On Apr 4, 2010, at 19:35 , Ivan Miljenovic wrote:

I would wonder _why_ anyone would refuse to do so. Are they that ashamed of their own software that they wouldn't want to be associated with it, or is there some legal reason that they don't want to be associated with it?

Some people are paranoid about such things, for example because it would allow people to google-mine for things they'd rather a random HR person not reading by linking names together. They use a consistent online name for non-official stuff but work hard to avoid this being linked to their real name. (Several people I know who do this are fairly active in the bi, poly, and/or BSDM communities and are justifiably worried that HR would take a dim view of it being possibly associated with their company.) -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allbery@kf8nh.com system administrator [openafs,heimdal,too many hats] allbery@ece.cmu.edu electrical and computer engineering, carnegie mellon university KF8NH

Exactly, it's not like the Hackage people are doing extensive background checks of everyone, they just want something consistent. You guys don't _really_ think my name is Joe Fredette, right? I'm actually Batman. /Joe On Apr 4, 2010, at 7:58 PM, Jesper Louis Andersen wrote:

On Mon, Apr 5, 2010 at 1:49 AM, Brandon S. Allbery KF8NH wrote:

...

Some people are paranoid about such things, for example because it would allow people to google-mine for things they'd rather a random HR person not reading by linking names together.

In addition, the concept is rather silly, as one can just take a pseudonym without any of us knowing:

Whats new: Thu Apr 1 13:37:00 UTC 2010 NicolasBourbaki algebre-1.0

History is ripe with examples of this.

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

Unfortunately, Ivan, it's not so much the Whenever-I-can-be-bothered and more the Joe-had-4-finals-in-2-weeks-and-3-papers-to-write. HWN should be back shortly. Come Summertime, I suspect all of these delays will stop, but with a 7 class semester, something's gotta give. /Joe On Apr 5, 2010, at 4:52 AM, Ivan Lazar Miljenovic wrote:

Joe Fredette writes:

...

You guys don't _really_ think my name is Joe Fredette, right?

I'm actually Batman.

Batman, Joe, whatever your name is...

I notice that the HWN has turned into the Haskell Whenever-I-can-be-bothered-getting-around-to-it News... >_>

-- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com IvanMiljenovic.wordpress.com

Joe Fredette writes:

Unfortunately, Ivan, it's not so much the Whenever-I-can-be-bothered and more the Joe-had-4-finals-in-2-weeks-and-3-papers-to-write. HWN should be back shortly.

Hang on, I thought your name was Batman, not Joe... -- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com IvanMiljenovic.wordpress.com

In addition, the concept is rather silly, as one can just take a pseudonym without any of us knowing:

When I registered I was prompted to verify my identity by means of my university email (as opposed to my gmail account), which would complicate using a pseudonym. This being said, I have no problem with this restriction. In fact, trying to determine the origin of code before agreeing to distribute it sounds like sound procedure. Perhaps a good compromise would be the ability to hide the uploader on the public website (thus preventing data mining)? From a users perspective, the "Uploaded by" field of Hackage packages is somewhat redundant in the presence of "Maintainer" and "Author" etc. /Jonas On 5 April 2010 01:58, Jesper Louis Andersen wrote:

On Mon, Apr 5, 2010 at 1:49 AM, Brandon S. Allbery KF8NH wrote:

...

Some people are paranoid about such things, for example because it would allow people to google-mine for things they'd rather a random HR person not reading by linking names together.

In addition, the concept is rather silly, as one can just take a pseudonym without any of us knowing:

Whats new: Thu Apr 1 13:37:00 UTC 2010  NicolasBourbaki   algebre-1.0

History is ripe with examples of this.

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

2010/04/05 Casey McCann :

Not to mention that pseudonymity is overwhelmingly the norm on the internet.

I suppose this is the collision of two cultures. Lambda the Ultimate also encourages (but does not require) real names. I think this has to do with academic values, really -- and the Haskell community has a lot of those. You don't submit papers under names like `solidsnack'. There certainly is a significant subculture of anonymity on the internet but maybe it has spread beyond its useful limits? There are places where it is helpful (Allberry's examples above come to mind) but I don't think contributing code to Hackage (or Cheeseshop or anything else) is like that. -- Jason Dusek

On Mon, Apr 5, 2010 at 6:00 PM, David House wrote:

You're coming at this from the wrong angle. Rather than saying, "why should we allow pseudonyms?" we should ask "why are we restricting the freedom of users that just wish to contribute code?"

Exactly. I don't understand the argument about not trusting code from anonymous people. If you don't want to depend on the code, don't. If somebody wants to publish it to Hackage, fine; you still have the choice to use or not use it as you would if they published it anywhere else.

If I'm honest, I'm really surprised so many people have replied in favour of the restriction. I've stated an explicit way in which it's hurting the community, and the only person to say anything in the policy's defence other that "well, why not?" has been Ross (and I hope I dealt with the flaky arguments he linked to in my reply).

I'm extremely surprised, too, which is why I responded.

(P.s., I certainly wouldn't describe the use of pseudonym anonymity a "subculture". Perhaps it's not the norm in academic circles, but virtually all websites requiring a registration allow you to use whatever you like as a username. As does email. As does IRC. I can't think of many bits of the internet that don't.)

Yep. On the internet, you get to be anonymous. Why don't we kick people that don't have their name in their email address or IRC nick? -- Jeff Wheeler Undergraduate, Electrical Engineering University of Illinois at Urbana-Champaign

This discussion makes me ponder whether someone like _why the lucky stiff would ever contribute Haskell packages, hehe. I can count on two hands people I know in various programming communities who have identity issues but are prolific creators. Are we missing out? Probably. But at least there is github.

How would enforcing a 'real names' policy affect a contributor like _why (http://en.wikipedia.org/wiki/Why_the_lucky_stiff)? I assume they would not join the community. I get the feeling that this discussion is somehow linked to haskell's type-system, but have no idea why...

I hardly think you can say that _why had a negative impact on the ruby community... On Tue, Apr 6, 2010 at 9:53 AM, Christopher Done wrote:

On 6 April 2010 01:52, Ivan Miljenovic wrote:

...

On 6 April 2010 10:48, Christopher Done wrote:

...

This discussion makes me ponder whether someone like _why the lucky stiff would ever contribute Haskell packages, hehe.

I think we can do without someone who hides behind anonymity and then suddenly decides to go and delete all of their work when they've had enough.

Yes, pseudonymous people tend to delete all their work, and non-anonymous people are incapable of deleting all their work. _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

On Apr 4, 2010, at 22:57 , Ivan Lazar Miljenovic wrote:

"Brandon S. Allbery KF8NH" writes:

...

(Several people I know who do this are fairly active in the bi, poly, and/or BSDM communities and are justifiably worried that HR would take a dim view of it being possibly associated with their company.)

I can understand wishing to be anonymous in these kinds of situations, but in terms of submitting open source software? Unless their employer is worried about them releasing proprietary software on Hackage, I don't see the potential for embarrasment there.

It's more about wanting to keep their non-work-related stuff under a *common* ID, but not one that can be tied back to their work persona. A sort of rigorously-policed double life. There are people who do this, and if they ever are able to release something work-related they'll ask for a separate account for that. (Be it noted that I don't work that way; anyone who wants to search for me in Usenet archives can determine that pretty quickly. :/ But I can understand the impulse.) -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allbery@kf8nh.com system administrator [openafs,heimdal,too many hats] allbery@ece.cmu.edu electrical and computer engineering, carnegie mellon university KF8NH

Well, Is the "real name" uniq enough? I mean if I google for "Marc Weber" many Haskell related posts show up. So yes, this is me - but there are also many false hits. So I for my part do no longer trust google results if I want to judge a person. It gives some hints - you can verify by asking the person. Different case: What happens if someone else chooses a pseudonym which happens to be your real name? Do those people change their real name then? I don't think you can protect against everything Marc Weber

David House wrote:

An issue came up on #haskell recently with Hackage accounts requiring real names. The person in question (who didn't send this email as he's wishing to remain anonymous) applied for a Hackage account and was turned down, as he refused to offer his real name for the username.

It appears to me that it's generally a good idea to adopt a pseudonym that looks like a real name anyway. The main benefit is that no one will notice that it's a pseudonym, thus avoiding such complications. Ivan Miljenovic wrote:

I would wonder _why_ anyone would refuse to do so. Are they that ashamed of their own software that they wouldn't want to be associated with it, or is there some legal reason that they don't want to be associated with it?

I'm sure they have their reasons, and who am I to judge them. Most likely, it's about googleability. Regards, Heinrich Apfelmus -- http://apfelmus.nfshost.com

On 5 April 2010 12:52, Ross Paterson wrote:

Basically http://meatballwiki.org/wiki/RealNameUserAdvantages, especially simplicity, trust and recognizability.

Allow me to respond to some of these points. I find none of them particularly convincing, especially not when compared to the disadvantage that it's holding back contributors to hackage.

Simplicity. It's the simplest thing. You need a name, you use your name.

Disagreed. For those people who consistently use an online pseudonym, the simplest thing is to continue that consistency, rather than remember a list of exceptions who had a real names policy. Moreover it makes things more difficult for everyone else. If someone uses their pseudonym on IRC, on the wiki, on the mailing lists, on their website and so on and so forth, that's how I know them. If I want to find their hackage contributions, now I need to know their real name. Where do I find this information, in general? (I presume this addresses your "recognisability" point as well.)

Trust. If a person doesn't use their RealName, there is a reason for it. There are many possible reasons, most of them mean problems. So the community will not trust people without RealName - except if there is a really credible explanation.

This is an incredible claim. The number of online communities that mandate real names is tiny. This article seems to imply that the vast majority of online communities would be rife with mistrust. This is simply not how the internet works, or has ever worked. The rest of that article is a list of barrel-scraping excuses, e.g., * "Authorship. Being recognized (and honored) as the author." (Why doesn't that apply to a pseudonym?) * "Reputation. Using a RealName is the most credible way to build a combined online and RealLife identity." (Some people don't want this, for whatever reasons.) * ... and so on.

I don't recall anyone else refusing to use their real name, though a number of people have not responded to enquiries I made of them. Of course some may have been put off by the User accounts page.

There was at least one other person in the conversation who mentioned they'd be put off by this policy. A few others chimed in with general support, if not a specific mention of boycott. IMO this policy is hurting the community in much greater weight than any purported advantages. I'd like to see the restriction lifted. -David

David House wrote:

* "Reputation. Using a RealName is the most credible way to build a combined online and RealLife identity." (Some people don't want this, for whatever reasons.)

I agree that the restriction should be lifted. A lot of very smart people do not want their real names connected to certain projects or be found on the internet at all. And I don't agree that "why not?" can be a valid argument, but even if it is, the above is a valid answer to it. So all in all there is no convincing argument for the restriction, but at least two convincing arguments against. Human identity is much more than just a file descriptor or a map key, and people from academia often don't get this, because they don't have to fear using their real names. Particularly in economically illiberal countries being known as the author of a certain Haskell package can get you into trouble either at work or even with the government. It can also prevent you from getting a job. Nobody should be forced to use their real name anywhere on the internet, because unlike a bulletin board in a university, lab or school, the internet can be searched by employers easily. Greets Ertugrul -- nightmare = unsafePerformIO (getWrongWife >>= sex) http://blog.ertes.de/

Luke Palmer wrote:

...

So all in all there is no convincing argument for the restriction, but at least two convincing arguments against.

When you say convincing, you are talking about yourself being convinced, right? So this paragraph means "The arguments against my position haven't convinced me, but the arguments for my position have."

Yes, of course, although I think that I can speak for some others as well. Greets, Ertugrul -- nightmare = unsafePerformIO (getWrongWife >>= sex) http://blog.ertes.de/

On 04/05/2010 11:32 PM, Ivan Miljenovic wrote:

4) The people who support the policy don't see why anyone has a problem with it.

I have seen no logical explanation of *why* anybody supports this policy. I've only seen vague hand-wavy statements like "people who use real names are more reliable." Really? Where's the proof? I bet there's a fairly large number of badly maintained projects on Hackage, and I bet it has little or no correlation with the accuracy of the maintainers' listed names. Even if it's true, what harm is there in allowing less reliable maintainers to upload packages? So we end up with a few extra packages that nobody uses. So what? - Jake

On Tue, Apr 6, 2010 at 7:32 AM, Ivan Miljenovic wrote:

5) No-one is convincing anyone else to their point of view, so we have a stale mate.

Can you really legally distribute your software under an open source license if you don't use your real name? At least the Linux kernel has a policy these days that contributors pretty much have to reveal their real name to avoid legal issues.

...

...

...

...

"Pekka" == Pekka Enberg writes:

Pekka> On Tue, Apr 6, 2010 at 7:32 AM, Ivan Miljenovic Pekka> wrote: >> 5) No-one is convincing anyone else to their point of view, so we >> have a stale mate. Pekka> Can you really legally distribute your software under an open Pekka> source license if you don't use your real name? Well, yes, you can. The FSF has always suggested that you can assign the copyright of your GPL'ed software to them, in some circumstances. Note I'm very much pro the real-name policy in principle, but I don't see how it can be enforced. -- Colin Adams Preston Lancashire () ascii ribbon campaign - against html e-mail /\ www.asciiribbon.org - against proprietary attachments

On 6 April 2010 05:32, Ivan Miljenovic wrote:

5) No-one is convincing anyone else to their point of view, so we have a stale mate.

Let me summarise the main arguments against the restriction: 1. It stops people from contributing to hackage. (It is immaterial that if you were in their position, you would have no problem with the restriction. Because of this policy, we have fewer libraries on hackage.) The reason this came up is because someone on IRC wrote a great implementation of which(1) as a Haskell library. I suggested they put it on hackage, and they told me they wouldn't because of this policy. The community loses out. 2. Inconsistency. If someone is known by their pseudonym on the mailing list, IRC, haskellwiki, blogs and so on, that is how I know them. How am I meant to find out their real name, in general? The rest of the internet works off pseudonyms and it is more convenient for everyone if hackage follows suit. 3. Privacy issues. Some people simply cannot reveal their real names. I've been over this thread and couldn't see anywhere where you'd made an attempt to refute these arguments, so I guess you take them as solid. On the other hand, every argument put forward by the pro-restriction group has been picked at and argued against by those against the restriction. That is not a stalemate.

On Tue, 06 Apr 2010 20:06:27 +1000, Ivan Lazar Miljenovic wrote:

...

I've been over this thread and couldn't see anywhere where you'd made an attempt to refute these arguments, so I guess you take them as solid. On the other hand, every argument put forward by the pro-restriction group has been picked at and argued against by those against the restriction. That is not a stalemate.

A stalemate occurs when people disagree rather than one side having more compelling arguments. For example, I understand and respect your arguments; I just don't find them compelling enough (since I find it a pain trying to match up different nicks, etc.; though this is to do with the inconsistency you mention above).

So, what's next? Ban anyone hiding behind a nickname on the mailing list? On IRC? On wiki? This will bring you consistency. As a side effect, it will bring to the Haskell community the fame of being one of the most inadequate open source communities. Let's conduct a thought experiment: suppose hackage has just launched and its policy is yet to be decided. Do you find your own arguments compelling enough to accept the policy of Real Name Requirement? Taking into account that this policy is _inconsistent_ with other Haskell resources policy. -- Roman Cheplyaka, who has nothing to hide but hates silly enforcements

On Tue, Apr 6, 2010 at 1:08 PM, Serguey Zefirov wrote:

http://lambda-the-ultimate.org is one lovely community that has that restriction: http://lambda-the-ultimate.org/policies#Policies

LtU has no restriction on user names. From LtU's policy:

Anonymity and the use of pseudonyms is discouraged.

I'm against *requiring* the use of a "real name" for hackage accounts. I have no problems with encouraging the use of given/everyday names. Regards, Roel

David House writes:

Let me summarise the main arguments against the restriction:

1. It stops people from contributing [..] 2. Inconsistency [..] 3. Privacy issues [..]

4. It inteferes with people's freedom - who has the right to dictate what name a person (or, for that matter, a group of people) should be known as? 5. It encourages dishonesty: if you want to contribute but not reveal your real name, you have the option to lie about it, and can be fairly confident your lie will never be called. +1 for allowing nicks. -k -- If I haven't seen further, it is by standing in the footprints of giants

...

Let me summarise the main arguments against the restriction:

...

1. It stops people from contributing [..] 2. Inconsistency [..] 3. Privacy issues [..]

4. It inteferes with people's freedom - who has the right to dictate what name a person (or, for that matter, a group of people) should be known as?

5. It encourages dishonesty: if you want to contribute but not reveal your real name, you have the option to lie about it, and can be fairly confident your lie will never be called.

+1 for allowing nicks.

Another +1 from me. I must admit that I had never really thought about this restriction, but the arguments against the restriction clearly convince me. I have heard no valid arguments in favour of the restriction. I can see that there are advantages to requiring real names, but that only makes sense if it is enforced (and I certainly don't advocate that). The way it is now, where some people who just silently use pseudonyms get accounts, and others, who are not willing to lie, are rejected, is very bad. If people are really worried about trust, then a comment/reviewing system for Hackage is a better solution. Cheers, Andres -- Andres Loeh, Universiteit Utrecht mailto:andres@cs.uu.nl mailto:mail@andres-loeh.de http://www.andres-loeh.de

On Tue, Apr 06, 2010 at 10:10:09AM +0100, David House wrote:

3. Privacy issues. Some people simply cannot reveal their real names.

I've already said I was prepared to make exceptions in such cases, but perhaps you missed that.

Ertugrul Soeylemez wrote:

Human identity is much more than just a file descriptor or a map key, and people from academia often don't get this, because they don't have to fear using their real names. Particularly in economically illiberal countries being known as the author of a certain Haskell package can get you into trouble either at work or even with the government. It can also prevent you from getting a job.

FWIW, I think this reason alone is enough justification to lift the restriction. Personally: after using a consistent pseudonym for years, I was eventually convinced[1] that real names are best when you're involved in contributing to online communities--- and I do mean *communities*, not mere interaction. But, while I feel it's probably in my best interest to have my community deeds associated with my real name, I'm under no delusions that it is in everyone's best interests that their deeds be so. I've never had to deal with illiberal governments. I have, however, worked with a number of excellent hackers who live with them. Also I don't believe there's anything sacrosanct about "real" names. Is the persona we have in the work place more "real" than the one we have with friends? To say nothing of the countless friends of mine who've legally changed their names for this or that reason. Our identity and the status accorded to us does not come from a name legally given at birth. It comes from the personae with which we participate in the world. [1] http://www.catb.org/~esr/faqs/hacker-howto.html#status -- Live well, ~wren

Out of curiosity: is there something wrong with my nickname "migmit"? I'm not gonna change it anyway. On 6 Apr 2010, at 09:52, Edward Z. Yang wrote:

This is a pretty terrible reason, but I'm going to throw it out there: I like real names because they're much more aesthetically pleasing. In my younger days, I once decided, "Hey, I should get a pseudonym" and I picked something fairly ridiculous, just because "everyone else was doing it". I would have appreciated someone to have conked me on the head earlier and said, "No, that pseudonym is stupid" and made me use something else.

That said, I think I'd be perfectly happy with a policy that preferred real names, but was willing to take an (obvious) pseudonym if the author insisted.

Cheers, Edward _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

2010/4/6 Jonas Almström Duregård :

Maybe users could choose between using a real name and being given a random one (like Anonymous<N>). This will (1) protect from data mining, (2) protect from government persecution and (3) keep the damned 1337 Haxxor names away from Hackage :)

I think this is a bad idea as it ruins recognisability. How am I meant to know that anonymous1 on hackage is the same person as mrfoo on haskellwiki, for example? P.s. if (3) is your real reason for supporting this restriction, please rethink your perspective. You are supporting a policy that is hurting the community for mere aesthetic reasons.

Miguel Mitrofanov writes:

Out of curiosity: is there something wrong with my nickname "migmit"?

How it was derived is at least apparent, as opposed to nicknames that have nothing to do with people's real names. -- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com IvanMiljenovic.wordpress.com

Edward Z. Yang wrote:

This is a pretty terrible reason, but I'm going to throw it out there: I like real names because they're much more aesthetically pleasing.

I agree, and this is why I phased out "apfelmus" in favor of the pseudonym "Heinrich Apfelmus". So, a more accurate policy would be to accept not only real names, but also names that look like they're real, i.e. aesthetically pleasing noms de plumes. Regards, Heinrich Apfelmus -- http://apfelmus.nfshost.com

Steve Schafer wrote:

Heinrich Apfelmus wrote:

...

I agree, and this is why I phased out "apfelmus" in favor of the pseudonym "Heinrich Apfelmus".

You mean your name isn't really "Applesauce"?

I would probably apply for a name change if it were. ;) Regards, Heinrich Apfelmus -- http://apfelmus.nfshost.com