Fwd: [Haskell-cafe] Filing bugs for haskell packages

Richard Eisenberg

7 Feb 2014 7 Feb '14

8:47 a.m.

Hi libraries, The message below came through Cafe. It reports a problem when emailing libraries@haskell.org, which is listed as the maintainer of several public packages. Yet, the list does not accept emails from non-subscribers. These two facts seem to be at odds, and turns people away from reporting bugs. Can we fix this? Or, do we want a separate, world-writable libraries-bugs@haskell.org? Thanks! Richard Begin forwarded message:

...

From: Szymon Pyżalski Subject: [Haskell-cafe] Filing bugs for haskell packages Date: February 7, 2014 3:08:20 AM EST To: haskell-cafe@haskell.org

Hello!

I am new to haskell and I found a bug in package http://hackage.haskell.org/package/regex-compat-0.92/docs/Text-Regex.html

On the hackage page it says that the maintainer e-mail of this package is: libraries@haskell.org. So i sent a bug report there, but bounced back to me because of quota. So apparently nobody reads these e-mails.

How can I file a bug report for this package? I don't feel good enough in haskell to try and fix it myself.

Greetings Szymon Pyżalski _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

Attachments:

attachment.html (text/html — 3.0 KB)

Show replies by date

Niklas Hambüchen

7 Feb 7 Feb

9:01 a.m.

Being able to post without subscription sounds sensible. I'm subscribed to > 10 mailing lists of projects for which I report less than a bug per year because of exactly this problem, and it is very annoying. Even if you take the extra effort and go through a sign up/off for every single bug you want to report, this doesn't really work because you might miss replies in between the signed-up periods, and while you are signed up you get all the clutter from the mailing list that is unrelated to your problem. I recall other projects implementing an open-mailing-list principle (posting without subscription), and I think it works great. On 07/02/14 13:47, Richard Eisenberg wrote:

...

Hi libraries,

The message below came through Cafe. It reports a problem when emailing libraries@haskell.org mailto:libraries@haskell.org, which is listed as the maintainer of several public packages. Yet, the list does not accept emails from non-subscribers. These two facts seem to be at odds, and turns people away from reporting bugs. Can we fix this? Or, do we want a separate, world-writable libraries-bugs@haskell.org mailto:libraries-bugs@haskell.org?

Thanks! Richard

Mateusz Kowalczyk

9:10 a.m.

On 07/02/14 14:01, Niklas Hambüchen wrote:

...

Being able to post without subscription sounds sensible.

I'm subscribed to > 10 mailing lists of projects for which I report less than a bug per year because of exactly this problem, and it is very annoying.

Even if you take the extra effort and go through a sign up/off for every single bug you want to report, this doesn't really work because you might miss replies in between the signed-up periods, and while you are signed up you get all the clutter from the mailing list that is unrelated to your problem.

I recall other projects implementing an open-mailing-list principle (posting without subscription), and I think it works great.

On 07/02/14 13:47, Richard Eisenberg wrote:

...
Hi libraries,

The message below came through Cafe. It reports a problem when emailing libraries@haskell.org mailto:libraries@haskell.org, which is listed as the maintainer of several public packages. Yet, the list does not accept emails from non-subscribers. These two facts seem to be at odds, and turns people away from reporting bugs. Can we fix this? Or, do we want a separate, world-writable libraries-bugs@haskell.org mailto:libraries-bugs@haskell.org?

Thanks! Richard

Libraries mailing list Libraries@haskell.org http://www.haskell.org/mailman/listinfo/libraries

I think a separate mailing list that you can post to without subscription would be the best. Spam is always a concern but at least it'd be contained to a single list. -- Mateusz K.

Nicolas Trangez

9:19 a.m.

On Fri, 2014-02-07 at 14:01 +0000, Niklas Hambüchen wrote:

...

I'm subscribed to > 10 mailing lists of projects for which I report less than a bug per year because of exactly this problem, and it is very annoying.

IMHO this yields a different question: is a mailing list a suitable method to report/track/discuss *bugs* discovered by end-users? Nicolas

Erik de Castro Lopo

4:19 p.m.

Nicolas Trangez wrote:

...

On Fri, 2014-02-07 at 14:01 +0000, Niklas Hambüchen wrote:

...
I'm subscribed to > 10 mailing lists of projects for which I report less than a bug per year because of exactly this problem, and it is very annoying.

IMHO this yields a different question: is a mailing list a suitable method to report/track/discuss *bugs* discovered by end-users?

+1 The libraries list has uses for discssion of library related issues but is not really a good venue for reporting bugs. We should probably use a *bug tracker* for bugs. Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/

Niklas Haas

4:49 p.m.

<52F4E749.2070900@nh2.me> <1391782771.2859.37.camel@chi.nicolast.be> <20140208081946.ed24903f9152bd862ce53de4@mega-nerd.com>

...

+1

The libraries list has uses for discssion of library related issues but is not really a good venue for reporting bugs. We should probably use a *bug tracker* for bugs.

Erik

+1 I'd also like to firmly request it be possible to submit bugs *without* registering first. I lost count of how many bugs I've neglected to submit just because I didn't want to go through the hassle of creating an account first. I really like a model like github's issues page, where you can at best just type a simple description of the bug. I think it's more important to enable people to make bug reports without hassle, than it is to force people to categorize them correctly etc. first.

Erik de Castro Lopo

5:42 p.m.

Niklas Haas wrote:

...

I'd also like to firmly request it be possible to submit bugs *without* registering first. I lost count of how many bugs I've neglected to submit just because I didn't want to go through the hassle of creating an account first.

The problem with being about to raise a ticket without creating an account is spam. Lots of it and its generated by automated tools.

...

I really like a model like github's issues page, where you can at best just type a simple description of the bug.

Yes, but even on github you need to be signed in. Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/

Gershom Bazerman

5:59 p.m.

On 2/7/14, 5:42 PM, Erik de Castro Lopo wrote:

...

The problem with being about to raise a ticket without creating an account is spam. Lots of it and its generated by automated tools. Agreed.

Folks have no idea how much spam we filter on the various haskell infrastructure things, be they wikis or tracs or email lists. We accidentally opened the lists up to unregistered email and got a wave of spam there too, immediately. Whatever we do, it'll need to require registration, whether its sticking with libraries@ as a place to report bugs, or doing something else... --Gershom

Ivan Lazar Miljenovic

6:03 p.m.

On 8 February 2014 09:59, Gershom Bazerman wrote:

...

On 2/7/14, 5:42 PM, Erik de Castro Lopo wrote:

...
The problem with being about to raise a ticket without creating an account is spam. Lots of it and its generated by automated tools.

Agreed.

Folks have no idea how much spam we filter on the various haskell infrastructure things, be they wikis or tracs or email lists. We accidentally opened the lists up to unregistered email and got a wave of spam there too, immediately.

Whatever we do, it'll need to require registration, whether its sticking with libraries@ as a place to report bugs, or doing something else...

As an alternative, maybe having a "report bug" page which doesn't require registration but instead uses a CAPTCHA or something to prevent bots?

...

--Gershom

_______________________________________________ Libraries mailing list Libraries@haskell.org http://www.haskell.org/mailman/listinfo/libraries

-- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com http://IvanMiljenovic.wordpress.com

Mateusz Kowalczyk

8:10 p.m.

On 07/02/14 23:03, Ivan Lazar Miljenovic wrote:

...

On 8 February 2014 09:59, Gershom Bazerman wrote:

...
On 2/7/14, 5:42 PM, Erik de Castro Lopo wrote:

...
The problem with being about to raise a ticket without creating an account is spam. Lots of it and its generated by automated tools.

Agreed.

Folks have no idea how much spam we filter on the various haskell infrastructure things, be they wikis or tracs or email lists. We accidentally opened the lists up to unregistered email and got a wave of spam there too, immediately.

Whatever we do, it'll need to require registration, whether its sticking with libraries@ as a place to report bugs, or doing something else...

As an alternative, maybe having a "report bug" page which doesn't require registration but instead uses a CAPTCHA or something to prevent bots?

...
--Gershom

_______________________________________________ Libraries mailing list Libraries@haskell.org http://www.haskell.org/mailman/listinfo/libraries

The problem with this approach is that there's only one way communication: someone fills in the form and relevant people get notified. If it's something simple, great, that can just be fixed. What happens if you need more information? You'd have to require the user to input contact information. Assume you do. What next? If you send them an e-mail, only you and they will see it: there's no public forum where more than one maintainer (or anyone that wants to) can pitch in and ask questions. At best you can forward the user's replies to a bigger audience but there's no way to go the other way around. In the end, to communicate with the user past the initial report, it's vital that the user can be reached somewhere where all interested parties can have a look at and communicate with each other. I think it's clear that there's a need for something with persistent communication channel, like an issue tracker. I think the best thing that can be done is to make the registration process easy: pop in your e-mail, solve captcha/verify e-mail and you're good. I think that requiring an account on an issue tracker is not too much. Requiring a subscription to a mailing list where all the communication gets sent to everyone _is_ too much and I think it's the only problem that needs working on here. -- Mateusz K.

Henning Thielemann

8 Feb 8 Feb

5:51 a.m.

Am 08.02.2014 02:10, schrieb Mateusz Kowalczyk:

...

I think it's clear that there's a need for something with persistent communication channel, like an issue tracker. I think the best thing that can be done is to make the registration process easy: pop in your e-mail, solve captcha/verify e-mail and you're good. I think that requiring an account on an issue tracker is not too much.

It is too much, since it requires (preferably) a new password that store safely and change frequently, as adviced by security experts. I guess a simple question like "what is Haskell's surname?" should keep the bots out.

Alain O'Dea

7 Feb 7 Feb

4:54 p.m.

New subject: [Haskell-cafe] Filing bugs for haskell packages

...

On Feb 7, 2014, at 21:19, Erik de Castro Lopo wrote:

Nicolas Trangez wrote:

...
...
On Fri, 2014-02-07 at 14:01 +0000, Niklas Hambüchen wrote: I'm subscribed to > 10 mailing lists of projects for which I report less than a bug per year because of exactly this problem, and it is very annoying.

IMHO this yields a different question: is a mailing list a suitable method to report/track/discuss *bugs* discovered by end-users?

+1

The libraries list has uses for discssion of library related issues but is not really a good venue for reporting bugs. We should probably use a *bug tracker* for bugs.

I'm new to reporting library issues in Haskell. It seems like the GHC Trac was the effective venue for reporting issues on platform libraries. Herbert filed the bug for me after I reported it here. It may be more effective for Hackage to link to a web page with bug reporting instructions rather than a maintainer email. It certainly reduces the burden on maintainers and provides clarity to reporters. Lowering the bar on bug reporting too far may facilitate non-constructive complaints. If there were endemic quality problems or a perception of quality problems in Haskell libraries, lowering the bar would be wise to address that. I believe that the quality of Haskell libraries and the perception of their quality generally is very high. Best, Alain

Alain O'Dea

5:10 p.m.

New subject: [Haskell-cafe] Filing bugs for haskell packages

...

On Feb 7, 2014, at 21:54, Alain O'Dea wrote:

...
On Feb 7, 2014, at 21:19, Erik de Castro Lopo wrote:

Nicolas Trangez wrote:

...
...
On Fri, 2014-02-07 at 14:01 +0000, Niklas Hambüchen wrote: I'm subscribed to > 10 mailing lists of projects for which I report less than a bug per year because of exactly this problem, and it is very annoying.

IMHO this yields a different question: is a mailing list a suitable method to report/track/discuss *bugs* discovered by end-users?

+1

The libraries list has uses for discssion of library related issues but is not really a good venue for reporting bugs. We should probably use a *bug tracker* for bugs.

It may be more effective for Hackage to link to a web page with bug reporting instructions rather than a maintainer email.

... so it is good that it actually does :D I'm in in rare form today.

Ganesh Sittampalam

8 Feb 8 Feb

11:29 a.m.

For what it's worth, I did exactly this for the http@projects.haskell.org list which is listed as the maintainer in the cabal file, and the list is overwhelmed by spam. Some subscribers even get bounced off it by mailman every so often because their own spam filtering is better than the pre-list filtering, and so they bounce enough of the messages that mailman sends them to trigger the "broken email address" check. I'm vaguely planning on giving up and going back to having myself only listed as maintainer as a result. I definitely wouldn't want people to have to subscribe to email the maintainer address though. On 07/02/2014 14:01, Niklas Hambüchen wrote:

...

Being able to post without subscription sounds sensible.

I'm subscribed to > 10 mailing lists of projects for which I report less than a bug per year because of exactly this problem, and it is very annoying.

Even if you take the extra effort and go through a sign up/off for every single bug you want to report, this doesn't really work because you might miss replies in between the signed-up periods, and while you are signed up you get all the clutter from the mailing list that is unrelated to your problem.

I recall other projects implementing an open-mailing-list principle (posting without subscription), and I think it works great.

On 07/02/14 13:47, Richard Eisenberg wrote:

...
Hi libraries,

The message below came through Cafe. It reports a problem when emailing libraries@haskell.org mailto:libraries@haskell.org, which is listed as the maintainer of several public packages. Yet, the list does not accept emails from non-subscribers. These two facts seem to be at odds, and turns people away from reporting bugs. Can we fix this? Or, do we want a separate, world-writable libraries-bugs@haskell.org mailto:libraries-bugs@haskell.org?

Thanks! Richard

Libraries mailing list Libraries@haskell.org http://www.haskell.org/mailman/listinfo/libraries

Bart Massey

11:53 a.m.

It is possible to set things up so that users can authenticate using an existing account on Facebook, GitHub, Google, etc. Not sure how much better this is than from-scratch registration, or how spam-proof it is, but more and more sites seem to be giving it a go. It at least amortizes password creation and maintenance. I think some kind of registration is a pretty low bar for reporting bugs. I do it all the time for various projects, and it's kind of expected if you're going to use a bug tracker. --Bart On Sat, Feb 8, 2014 at 8:29 AM, Ganesh Sittampalam wrote:

...

For what it's worth, I did exactly this for the http@projects.haskell.org list which is listed as the maintainer in the cabal file, and the list is overwhelmed by spam.

Some subscribers even get bounced off it by mailman every so often because their own spam filtering is better than the pre-list filtering, and so they bounce enough of the messages that mailman sends them to trigger the "broken email address" check.

I'm vaguely planning on giving up and going back to having myself only listed as maintainer as a result. I definitely wouldn't want people to have to subscribe to email the maintainer address though.

On 07/02/2014 14:01, Niklas Hambüchen wrote:

...
Being able to post without subscription sounds sensible.

I'm subscribed to > 10 mailing lists of projects for which I report less than a bug per year because of exactly this problem, and it is very annoying.

Even if you take the extra effort and go through a sign up/off for every single bug you want to report, this doesn't really work because you might miss replies in between the signed-up periods, and while you are signed up you get all the clutter from the mailing list that is unrelated to your problem.

I recall other projects implementing an open-mailing-list principle (posting without subscription), and I think it works great.

On 07/02/14 13:47, Richard Eisenberg wrote:

...
Hi libraries,

The message below came through Cafe. It reports a problem when emailing libraries@haskell.org mailto:libraries@haskell.org, which is listed as the maintainer of several public packages. Yet, the list does not accept emails from non-subscribers. These two facts seem to be at odds, and turns people away from reporting bugs. Can we fix this? Or, do we want a separate, world-writable libraries-bugs@haskell.org mailto:libraries-bugs@haskell.org?

Thanks! Richard

Libraries mailing list Libraries@haskell.org http://www.haskell.org/mailman/listinfo/libraries

_______________________________________________ Libraries mailing list Libraries@haskell.org http://www.haskell.org/mailman/listinfo/libraries

4119

Age (days ago)

4120

Last active (days ago)

List overview

Download

14 comments

12 participants

participants (12)

Alain O'Dea
Bart Massey
Erik de Castro Lopo
Ganesh Sittampalam
Gershom Bazerman
Henning Thielemann
Ivan Lazar Miljenovic
Mateusz Kowalczyk
Nicolas Trangez
Niklas Haas
Niklas Hambüchen
Richard Eisenberg