Note: I clarified the text a bit due to the input.
So I wanted to submit a small patch to Parrot. Where do I start? Possibly its Bug Tracker. However, there isn't any link there on how to post bugs. None whatsoever. Not even a link to an explanation. None.
So I had to ask the #parrot channel. They referred me to this document. After reading it, it said to prepare the patch, and send it to "parrotbug at parrotcode dot org" (a spam-obfuscated email "address" of parrotbug@parrotcode.org, which is the address for sending patches to parrot). However, since I could not press the link in my browser to invoke the email client (or right-click+copied it), I had to type it from memory. So I typed it as parrotbug@parrot.org which is the incorrect address. I received a bounce which I didn't notice, and naturally the mail did not arrive there.
Eventually, after the patch recipient complained that the email did not arrive, I sent it again (using my mailer's "Send Again" feature) while CCing her. She received the email with the patch, but the patch was still not registered in the issue tracker, because it was still sent to the wrong address.
I see several usability problems here:
I hope both these issues will be corrected.
Re:Life is tough!
Shlomi Fish on 2007-05-18T16:50:03
For the record, I included the patch in that message. It's just that I still sent it again to the same address believing it was correct.
And for the record, if I could not get something right, and had to consult people, and then the admins on something (and I'm very intelligent and computer savvy) then the blame is on the system. It's bad usabilty as I demonstrated.
If I designed a program, and someone had a problem using it, and he asked me what to do, then I should first explain him how to do it, and then restructure the program in a way that will eliminate future people not finding their way around the program. Likewise for a web-site or any other computerised system. It should not make you ask questions.
For more information consult:
- User Interface Design for Programmers
- Seven Steps to Remarkable Customer Service
- "The Design of Everyday Things"
So please don't throw the old "You're an idiot." card. If someone intelligent has a problem using a system someone else designed, then the blame is usually on the implementor.
Re:Life is tough!
chromatic on 2007-05-19T01:28:44
And for the record, if I could not get something right... then the blame is on the system. Sometimes I try to pull doors clearly marked "Push", even if they have the proper type of affordances (that means appropriate door handles). I take from that lesson two possibilities. One, I could be a complete fool. Two, people sometimes make mistakes no matter how well designed things are.
Maybe you can only insert your house key with the correct alignment, but that doesn't prevent you from cleaning your ears with it.
Re:Life is tough!
Shlomi Fish on 2007-05-19T06:03:56
And for the record, if I could not get something right... then the blame is on the system. Sometimes I try to pull doors clearly marked "Push", even if they have the proper type of affordances (that means appropriate door handles). I take from that lesson two possibilities. One, I could be a complete fool. Two, people sometimes make mistakes no matter how well designed things are.
Actually, Norman Walsh talks about that in "The Design of Everyday Things". He shows how some everyday objects have serious usability problems, and the signs of them don't help at that. So these doors are badly designed.
I demonstrated two easily fixed things in which the parrot bug and patch submission process got it wrong:
There wasn't a way to submit bugs from the bug tracker, not even a referral to a document that explains how. Yet, people expect it to be there, because that's normally the way to do it.
So I had to ask the people on #parrot (twice) and they referred me to the document. But any such request ticket should also be fixed by permanently fixing the problem, as Joel says
The Parrot help page had the address, and a non-trivial one in obfuscated text, as people can mistake it with parrotbugs@parrot.org or whatever, instead of the more convenient click on a mailto: link.
It will save a lot of time of the developers and webmasters into the future to fix these problems once and for all.
Re:Life is tough!
chromatic on 2007-05-19T14:46:20
There wasn't a way to submit bugs from the bug tracker, not even a referral to a document that explains how.Completely wrong. See bugs6.perl.org as well as rt.perl.org. You said that you visited the latter.
I'm still waiting for an explanation as to how people have the magic knowledge of the URL of the Parrot bug tracker such that they can type it from institutional memory without error without having to use a search engine or ask someone, yet are completely incapable of copying and pasting an e-mail address.
Somehow, you visited the wrong page and didn't see the directions posted prominently elsewhere on every entry point we can reasonably expect a novice to see. You made a bad assumption based on knowledge we don't reasonably expect users to have, and hide that bad assumption and the resulting mistakes behind a veneer of logic errors while waving the flag of bad usability.
You screwed up. That's okay. Everyone does. Yet ranting and raving isn't the way to get what you want here.
Re:Life is tough!
Alias on 2007-05-19T05:55:37
The thousands of pointless installations of junoscript suggest that stupidity is common and even normal, and should be designed for.
The bug tracker, which is where people expect to be able to submit bugs...
How do people find the bug tracker? Magic powers?
(My theory is that they look for suggestions on how to report a bug on the Parrot website.)
Re:Hasty Generalization
pudge on 2007-05-18T14:56:46
I have a theory too.
People who complain about how the Perl developers are rude, and then intentionally post de-obfuscated developer e-mail addresses on a public web site that are certain to get picked up by spammers, have very little basis for their complaint.
Re:Hasty Generalization
Juerd on 2007-05-18T15:34:16
As if spammers don't know how to deobfuscate that.
Such obfuscation is ineffective and indeed does make using the email address harder.
If you want people to report bugs, MAKE IT EASY AND OBVIOUS. If it takes longer than just typing the bug report itself, there will be people who no longer bother.
It's still unnecessary and rude to deobfu it in a journal, though.And... the spammers already have it.
Juerd on 2007-05-18T15:37:02
http://www.google.nl/search?q=parrotbug-parrotcode.orgRe:Hasty Generalization
pudge on 2007-05-18T15:59:50
It's still unnecessary and rude to deobfu it in a journal, though.Right. That is my only point. He likes to complain about how rude Perl developers are, and then he goes and does that. I think such obfsucation is pointless, but the people who did it don't, and I would respect their wishes, unless I was holding some sort of a grudge and I wanted to childishly "get back" at them somehow...
Re:Hasty Generalization
Shlomi Fish on 2007-05-18T17:07:18
Let me tell you something: if you're living today and think that you're protecting your addresses from spam by obfuscating them, then you're really deluded. Spammers can and do:
- Use web crawlers that read all addresses off web-sites.
- Install malware on millions of exploited computers and periodically keep record of all the addresses in their address books or received or sent email messages.
- Subscribe to mailing lists and listen to the traffic.
- Etc.
If you think you can keep an obfuscated address from the clatches of the spammers for long, you're wrong.
People who complain about how the Perl developers are rude, and then intentionally post de-obfuscated developer e-mail addresses on a public web site that are certain to get picked up by spammers, have very little basis for their complaint.
Just for the record, this is the "Disqualifier disqualifies based on his own fault" or Ad-hominem fallacy. And I never complained that Perl developers are rude, just that some of the resources available for beginners were sub-optimal.
Re:Hasty Generalization
pudge on 2007-05-18T17:25:00
Let me tell you somethingNo, I will not let you do any such thing. This is not your decision to make. They want to obfuscate it. It is their address. Taking it upon yourself to de-obfuscate it is rude and disrespectful, and you have no argument to make on that matter.Just for the record, this is the "Disqualifier disqualifies based on his own fault" or Ad-hominem fallacy.It's not a fallacy. It is legitimately calling into question whether your subjective judgment about rudeness is a reasonable one.And I never complained that Perl developers are rudeThat is, of course, a lie.
Re:Hasty Generalization
grinder on 2007-05-19T13:03:58
Obfuscated e-mail addresses are like copy protection systems. They annoy the legitimate user and pose absolutely zero problems to a spammer gang. I mean, how hard is it to write s/ AT
/@/g and see what falls out? And how much effort do I have to go through to decode the latest trick to fool the harvesters? For all I know, there's a Mozilla filter that deobfuscates addresses on the fly, which means the code is out there to do this. In that light, Shlomi posting a normal e-mail address is neither here nor there.
Re:Hasty Generalization
phaylon on 2007-05-19T15:09:52
Sure, there's code out there to do this. But usually (at least for me) de-obfuscated email adresses work pretty good. Since the net is full with non-obfuscated mail adresses, why should the spammers bother with the small part that is?
I second the thought that deobfuscating it was rude and childish, and Shlomi owes a big apology.
Re:Hasty Generalization
Shlomi Fish on 2007-05-18T16:53:23
Well, I'm used to rt.cpan.org where I have a convenient "New Bug" link at the top. And a bug tracker is the first place I'd expect to find that, because that's where every other project expects you to submit bugs.
So it is bad usability.
Re:Hasty Generalization
chromatic on 2007-05-19T01:36:50
I'm used to rt.cpan.org...Not everyone is. Not every Parrot contributor is a CPAN contributor or even a Perl user. I'm not sure why you think this matters.
... a bug tracker is the first place I'd expect to find that... Again, how do you expect people to find the bug tracker? Magic powers? Innate knowledge? Is it part of Joseph Campbell's monomyth, or some sort of Jungian archetype that me in my pushes-pull-doors-occasionally cluelessness have failed to integrate into my worldview?
Me, I think that the link to the Parrot Bugs Page on the Parrot homepage trumps knowledge mystically encoded into our DNA, because not everyone in the world is Shlomi Fish.
... that's where every other project expects you to submit bugs. Every other project? In the world? All 6,784 of them? You counted and surveyed all of them?
So it is bad usability.Shlomi Fish is mortal. A cat is mortal. Therefore, all of the cats in the world are Shlomi Fish.
Re:Hasty Generalization
Shlomi Fish on 2007-05-19T06:32:34
chromatic, most of the large or semi-large projects I know allow you to submit bugs using the bug tracker: KDE, Mozilla, GNOME, all the Various SourceForge projects, CPAN. Etc. This is the default behaviour of most bug trackers, and it meets the users' expectations.
And as for my generalisation, let me invoke Joel's Generalisation here: if it applies to me, it applies to everybody. I'm probably not the first one to make this mistake, and I won't be the last.
When someone points a usability problem in your system, the easy-way-out is to claim he's an idiot or an exceptional user. But this is also the most costly way, that will make many people unhappy, clog your support system with a lot of constant questions, and in general doesn't scale well.
chromatic, have you read any resources on UI-Design or human engineering?
Re:Hasty Generalization
chromatic on 2007-05-19T14:38:28
... most of the large or semi-large projects I know... Now it's merely most, not every. A good UI is consistent. Please be consistent.
... if it applies to me, it applies to everybody. Thus clearly we should publish all of our documentation in Hebrew first, as it's your native language. It's easier for everyone to understand their native languages than secondary or tertiary languages.
My corollary to Joel's law is: Joel is wrong.
When someone points a usability problem in your system, the easy-way-out is to claim he's an idiot or an exceptional user.Right, because the user is always right. Dear Mrs. Crabapple, We will miss you. Love, Herb.
chromatic, have you read any resources on UI-Design or human engineering?Shlomi, do you know what a hasty generalization is?
Re:Hasty Generalization
Aristotle on 2007-05-19T21:00:39
The wiki page you linked says this:
Now a generalisation of this generalisation is “If it happens to you, it happens to everybody.” or “if it applies to you, it applies to everybody”, which may or may not be true depending on the context.
You’ll note that the first tenet of usability is “you are not your user.”
Shlomi, I've actually defended you before in some forums because I think you're actually trying to be helpful. But frankly, at some point people have to consider whether the annoyance of dealing with you is worth the potential benefit of receiving your patches or other input.
I mean, you're bitching about having to correctly type in an email address, for fuck's sake. Think about it objectively: do you think it's worth someone's time to look at patches from someone who is unable to type in an email address? "From memory"? It is possible that the process of submitting patches should not be so easy that random lusers can do it? That they have to put in a little bit of effort as a kind of minimal filter?
However, since I could not press the link in my browser to invoke the email client (or right-click+copied it), I had to type it from memory.
Does not compute. The conclusion does not follow from the premise.
Re:Non-sequitur
chromatic on 2007-05-19T01:41:01
Deductive reasoning has bad usability because it is not intuitive.
Re:Non-sequitur
Shlomi Fish on 2007-05-19T05:40:58
Aristotle, for the last time - please stop being so laconic and cryptic. Explain what you want to say in meaningful and understandable English.
Regards, Shlomi Fish.
Re:Non-sequitur
Aristotle on 2007-05-19T08:42:39
for the last time
When was the first time?
Explain what you want to say in meaningful and understandable English.
My comment is perfectly meaningful and understandable English. The conclusion (“I had to type it from memory”) does not follow from the premise (“I could not press the link in my browser to invoke the email client (or right-click+copied it)”). Just because it wasn’t a
mailto:-link doesn’t mean you couldn’t copy-paste the obfuscated email address and correct it.This is a lesson I learned from another programmer long ago (as a fledgling console user, but the lesson applies to many other things): do not typewrite a name (such as a filename) that you see on the screen, as that is too prone to errors – either copy-paste it or use some form of completion.
