Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-21 Thread Erwan


Hi Tobias,

Thank you for this, that helps me greatly with getting a more general 
perspective.


In fact, I originally developed the aggregator without thinking further 
than aggregating content for users to search in it. It's only recently 
that I slowly started realizing that the main idea can be used for 
various applications. So currently it's difficult to distinguish the 
parameters from the process, in particular in my code where some 
parameters are hard-coded, simply because they didn't make sense as 
"parameters" in the original application.


But even if it takes a long time, my brain is slowly integrating this 
potential for various applications. this is actually the reason why I 
decided to separate the aggregator /code/ from the community-search 
/wiki/: although the former is still 99% about generating the latter, 
it's now not the sole purpose anymore.


Regards,
Erwan



On 20/09/15 16:54, Tobias Beer wrote:

Hi Erwan,

First of all, thanks for your amazing contribution. Needs to be said, 
again, and again. :-)


There are many possibilities and I don't know what is best, so of
course suggestions are welcome!


In general, I would design that application / workflow in a way that 
everyone could use it in their environment just the way you do, 
however easily customized to personal preferences.


As I see it, right now, you the following workflow...

 1. you maintain some sort of "channel-list" / "subscription list"
 2. you maintain certain filters, either global or with respect to
every channel
 3. via commandline you batch-download all channels
 4. via commandline you extract a channel's data according to either a
global or per-channel filter or both into tid files
 5. via commandline you build an output wiki from those tid files
alongside a set of dedicated template components

All that is good and working. The required options would correspond to 
each step. Right now, you are possibly the only one who knows what 
needs configuration or what can be configured and how. Unless you 
already have, can start by publishing documentation on what you (can) 
do, step-wise, at the moment... and then we can give feedback on how 
that can be utilized or enhanced or whichever workflow to best 
construct around all that. Parameters should be simple text-based, 
preferably tiddlers. No fancy UIs or anything... except for what 
TiddlyWiki delivers out of the box. But maybe all that doesn't even 
need TiddlyWiki in the beginning, just plain text-editing.


One thing I would like to see around all this is the ability to adapt 
_the entire workflow_ to "collections", i.e.  being able to do 1..5 
for different collections of configurations rather than just one. So, 
for any "collection" you could do 1..5 independent from other 
collections, perhaps with a fallback to some global defaults.


Eventually, you may have...

 6. via commandline execute the update process for all collections 
at once


It would be good if we can reserve the term "collection" or something 
equivalent for meaning that: running the aggregator process against a 
distinct set of necessary configuration options. So, right now, your 
aggregator only caters for one "collection", being an 
"all-subscriptions", "all-channels" aggregator.


Although later on of interest, what I would not discuss or develop 
right now is some sort of community process / workflow to design 
"shared aggregator configurations". What I can perhaps see is that you 
wish to expose your configuration for "community search" or "community 
news" more so that others know where and how to contribute. I don't 
know if that needs actual development, rather than communication / 
documentation / a suitable workflow.


Best wishes,

— tb
--
You received this message because you are subscribed to the Google 
Groups "TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send 
an email to tiddlywiki+unsubscr...@googlegroups.com 
.
To post to this group, send email to tiddlywiki@googlegroups.com 
.

Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/5a9c2bfe-211c-4736-ace3-65e59fba7859%40googlegroups.com 
.

For more options, visit https://groups.google.com/d/optout.


--
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 

Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-20 Thread Erwan


Hi Tobias,

With Mat we had a discussion on github which is related to this kind of 
idea: https://github.com/erwanm/tw-aggregator/issues/77


It would indeed make sense to have this as a local application, but 
first I'd like to try something based on reading input parameters from 
the indexed wikis, and creating the collection/channel/customized wiki 
as usual, either as a specific tiddler in the community-search (so that 
it can be bookmarked) or maybe even as an independent wiki. The idea in 
this case is that many users can be interested in following the same 
"channel".


It seems to me that there is no big technical obstacle with this kind of 
feature/application, however the exact design is not clear at all in my 
mind: for example what would be the input parameters that a user can 
choose? set of wikis, tags, filters, ... How are these parameters 
represented? How is the result represented? As a tiddler in the 
CommunitySearch wiki (similar to the tiddler created for every tag), in 
another custom wiki... There are many possibilities and I don't know 
what is best, so of course suggestions are welcome!


Erwan


On 18/09/15 06:33, Tobias Beer wrote:

Hi Erwan,

I don't see what is missing in TW functionality for this kind of
scenario, it's already perfect :-)


You sure have done a great job here, no question. :-)

What I was suggesting was, that if this ever became some sort of 
desktop application,
it would be good to not only have a single "channel subscription list" 
extracted from tiddlers from which to construct,
but to rather be able to define "collections" or "interests", i.e. 
different sets of subscriptions for different purposes,
e.g. one for the latest "Plugin News" and another for "My Project Team 
Wikis",

which could, theoretically, even overlap.

Not sure if you're familiar with it, but think of "interests" in Facebook.
There you can add users or pages to special subscription lists for 
which you can then see a dedicated news feed.


Best wishes,

— tb
--
You received this message because you are subscribed to the Google 
Groups "TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send 
an email to tiddlywiki+unsubscr...@googlegroups.com 
.
To post to this group, send email to tiddlywiki@googlegroups.com 
.

Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/5f9a258b-99c7-4bd7-b482-029ab99f22a5%40googlegroups.com 
.

For more options, visit https://groups.google.com/d/optout.


--
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/55FEB11C.4020108%40gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-20 Thread Tobias Beer
Hi Erwan,

First of all, thanks for your amazing contribution. Needs to be said, 
again, and again. :-)
 

> There are many possibilities and I don't know what is best, so of course 
> suggestions are welcome!
>

In general, I would design that application / workflow in a way that 
everyone could use it in their environment just the way you do, however 
easily customized to personal preferences.

As I see it, right now, you the following workflow...

   1. you maintain some sort of "channel-list" / "subscription list"
   2. you maintain certain filters, either global or with respect to every 
   channel
   3. via commandline you batch-download all channels
   4. via commandline you extract a channel's data according to either a 
   global or per-channel filter or both into tid files
   5. via commandline you build an output wiki from those tid files 
   alongside a set of dedicated template components

All that is good and working. The required options would correspond to each 
step. Right now, you are possibly the only one who knows what needs 
configuration or what can be configured and how. Unless you already have, 
can start by publishing documentation on what you (can) do, step-wise, at 
the moment... and then we can give feedback on how that can be utilized or 
enhanced or whichever workflow to best construct around all that. 
Parameters should be simple text-based, preferably tiddlers. No fancy UIs 
or anything... except for what TiddlyWiki delivers out of the box. But 
maybe all that doesn't even need TiddlyWiki in the beginning, just plain 
text-editing.

One thing I would like to see around all this is the ability to adapt *the 
entire workflow* to "collections", i.e.  being able to do 1..5 for 
different collections of configurations rather than just one. So, for any 
"collection" you could do 1..5 independent from other collections, perhaps 
with a fallback to some global defaults.

Eventually, you may have...

 6. via commandline execute the update process for all collections at 
once

It would be good if we can reserve the term "collection" or something 
equivalent for meaning that: running the aggregator process against a 
distinct set of necessary configuration options. So, right now, your 
aggregator only caters for one "collection", being an "all-subscriptions", 
"all-channels" aggregator.

Although later on of interest, what I would not discuss or develop right 
now is some sort of community process / workflow to design "shared 
aggregator configurations". What I can perhaps see is that you wish to 
expose your configuration for "community search" or "community news" more 
so that others know where and how to contribute. I don't know if that needs 
actual development, rather than communication / documentation / a suitable 
workflow.

Best wishes,

— tb

-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/5a9c2bfe-211c-4736-ace3-65e59fba7859%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-17 Thread Tobias Beer
Hi Erwan,

I think a local version of your aggregator has quite more potential than 
just being the basis for a great community resource.

With a basic ui, even in TW, mayybe some extension to TiddlyDesktop, users 
could define feed collections in the context of which they add channels, 
essentially other TiddlyWikis, e.g. from a predefined list like the ones 
indexed by community search, or add new channels, or delete / deactivate 
others to ones interests.

Then all you need to know is the right commandline, perhaps even OS 
integration via some addon, that lets you pull updated contents onto your 
local machine, at the desired location corresponding to the collection, 
e.g. "dev resources", "my school project", custom to your preferences. No 
need for a server at all. All you want and get is your custom TiddlyWiki 
"news feed" with each "news collection" being the result of your 
aggregators output.

One step would be to levererage TiddlyWiki's save or export mechanisms to 
export the collection / subscription lists and preferences so that a given 
commandline or other executable environment can pick it up and construct 
the output.

Best wishes,

— tb

-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/42e40456-5626-47a8-a13c-129cfd84e115%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


[tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-17 Thread PMario
On Tuesday, September 15, 2015 at 11:53:58 PM UTC+2, Mat wrote:
>
> - thank you for your answers! (And I will from now on write twcom! Good 
> point Mario)
>

imo use tiddlywiki.com .. not relly much more to type, but that's the name
-m


-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/618ae94b-b23d-45f8-a97b-afac93ae5755%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-17 Thread Erwan


Hi Tobias,

Yes, I agree that it would make sense for the kind of application that 
you describe.


I only meant that for this application alone (generating a customized 
edition) it seems like an overkill, since I don't see any big technical 
difficulty in doing this server-side.


One step would be to levererage TiddlyWiki's save or export mechanisms 
to export the collection / subscription lists and preferences so that 
a given commandline or other executable environment can pick it up and 
construct the output.




I think this is more or less what my aggregator does already:

1) the wiki addresses are defined in the "skeleton" wiki (as node.js, 
but anyway conversion can be done both ways)
2) a command asks the node.js TW to render a list of the wiki tiddlers 
which match a filter (e.g. those for which the author agreed to have 
their content indexed); this step could be done from .tid files as well.

3) download these wikis as standalone html
4) convert the individual wikis to node.js
5) process tiddlers as .tid files, build the output wiki
6) convert the output wiki back to standalone html

I don't see what is missing in TW functionality for this kind of 
scenario, it's already perfect :-)
The only minor issue is that any command-line call to tiddlywiki is 
quite long to execute, I assume that this is because some node.js 
initialization must take place every time (?).


Regards
Erwan


--
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/55FB5942.2000305%40gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-17 Thread Tobias Beer
Hi Erwan,
 

> I don't see what is missing in TW functionality for this kind of 
> scenario, it's already perfect :-)
>

You sure have done a great job here, no question. :-)

What I was suggesting was, that if this ever became some sort of desktop 
application,
it would be good to not only have a single "channel subscription list" 
extracted from tiddlers from which to construct,
but to rather be able to define "collections" or "interests", i.e. 
different sets of subscriptions for different purposes,
e.g. one for the latest "Plugin News" and another for "My Project Team 
Wikis",
which could, theoretically, even overlap.

Not sure if you're familiar with it, but think of "interests" in Facebook.
There you can add users or pages to special subscription lists for which 
you can then see a dedicated news feed.

Best wishes,

— tb

-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/5f9a258b-99c7-4bd7-b482-029ab99f22a5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-16 Thread Erwan


Hi Mat,

On 15/09/15 22:53, Mat wrote:


The question is where to download /from /and, where the current 
aggregator work is performed, i.e to split up the downloaded TWs and 
then concoct the good parts into the desired TW. Maybe some server 
hosting an aggregator could offer a /temporary /space for this baking 
to take place



I don't understand everything, but I think there is an easy answer to 
this particular question, *assuming*:


1) that the user doesn't mind waiting a long time "for the cake to bake" 
(delayed service)


2) that the service is not abused or overloaded

The way I see it, a request would be processed when the update takes 
place (e.g. you might have to wait up to 24h in the current setting), 
and the resulting wiki would simply be created like the CommunitySearch 
wiki on github (with a predetermined name indicated in the request, I 
guess). Basically github already plays half the role of a server in my 
system: the computation takes place on a local computer which transmits 
the result to github at the end, from where users download it.


Erwan

--
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/55F9C8F5.6%40gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-16 Thread Erwan




Again, I wasn't talking about your particular instance (simply because 
it'd be rude of me to put some kind of "public pressure" on you 
personally. You're incredibly generous in creating this to begin with!).


Thank you for your kind words, but technically my contribution is very 
small compared to what Jeremy, the other contributors and the plugin 
developers do!


So, regarding the "delayed service", in another instance of a 
tw-aggregator this might not be an issue. For example, one can imagine 
a user triggered initiation.


Another idea is for users to have a /personal /aggregator set up so it 
does things locally. I don't know what this would require but I'm 
guessing something akin to those bash scripts 
 
should be functional also locally. Then a user would go to the 
community aggregator to produce the filtered list of updated titles 
and then use this list locally. Or he'd download the full skeleton 
list and create his filter locally.




Even though I agree that the idea of a personal aggregator is indeed 
completely feasible (btw anybody on linux can try mine), In my opinion 
the creation of a customized TW is not an application worth the bother 
of installing this kind of program for a user, even assuming a good 
packaging and user-friendly interface. To me this is typically much more 
of a server-side service, like Andreas' customizer that you mentioned 
(http://twguides.org/customizer.html). I said that with my current 
aggregator it would be delayed simply because I don't have a real 
server, but otherwise the creation process is certainly fast enough to 
be carried out on demand (as opposed to the aggregation process with 
multiple wikis, which is harder on the CPU and bandwidth).




Yes, talking is easier than coding ;-)

...

Another fantasy; if those user created filters are saved, they could 
be /analyzed /to eventually get a picture of which "combos of 
tiddlers" are popular. And if the users provided some meta-data info 
such as what the generated TW is intended for, we would have 
indications for what should be /premade /"suggested editions". For 
example, the combo "empty+pluginX+pluginY+themeZ" meta-tagged 
"collections" and "wines" is a basis for a "wine connoisseur" edition.


This is another good reason why the process should be done on the server 
side. I'm not sure however that a lot of users would bother filling the 
meta-data info... or even know exactly what they want and how to 
describe it! ;)


Erwan


--
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/55F9FBCA.7030904%40gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-16 Thread Mat
  Hi Erwan


Again, I wasn't talking about your particular instance (simply because it'd 
be rude of me to put some kind of "public pressure" on you personally. 
You're incredibly generous in creating this to begin with!). So, regarding 
the "delayed service", in another instance of a tw-aggregator this might 
not be an issue. For example, one can imagine a user triggered initiation.

Another idea is for users to have a *personal *aggregator set up so it does 
things locally. I don't know what this would require but I'm guessing 
something akin to those bash scripts 
 
should be functional also locally. Then a user would go to the community 
aggregator to produce the filtered list of updated titles and then use this 
list locally. Or he'd download the full skeleton list and create his filter 
locally.

Yes, talking is easier than coding ;-)

...

Another fantasy; if those user created filters are saved, they could be 
*analyzed 
*to eventually get a picture of which "combos of tiddlers" are popular. And 
if the users provided some meta-data info such as what the generated TW is 
intended for, we would have indications for what should be *premade *"suggested 
editions". For example, the combo "empty+pluginX+pluginY+themeZ" 
meta-tagged "collections" and "wines" is a basis for a "wine connoisseur" 
edition.

<:-)

-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/1933e2e2-d387-4316-941b-954c53198ba8%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


[tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-15 Thread Mat
Ben, Mario

- thank you for your answers! (And I will from now on write twcom! Good 
point Mario)


My thoughts arose apropos Erwans tw-aggregator - i.e if it would be 
possible to let the* user design his own version* before downloading it. 
Your replies make me think this might well be possible.

A bit of fantasy:

...but first I should mention that the only instance currently available of 
tw-aggregators is of course Erwans own so the following thoughts, while 
based on a tw-aggregator mechanism, are not particularly about his personal 
instance.

It seems reasonable to let the user filter+manually select tiddler *titles* in 
a community aggregator... and then have the result be used as a download 
*filter 
somewhere!*  To get a mere *filter *would be possible even if an aggregator 
is set up like Erwans to only display "skinny tiddlers" (i.e not the actual 
tiddlers). 

The filter could then be used for the downloading, like pmario mentions.

The question is where to download *from *and, where the current aggregator 
work is performed, i.e to split up the downloaded TWs and then concoct the 
good parts into the desired TW. Maybe some server hosting an aggregator 
could offer a *temporary *space for this baking to take place 

Anyway, I brought up part of this issue 
on 
the tw-aggregator github.


<:-)


-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/a3b6a2fb-edb8-4b0a-aa0f-f0307a495f73%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


[tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-15 Thread PMario
hi everyone,
 
please don't use tw(.)com in posts. It creates a link, that has nothing to 
do with tiddlywiki.com

-m

-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/6b1d7912-f30d-4942-b511-da46e38e1130%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


[tw] Re: Are the downloadable TWs from tw.com assembled at download or pre-made?

2015-09-15 Thread PMario
The $:/editions/tw5.com/download-empty contains a filter, that is used by: 
http://tiddlywiki.com/#%24%3A%2Feditions%2Ftw5.com%2Fsnippets%2Fdownload-empty-button

The download full version uses a different filter. So if you modify the 
filters, you can create customized downloadable versions. 

mario 

-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/10849d01-c159-4105-a08e-828ae290f6e6%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.