[gcd-software] You (yes, you!) have a homework assignment!

[Henry Andrews]

On second thought, let's try something a bit more specific, picking
stuff I've at least sort of had a conversation about with each of you
before. You can all comment on something else if you'd rather, but if
you're named here I expect a comment on something :-D

Alexandros: Say something about about other comic sites in languages
we haven't supported well or at all in the past. Bonus points for
mentioning anything that would be particularly good for us to try to
attract their users, or anything particularly bad or pointless that we
should avoid.

Jason: Comment on how well or poorly we present our data to the end
users. Positive or negative, general or specific. Bonus points for
suggesting any related changes, large or small.

Phil: Let's see- we've been talking more about gcd-policy than tech
stuff recently. How about site navigation and docs, whether that's
getting from the front page to where you want to go, or how the
documentation wiki is (or isn't) integrated with the site. Should we
consider more links between the two, or moving docs from one location
to another?

Will: You've been with the project a long time. What's the single
biggest improvement you've seen over the years (whether it's part of
this latest site or an earlier change), and what's a major thing you'd
like to see done in the next five years? Nothing small and specific
like "implement deletions" :-) But the main site and all of the
support sites / applications / systems are all fair game.

(I'm leaving Jochen, Lionel and myself off the hook because we've all
said a lot at times and part of the reason for this committee is to
get some alternate views and ideas).

thanks,
-henry

[Jochen Garcke]

> (I'm leaving Jochen, Lionel and myself off the hook because we've all
> said a lot at times and part of the reason for this committee is to
> get some alternate views and ideas).

Besides I am busy for another week. Just checking in between work trips.

I think I have email next week, but then the conference hotel is on the
beach as far as I understood...

Jochen

[Henry Andrews]

Beach wins over comic book software, IMHO :-)
I need plenty of time to digest your reprint ideas anyway, and maybe even go back and put more thought into your variant issue proposal. This discussion will still be here when you return!
thanks,
-henry

[Will Allred]

Sorry it's taken me so long to respond to this, but I was out yelling at the
kids to get off my lawn while shaking my fist. :)
Geez Henry, now I feel old.
But you're right, I've been around the project since 1996, back in the flat
file days. In fact, I indexed in Excel and Notepad you young
whippersnappers. It's all we had, and we liked it. Actually, most of my
indexing was done during these days. I'm not just bringing this up to reap
the rewards that seniority in the GCD confer (waitaminute, there aren't
any...aw heck). Unfortunately, I think the project is still held in the grip
of the flat file format, and it has stifled our growth. Early on, it was
easy, got a new piece of information, just add a field. But wait, where does
painter go? Shove it in colors...yeah, that works. Wait, how about this ad?
Uh...let's add another type, but do ads have a feature? And we get to where
we are now...round holes, dodecahedral pegs.
Jon implementing on-line indexing was a big step forward for the project,
but our over-dependence on Jon caused other problems. Don't get me wrong,
Jon single-handedly carried the technical side of the GCD for many years at
no small cost to himself, but we became stagnant.
To be honest, the best thing to happen to project was the organization of
the latest tech team. Henry brought in a lot of enthusiasm which helped kick
a lot of us out of the same ol' same ol' doldrums. When Jon requested we
find another hosting solution, we finally got control of the site which was
unimaginably wonderful from my perspective. And here we are.
I think our primary goal should be to make it easier to put data in and take
data out. This, of course, means that our UI needs some serious work, but I
think an API (perhaps an XML web service, or something better...I am
definitely NOT a developer, so be kind) needs to be done quickly, as well. I
believe that I've seen an iPhone or Droid app mentioned. These would be
freakin' awesome. Imagine being at a con and using your phone to scan the
barcode of the comic. It's not indexed, so you quickly input the credits for
the cover...or maybe just the price. How about creating an XML standard for
comics (we are the experts, right) so that Realms of Wonder, ComicBase, or
any other collection tracking app can add data. Or, even a Facebook app. We
really need to be thinking about integrating Web 2.0 into the project...of
course, we should already be thinking about what's next, too.
I know, it's the 900-pound gorilla, but we have to normalize our data. The
creator and character fields are a mess, and we're still doing goofy crap
with parentheses and square brackets that I have never quite understood. The
documentation on the wiki and the recent establishment of the gcd-policy
group are going to help us make decisions and hold our indexers and editors
to these new decisions. It'll be fun...you should join...there's s nothing
like a good ol' formatting flame war on the GCD. :)
And, in my heart, I have this wish (and you will all call me crazy...blame
it on senility). I would love to have the GCD completely distributed to the
point that if you have a static IP and an OK'd OS, follow these steps, and
BAM...you just became a mirror...almost like a peer to peer network...a GCD
cloud. Now, the routing alone for something like that makes my head hurt,
but thankfully there are guys here a heck of a lot smarter than me, so I can
still dream.
Now you kids get off my lawn! ;)

*For the record, I hit 39 in about 7 months...so I'm not old yet, but having
two, fourteen-year-old boys sure makes me feel that way.

Will
______________________________

Will Allred <wal...@gmail.com>
http://www.quantumzone.org/ http://www.comics.org/

[Alexandros Diamantidis]

* Will Allred [2010-05-25 00:35]:

> In fact, I indexed in Excel and Notepad you young
> whippersnappers.

By the way, indexing in Excel might be easier for some indexers or for
some tasks. Currently one might be able to do something like this by
exporting from Excel as a tab-separated file and importing it in the OI,
but it's a bit fiddly. I was thinking about offering an alternative
OI scheme presenting the issue details in a Javascript/DHTML
spreadsheet-like grid. There are libraries (either on top of jQuery UI
or stand-alone) that would make this not too difficult to implement.

Of course, if we can create an API to contribute data from ouside
applications, anybody will be able to create alternative indexing
interfaces which would be a big win...

> Jon implementing on-line indexing was a big step forward for the project,
> but our over-dependence on Jon caused other problems.

Right... Through the years, various technically-minded people appeared
in the GCD lists, but very few tried to do some substantial work. I'm
certainly one of them, also becoming aware of the project in the
flat-file days, and not having done much since then :-(

> I think our primary goal should be to make it easier to put data in and take
> data out. This, of course, means that our UI needs some serious work, but I
> think an API (perhaps an XML web service, or something better...I am

OK, so you agree that an indexing / data contribution API should be
given higher priority... As for a XML schema for our data, there have
been various discussions in gcd-tech since at least 2002 ;-) J�mal even
set up a Yahoo group for this purpose, but I don't know if it went
anyhwere and I can't find it now.

> I believe that I've seen an iPhone or Droid app mentioned. These would
> be freakin' awesome. Imagine being at a con and using your phone to

I mentioned that possibility, and Henry replied that these apps have
already been created by others, so this spares us the effort ;-)
It would be good, though, to have their creators speak up in gcd-tech
about how an API convenient for them would look like.

> It's not indexed, so you quickly input the credits for the cover...or
> maybe just the price.

This presents a small problem, that applies not only to such a scenario,
but is also a source of frustration for current users of the OI: Having
to wait for approval after adding a new publisher, series, or issue,
before you can use it in indexing.

Maybe we should put some priority on thinking about solving this...

Since I mentioned MusicBrainz a couple of messages ago, you can see how
this works for them here:

http://musicbrainz.org/doc/Voting_FAQ

In short, they don't have an absolute approval process for edits, but
rather positive/negative votes, edits are added as pending in the db,
and edits can fail either due to negative votes, or if their
dependencies fail (e.g. adding a record label, then adding a release in
this label, then the new label is voted down (for example because it's a
duplicate of an existing one) and is rejected, so the release adition
also fails. Would something like that work for us? Any other ideas?

> And, in my heart, I have this wish (and you will all call me crazy...blame
> it on senility). I would love to have the GCD completely distributed to the
> point that if you have a static IP and an OK'd OS, follow these steps, and

It sounds nice in theory, but distributed databases and sites take much
work to get right - and are, in fact, usually research projects and not how
production sites operate.

> *For the record, I hit 39 in about 7 months...so I'm not old yet, but having
> two, fourteen-year-old boys sure makes me feel that way.

Congratulations! I'm 36 and can't see myself starting a family for
myself in the forseeable future...

Alexandros

[Lionel English]

On Tue, May 25, 2010 at 3:58 AM, Alexandros Diamantidis <ad...@hellug.gr> wrote:

> It's not indexed, so you quickly input the credits for the cover...or
> maybe just the price.

This presents a small problem, that applies not only to such a scenario,
but is also a source of frustration for current users of the OI: Having
to wait for approval after adding a new publisher, series, or issue,
before you can use it in indexing.

Maybe we should put some priority on thinking about solving this...

I think there is a bug filed for allowing people to submit multiple levels of edits at one time.  It's obviously a bit more complicated, as it would required a nested series of approvals, and you'd have to work out what to do with the dependent approvals if a higher level change gets rejected, but in theory we should be able to do this.

> *For the record, I hit 39 in about 7 months...so I'm not old yet, but having
> two, fourteen-year-old boys sure makes me feel that way.

Congratulations! I'm 36 and can't see myself starting a family for
myself in the forseeable future...

 

I didn't get married till I was 38, and I was 40 when I had my daughter (who is now three).  I think that makes me the old man of the group.

--
Lionel English
lio...@beanmar.net

[Philip Rutledge]

On Thu, May 20, 2010 at 5:55 PM, Henry Andrews <hh...@cornell.edu> wrote:

Phil:  Let's see- we've been talking more about gcd-policy than tech
stuff recently. How about site navigation and docs, whether that's
getting from the front page to where you want to go, or how the
documentation wiki is (or isn't) integrated with the site.  Should we
consider more links between the two, or moving docs from one location
to another?

I have a separate post about site navigation so let's tackle the docs question.  

When I think about wiki's I tend to think about many-to-many collaboration where many people are actively working on topics that are viewed by many people.  However the GCD wiki isn't really being used for that (as far as I can perceive) it really is just a convenient CMS system to publish information to the GCD audience (few to many relationship).  If that's the case then I'd think there are many other CMS like systems that could be better integrated with the main GCD site (of course Django is itself a CMS).

Looking at some of the specific areas.

Formating documentation, these pages are intended to be the rules which govern how users index books, however when I started I really just jumped into the OI and used my best judgement assisted by my mentor.  I did eventually read through the wiki but even though there is a help link the information isn't as readily consumable as I think it should be.  I would suggest that you really want a system similar to an online help infrastructure - when you're on a specific field you can get the rules for that field to open up in an adjacent panel.  You want to ensure that any information that can help the user is front and center.  (Along with enhancements to the UI to catch simple errors and auto-correct when needed (I still occasionally have an extra semi-colon at the end of my character entries!)  Such a CMS system would make it easier to flag "new rule" when a new rule has been added for a field as well.  Also I'm sure there are other workflow enhancements that could be implemented easier than via the wiki.

Much of the other information related to how to contribute, e-mail lists etc appears to be quite static and could just be folded into the main site with appropriate portals to allow the information to be edited by the editors/admins.  

I know there was a recent discussion of a CMS like portal for the site "news" on the main page, I would see that all of this could be handled by the same sort of system.

The proposal system is already moving to the "voting" app so we're already heading in this direction.  Some of the archival records perhaps may actually be easier to maintain as docs/files in the groups or again within the CMS system.

Don't get me wrong, I love wikis and I'm a bit advocate of them at work (although I prefer Confluence over MediaWiki) but I think they have a better role in allowing adhoc collaboration and aren't the best choice for simple information presentation (due to their inherent adhoc structure)

I do think that the technical design docs and the "other" resources should stay on the wiki.

WRT the current links - the "Documentation Wiki" link drops you off at the top level and "Become an Indexer" at the Indexing page - both of these destinations have a lot of info and might be a bit overwhelming.  I think that we want to encourage people to jump in and start contributing with some simple documentation integrated into the main site.  For example in the search results page have a note saying "didn't find what you were looking for? perhaps we don't have it indexed, become an indexer and help us... here's how..."  Also, I thought it was interesting that the "How to Contribute" page wasn't directly linked from the main page.

 

In summary, I'm suggesting using the appropriate tool for the information at hand which I think would imply moving away from the wiki for some of the information currently there to a CMS system more tightly integrated into the main site to allow integrated presentation and enhanced workflow features.

Phil 

[Henry Andrews]

Thanks, Phil.  I largely agree with what you're saying here.  The reason that the wiki and the error tracker are completely separate applications is primarily that we had no control over or access to the last server.  So doing anything integrated was flatly not an option.  The error list, in particular, used to just be an email list and things got dropped frequently.  We're kind of overdue for a reconsideration now that things are directly under our control- all of the apps run on the same server already (just with different virtual hostnames).

thanks,

-henry

[Jason Sacks]

I apologize for taking so long to join in on the discussions around here.
Let me try to make up for that by giving my thoughts on the above, many of
which will echo comments made by others around here, especially Phil
Rudledge's comments on 5/27 on the User Interface.

I'll start in this mail with the start page, then send further comments in
future mail. I know this mail is awfully detailed, but I hope it's helpful.

START PAGE IN GENERAL
Look at all the clutter on this page. As a first-time user, how do I know
where my eyes should go? There's so much text on this page, so many
disparate and confusing elements that seem to be competing for my attention.
This page just feels old-fashioned and cluttered when compared with most web
portals.

There are three disparate boxes on the left nav, three other disparate boxes
on the right nav, the Search box, a promo box in the middle, then a giant
cluster of text at the bottom along with an ad. On my 20 inch monitor, I
have to scroll to see the entire page. There's a lot of information being
thrown at me on this page, and my eyes aren't sure where to go or why things
are organized as they are.

On a high level approach to site design, wouldn't this page be more
interesting and feel more modern if most of the important information on
this page were either highlighted or moved to subpages that are linked-to
from this page?

The most exciting facts about this page to me are the Statistics. Like GMail
and its note about ever-increasing storage space, the Statistics widget
tells the user that the GCD is a uniquely useful site. Where else can a user
get information on nearly 600,000 comics - and growing? That fact should be
highlighted.

>SEARCH BOX
I think it's clear that the Search box is intended to be the point of entry
to the entire site. But its placement is not optimized. The Search box is
hidden at 0.25em from the top of the page, with text all around it. Compare
the placement of the Search box on Google or Bing. Notice how the box on
those pages is clean and simple and is placed right at eye level? We're
hiding the biggest asset of the site by having the Search box in a confusing
location, and this seems like an easy fix to make with significant gains.

Yes, the Search box is embedded in a purple table, which makes it stand out
from the page somewhat, but there is a lot of text even inside the Search
box. For a first-time user, are the drop-down selections a help or a
hindrance? How often do new users even use the drop-downs? To me, having two
different drop-downs just adds to the confusion of the box and makes me
scratch my head a bit about the results. I know I'm conditioned by the
search engines to expect appropriate results when typing any term into a
search box - whether typing "Thor" or "Perez" or "This Man This Monster" I
would expect to get valid results back. So why the multiple drop-downs?

Why is "Advanced Search" located directly below the Search box? Why isn't it
placed to the right of the box, since you have so much real estate on the
box? Also, placing the Search button in the middle of the Series Name and
Advanced Search boxes is confusing. Again, you're hiding a key site asset
for no good reason.

It's unclear at first what the user gets out of Advanced Search when the
Search box itself seems to provide so many choices.

Why is this box purple, and the Advanced Search box gold? Aside from the
fact that these are the colors of the University of Washington, why these
colors and not more traditional comics colors?

This box is also camouflaged a bit by the box on the top left nav on the
site. Having two boxes colored the same in close proximity really detracts
from the attractiveness of the Search box.

So my recommendations for the Search box:
- Lower it to at least 1em
- Remove the coloring or choose a more attractive coloring that grabs the
eye better
- Remove or at least simplify the drop-down - and consider just having one
drop-down (two is confusing - save that for Advanced Search)
- Place the Search button next to the free-text Search box so it's not
hidden
- Remove the contrasting colors for the Advanced Search box; place the
Advanced Search button in a more appropriate place (plain-text link under
the free-text Search box?)

>LEFT AND RIGHT NAV
Definitely the Donate box should stay on the front page, perhaps moved to
the bottom center of the page, above the legal disclaimers.

To me, all the other widgets on this page should be moved to sub-pages as
they only serve to distract the user.

The six items in the "About the GCD" likely will only be of interest to more
serious users (do you have stats on click-throughs from those links?); can
they be moved to the "About the GCD" page? How many people send email from
the Contact Us link, and might that be better handled as a web form?

Login is distracting as it adds two more blank text boxes to the page. Can
the start page just have a link to log in, and then have the user log in on
a separate page? How many users log into the site per month? Also, why is
there no "remember me" checkbox?

The disclaimer is useful, but why is it not on the page footer where it
belongs? Much of the content in that box duplicates the information at the
top of the page. Does it address real-world concerns about adult content on
the site or is it there for safety? Wouldn't such text belong in the site
wiki?

Why are the four links in the Useful Links box useful? The English start
page link is redundant (or the German link is redundant); can you manage
those links more elegantly by using flags as many sites do? Why are Today's
Covers and Covers to replace useful for site visitors? Those would only be
interesting to GCD admins and indexers, it seems to me, and should be moved
to a sub-page.

Last Updated Issues is cool, but doesn't strike me as especially useful. How
many people will visit GCD wondering what content was updated today? That
info also seems redundant with the Statistics directly below it.

>MIDDLE OF THE PAGE
I also really like the blurb that states "We're a nonprofit, Internet-based
organization..." but that statement is awfully verbose. How many readers get
to the "Give it a try" line, really? I would guess not very many. My
suggestion is to change that line to be a compelling mission statement for
the site. Get users excited about the non-profit, community based aspects of
the site. GCD stands out as being a community-based project, and it would be
cool to emphasize that.

The ad in the middle of the page will only apply to a small subset of users.
I know the site needs more web devs, but for the vast majority of users,
this is just white noise. Can you recruit in other ways or in other places
on the site? (Maybe just a highlighted box on the start page saying
"Volunteer for the GCD"?)

>FOOTER
The comment about "If you believe any of this data is incorrect" is
redundant to "Report an Error" in the top left box.

The disclaimers and ad should stay, obviously.

So overall recommendations above and beyond the comments about the Search
box above
- Get as much text as possible off of this page. Keep it simple, clean and
focused, and use subpages to connect your power users to the information
they need.
- Focus should be on helping the users get to the data they need as easily
and quickly as possible.
- GCD logo at the top center of the page
- Below the logo a note about the number of issues indexed (591,467 issues
indexed - and more every day!)
- Search box about 1em down from the note about pages indexed
- 4 to 6 links arrayed to the left, right or beneath the Search box (ABOUT
US - LOG IN - HELP THE GCD - STATISTICS - COMMUNITY)
- Mission statement below that
- Donate link below that
- Then the footer
- Have everything on the main screen so most users don't have to scroll.

I'll try to mock something up tonight for this. I hope this is helpful and
isn't too presumptuous of me. Let me know if this is the kind of stuff you
were looking for.

Jason

[Lionel English]

Either I mislaid it, or Henry didn't introduce people in this thread--great insights from Phil and Jason so far, could you be kind enough to share your backgrounds with us?

My thoughts on Jason's thoughts:

On Tue, Jun 1, 2010 at 4:55 PM, Jason Sacks <jason...@hotmail.com> wrote:

START PAGE IN GENERAL

There are three disparate boxes on the left nav, three other disparate boxes on the right nav, the Search box, a promo box in the middle, then a giant cluster of text at the bottom along with an ad. On my 20 inch monitor, I have to scroll to see the entire page. There's a lot of information being thrown at me on this page, and my eyes aren't sure where to go or why things are organized as they are.

Good general observations, and one of the reasons we need someone or someones to really take charge of that page.

SEARCH BOX

So my recommendations for the Search box:
- Lower it to at least 1em

- Remove or at least simplify the drop-down - and consider just having one drop-down (two is confusing - save that for Advanced Search)

Universal search is still being worked, but the goal is to get it working and to replace the clunky search box with a universal search box, which I think we all agree is what most people would expect.

 

LEFT AND RIGHT NAV

Definitely the Donate box should stay on the front page, perhaps moved to the bottom center of the page, above the legal disclaimers.

To me, all the other widgets on this page should be moved to sub-pages as they only serve to distract the user.

I am by no means a UI designer, so people should definitely take my comments with healthy handfuls of salt. I think the current three-column layout was a quick attempt to emulate the layout of the old site, but it's not optimal.  As I've thought about it over the last few months, I've come to think we should have an inverted "L" layout, like gMail, Amazon, eBay, and many other sites.  Search/login bar spanning the top, browse/quick link nav bars on the left, with site logo where they intersect.  All other content should be below and to the right of this frame; the right column should be eliminated if possible.  Disclaimers, licensing, legalese, donate links, could all be part of a standard footer.

Of course I'm not dong the work, so it's easy for me to spout off :-)

 

The six items in the "About the GCD" likely will only be of interest to more serious users (do you have stats on click-throughs from those links?); can they be moved to the "About the GCD" page? How many people send email from the Contact Us link, and might that be better handled as a web form?

I'd love to have the contact us moved to a web form with captcha.  But I don't imagine it's a priority at the moment.

 

Why are the four links in the Useful Links box useful? The English start page link is redundant (or the German link is redundant); can you manage those links more elegantly by using flags as many sites do? Why are Today's Covers and Covers to replace useful for site visitors? Those would only be interesting to GCD admins and indexers, it seems to me, and should be moved to a sub-page.

These are all good points.  Perhaps a login link could take logged in users to a second main page, with the indexer/approval links on them.  As you note, 95% of site visitors probably won't find those useful.  Current front page is probably too heavily designed around the indexing community rather than visitor-oriented.

 

MIDDLE OF THE PAGE

The ad in the middle of the page will only apply to a small subset of users. I know the site needs more web devs, but for the vast majority of users, this is just white noise. Can you recruit in other ways or in other places on the site? (Maybe just a highlighted box on the start page saying "Volunteer for the GCD"?)

The want of tech resources (i.e. people) is really slowing down our progress at the moment, so there is a plan to replace a lot of the current verbiage with some text that emphasizes our needs.  I think it's to be deployed shortly.  But beyond that I think there's a desire to recruit a few volunteers to handle PR, and one of their prime responsibilities would be to keep the front page fresh and relevant.  Your ideas here are very helpful, and are probably the direction we need to be looking at for that.

--
Lionel English
lio...@beanmar.net

[Jochen Garcke]

couple of short things:

Am 02.06.2010 01:55, schrieb Jason Sacks:
> Why are the four links in the Useful Links box useful? The English start
> page link is redundant (or the German link is redundant); can you manage
> those links more elegantly by using flags as many sites do? Why are
> Today's Covers and Covers to replace useful for site visitors? Those
> would only be interesting to GCD admins and indexers, it seems to me,
> and should be moved to a sub-page.

Today's Covers is after pending/editing queue and before advanced search
our most often accessed page. Arguably it counts accesses of further
navigation on these as well, but quite a few people use it just to look
at pretty pictures.
The covers to replace was there so that the daily covers one didn't feel
lonely.

We only need the link to other language pages, not the current one, so
yes, English and German are redundant and I intended to change that in
any case.

Flags don't work, countries don't correspond to languages. It is often
recommended to have the name of the link in the language it leads to.
E.g. google doesn't use flags either.

Jochen

[Jochen Garcke]

Am 02.06.2010 01:55, schrieb Jason Sacks:

> Yes, the Search box is embedded in a purple table, which makes it stand
> out from the page somewhat, but there is a lot of text even inside the
> Search box. For a first-time user, are the drop-down selections a help
> or a hindrance? How often do new users even use the drop-downs? To me,
> having two different drop-downs just adds to the confusion of the box
> and makes me scratch my head a bit about the results. I know I'm
> conditioned by the search engines to expect appropriate results when
> typing any term into a search box - whether typing "Thor" or "Perez" or
> "This Man This Monster" I would expect to get valid results back. So why
> the multiple drop-downs?

Because we need them ? Once the new search is implemented we can drop
the search type one. People use both alphabetical and chronological
search, when I suggested to only have alphabetical we got complaints,
would be the same the other way around I assume.

> Login is distracting as it adds two more blank text boxes to the page.
> Can the start page just have a link to log in, and then have the user
> log in on a separate page? How many users log into the site per month?
> Also, why is there no "remember me" checkbox?

Because it always remembers you ?

Best way probably would a javascript based login.

There still will need to be a direct link to the indexing functionality
from the main page.

Jochen

[Philip Rutledge]

On Wed, Jun 2, 2010 at 1:18 AM, Jochen Garcke <gcd...@garcke.de> wrote:

Am 02.06.2010 01:55, schrieb Jason Sacks:

> Yes, the Search box is embedded in a purple table, which makes it stand
> out from the page somewhat, but there is a lot of text even inside the
> Search box. For a first-time user, are the drop-down selections a help
> or a hindrance? How often do new users even use the drop-downs? To me,
> having two different drop-downs just adds to the confusion of the box
> and makes me scratch my head a bit about the results. I know I'm
> conditioned by the search engines to expect appropriate results when
> typing any term into a search box - whether typing "Thor" or "Perez" or
> "This Man This Monster" I would expect to get valid results back. So why
> the multiple drop-downs?

Because we need them ? Once the new search is implemented we can drop
the search type one. People use both alphabetical and chronological
search, when I suggested to only have alphabetical we got complaints,
would be the same the other way around I assume.

Ultimately, I would see handling user preferences like this via a user profile/preference or a cookie.  Set the default to alphabetical and allow people to change the results to chronological and "save as default" if desired.  Of course just having a single toggle option to flip the results back and forth (or a sort by on the column header) would probably make people just as happy.

In general, I think we can all see the value of a default universal search which when coupled with some additional sorting. column manipulation (add/delete different fields) and refine result options in the results views is likely to satisfy the bulk of the visitors.  (This would do away with the drop downs)  For the advanced users that really want the fine control of creating advanced boolean searches on specific fields (which I for one would use) then I think that should also be available.  I think the presentation would be similar to the google search we've all grown accustomed to, a single search field with a small "Advanced Search" link. 

 

> Login is distracting as it adds two more blank text boxes to the page.
> Can the start page just have a link to log in, and then have the user
> log in on a separate page? How many users log into the site per month?
> Also, why is there no "remember me" checkbox?

Because it always remembers you ?

Best way probably would a javascript based login.

There still will need to be a direct link to the indexing functionality
from the main page.

I love the direct link to my "Editing" queue from the front page however this goes back to the discussion of different user groups.  For the user that is just browsing they don't want to be distracted by the login fields and a javascript login would be great.  For the indexer then the site should recognize me and the "login" prompt becomes "Logout | Editing | Profile" etc and we could also customize the front page to provide more indexing related info, eg: site news, stats on pending approvals etc.

I guess I am advocating for having custom presentation based on who is accessing the site to optimize it for their usage.  However I know this is more work!

Phil

 

Jochen

[Philip Rutledge]

On Wed, Jun 2, 2010 at 12:52 AM, Jochen Garcke <gcd...@garcke.de> wrote:

couple of short things:

Am 02.06.2010 01:55, schrieb Jason Sacks:

> Why are the four links in the Useful Links box useful? The English start
> page link is redundant (or the German link is redundant); can you manage
> those links more elegantly by using flags as many sites do? Why are
> Today's Covers and Covers to replace useful for site visitors? Those
> would only be interesting to GCD admins and indexers, it seems to me,
> and should be moved to a sub-page.

Today's Covers is after pending/editing queue and before advanced search
our most often accessed page. Arguably it counts accesses of further
navigation on these as well, but quite a few people use it just to look
at pretty pictures.
The covers to replace was there so that the daily covers one didn't feel
lonely.

I do think that there is a lot of value in "eye candy" as a way of engaging new users and demonstrating the breadth of the content in the site.  

From a usability perspective I think you have two points of view to consider.  One is the searcher who arrives at the home page and wants to find something specific which is why putting the search front and center and making some of the enhancements mentioned in my last post are important.

The other user type is the browser or discoverer who doesn't know what they want to find and is looking to find "cool stuff" so one way of addressing this is to couple it with the site stats.  If we keep the stats as a major element on the front page a link to "browse our vast content" could take the new user to a view to see covers based on most recent, based on genre, character, company or based on date views (like the calendar view on dcindexes)

WRT to the "Covers to be replaced" I think these should be presented to the indexer user once they log in as part of their dashboard.

 

We only need the link to other language pages, not the current one, so
yes, English and German are redundant and I intended to change that in
any case.

Flags don't work, countries don't correspond to languages. It is often
recommended to have the name of the link in the language it leads to.
E.g. google doesn't use flags either.

Jochen

I remember some of this discussion on the tech list a while back but didn't have time to contribute at the time.  There are various best practices on I18N that deserve discussion which were started separately but I do think user preferences is the ultimate best way of addressing this coupled with IP address geo-location or browser preferences as a way of getting the best guess initially.

[Philip Rutledge]

On Tue, Jun 1, 2010 at 10:22 PM, Lionel English <lio...@beanmar.net> wrote:

Either I mislaid it, or Henry didn't introduce people in this thread--great insights from Phil and Jason so far, could you be kind enough to share your backgrounds with us?

Professionally, I've been a software developer/architect/manager for 20+ years primarily in the telecommunications, networking and wireless industries.  This has primarily been developing embedded code in C/C++ however I've also been responsible for the associated management systems on Unix/PC platforms.  Also for the past 10+ years most of my projects involve some sort of web interface for user access to capabilities or for administrative access and this (along with some PC app projects) have driven my investigations into usability and allowed me to work with some top notch usability consultants.  I've also been involved in creating third party interfaces into systems initially through proprietary APIs, TAPI and most recently REST based APIs.   As my career progressed and I moved more into management I focused more on project and process management and architectural discussions.  However I still dabble in code and I'm have some fun coming up to speed on Django and Python through the GCD and hope to be able to contribute some code eventually.

On the comic side, I've been collecting (or as my wife say's "hoarding" :-) ) comics most of my life.  I started cataloging my collection on recipe cards as a kid, then wrote TRS-80 Basic programs to track it and then moved through Lotus 1-2-3, dBase, Excel etc and now have a set of perl scripts and tools that let me generate the reports I'm interested in.

I've been an observer of the GCD since the 90's when I heard about it on rac.misc and used it as a reference tool but was always frustrated by the speed. I started using it much more last fall when the new servers were put in place.  Since I've been using GCD more and more as a reference in my collection tracking to find holes etc, I decided to jump into indexing and to get more involved.  I've watched the dozens of different comic database and tracking sites come and go over the years and was drawn to the GCD simply since it has been around the longest and is the largest.  There is a ton of great data spread across the net and I would love to see the GCD become equivalent to wikipedia in terms of awareness as the go to place for comic info.

Phil

[Henry Andrews]

Hi Jason,
I haven't had time to write up a full reply, but this is definitely the sort of feedback and discussion material we're looking for. The details are good! We'll pull high-level stuff out of this later. I'd love to see that mockup you mentioned, and don't worry about being "presumptuous". The current site UI was a prototype I hacked together to demonstrate the viability of Django as a development platform. It was not actually intended for productions. And then we lost our old server, so... Anyway, I'm certainly not all that attached to any part of it.

At some point we can send a lot of this over to gcd-tech for immediate (more or less :-) work, and keep looking at the large implications for the site on this list.

thanks,
-henry

----- Original Message ----
> From: Jason Sacks <jason...@hotmail.com>
> To: gcd-software-committee <gcd-softwar...@googlegroups.com>
> Sent: Tue, June 1, 2010 4:55:25 PM
> Subject: Re: [gcd-software] You (yes, you!) have a homework assignment!
>

> I apologize for taking so long to join in on the discussions around here. Let me
> try to make up for that by giving my thoughts on the above, many of which will
> echo comments made by others around here, especially Phil Rudledge's comments on
> 5/27 on the User Interface.
>
> I'll start in this mail with the start page,
> then send further comments in future mail. I know this mail is awfully detailed,
> but I hope it's helpful.

[edit]

[Jason Sacks]

Either I mislaid it, or Henry didn't introduce people in this thread--great insights from Phil and Jason so far, could you be kind enough to share your backgrounds with us?

 

I started working professionally in the computer industry in 1999 at Microsoft, initially doing production work for the MSDN online and offline Quarterly Libraries. I spent the next seven years at MSDN, building, designing, managing and publishing millions of pages for the Libraries. The team did a lot of work around improving presentation for users, endlessly (it seemed) debating the best ways to deliver quality content to users. From there I moved into program management, working on the Windows SDK team (a merger of the old Platform SDK and .NET Framework SDK), delivering an integrated modular setup for the application and later managing tools and C++ headers and libraries for the SDK. In 2008 I left Microsoft to work for a small startup that delivered an online prescriptive guidance security system for users. I worked with some great UI designers and database experts to deliver a highly performant and targeted documentation system. Since leaving that company, I've returned to Microsoft, where I manage release of the consent decree documentation to satisfy the EU and DOJ injunctions against the company.

 

I've been reading comics since I was a kid, and have a collection of around 12,000 comics, graphic novels and other ephemera. One of the great moments in my collecting life was in buying a copy of ComicsBase, which allowed me to catalog my comics without having to manually enter them, and which added a real touch of sanity to my collection.

 

I am the editor in chief of Comics Bulletin, "the internet's most diverse comics webzine" and manage a staff of about 30 people on the site. That project's a lot of fun, and I really enjoy putting together our Top 10 lists. But if Program Management is all about herding cats, site management with a team of volunteers is a quantum level more difficult!

 

I've been using GCD for years and have been frustrated by some of the issues I've seen in it. After attending Henry's presentation about the GCD at Wondercon this year, I decided it was time to move from being a complainer to a problem solver. <g>

 

I'm married, with three kids (19, 17 and 10) and live in the Seattle area.

 

Jason

[Jason Sacks]

Hi all,

In an effort to continue this discussion with a "stalking horse" image, I
whiteboarded this very simple mockup of a potential new start page for the
site.

http://www.comicsbulletin.com/images/1006/photo.jpg

I realize it's very stark and simple, but it's also a clean and contemporary
style that I think will resonate with users.

What are the "must-have"s that are missing from it - and what did I get
wrong?

Jason

--------------------------------------------------
From: "Henry Andrews" <andrew...@yahoo.com>
Sent: Wednesday, June 02, 2010 9:54 AM
To: <gcd-softwar...@googlegroups.com>

[Jochen Garcke]

One concern with all these I have is that we are considering the main
page as a search entry page. But it is more than that. There is a
community of people behind it.

The other site people compare us with conceptually is imdb.com.
Or some think in the way of wikipedia.

That the GCD is loosing parts of it community feel and aspect was
mentioned by others on other lists, a bare bone entry page might be a
step in the wrong direction under these aspects.

I am not sure what is the better way to structure the front page, just
thinking out loudly.

Jochen

[Jochen Garcke]

On further thought, I think I might like your mockup as the top part of
our site with additional content below. Easy access to search and
editing, with more (e.g. detailed stats, last updated, what did we
change recently,...) further below with maybe need for scrolling.

Jochen

[Henry Andrews]

Thanks, Jason- this whiteboard snapshot is a great focal point for discussion. I'll respond to it, some of your other questions, and some of Phil's recent comments in this thread below. I ended up with a lot of comments, but this was really useful for forcing me to break down exactly what's on the front page and why.

General Front Page Stuff:
-------------------------------------
One concern that is also a problem with the current design is that I'd really like to see a graphic element on the site. Ideally, we should have a mascot character of some sort as that would be better than appropriating specific public domain characters. Kind of a side point, but one I wanted to make while I was thinking of it. Like I said, it's a problem now, but the sparse text design makes the lack all the more glaring.

Search:
----------
As Jochen pointed out, our search box is complicated because our search code is lame (I wrote it, so I can say so :-) Fortunately, as you saw today on gcd-tech Tim is making good progress with a new search solution. I'm not sure what the advanced search page will look like with that solution, but there will likely still be one. So I'd say the design you have for search is what we'll want, but not something we can do until Tim's Solr work is complete.

Login:
--------
It might take me a while to sell me on not having a login box up front. I see the design benefits. But for heavy indexing users, an extra click to login is annoying. I assume this is supposed to be done with a JavaScript panel? Probably the right option, as heavy indexing users aren't the people whose attention we need to attract, but it feels weird :-P

I *would* want the edit link to appear in that list if a person is logged in. Having the edit link on the front page was a recent and popular change.

Also, as Jochen says, the system always remembers you, so no point in having a checkbox. That's probably not really a good idea, though. I'm not sure how likely we are to be "attacked" as a result of someone with approver privileges leaving the public library computer logged in, but a little caution is probably worth implementing.

Statistics:
--------------
I also have trouble letting go of the detailed stats :-) I agree that one number is much more compelling than a suite of numbers. I'm not sure that a "Statistics" link is all that compelling, though. What about just making the "591,000 issues..." thing a link? Possibly too subtle, I suppose. Some sort of "Learn more!" tag? Hmm... not sure here.

Disclaimer:
----------------
There was a lot of debate about how prominent the disclaimer should be, and whether it should actually be near the daily covers link, as that's where you're most likely to accidentally stumble across porn. The legal disclaimer about copyright definitely belongs in (and is in) the footer. But the adult content warning, I think, needs to be fairly prominent. We don't censor or flag (because then you are liable for the decisions of what to censor or flag). So I think we need to be pretty clear about that. But I'm not too certain what the real requirements are here as I've usually tried to stay out of this area of discussion.

Help Wanted:
-------------------
I'd like to have some way to emphasize that we are looking for web developers as well as content contributors. It's not necessarily obvious to folks that we would love someone to show up and overhaul our UI design, for instance :-) I know it targets a very small percentage of visitors, but it's very important (to my long-term mental health) that we maximize our chances of attracting the attention of those visitors. Maybe that doesn't need to be a permanent part of the main design, but we really need it right now. We attract data contributors naturally through the "Login to Edit" buttons. We don't have any such way to attract developers / designers.

Community:
-----------------
What's behind this link? What makes the casual user interested in clicking on this?

"The GCD is a non-profit..."
----------------------------------------
Would you believe that's an *edited down* version of the original statement? I basically snagged the first line of the Charter after Jochen complained that I removed everything that explained who we were. I did it rather quickly and I'm sure we can fashion a more concise version.

Colors:
----------
(Not on the mockup, but you asked before): Purple and gold... well, on my Mac, it's a much softer blue with purple tinges. I notice in Windows it's a rather vivid blueish purple. In the grand scheme of things when we were frantically trying to get the prototype deployed, messing with the color scheme just wasn't on anyone's radar. That particular shade was from me adjusting the blue that the old site had to something that I thought looked better (on my monitor). As I've mentioned, this UI was never actually intended for production, so there wasn't much thought put into colors.

As for why the gold? It stood out from the purple. That's all. I think someone said that the Advanced Search link was too easy to overlook, and we really wanted to promote it, so I picked some high-contrast color. By that time, sleep-deprivation was no doubt part of the equation. This is also why the OI color scheme is plain white, with black lines and a gray control bar, and all highlighting is done with CSS "yellow". First thing that came to mind, that's all :-)

Browse:
-----------
Phil mentioned a browse option. This was actually in the design proposal for the New Fun release. Although I think it was an optional part of that UI (the doc in question is at http://docs.comics.org/wiki/Release_New_Fun if anyone's interested- it was out of date even before the server crisis as we'd shifted to a database-centric approach and I hadn't gone back to update things, but it's somewhat interesting as an historical artifact).

comicbookdb.com has a simple by-letter browse system which is pretty typical of any site with structured content. I'd intended something vaguely similar.

OK, enough from me for now. I'll write more on the "different behavior for different people" idea from Phil soon, hopefully. I'm trying to go through the messages so far and collect things but people keep writing cool code for me to review :-)

thanks,
-henry

----- Original Message ----
> From: Jason Sacks <jason...@hotmail.com>
> To: gcd-softwar...@googlegroups.com
> Sent: Wed, June 2, 2010 1:09:24 PM
> Subject: Re: [gcd-software] You (yes, you!) have a homework assignment!
>

[Lionel English]

I think there might be some benefit to having two front pages.  A plain vanilla one like Jason suggests, and another for people who are logged in, with more "useful links" and less "about us".  Community members will normally be logged in, so they'd see the "community" front page most often.

On the other hand, I've long fantasized about a blog-like front page where content is rotated every couple of days and new blurbs highlight various creators/publishers/series/issues/runs/roles that are of historical or personal (to the blogger, whoever they may be) interest.  Things that highlight the depth and breadth of the database, rather than looking at current/upcoming stuff, which is the focus of so many news sites.

The advantage of Jason's design, of course, is that it would not require updating often.  Whereas any design that features more content (especially mine) probably would require freshening that content on a regular basis.

--
Lionel English
lio...@beanmar.net

[Philip Rutledge]

On Thu, Jun 3, 2010 at 8:28 AM, Lionel English <lio...@beanmar.net> wrote:

I think there might be some benefit to having two front pages.  A plain vanilla one like Jason suggests, and another for people who are logged in, with more "useful links" and less "about us".  Community members will normally be logged in, so they'd see the "community" front page most often.

This is definitely the way I was thinking as well.  

A generic front page focused on the new user with the search box as the prominent element with supporting lower areas providing information on how to help the GCD, GCD background etc.  So the focus is on utility without logging in for the new user and providing information for the un-indoctrinated.   This is the way Jason's mockup and the discussions have been heading

For the indexer (and perhaps slightly different for the editors) I think a dashboard type scheme would be useful.  You are presented with modules that have your editing queue, your pending approval queue, the most recent updates, saved searchers, site alerts, site news, community news, stats etc.  One of these modules could be the blog like content Lionel suggests.  

Editors would have access to different "modules", mentoring, approval queue etc (I'm not an editor so I don't know all of the possibilities)

For users that have accounts but don't index they could minimize/remove the indexing centric modules but still use saved searches, get the site news, stats etc

I think of sites like iGoogle, Netvibes, Jira etc with their library of modules that can be arranged on demand by the user.  Perhaps OpenSocial could form the basis of this?  Benefits would be that it could perhaps be built on the same web apis used for other purposes.  Also it might allow others to contribute "modules" and that modules could be created independently over time.

Not sure what open source resources are available for creating these sort of custom dashboards but I've seen enough of them around the web that I'm sure their are frameworks that can be integrated in.

Phil

[Jason Sacks]

I agree, that's an outstanding idea. Give the logged-in editors and indexers appropriate widgets to manage their own resources in their own custom landing page while giving the non-logged-in users the more generic start page. We tend to follow that approach at work and I love hitting the "Hello Jason" pages that give me all my information at a glance. It's also intuitive for users, since we all have experience viewing those widget-based pages.

 

The whole idea is to get people quickly and directly to exactly the resources that they need most, and this idea accomplishes that beautifully. And if we can use reusable widgets for the project, it would be a low cost/high gain project.

 

Jason

From: Philip Rutledge

Sent: Thursday, June 03, 2010 9:54 AM

To: gcd-software-committee

Subject: Re: [gcd-software] You (yes, you!) have a homework assignment!

[Henry Andrews]

While I'm sure this is, and will continue to be, a source of irritation to a number of users, folks seem to have more-or-less adjusted their workflow.  I would tend to rate a number of structural data problems and basic usability limitations as more urgent than this workflow issue.

thanks,

-henry

[Jochen Garcke]

At least when generating a series one should be able to generate the
issue records at the same time, this is the most common and most
annoying part when entering new series.

Jochen

Am 06.06.2010 01:31, schrieb Henry Andrews:
> While I'm sure this is, and will continue to be, a source of irritation
> to a number of users, folks seem to have more-or-less adjusted their
> workflow. I would tend to rate a number of structural data problems and
> basic usability limitations as more urgent than this workflow issue.
>
> thanks,
> -henry
>
>

> *From:* Lionel English <lio...@beanmar.net>
> *To:* gcd-softwar...@googlegroups.com
> *Sent:* Tue, May 25, 2010 6:52:18 AM
> *Subject:* Re: [gcd-software] You (yes, you!) have a homework

[Alexandros Diamantidis]

* Jochen Garcke [2010-06-06 23:45]:

> At least when generating a series one should be able to generate the
> issue records at the same time, this is the most common and most
> annoying part when entering new series.

The way I read Henry's reply is not that we should forget about fixing
this, just put it on hold for the moment.

For what it's worth, I too find it annoying not being able to enter
data immediately, but also agree that issues like deletions and moves,
which affect our ability to correct mistakes in our data, should be
given higher priority. On the other hand, a more convenient workflow for
indexers is important, since it reflects directly on the effort needed
to submit data, an important factor in attracting new contributors who
aren't used to the quirks of the existing system.

Thinking about how this could be solved, I suppose it wouldn't be too
difficult: how about adding extra fields to revisions pointing to other
revisions? For example, for the case of adding a series together with
its issues, the issue revision objects could have a field pointing to
the series revision that will create the series. Then, have a mechanism
that forces approvers who want to assign the issue changeset, to first
consider the series changeset and when that is approved, all references
to the series changeset will be converted to references to the newly
created series.

Similar provisions can be made for publishers, brands, and issues.

Does this seem workable? Or will it lead to unmaintainable spaghetti
logic in the code?

[Henry Andrews]

There are a couple of approaches. It's fairly complex, and it's really not what this group is for. I would be in favor of attacking this problem after deletions are implemented, after the underlying designs and core implementations for moves and complex fields (like reprints) are complete, and after the revision history and member accounting code (of critical importance for voter eligibility) are all complete. As I outlined on gcd-tech, those are the priorities that I see right now.

If folks want to discuss the design for this, it should go on gcd-tech rather than here, but I would respectfully request that no one start this thread now, as I am having trouble keeping up with the discussion on the complex design decisions we need to complete for our top priority items.

thanks,
-henry

[Alexandros Diamantidis]

* Henry Andrews [2010-06-06 16:06]:

> I would be in favor of attacking this problem after deletions are implemented, after the underlying designs and core implementations for moves and complex fields (like reprints) are complete, and after the revision history and member accounting code (of critical importance for voter eligibility) are all complete. As I outlined on gcd-tech, those are the priorities that I see right now.

Right, all of the above are indeed much more important. I was just
thinking out loud, but I agree, this is not the place nor the time.

Thanks,
Alexandros