hard to correctly quote when replying to lots of things in one post. You're probably used to it, but...
Forum admins can give people badges and/or titles to make it clear who's a developer. And you can see anyone's posting history. On a mailing list you can only sometimes tell by email address or signature.
Does it really matter if the person who fixes your problem is a developer? We're talking about a user forum. For wine-devel I care who's a developer. But if you're reading wine-patches/wine-cvs and wine-devel, that's easy to find out, too.
Forum admins can lock (prevent replies to) a topic that's turned into a really bad flamewar.
I agree that this feature can be handy.
A lot of pro-forum arguments can be used as pro-mailing list arguments, depending on what the author prefers to use.
Cheers, Kai
Btw,
comp.emulators.ms-windows.wine still exists and is used as a forum just fine.
http://groups.google.com/group/comp.emulators.ms-windows.wine
Instant Web Forum, just add water.
ciao, Marcus
You see Google Groups as something very Google-specific and you don't like it. This is maybe because you don't know what Google Groups is? I will try demystifying Google Groups a bit for our readers. I guess you know what newsgroups are, and that they existed before Google. What Google did is to index them into their database. They did not only that, they also created a great web interface to browse them. That's Google Groups. That's all there is to say for our current concern. True, only registered users may post through Google Groups, but anyone can browse them and anyone can post into them through a newsreader. Heck, chances are there is a newsreader integrated in your current mail client.
2006/4/22, Segin segin2005@gmail.com:
Hello, I have noted below problems that we can take care of [...]
- Missing Googlish search function
[...]
Well, that's not a real problem because it is truthful for some poor people that don't use Yahoo!/GMail/Hotmail/Excite/etc. (those services offer large email boxes)
This isn't very serious. It's like you're saying there is no point searching archives because you can leave everything in your inbox. Well I don't want to download months of archives in my mailbox to be able to search them.
Solution 2: Point people at Gmane or Google Groups in a prominent place
========================================================================
Pros: ) We don't have to do anything :-).
You may want to try to transfer archives into the newsgroup, if that's feasible. You may also need to learn more about how newsgroups are registered etc. I don't know anything about that myself.
Cons:
) If it's not directly on winehq.org, it's probably not official enough that people are going to use it.
Oh come on. A project like the Eclipse IDE could have their official newsgroup listed on their website but Wine could not?
) Not everyone has a Gmail account (required to post through Groups)
) Not everyone is adept to using Gmane
Right. But not everyone is adept of (mailing lists|newsgroups|forums|you name it). That's not the question. The question was how we thought things could be improved. Maybe you should propose status quo, if it is what you have in mind.
Here's my own evalutation of the newsgroup suggestion.
Advantages (most important first in my opinion): - The fact is a newsgroup is the closest format to the actual mailing list. - Chances are you will retain the features of your current mail client. - Chances are it is not too much work to maintain once deployed, compared to other suggestions. - You instantly get a very good search. - You save yourself coding a web interface which has a fraction of the features found in a newsreader. I guess you can always code one if you want though. - You really are doing yourself a favor by using format specifically aimed by Google.
Costs (worse first in my own opinion): - Transferring the archives, registration hassles, etc. - Convince existing users to configure their newsreader.
It really comes down to priorities. Easier submission, better search, better structure (directory-like structure of a forum)?
Regards!
Hi Philippe,
Interesting post :-). I don't have time right now to answer all the pros you mention for newsgroups, so I'll just throw in a couple of things that I think is flawed in your suggestion (rude, perhaps! sorry!).
Philippe A. wrote:
I guess you know what newsgroups are, and that they existed before Google. What Google did is to index them into their database. They did not only that, they also created a great web interface to browse them.
Actually DejaNews did that and then Google bought DejaNews, AFAIR.
) Not everyone has a Gmail account (required to post through Groups) ) Not everyone is adept to using Gmane
Right. But not everyone is adept of (mailing lists|newsgroups|forums|you name it). That's not the question. The question was how we thought things could be improved.
The point was that everyone knows how a browser works.
Newsgroups might be too technical to engage in for newcomers, especially if they're already *very* busy trying to make this new software they've downloaded (Wine) work and everything's failing right and left around their ears with odd 'FontForge' messages, seh_exception overflows and what not.
Costs (worse first in my own opinion):
- Convince existing users to configure their newsreader.
I think the above point is enough to be a showstopper for the newsgroup approach.
Hi Philippe,
Interesting post :-). I don't have time right now to answer all the pros you mention for newsgroups, so I'll just throw in a couple of things that I think is flawed in your suggestion (rude, perhaps! sorry!).
Philippe A. wrote:
I guess you know what newsgroups are, and that they existed before Google. What Google did is to index them into their database. They did not only that, they also created a great web interface to browse them.
Actually DejaNews did that and then Google bought DejaNews, AFAIR.
) Not everyone has a Gmail account (required to post through Groups) ) Not everyone is adept to using Gmane
Right. But not everyone is adept of (mailing lists|newsgroups|forums|you name it). That's not the question. The question was how we thought things could be improved.
The point was that everyone knows how a browser works.
Newsgroups might be too technical to engage in for newcomers, especially if they're already *very* busy trying to make this new software they've downloaded (Wine) work and everything's failing right and left around their ears with odd 'FontForge' messages, seh_exception overflows and what not.
Costs (worse first in my own opinion):
- Convince existing users to configure their newsreader.
I think the above point is enough to be a showstopper for the newsgroup approach.
Well, you may think that, but really, it's not. They had to configure their mail client for email, and since most (if not all) mail clients still used double as a news reader, it shouldn't be too much of a hassle... Then again, they need to configure their newsreader if they plan on downloading all the warez on USENET (just kidding)
- Segin
Newsgroups might be too technical to engage in for newcomers, especially if they're already *very* busy trying to make this new software they've downloaded (Wine) work and everything's failing right and left around their ears with odd 'FontForge' messages, seh_exception overflows and what not.
Costs (worse first in my own opinion):
- Convince existing users to configure their newsreader.
I think the above point is enough to be a showstopper for the newsgroup approach.
Well, you may think that, but really, it's not. They had to configure their mail client for email, and since most (if not all) mail clients still used double as a news reader, it shouldn't be too much of a hassle... [...]
Segin is right saying it's not much harder to configure a newsgroup than it is to configure regular mail.
When I started using the net in 1998 or so, I really didn't like newsgroups. I stayed away from them for a very long time, until I came across Google Groups. Google taps so well into newsgroups that now I think all mailing lists of the world should be turned into a newsgroup :-)
(Note I keep advocating this suggestion because I'm quite convinced of its advantage but I'm not making that personal a personal matter at all.)
Philippe A. wrote:
I stayed away from them for a very long time, until I came across Google Groups. Google taps so well into newsgroups that now I think all mailing lists of the world should be turned into a newsgroup :-)
(Note I keep advocating this suggestion because I'm quite convinced of its advantage but I'm not making that personal a personal matter at all.)
I don't think a newsgroup is flexible enough that we can just send people there and imagine that all problems are solved, even with Google Groups existing.
My arguments are that a.) people might want to use their own e-mail instead of a Gmail, and b.) it's easy to get lost on Google Groups and end up in some other newsgroup.
b.) could be solved by adding a NNTP web interface such as NewsPortal @ http://florian-amrhein.de/newsportal/ to WineHQ, although I don't think it's as flexible as fx. a phpBB forum. For example, it doesn't allow posting or following conversations.
a.) does not seem to be easily solvable, which is (still) why I do not think the NNTP solution is good enough...
On 4/23/06, Molle Bestefich molle.bestefich@gmail.com wrote:
Solution 3: WebForum-on-top-of-List
Pros: ) We concentrate people, allowing web and email users to communicate. ) We concentrate (archived) knowledge in one place.
Cons: ) The reason that forums are often filled with spam might be that it's too easy to post to them. We'd be duplicating that.
A solution to this is to have moderators, and for new users have a delay between posts showing up on the forum and them being sent to the list. If a moderator deletes such a post within this time, it's not ever sent to the mailing list. Once a user has been on the forum for a week and has made atleast a few non-spammy posts, they no longer get moderated. You still get some spam, but for problem users you can warn them (and maybe ban them if it's bad.)
) We need to code it all ourselves!
I think solution 1 is good, but I prefer the extra icing on the cake that comes with solution 3.
Now, I'm not high on crystal meth or anything, but I'll just assume for a moment that we all want better accessibility for the newcomers (even with the rise in volume that they bring), and that all agree to go with solution 3 :-). Thus I'll try to outline a battle plan:
Battle-plan for solution 3
<snip>
Any holes or oversights in that plan?
Let's not make this so complex. Keep things as seperate as possible, but don't go out of our way to make it suit every possible purpose. If we make a complete list of things we want to implement before we start, the development should be easy and nothing will need bolting on after.
n0dalus.
Molle Bestefich wrote:
Hi
I've tried to sum up the problems with the wine-users mailing list that are urging many people to call for a forum.winehq.org.
I hope it's useful.
Problems
- Cannot post without configured mail client
- Browsing old topics and replying/posting new ones happen in two
different places (web archive vs. mail client) 3) Current archive web interface unwieldy (many clicks needed to browse, always sorted by month) 4) Missing Googlish search function 5) "Subscribe" wording and web pages suggest your inbox will fill up when doing so 6) Subscribing will cause your inbox to fill up; it's not obvious how to change it; neither how to receive mail on specific topics you DO want once you've changed it 7) Missing "forward all postings re this topic to my inbox" feature 8) Very difficult to post under pseudonym 9) Missing 'sticky' feature (?)
Solution 1: A forum, fx. phpBB
Pros: ) Easy to set up. We might even be able to sneak it in before AJ returns :-). (Sorry big guy, couldn't resist pulling one on your expense =)..) ) Might attract away from wine-users a high quantity of dumb questions that people ask when they can't be bothered to search the archives. ) Adds a structural approach by categorizing users' various problems, which encourages searching instead of asking.
Cons: ) Dillutes knowledge; some ends up in forum and some on wine-users, unless we nuke wine-users. ) Need to hack the code to get visual integration with WineHQ (menus etc). ) To get single-sign-on, we need either a.) reverse proxy or b.) something LDAPish and some amount of phpBB hacking
Solution 2: Point people at Gmane or Google Groups in a prominent place
Pros: ) We don't have to do anything :-).
Cons: ) If it's not directly on winehq.org, it's probably not official enough that people are going to use it. ) Not everyone has a Gmail account (required to post through Groups) ) Not everyone is adept to using Gmane
Solution 3: WebForum-on-top-of-List
Pros: ) We concentrate people, allowing web and email users to communicate. ) We concentrate (archived) knowledge in one place.
Cons: ) The reason that forums are often filled with spam might be that it's too easy to post to them. We'd be duplicating that. ) We need to code it all ourselves!
I think solution 1 is good, but I prefer the extra icing on the cake that comes with solution 3.
Now, I'm not high on crystal meth or anything, but I'll just assume for a moment that we all want better accessibility for the newcomers (even with the rise in volume that they bring), and that all agree to go with solution 3 :-). Thus I'll try to outline a battle plan:
Battle-plan for solution 3
We'd like non-Wine developers to be able to join in; so let's make sure that this system is generically usable. That means:
- Standard user authentication, fx. using LDAP.
- Componentize a bit, fx. keep the web interface itself separate from
WineHQ menu structure and logo.
We'd need to:
- Prioritize features
- Setup a mailing list
- Setup a simple project web page
- Setup a winehq-like staging area for development (follows
HEAD/TIP/origin of web repository so everyone can see what the others are doing)
- Find developers/volunteers!
- Get started on the coding
When prioritizing features, we should take a look around and see if there are software components available that could be reused, fx.:
- PHP LDAP auth bindings
- Existing, good web interfaces for the wine archives (mailman?)
- ?...
Any holes or oversights in that plan?
If solution 3 ever takes off, I will gladly contribute with coding. Realistically, though, i cannot implement something like this all by myself, especially not in any kind of a timely fashion.
I can contrib some code (need to start learning php anyways), and as I said I will be more than happy to setup a shell with a hosting provider for this and irc............
Tom Spear (Dustin Booker, Dustin Navea) wrote:
If solution 3 ever takes off, I will gladly contribute with coding. Realistically, though, i cannot implement something like this all by myself, especially not in any kind of a timely fashion.
I can contrib some code (need to start learning php anyways),
A single brave soldier. Not exactly the number I was hoping for. On the bright side we won't need a mailing list to communicate ;-).
I wonder about two things right now:
1.) === Do we agree on the way forward?
I'll quote in a more concrete fashion what I think would be neat in a moment.
I have this idea in my head that you'd like a phpBB forum and then implement a mailing list on top of each category. I think that will lead to somewhere near zero subscribers on those lists. Am I completely off the track here?
If we (the two of us, for a start) seem to agree, then we can try persuading Mike Hearn to hold the implementation of forum.winehq.org via phpBB off for a while while we try to make something work.
2.) === Is two people enough to implement the labourious parts, fx. a forum-like web interface?
I don't know.
Concrete proposal (also attached, in case Gmail linewraps too badly) =================
* 'Register' web page - Same functionality as Mailman subscribe (piggyback Mailman list configuration) - Email delivery is turned off when registering - Wording in emails should be altered ("forum" vs. "mailing list", "register" vs. "subscribe")
* Forum login web page - Piggyback Mailman user database. We'll automagically achieve single sign-on whenever Mailman @ WineHQ does ;-).
* Threading and import daemon - Imports raw .eml files - Stores in DB - Assigns internal thread/topic ID to each mail based on Message-ID, References and Subject - Inspiration?: weaverd @ http://weaver.gmane.org/
* Forum web interface - Sorted by Category, assignable to each thread/topic (Also wanted: subcategories) - Threaded view - Search engine (use DB fulltext index) - Allow posting with your registered email address (automagically sets "Reply-To:list" !) - Allow ticking of topics you're interested in ("follow this topic" checkbox when posting or so) - Inspiration?: http://forums.mysql.com and phpBB
* Notification daemon - Pick up a list (DB) of which topics users want notifications for - For each new message in DB, - Send an email to interested users (if any) - Flag as 'notification complete'
Labor requirements guesstimate ==============================
* 'Register' web page: Low * Forum login web page: Low * Threading and DB import daemon Medium * Notification daemon Medium * Forum web interface High
Futureware ==========
The following can be done later, or not at all:
* Additions to Forum web interface - Keep track of which threads and/or messages you've read (does a forum?) - Rating system, fx. 'Sticky', 1-5 stars or Slashdotish
* Mail gateway allowing to post with a pseudonym <random>@obscure.winehq.org
* Additions to Threading and DB import daemon - Try to guesstimate category for new list-originated topics
Molle Bestefich wrote:
Tom Spear (Dustin Booker, Dustin Navea) wrote:
If solution 3 ever takes off, I will gladly contribute with coding. Realistically, though, i cannot implement something like this all by myself, especially not in any kind of a timely fashion.
I can contrib some code (need to start learning php anyways),
A single brave soldier. Not exactly the number I was hoping for. On the bright side we won't need a mailing list to communicate ;-).
in response to below, keyword is start learning, I dont know jack about php, so we will definitely need more than 2 people, that and I work a full time job that doesnt let me do much else, so I can do some small things, but as for writing half or even 1/8 of the implementation myself (even if i did know php), I dunno
I wonder about two things right now:
1.)
Do we agree on the way forward?
I'll quote in a more concrete fashion what I think would be neat in a moment.
I have this idea in my head that you'd like a phpBB forum and then implement a mailing list on top of each category. I think that will lead to somewhere near zero subscribers on those lists. Am I completely off the track here?
Well, yes and no.. You are right about 0 subs, and I'm not quite sure why i even said that, I should have just said we need to make sure certain (or all as the demand arises) people can subscribe to an entire category, even if they dont actually take advantage of that.. I would subscribe to all categories, set my settings up to only send me the headers of new topics, and then if the topic is interesting to me, or something I can help with, then I click the link to get to that topic, and read thru the messages already there, replying if necessary. Plus if I dont read a topic, and someone else does and they think I could help out with it, they could just send me the link, instead of me being CC:ed and getting new emails for each post
If we (the two of us, for a start) seem to agree, then we can try persuading Mike Hearn to hold the implementation of forum.winehq.org via phpBB off for a while while we try to make something work.
I like the idea of at least getting it started while we (the project) write our implementation. We just need to make sure that we can parse the phpbb databases when the time comes, so that we can xfer the topics and messages over to the new forums when they go live.
2.)
Is two people enough to implement the labourious parts, fx. a forum-like web interface?
I don't know.
I doubt it seriously, especially with 1 having no experience in php (me)
- Allow posting with your registered email address (automagically
sets "Reply-To:list" !)
If this is to be part of the forums on top of mailing lists idea, I thought of a way we could possibly do it, but it might be more laborious than even the web interface
- Notification daemon
- Pick up a list (DB) of which topics users want notifications for
- For each new message in DB,
- Send an email to interested users (if any)
- Flag as 'notification complete'
Make the notification settings user configurable so that if they only want to receive 1 email per thread, that is all they get (an email with the thread topic as the subject and a link to the thread in the body is how I want to set mine up)
Labor requirements guesstimate
- 'Register' web page: Low
- Forum login web page: Low
- Threading and DB import daemon Medium
- Notification daemon Medium
- Forum web interface High
Futureware
The following can be done later, or not at all:
Additions to Forum web interface
- Keep track of which threads and/or messages you've read (does a forum?)
- Rating system, fx. 'Sticky', 1-5 stars or Slashdotish
Mail gateway allowing to post with a pseudonym <random>@obscure.winehq.org
Additions to Threading and DB import daemon
- Try to guesstimate category for new list-originated topics
There are actually a couple of ways we could do this, none of them seem easy to me..
1) wine-forums@winehq.org : users subscribe to there to get every posting to the forums, only replying to the ones they want, just like current mailing lists are but its the forums emailing them instead of mailman.. Here is the key to #1.. Each thread has it's own pseudo email address that users respond to (instead of responding to the list).. The problem with this is that we end up eating lots of memory on the server with all of these pseudo emails when the forums get huge..
2) wine-forums@winehq.org : same as 1, except different key.. Each thread is replied to by responding to the list, and we parse the subject to determine which thread the email goes to.. What about when we end up with 2 threads named exactly the same? Should we do wine-forums-(category)@winehq instead? That would cut down on the likelyhood of identical thread names, but it still doesnt eliminate the problem.. How to fix this issue im not sure..
There is another problem with forums on top of a ml.. The order messages are posted, and auto quoting that most clients do.. When someone posts to a thread, and then someone else receives it in their inbox, they hit reply, the client quotes the entire last message. The forums need to strip that quote out before applying it to the database, and then look for a special keyword to mark what should be left in (such as [quote] blah [/quote] just like phpbb does but this is in an email)... Then there is when someone doesnt check their email right away, and still hits reply.... We need to put something in the headers that gets sent with a reply for the forums to determine where to put the message on the web layout (so that what would be message 8 looks like message 4 in the thread) If I need to draw a diagram for this one (because I know my wording is weird) let me know.
Of course now that I look at all I have said.. I still like the idea of forums on top of ml, but it doesn't look very feasible, even for the most advanced coders.. I'm still willing to give it a shot, but if it doesn't work out, I'd like to fall back to a phpbb at the very least...
Now, I'm not high on crystal meth or anything, but I'll just assume for a moment that we all want better accessibility for the newcomers (even with the rise in volume that they bring), and that all agree to go with solution 3 :-). Thus I'll try to outline a battle plan:
Well, my own opinion is that this is aiming for perfection where the last 10% just isn't worth that much.
Simply setting up a stock forum program somewhere, doesn't even have to be on the HQ servers as long as it has an official looking address, would be have a good effort:reward ratio, whereas writing some new forum/list interface software would not.
I rather think you'd find that part of the reason people like web forums is that they do offer a lot of features that a pure mailing list doesn't .... eg the ability to have a buddy icon, a personal profile, etc. If you start from scratch you'd lose a lot of that and then wonder why not many people were using it.
Look at it this way. There have been unofficial Wine forums before. People already discuss it in distro-specific web forums. Yet, wine-users is not dead, we are not suffering badly due to split resources, and I don't see developers (well, except me sometimes) going onto these forums to talk to people. So we appear to have nothing to lose.
Single sign-in etc would be nice but it's _not_ essential and given the choice between another account (really not that bad given the N^2 we all already have) or not having some official forum, I'd go for the former.
thanks -mike
I don't see developers (well, except me sometimes) going onto these forums
Should have read everything first .... I should have said developers except me and Scott :)
thanks -mike
-
James Trotter -
Jesse Allen -
Kai Blin -
Marcus Meissner -
Mike Hearn -
Molle Bestefich -
n0dalus -
Philippe A. -
Segin -
Sterling Christensen -
Tom Spear (Dustin Booker, Dustin Navea) -
Vitaliy Margolen