BrickLink Discussion Forum

In Suggestions, revfds writes:

I really wish I didn't have to click into a specific order to see the tracking
#.

It would make a lot of sense, and be really helpful if there was a column on
the Orders page that listed any available tracking number along with the order.
Bonus points if clicking on the tracking number takes me to viewing the tracking
history (via usps website or google, or wherever).

Hmm you have a point, the tracking number field can be switched on and off for
the seller's Orders Received, but oddly it cannot be switched on for the
buyer's Orders Placed list. No idea why not, it could be added.

Also, incentives should be made to encourage tracking numbers. 1st class postage
with tracking can be had for around $3 in the US for small items, most places
have charged me at least that regardless of whether they include tracking or
not. It should be prioritized, and required for purchases over a certain amount
as it alleviates a lot of headache and prevents a lot of bad interactions.

Definitely no. Now you're just talking about America, over here I strongly
prefer shipping without tracking and I wouldn't let a site force my hand
because it has rules that are based on how things work in America.

Author:

Teup

Posted:

May 28, 2020 12:57

Subject:

Re: Why are these Hinges?

Viewed:

29 times

Topic:

In Catalog, SylvainLS writes:

In Catalog, Teup writes:

[…]
As far as I can see I have understood everything you said but it seems that you
honestly misunderstand what I try to say.

What I understand is that you’re sometimes conceding there are problems and then
you go harping on again about how their identified¹ causes are only theoritical,
logical or conceptual considerations that doubtfully cause real problems.
“Oh yes, those are problems. But causes are not problems. And these causes
are only conceptual. So there’s no real problem.”
Guh.

Inconsistency is a no-no in categorization theory. It doesn’t mean the problems
caused by inconsistencies are theoretical.

¹ Yes, the causes are identified:
If the problem is “I can’t identify this part” (which is only one example of
the problems we have), it’s because “I” (the “I” who has the problem, not necessarily
me) looked in categories (plural, “I” is not obtuse and knows they can be wrong
or confused) “I” thought the part would be. And “I” think a part would be in
maybe this or that category because of the names of this and that category and
because the part “I”’m trying to identify looks like the parts that are in this
and that categoriy, or because the part is used like (*brr, shivers*) the parts
that are in this and that category.
So “I” was misguided and confused by the category names and the parts that are
in the categories, by what “I” thought the parts shared.
Same (or worse) with browsing by category: “I” look for parts like such other
parts in the categories these other parts are in, if the parts “I” should see
aren’t in these categories, it’s because “I” didn’t get what the categories “meant”
(what their names meant, what features the parts that are in them share).
Causes: names don’t mean what “I” think they mean, parts don’t share what “I”
think they share.

And what the current project is doing is only addressing the names of the categories,
and then, only by tweaking their definitions, not by changing their names (or
maybe just a little, eventually, if really necessary) and certainly not by regrouping
parts and identifying why people are looking in the wrong categories.
IOW, they are writing down the exceptions to the definitions so that parts will
only be in one category. But if “I” don’t learn these definitions by heart,
“I”’ll still look for them in the wrong places because the names will be the
same and the parts will be in the same place.

Changes to the categories and parts is always on a case by case basis. Sometimes,
one person proposes to split a category or move a couple parts. And then people
discuss this little bit of the catalogue and, maybe, the consequences on another
little bit.
No global view.
“Rearranging chairs” as mfav once put it (well, more than once actually).

[…]
And it seems the story is inconsistent on categories like hinges and boat will
or will not change - right now it's a bit of eating the cake and having
it too, because on the one hand everything will be logically restrucutured but
on the other hand it will not result in major mismatches in jargon or catalogs
across platforms.

It’s simply because we don’t know: we haven’t listed the attributes, we haven’t
prioritized them, we haven’t defined the categories.

You didn't say it would cost nothing, but still, don't assume you oversee
everything. There are a lot of implications for a lot of people. These need to
be assessed.

But we can’t know. The only question we can ask now is “would you be favourable
to a change in the categories.” But we can’t say how important the change.

100% agree that you guys made a very good analysis of what would be the reasons
behind it if there are problems.
Also agree there are these problems.
Not yet sure about the magnitude of the problems, ranging from on the
one hand 'a few guys and gals a day not getting to the part the first try
but reaching it the second time and then memorising it for the next time with
no further difficulty' (the best scenario) to 'lots of people constantly
confused, giving up on placing orders and throwing heaps of parts in the will-do-later-aka-never
bin' (worst scenario)

And yeah, the implications of the solution also depend on how dramatic the rearranging
will be, again ranging from on the one end '99% the same result as we have
now but with perfect definitions' (62Bricks also mentioned this is still
a possibility) to 'forget everything you thought you knew about lego'

on the other end.

That's the prize dimension and the cost dimension.

If changes like this would ever be in the works for real, I guess I would need
to wait with forming an opinion until more would be clear about where we are
on these two scales. I want things to be within reason and proportion (because
there are consequences beyond just getting used to new things), but I would surely
give it an honest chance.

Author:

Teup

Posted:

May 28, 2020 10:21

Subject:

Re: Why are these Hinges?

Viewed:

35 times

Topic:

In Catalog, SylvainLS writes:

In Catalog, Teup writes:

[…]
OK, you're right, here is a problem. If you have difficulties, then
that is certainly important.

There you go again, mocking the reality of problems with the categories because
I avoided repeating (again) the list of troubles people have by summing them
up as “difficulties.”

As far as I can see I have understood everything you said but it seems that you
honestly misunderstand what I try to say. Maybe that's just my fault because
it seems I can't express things well in English, but anyway, you don't
understand what I mean. Wasn't belittling anything, it was the opposite.
I was in fact pointing out that that is the major deal - the difficulties
are THE important thing in all of this (no idea that word was a euphemism, my
bad). It seems to me that you don't see that pointing out an inconsistency
on an intellectual level and figuring out whether or not it is a problem are
two separate steps, it looks like you see it as one automatic, impartible step.
The first can cause the second, but it is not a given - you start with the problem,
then trace its causes.

I think your talents and attention to detail are invaluable for improvements
once those are planned, but I think you seriously miss the point on some things
in the greater picture, and I don't manage to make you see it. Which is OK,
because if there will be changes, it is going to require a combined community
effort anyway.

And it seems the story is inconsistent on categories like hinges and boat will
or will not change - right now it's a bit of eating the cake and having
it too, because on the one hand everything will be logically restrucutured but
on the other hand it will not result in major mismatches in jargon or catalogs
across platforms.

You didn't say it would cost nothing, but still, don't assume you oversee
everything. There are a lot of implications for a lot of people. These need to
be assessed. For me: Rearranging 1 million parts because of a new catalog, sure,
it's a business risk - give me a week and I will happily do it (this is NOT
a strawman - ALL bins will move). Or writing software so that I can keep selling
on both platforms with the same system, sure I will do it. Each time I add parts,
I am going to have to download my inventory, run it through some software to
generate the categories that match BO, then feed it back into BL in XML mass
inventory update chunks (also NOT a strawman, believe me, I know my systems).
Extra work every time, but I will do this. (I have no idea how others will solve
it. tell me if you do)
But I will do it because it improves the user experience for many people in terms
of measurable UX quality. I am not enthusiastic about complicating the tools
I constantly use every day because some users just feel good about logical systems
for their own sake. We need a survey, then I'm going to be impressed.

Author:

Teup

Posted:

May 28, 2020 04:47

Subject:

Re: Why are these Hinges?

Viewed:

28 times

Topic:

In Catalog, StormChaser writes:

In Catalog, Teup writes:

We need the community.

We do. And one of the problems there is that historically much of the work on
BrickLink's catalog, from major decisions to the most minor contributions,
has been done by an incredibly small number of people compared to the overall
membership of BrickLink.

If you have suggestions regarding how to engage a larger portion of the community
in catalog decision-making, we will gladly listen.

Good question. I feel like a survey could be useful. If the Bricklink main page
would display it in that big banner, I would think a lot of people will participate.
(I think it's more likely to attract people who are dissatisfied with the
category than people who are satisfied, so that would be erring on the safe side
for those who want to go and change things.)

If this would happen, I'd be happy to help putting together the questions
- I'm thinking a handful of statements with a 5 or 7 point "strongly agree
- strongly disagree" scale setup, and an open question to gather some input on
possible consequences that weren't thought of yet.

Author:

Teup

Posted:

May 28, 2020 03:57

Subject:

Re: Why are these Hinges?

Viewed:

37 times

Topic:

In Catalog, SylvainLS writes:

In Catalog, Teup writes:

In Catalog, SylvainLS writes:

In Catalog, Teup writes:

[…]

“Yet”

Not in a temporal meaning

Come on, adverbs can refer to either time or space, you're a language guy,
you must've figured this "yet" referred to something a bit further on in
the context and not to "now" vs "another time"

I still don’t get the original sentence in any other way than temporal. I don’t
see it applying to something placed further (and later) in the message.

Aw man really?

Here we go:

Saying "the problem is that categories are inconsistent" to me isn't good
enough, because it is not in and of itself a problem yet. Having multiple options
for categories
to put a part in, is not a problem yet. Things like people searching endlessly
or giving yup, yeah, those are problems.

You're a smart guy so you must see the point I keep trying to make..... that
things are not problems just because they violate how we imagine things.. if
I put 10 pens in a cup and one is the other way around from the other 9, you
can perceive it as a problem because it violates some conception of how it "should"
be. But is it a problem in the cup or in the mind? Only if we're talking
about sharp knives, is that really a problem. I'm interested in results
of the way the catalog is, not in intrinsic properties.

Something that upsets people is a problem.

(...)

Do I have a problem with the categories as they are?
Yes.
Proof: I say so

Surely we are changing the catalog to improve its practical usefulness, and not
because its intrinsic properties are upsetting?

Streetmakers don't lay
a street a certain way to avoid people with OCD being upset with how the bricks
are distributed, they lay it so it's comfortable to walk on.

Do other people have a problem?
Yes.
Proof: they say so in forum posts, either directly saying so or having difficulties
with the categories.

OK, you're right, here is a problem. If you have difficulties, then
that is certainly important. (I assumed you were one of those veterans who knew
the catalog by heart.)

Do admins see a problem with the categories?
Yes.
Proof: https://www.bricklink.com/message.asp?ID=1117709
and in https://www.bricklink.com/message.asp?ID=1190847
“””
Do the definitions do their job?

Imagine a part you'd like to see placed into another category. Do the definitions
of the existing category for the part and the category where you'd like to
see the part moved fully explain the necessity for the move?
”””

Their solution is to make a very, very long dictionary (that no one will ever
read and even less learn but them¹) with lots of convoluted phrasing and exceptions
because they don’t want to touch the categories themselves, yet (in the temporal
sense

).

(¹ It’s okay, it’s their main goal: guidelines for the catalogue to make it less
person-dependant. But it can’t really serve as a guideline for simple users.)

The approach for now has been cautious: move parts on a case by case basis, try
to empty the (Other) category, create a few new categories.
Patches, patches, patches.
Plaster on a wooden leg.

But I get it, that’s not enough people, you want a “survey.”

If this plaster on the wooden leg is cause for reduced ease of use (which is
plausible, I just want to know how if it is a small or a big problem), then it
is a problem. If it does the job, then all of what you mention above is fine.

[…]

You had and have no problems but you made your own categorization for your shop….

No, I follow Bricklink's. My 2 Boat bins and my 4 Hinge bins are right here
behind me.

“I've already built my own catalog for my own shop which I can apply to my
Bricklink
shop by automatically assigning my categories to the remark fields.”
— https://www.bricklink.com/message.asp?ID=1200018

Exactly. If things change, then I could create a way of keeping it the way it
is for myself, so this is not going to affect me.

[…]
I just need to see that either A. the solution comes
at no cost,

There’s always a cost.
For the ones who’ll do the job.
For the ones who know the old tricks and will need to learn new ones.

You're a really thoughtful and meticulous guy. But this is almost going over
things quick and carelessly. Seems the story is the catalog is all bad, it
needs to be rewritten because well, using logic you can see it without needing
a survey, and it doesn't really cost anything, just gotta relearn it.
Hey, slow down man! It's great you're contributing to improving things,
but please, please, please let's be a little bit more meticulous in each
of these steps. The catalog is not a brain game. We are talking about things
that affect thousands of people, people for who LEGO is a serious passion and
people who depend on this system for their families livelihoods.

The cost part: Yeah relearning, OK. Now let's think a little bit deeper.
Regardless of what some people say bringing up details, the BL and BO catalogs
are incredibly similar (the hinge and plate modified categories are the same,
so is boat). There is also a third platform that too has the same catalog, and
possibly more (I also see the terminology used outside of it). Changing it means
for example these two things:

- We move further away from a universal or at least mutually intelligible language
for Lego parts. With all platforms having fundamentally their very own way of
classifying things, we're going to have islands rather than 1 strong community.
People will be more focussed on 1 particular platform. That essentially reduces
our power. Also, in the flesh world, people use the English Bricklink category
names in Dutch, and I'm sure that happens all around the world. It's
become jargon.

- Category based sellers are going to be in trouble selling on both platforms
because the picking lists will be inconsistently ordered. I can work around it
because I can do programming, but if I couldn't, I am not sure if I would
want to continue selling on both sites. Most sellers want to start on 1 end of
the stock, and end on the other. Running around in random directions with increased
chance of errors is not appealing.

Now, these things aren't holy - they can be given up. But they surely are
costs. These are my 2 examples, maybe there are more. Let's recognise them.
Putting it polemically: If we're achieving huge usability improvements then
let's do it. If we're achieving the peace of mind of a few people who
just like to see things in logical places when they think about it, at the cost
of sellers making mistakes running around with randomly ordered picking lists,
that is just not worth it. The reality seems to be somewhere in between, but
we don't know where.

IMO We need to know what we gain and what we lose before we would ever make fundamental
changes to the catalog. It takes research that we can't do in an armchair
with just a few people. We need the community.

Author:

Teup

Posted:

May 27, 2020 16:42

Subject:

Re: Why are these Hinges?

Viewed:

38 times

Topic:

In Catalog, SylvainLS writes:

In Catalog, Teup writes:

[…]

“Yet”

Not in a temporal meaning

Come on, adverbs can refer to either time or space, you're a language guy,
you must've figured this "yet" referred to something a bit further on in
the context and not to "now" vs "another time"

Those are problems now.

This is not a problem,

Indeed not. It’s an introductory sentence for the following sentences that explain
the problem.

No, really. It is too quick to immediately label things as problems without us
knowing anything about the effects. If a part could be two categories and it's
up to randomness in which category it belongs, then that in and of itself is
not the problem. The potential problem is difficulty shopping, but so far we
don't know anything how big that is. The catalog is for humans, not for robots.
You could look at a human language and spot tons of "problems" in it, yet they
don't prove problematic. Inconsistency in and of itself is not a problem,
we're surrounded by it. If the solution is possible and free, then sure.
If the solution is not free, then we need to be able to weigh the problem's
magnitude vs the sacrifice.

and there too people are asking. It's just within reasonable
limits. Is the BL situation within reasonable limits? We would need to test it.

If first time findability is a huge problem, then yeah, we need to put
all efforts in designing a new catalog and give that 100% importance. If it is
a small problem, then we can afford to balance it with attention to categories
that are nice sets of parts that are relevant to browse.

Then there’s no way to measure the extent of the problem to your satisfaction.

My satisfaction... wish it was everyone's satisfaction. I would expect
anyone interested in this topic would be interested in finding out if, well,
they're problems. Not liking it from a logical perspective isn't a problem..
unless we were gonna print the catalog and hang it on a wall to enjoy it. I think
we really need to research this. Bricklink could easily create a survey about
this and then we would know a lot more. I don't really care about how many
things we could think of that "don't make sense if you think about it". I
just care about the shopping experience.

I know, just saying it worked very conveniently for me. Sure, you're right,
a lot of those parts can very well be in other categories. But I had no problems.

You had and have no problems but you made your own categorization for your shop….

No, I follow Bricklink's. My 2 Boat bins and my 4 Hinge bins are right here
behind me.

As long as whoever fleshes out the improvements says that they keep a balance
of those two perspectives in mind, then I really am fine with any outcome. […]

Then you’re fine because we’ve said it from the beginning.

I never heard anyone about creating meaningful categories because there are no
examples. So I have no idea if that example would be a strawman. But sure, in
that case we're cool. I just need to see that either A. the solution comes
at no cost, or B. that its cost weighs up against the magnitude of the problem
- for which we need a survey.

I really hope Bricklink can make this happen. Not because I want to prove any
particular point, I am just genuinely interested in the catalog user experience.
If that would turn up a great degree of dissatisfaction then drastic changes
will probably receive wide support, also from me.

Author:

Teup

Posted:

May 27, 2020 11:34

Subject:

Re: Why are these Hinges?

Viewed:

39 times

Topic:

In Catalog, SylvainLS writes:

In Catalog, Teup writes:

[…]
I won't deny there are these problems, but I would really like to have some
kind of data, say a survey, so that we have something to talk about. Saying "the
problem is that categories are inconsistent" to me isn't good enough, because
it is not in and of itself a problem yet. Having multiple options for categories
to put a part in, is not a problem yet.

“Yet”

Not in a temporal meaning

Those are problems now.

This is not a problem,

People can’t find parts without looking everywhere now.

this is. And I would like to know more about the size and the nature of this
problem. We're now concluding things are problems because it sounds unsatisfactory,
and that's not enough (it always will be).

Parts are put into categories (and sometimes moved from one category to another)
arbitrarily (on whims) now.

This is not the problem. The problem is the implicitly assumed result: Difficulty
finding things. And that I just want to know more about the problem, not by reasoning
about what logically are problems, but by measuring the actual
problem.

The supermarket example is not just an analogy, it's an example of another
business where categorising is an issue. Is it an issue? Yes, because there too
people are searching, and there too people are asking. It's just within reasonable
limits. Is the BL situation within reasonable limits? We would need to test it.

If first time findability is a huge problem, then yeah, we need to put
all efforts in designing a new catalog and give that 100% importance. If it is
a small problem, then we can afford to balance it with attention to categories
that are nice sets of parts that are relevant to browse.

[…]
I'm responding based on my buying experience, which was the very first years
of me being a (casual) member (pretty soon I realised it was a too expensive
hobby for me

But LEGO is all about using parts for anything and everything.

“Boat stuff”? Parts for a boat? What’s that? All the parts you can use to
build a boat (bricks, plates, etc.)? They are all in different categories now.
The giant boat hulls? The sails? The oars? The windscreens? The porthole
windows? The hatches? The rigging? They are all in different categories now.

You’ve never been able to find or browse for all the “boat stuff” in one category.

I know, just saying it worked very conveniently for me. Sure, you're right,
a lot of those parts can very well be in other categories. But I had no problems.
If others do, then sure. It's just that in my work I have encountered so
many people saying "well I don't have a problem, but probably other people
do" and then it turned out those people never existed. That is why I am this
kind of skeptical and want to hear it from the people who have the problem. The
signals I'm picking up are limited, but they're more "just take a moment
to learn it" than "omg the BL catalog is so difficult".
And this discussion all started pretty much with everyone in agreement that those
items are not hinges because they have clips. I don't think all is terrible.

I just want
a buyer-friendly catalog that takes into account but not overfocusses on first-time
identification of a part, but also on producing relevant categories that are
pleasant and supporting to browse around. If a more principal system could still
do that, then that's alright.

At least, it shouldn’t prevent that anymore than the current unprincipled system
does

It would if it would be something too technical like based on dimensions
only. It is not a guaranteed improvement - there is also a lot to lose here,
and we forget that if we bash everything about the catalog. We just need to be
careful.

Well, let's wrap it up. There are lots of things that could be both better
or worse about the catalog. The reason I'm posting is that all the time I
see it is being talked about from the perspective of the part - where
it is and if it is easy to find. But I never hear about the perspective of the
category - what assortment of parts it contains and if that is a meaningful
set of parts to browse and if it is nicely sized (this is definitely not perfect
now, but it could be worse). The tag system is a good example too: While it is
a good idea, it's exemplary of the 100% focus on that individual part and
how that can be found, but it says nothing about what browsing a store will look
like (you will see the same things many times if they don't have 1 category
location) or what storage is going to look like (both sellers and buyers may
adopt the categories).

As long as whoever fleshes out the improvements says that they keep a balance
of those two perspectives in mind, then I really am fine with any outcome. I
just don't want super-easy-to-identify categories if they are going to be
"everything with 3 studs, everything with 2 clips, .." etc. In that case I prefer
a common-sense hinge category with 3 parts you expected elsewhere that you just
need to memorise. But if it will be better, then go ahead. Maybe we end up in
total agreement once an example is put forth, or when we have data on the problem,
so I'd say let's continue this when it's less hypothetical

Author:

Teup

Posted:

May 27, 2020 09:29

Subject:

Re: Why are these Hinges?

Viewed:

49 times

Topic:

In Catalog, SylvainLS writes:

In Catalog, Teup writes:

[…]

What other aspects? You still haven’t given examples.

Well here's what comes to my mind right now.

Meaningful categories: You could sort a supermarket based on size of the products
of based on alphabet, and it would be extremely easy to find a cauliflower. But
the aisles would not have meaningful categories. The result is that it is easy
to find a single product while at the same time browsing the store is very inconvenient.

As you define “meaningful,” shape is meaningful and function isn’t, or is less
meaningful.
Function is like alphabetical order: you need prior knowledge, arbitrary prior
knowledge, prior knowledge that you can’t guess by looking at a part.

Navigability: You can know where a part is, but if it would take many clicks
or staring at the screen, it is still inconvenient to use. For example if the
system would have many subcategories, which we don't at this point, or if
the category would be so large that it has many pages, which we do have.
[…]
So what I'm saying is that these are separate interests, that can sometimes
can conflict. […]

That’s what the prioritization of the attributes does.
You first find the attributes, then you prioritize them to have categories of
reasonable sizes while keeping simpler, more basic, attributes first (because
they simpler, more direct, to see in a part), and avoid conflicts.

It’s still arbitrary but systematic, needing less arbritrary / specialized prior
knowledge.

The problem with the catalogue now is that it’s inconsistent: there’s no system,
or rather, there are several conflicting systems. Some categories are shape-oriented,
others are function-oriented. Some categories are almost empty, others are full
to the gills.
When a new part arrives or when you’re searching for a part, there often are
several categories in which it can fall, because they are orthogonal categories
(one is shape, one is function) or because there’s no priority (the part mixes
two shapes with the same priority (is 43093 a pin with an axle or an axle with
a pin?)).

What the current project is trying to do is to first refine the categories’ definitions,
but not the categories themselves, hoping it will allow to define clear boundaries.
It might solve the Plate vs. Tile problem but it won’t solve the problem that
a hinged-plate is both a plate and a hinge, it will just put a lot of exceptions
in the definitions and the need of prior knowledge to prevent putting it and
looking for it in the Plate, Modified category.

We already know most of the attributes because they are already used in the category
names: brick, plate, round, hose, decorated, etc. There even is a prioritization
(brick ≫ round ≫ decorated).
What should be done is determine which attributes conflict and need prioritization
or deletion. That means the attributes need to be defined (what’s “hinge”?
it’s a function, it’s orthogonal with primary shape-attributes like brick and
plate, how do we prioritize them?) not the categories (what parts are in “Hinge”?
oh, they could also be elsewhere, let’s massage all the definitions to keep
(most of) them there).

By the way, I also think we mustn't forget which problem we are trying to
solve. Is it about solving practical problems with its use, or is it about solving
our feeling of dissatisfaction on some intellectual level? I feel a bit of both
are involved.

True.

So how big are the practical problems with the catalog really?
I'm honestly curious. I wonder what info we have that suggests dramatic overhauls
are needed. You'd have to watch people use Bricklink in order to find out
I guess. But if I had to go by the number of ID topics, the people who ask are
vastly outnumbered by the people who answer.

Not true. It’s always the same people who answer.
(Old timers who know the parts, and how they are arranged in the catalogue, or
have developped tricks to circumvent its inconsistencies, or have simply memorized
those inconsistencies.)

But of course, that doesn't count the times someone had to search for too long.

Or abandoned and didn’t come to the forum to ask, putting the part in their “what’s
that?” bin for later (a.k.a. never).

I won't deny there are these problems, but I would really like to have some
kind of data, say a survey, so that we have something to talk about. Saying "the
problem is that categories are inconsistent" to me isn't good enough, because
it is not in and of itself a problem yet. Having multiple options for categories
to put a part in, is not a problem yet. Things like people searching endlessly
or giving yup, yeah, those are problems.

Supermarkets btw have the same inconsistency, the categories are also a hybrid
there. It's not the end of the world (but sure, also not an argument to defend
inconsistencies)

I'm responding based on my buying experience, which was the very first years
of me being a (casual) member (pretty soon I realised it was a too expensive
hobby for me

). I would look around thinking "so what kind of boat stuff does
this shop have?" and it would not have been convenient for me if that whole category
was broken down into a too technical approach to part attributes - but yeah,
that is going to depend on what precisely the proposed catalog would look like.

As for my selling experience, I don't care either way anymore, because I've
already built my own catalog for my own shop which I can apply to my Bricklink
shop by automatically assigning my categories to the remark fields. I just want
a buyer-friendly catalog that takes into account but not overfocusses on first-time
identification of a part, but also on producing relevant categories that are
pleasant and supporting to browse around. If a more principal system could still
do that, then that's alright.

Author:

Teup

Posted:

May 27, 2020 09:01

Subject:

Re: Why are these Hinges?

Viewed:

40 times

Topic:

In Catalog, 62Bricks writes:

In Catalog, Teup writes:

That is an extreme example, but it is a good point. So the conclusion might be
that "First letter of the common English name" is not a suitable attribute for
grouping items. It tells you nothing about the item itself and so leads to categories
that are not meaningful.

And while size is a physical attribute of the item itself, it is not meaningful,
because people do not generally select food by size. It is in the selecting of
the attributes that the final groups are shaped.

What a supermarket actually does is a good real-life example of what a part catalog
could do. It uses meaningful attributes and the end result is meaningful organization.

Usually, the most general attribute is whether the item is fresh or packaged.
Produce, meat, flowers, fresh-baked goods all have their own sections separate
from the packaged goods.

The next level of attribute you find applied in, say, the produce section is
"Fruit or vegetable?" Fruits are grouped with fruits, vegetables with vegetables.
That is based on some foreknowledge on the part of the shopper of the difference
between the two, but it is common knowledge, not specialized knowledge.

Finally, an attribute based on how the item grows is used - roots, bulbs, vines,
leafy plants, trees, etc.

And so running everything through those levels you find the apples, pears and
oranges near each other. You find the turnips and potatoes together. Onions,
garlic and shallots are usually close by. Tomatoes and peppers are together.
You'll find the cauliflower by the broccoli, celery and greens, usually.
By choosing meaningful attributes you get useful and meaningful categories.

But now imagine the supermarket added a category called "Ingredients for pizza."
Now the top-level attribute is no longer a general one like whether it is fresh
or packaged, it is based on a specific usage: whether it can be used to make
pizza. Now the tomato sauce, cheese, flour, yeast and salt are pulled out of
their respective aisles, where they were previously grouped with like items based
on the general-to-specific model, and put together in a separate part of the
store.

Is that a meaningful category? It is for someone who wants to make pizza. Is
it meaningful for someone who wants to make pasta sauce?

That person might reasonably go the aisle that has all the canned tomato products
expecting to find tomato sauce. But it won't be there. To track it down he
has to know two things - this supermarket has a "pizza ingredient" section, and
tomato sauce is a pizza ingredient. That it is also an ingredient in countless
other dishes makes no difference, this supermarket has determined that "pizza"
is its primary purpose and so has stuck it away in a different spot.

He might ask the manager, "Why is your supermarket organized in such a strange
way?"

"Good point," the manager might reply. "Here let's fix that - we've re-written
our definition of 'canned tomato goods' so now it reads 'canned tomato
goods that are not also pizza ingredients.' We'll be posting this on
the bulletin board at the back of the store by the restrooms so everyone will
know where to find the tomato sauce."

True, even though actually they have such things - there is/was an isle for wrap
ingredients here. There's an isle for Mexican dishes, for Asian dishes..
and you're right that's where it gets tricky. Because jalapenos are in
the Mexican section and not in the pepper section with the chilis. Greek peppers
may be somewhere else. Conserved fruit cans are in the section with the conserved
vegetables, but conserved fruit cans that are intended for pies are in the baking
section. So yeah, this is definitely where it gets tricky.

Supermarkets are highly professional and adapted businesses where everything
down to what exact songs are being played is laid out to get the most out of
everything. If even they struggle with these issues, I think we are doing really
well here. Not as an excuse not to do anything, but having some dilemmas doesn't
mean we failed.

I do think some degree of less technical and more thematic categories are good,
even if it causes a bit of gray area. A supermarket couldn't do without a
baking section, or a section for different kinds of lunch snacks, a section for
stuff to put on a sandwich, etc. You're definitely right it shouldn't
go overboard. But I do think it's great we have a "windscreen" and a "panel"
section even if there are some panels that work as windscreens and some opaque
windscreens. The general concepts of windscreens and panels are meaningful when
you're building and shouldn't be tossed away because we can't reach
a 100% logical consistency there.

All in all, I think the important thing is that when a customer goes to some
aisle/category with a certain thought, they should be surrounded by items in
that area that are relevant and that allow them to choose or pick a few more
things. If you're thinking about putting egg salad on a sandwich and you
see there's also hummus and slices of this and that, that is really helpful
and you might also buy more. Whatever catalog system still provides categories
that are helpful for the buyer and give them some overview to brainstorm, is
fine with me.

Author:

Teup

Posted:

May 27, 2020 04:40

Subject:

Re: Why are these Hinges?

Viewed:

52 times

Topic: