Menu
▾
▴
fofredux-devel — Developers discussion, bug fixing, and coding of fofredux
You can subscribe to this list here.
| 2005 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(92) |
Dec
(142) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2006 |
Jan
(33) |
Feb
(65) |
Mar
(76) |
Apr
(172) |
May
(124) |
Jun
(45) |
Jul
(76) |
Aug
(78) |
Sep
(1) |
Oct
(20) |
Nov
(6) |
Dec
(2) |
| 2007 |
Jan
(12) |
Feb
(7) |
Mar
(17) |
Apr
(3) |
May
|
Jun
(2) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(1) |
|
From: Carlos K. <ck...@gm...> - 2005-11-29 12:28:39
|
I agree with Benjy, allowing the users run fof and fofredux simultaneously with the same db and tables, would allow them to choose more easily. I guess that at fofredux installation time, we should only add tables and keep the original ones unmodified. To summarize, what features we want to include in the 0.2 release ? categories, feeds and search, anything else ? I order to get to users to migrate to fofredux we'll have to offer a significant set of new features. Nonetheles, having an active development team backing up the project is a "feature" strong enough to convince a lot of people. Carlos. |
|
From: Carlos K. <ck...@gm...> - 2005-11-29 11:58:43
|
Kevin said: > How many levels should the category tree support? > Can a category containfeeds and other categories at the same time? Im actually using aKregator (KDE) as my feed reader and the feeds are organized initially with a "root" category and allowing the user to create their own nested categories. I dont really know if there is a limit to the depth level of that nesting, but i think there is not. We will have to chech the opml definition to make sure of that. Regarding to "a tag containing another tag", well... that's exactly what a category is. I think we sould stick to a tag definition similar to the ones we find at del.icio.us and flickr. Supporting both categories and tags will confuse users only if we don't pay attention and create a messy UI. IIRC, one of you guys was some kind of specialist on UIs, wasn't you ? About tagging items vs. tagging feeds, I can only express my subjective point of view about that. *I'd really like to tag feeds, anyway I wouldn't spend much time tagging each item.* I'd rather like a link besides each item with the option "del.icio.us this!" to save and tag the items i'd want to preserve in my personal bookmark. Carlos. |
|
From: Kevin <ke...@dr...> - 2005-11-28 21:31:01
|
>> So, what have we done so far that breaks FoF compatiblity? I guess I'= ll >> have to install it and see. Removing or renaming the "read" column >> definitly will break it. We could set a "compatiblity flag" that will >> keep the "read" column in sync with the read/unread tags. There are >> very >> few places where the read column is altered, so it's not too difficult >> to >> do. > > The problem is, if we switch to tags for state, FOF won't update that > and will *only* update 'read,' which means changes made in FOF won't be > picked up by us. Okay, so it's not just supporting a downgrade path back to FoF, but also running them in parallel. So, the plus for doing this is making it easie= r for existing FoF users to try out FoFr. The downside is we wait a few revs before converting the read column to a tag. --=20 Kevin |
|
From: Benjamin S. <bs...@cr...> - 2005-11-28 21:09:03
|
> So, what have we done so far that breaks FoF compatiblity? I guess I'll > have to install it and see. Removing or renaming the "read" column > definitly will break it. We could set a "compatiblity flag" that will > keep the "read" column in sync with the read/unread tags. There are very > few places where the read column is altered, so it's not too difficult to > do. The problem is, if we switch to tags for state, FOF won't update that and will *only* update 'read,' which means changes made in FOF won't be picked up by us. --Benjy |
|
From: Kevin <ke...@dr...> - 2005-11-28 21:03:41
|
>> Ben, Do you really think retaining backwards compatiblity with FoF at >> the >> db schema level is a good idea? What about if we provide a decent >> import >> function that pulls the feeds and items from the FoF tables but does n= ot >> modify them? > > I don't think it's a good idea *long* term, but I don't think it'd be > that hard to do for just one release. While on the one hand we have to > bite the bullet at some point and release 1 seems like as good a time a= s > any, I'm not really sure it's *necessary* for v1 unless some feature > really calls for it. > > The advantage of this is it'd let people switch back and forth between > the two and let them decide which one they like. If nothing else, > hopefully this would earn us some points with whatever users decide to > check us out. If we have even one compelling feature that FOF doesn't, > then why wouldn't they switch? > > But, I could be pursuaded otherwise. For example, if the name 'read' > causes us too many problems, then sure, we could get rid of it. And > whoever suggested that it'd be easy enough to write a script to flip th= e > database back to an FOF database is probably right. > > I just thought it'd be a "nice to have." So, what have we done so far that breaks FoF compatiblity? I guess I'll have to install it and see. Removing or renaming the "read" column definitly will break it. We could set a "compatiblity flag" that will keep the "read" column in sync with the read/unread tags. There are very few places where the read column is altered, so it's not too difficult to do. --=20 Kevin |
|
From: Kevin <ke...@dr...> - 2005-11-28 20:37:59
|
> For storing searches, do we just have a table of SQL "WHERE" clauses, > or do we attempt to store more complex columns? I will want to do > searches/aggregations like: > > terms: > feeds: > tags: unread, mac, programming > when: 04/01/04 to 05/320/05 > > terms: illinois governor OR congress > feeds: > tags: read > when: > > Are these columns enough? That should do. > What are "Tags in tags"? Do you want to create a tree (politics -> > republican, wig, independent; computers -> mac -> programs, windows), > or just sets (computers -> mac, windows; technology -> mac -> > programs, palm -> programs, gadgets) > > Trees/sets could get fairly ugly trying to maintain, don't you think? > Categories could just be "collections" of tags? Again, seems like a > complicated matter. A tag that contains other tags is probably too much to deal with. If we support grouping tags together within categories or saved searches, that should cover most needs. > On 11/28/05, Kevin <ke...@dr...> wrote: >> >> > I think a tag is like a "categorization" of feeds and items. This ha= s >> > been the 'traditional' meaning and the one that most users would com= e >> > in understanding from Flickr, del.icio.us, etc. >> > >> > A "category" would mean a genre and is more general. >> > >> > - tags are specific classifications (php, toread, code, political, >> > republican, article) >> > - categories are broad classificiations (programming, politics, loca= l) >> > >> > One of the discussion lines awhile ago was to not do categories, and >> > just do tags, since tags should be able to encapsulate anything a >> > category is. However, perhaps they are "too specific", and a hierach= y >> > in an OPML file would be difficult, since feeds have multiple tags. >> > Now an OPML file can contain tag markups of feeds and items, but thi= s >> > wouldn't be an outline form, just more metadata within the feed tags= . >> > >> > Categories are pretty much done, just some things to clean up in the >> > code (initial install when no categories exist or imported feeds w/o= a >> > category - I guess just add an initial "None" category) But do you >> > think tags vs. categories as two organization schemes will confuse >> > users? >> >> The differences between categories and tags are blurry in places. As >> you >> can see with all my emails today, I'm still trying to iron out what a >> tag >> should be. I guess it's all about how the user wants to use them. I'= ll >> get alot of value out of having only tags as long as they we can >> aggregate >> them together.(tags that contain other tags) Or, maybe that's where we >> start calling them "categories". >> >> >> > You brought up a good point about "Tagging" searches. I definitely >> > think saved searches ("smart folders") are great. These searches wou= ld >> > be available for reading or for providing an aggregated feed. Taggin= g >> > searches is interesting, can you provide an example of how it would >> > work? >> >> The tagged search would be stored as a list of search criteria. Which >> then would be compiled into a WHERE clause for an SQL query. >> >> The criteria would be defined like: >> >> feed =3D 2 >> tag =3D unread >> title like '%something%' >> >> -- >> Kevin >> >> >> >> ------------------------------------------------------- >> This SF.net email is sponsored by: Splunk Inc. Do you grep through log >> files >> for problems? Stop! Download the new AJAX search engine that makes >> searching your log files as easy as surfing the web. DOWNLOAD SPLUNK= ! >> http://ads.osdn.com/?ad_idv37&alloc_id=16865&opclick >> _______________________________________________ >> Fofredux-devel mailing list >> Fof...@li... >> https://lists.sourceforge.net/lists/listinfo/fofredux-devel >> > > > -- > Andrew Turner > ajt...@hi... 42.4266N x 83.4931W > http://highearthorbit.com Northville, Michigan, USA > idealistic technocrat > > Photos - http://flickr.com/photos/ajturner > Travel - http://highearthorbit.com/projects/location/ > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log > files > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://ads.osdn.com/?ad_idv37&alloc_id=16865&op=3Dclick > _______________________________________________ > Fofredux-devel mailing list > Fof...@li... > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > --=20 Kevin |
|
From: Andrew T. <ajt...@hi...> - 2005-11-28 20:05:37
|
For storing searches, do we just have a table of SQL "WHERE" clauses, or do we attempt to store more complex columns? I will want to do searches/aggregations like: terms: feeds: tags: unread, mac, programming when: 04/01/04 to 05/320/05 terms: illinois governor OR congress feeds: tags: read when: Are these columns enough? What are "Tags in tags"? Do you want to create a tree (politics -> republican, wig, independent; computers -> mac -> programs, windows), or just sets (computers -> mac, windows; technology -> mac -> programs, palm -> programs, gadgets) Trees/sets could get fairly ugly trying to maintain, don't you think? Categories could just be "collections" of tags? Again, seems like a complicated matter. Andy On 11/28/05, Kevin <ke...@dr...> wrote: > > > I think a tag is like a "categorization" of feeds and items. This has > > been the 'traditional' meaning and the one that most users would come > > in understanding from Flickr, del.icio.us, etc. > > > > A "category" would mean a genre and is more general. > > > > - tags are specific classifications (php, toread, code, political, > > republican, article) > > - categories are broad classificiations (programming, politics, local) > > > > One of the discussion lines awhile ago was to not do categories, and > > just do tags, since tags should be able to encapsulate anything a > > category is. However, perhaps they are "too specific", and a hierachy > > in an OPML file would be difficult, since feeds have multiple tags. > > Now an OPML file can contain tag markups of feeds and items, but this > > wouldn't be an outline form, just more metadata within the feed tags. > > > > Categories are pretty much done, just some things to clean up in the > > code (initial install when no categories exist or imported feeds w/o a > > category - I guess just add an initial "None" category) But do you > > think tags vs. categories as two organization schemes will confuse > > users? > > The differences between categories and tags are blurry in places. As you > can see with all my emails today, I'm still trying to iron out what a tag > should be. I guess it's all about how the user wants to use them. I'll > get alot of value out of having only tags as long as they we can aggregat= e > them together.(tags that contain other tags) Or, maybe that's where we > start calling them "categories". > > > > You brought up a good point about "Tagging" searches. I definitely > > think saved searches ("smart folders") are great. These searches would > > be available for reading or for providing an aggregated feed. Tagging > > searches is interesting, can you provide an example of how it would > > work? > > The tagged search would be stored as a list of search criteria. Which > then would be compiled into a WHERE clause for an SQL query. > > The criteria would be defined like: > > feed =3D 2 > tag =3D unread > title like '%something%' > > -- > Kevin > > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://ads.osdn.com/?ad_idv37&alloc_id=16865&opclick > _______________________________________________ > Fofredux-devel mailing list > Fof...@li... > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > -- Andrew Turner ajt...@hi... 42.4266N x 83.4931W http://highearthorbit.com Northville, Michigan, USA idealistic technocrat Photos - http://flickr.com/photos/ajturner Travel - http://highearthorbit.com/projects/location/ |
|
From: Kevin <ke...@dr...> - 2005-11-28 19:46:44
|
> I think a tag is like a "categorization" of feeds and items. This has > been the 'traditional' meaning and the one that most users would come > in understanding from Flickr, del.icio.us, etc. > > A "category" would mean a genre and is more general. > > - tags are specific classifications (php, toread, code, political, > republican, article) > - categories are broad classificiations (programming, politics, local) > > One of the discussion lines awhile ago was to not do categories, and > just do tags, since tags should be able to encapsulate anything a > category is. However, perhaps they are "too specific", and a hierachy > in an OPML file would be difficult, since feeds have multiple tags. > Now an OPML file can contain tag markups of feeds and items, but this > wouldn't be an outline form, just more metadata within the feed tags. > > Categories are pretty much done, just some things to clean up in the > code (initial install when no categories exist or imported feeds w/o a > category - I guess just add an initial "None" category) But do you > think tags vs. categories as two organization schemes will confuse > users? The differences between categories and tags are blurry in places. As you can see with all my emails today, I'm still trying to iron out what a tag should be. I guess it's all about how the user wants to use them. I'll get alot of value out of having only tags as long as they we can aggregat= e them together.(tags that contain other tags) Or, maybe that's where we start calling them "categories". > You brought up a good point about "Tagging" searches. I definitely > think saved searches ("smart folders") are great. These searches would > be available for reading or for providing an aggregated feed. Tagging > searches is interesting, can you provide an example of how it would > work? The tagged search would be stored as a list of search criteria. Which then would be compiled into a WHERE clause for an SQL query. The criteria would be defined like: feed =3D 2 tag =3D unread title like '%something%' --=20 Kevin |
|
From: Benjamin S. <bs...@cr...> - 2005-11-28 19:30:51
|
> For me, just running install.php and doing an OPML import was a *good > enough* upgrade path. ;) Heh. > Ben, Do you really think retaining backwards compatiblity with FoF at the > db schema level is a good idea? What about if we provide a decent import > function that pulls the feeds and items from the FoF tables but does not > modify them? I don't think it's a good idea *long* term, but I don't think it'd be that hard to do for just one release. While on the one hand we have to bite the bullet at some point and release 1 seems like as good a time as any, I'm not really sure it's *necessary* for v1 unless some feature really calls for it. The advantage of this is it'd let people switch back and forth between the two and let them decide which one they like. If nothing else, hopefully this would earn us some points with whatever users decide to check us out. If we have even one compelling feature that FOF doesn't, then why wouldn't they switch? But, I could be pursuaded otherwise. For example, if the name 'read' causes us too many problems, then sure, we could get rid of it. And whoever suggested that it'd be easy enough to write a script to flip the database back to an FOF database is probably right. I just thought it'd be a "nice to have." --Benjy |
|
From: Andrew T. <ajt...@hi...> - 2005-11-28 19:30:03
|
On 11/28/05, Kevin <ke...@dr...> wrote: > > For me, just running install.php and doing an OPML import was a *good > enough* upgrade path. ;) Me too - which is why I didn't notice missing tables, etc. in applying FoFRedux to an existing DB. > 30 day old items are by default deleted currently anyways? > > Ben, Do you really think retaining backwards compatiblity with FoF at the > db schema level is a good idea? What about if we provide a decent import > function that pulls the feeds and items from the FoF tables but does not > modify them? We could create a new set of tables (ergo the table prefix change by default) and migrate across for the first time. Does this scale for the future? |
|
From: Kevin <ke...@dr...> - 2005-11-28 19:22:12
|
> Ok - so read/unread in tags it is. Sounds reasonable. Do we filter > read/unread when we display an item's tags? If so, do we use > read/unread as the "magic tags" regardless of language (I think we > will definitely want to support multiple languages in the UI) > > Also, as for backwards compatability, you mean being able to switch > back and forth b/w FoF v0.19 and FoFRedux v0.2? We will need to > reconfigure tables at some point, and this may be the best point to > provide an "upgrade" script that would migrate the "read" column to a > "read/unread" flag. ADODB is nice in that it supports table format, > but we could write a script for this first release. Do you think that > covers the compatability enough? For me, just running install.php and doing an OPML import was a *good enough* upgrade path. ;) Ben, Do you really think retaining backwards compatiblity with FoF at the db schema level is a good idea? What about if we provide a decent import function that pulls the feeds and items from the FoF tables but does not modify them? > On 11/28/05, Benjamin Stewart <bs...@cr...> wrote: >> Whatever we do, I would like to reiterate the desire to maintain >> backwards compatability for the first release, which means we can't ge= t >> rid of 'read' yet. >> >> Ultimately, I think I like the idea of using tags for all status, and = we >> might have some "reserved" tags (read, unread, etc) that don't show up >> as normal tags in the UI, but act just like tags. >> >> --Benjy >> >> >> >> On Mon, Nov 28, 2005 at 01:50:42PM -0500, Andrew Turner wrote: >> > Do we want to support some enumerated "Flag" that is separate from >> > tags/categories for things like read, unread, "flagged", "delete" ? >> > >> > I guess incorporating this into tags is the most flexible means. Wou= ld >> > require less coding to do a search on "all items with tags windows, >> > programming, unread" and completely flexible "flags" >> > >> > What say anyone else? >> > >> > Andy >> > >> > >> > On 11/28/05, Kevin <ke...@dr...> wrote: >> > > Benjamin Stewart wrote: >> > > >> > > <snip> >> > > >> > > >(Oh, this does bring up the question of whether or not we want to >> change >> > > >the name of the 'read' column at some point, so we can stop >> confusing >> > > >mysql) >> > > > >> > > > >> > > +1. Using reserved words as column names is a bad idea. The >> backtick >> > > workaround is not cross DB compatible. This column could even be >> > > eliminated by using a "read" or "unread" tag. >> > > >> > > --Kevin >> > > >> > > >> > > ------------------------------------------------------- >> > > This SF.net email is sponsored by: Splunk Inc. Do you grep through >> log files >> > > for problems? Stop! Download the new AJAX search engine that mak= es >> > > searching your log files as easy as surfing the web. DOWNLOAD >> SPLUNK! >> > > http://ads.osdn.com/?ad_id=3D7637&alloc_id=3D16865&op=3Dclick >> > > _______________________________________________ >> > > Fofredux-devel mailing list >> > > Fof...@li... >> > > https://lists.sourceforge.net/lists/listinfo/fofredux-devel >> > > >> > >> > >> > -- >> > Andrew Turner >> > ajt...@hi... 42.4266N x 83.4931W >> > http://highearthorbit.com Northville, Michigan, USA >> > idealistic technocrat >> > >> > Photos - http://flickr.com/photos/ajturner >> > Travel - http://highearthorbit.com/projects/location/ >> > >> > >> > ------------------------------------------------------- >> > This SF.net email is sponsored by: Splunk Inc. Do you grep through l= og >> files >> > for problems? Stop! Download the new AJAX search engine that makes >> > searching your log files as easy as surfing the web. DOWNLOAD >> SPLUNK! >> > http://ads.osdn.com/?ad_idv37&alloc_id=16865&op=3Dclick >> > _______________________________________________ >> > Fofredux-devel mailing list >> > Fof...@li... >> > https://lists.sourceforge.net/lists/listinfo/fofredux-devel >> > > > -- > Andrew Turner > ajt...@hi... 42.4266N x 83.4931W > http://highearthorbit.com Northville, Michigan, USA > idealistic technocrat > > Photos - http://flickr.com/photos/ajturner > Travel - http://highearthorbit.com/projects/location/ > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log > files > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://ads.osdn.com/?ad_idv37&alloc_id=16865&op=3Dclick > _______________________________________________ > Fofredux-devel mailing list > Fof...@li... > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > --=20 Kevin |
|
From: Andrew T. <ajt...@hi...> - 2005-11-28 19:00:27
|
Ok - so read/unread in tags it is. Sounds reasonable. Do we filter read/unread when we display an item's tags? If so, do we use read/unread as the "magic tags" regardless of language (I think we will definitely want to support multiple languages in the UI) Also, as for backwards compatability, you mean being able to switch back and forth b/w FoF v0.19 and FoFRedux v0.2? We will need to reconfigure tables at some point, and this may be the best point to provide an "upgrade" script that would migrate the "read" column to a "read/unread" flag. ADODB is nice in that it supports table format, but we could write a script for this first release. Do you think that covers the compatability enough? On 11/28/05, Benjamin Stewart <bs...@cr...> wrote: > Whatever we do, I would like to reiterate the desire to maintain > backwards compatability for the first release, which means we can't get > rid of 'read' yet. > > Ultimately, I think I like the idea of using tags for all status, and we > might have some "reserved" tags (read, unread, etc) that don't show up > as normal tags in the UI, but act just like tags. > > --Benjy > > > > On Mon, Nov 28, 2005 at 01:50:42PM -0500, Andrew Turner wrote: > > Do we want to support some enumerated "Flag" that is separate from > > tags/categories for things like read, unread, "flagged", "delete" ? > > > > I guess incorporating this into tags is the most flexible means. Would > > require less coding to do a search on "all items with tags windows, > > programming, unread" and completely flexible "flags" > > > > What say anyone else? > > > > Andy > > > > > > On 11/28/05, Kevin <ke...@dr...> wrote: > > > Benjamin Stewart wrote: > > > > > > <snip> > > > > > > >(Oh, this does bring up the question of whether or not we want to ch= ange > > > >the name of the 'read' column at some point, so we can stop confusin= g > > > >mysql) > > > > > > > > > > > +1. Using reserved words as column names is a bad idea. The backti= ck > > > workaround is not cross DB compatible. This column could even be > > > eliminated by using a "read" or "unread" tag. > > > > > > --Kevin > > > > > > > > > ------------------------------------------------------- > > > This SF.net email is sponsored by: Splunk Inc. Do you grep through lo= g files > > > for problems? Stop! Download the new AJAX search engine that makes > > > searching your log files as easy as surfing the web. DOWNLOAD SPLUN= K! > > > http://ads.osdn.com/?ad_id=3D7637&alloc_id=3D16865&op=3Dclick > > > _______________________________________________ > > > Fofredux-devel mailing list > > > Fof...@li... > > > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > > > > > > > > > -- > > Andrew Turner > > ajt...@hi... 42.4266N x 83.4931W > > http://highearthorbit.com Northville, Michigan, USA > > idealistic technocrat > > > > Photos - http://flickr.com/photos/ajturner > > Travel - http://highearthorbit.com/projects/location/ > > > > > > ------------------------------------------------------- > > This SF.net email is sponsored by: Splunk Inc. Do you grep through log = files > > for problems? Stop! Download the new AJAX search engine that makes > > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > > http://ads.osdn.com/?ad_idv37&alloc_id=16865&op=3Dclick > > _______________________________________________ > > Fofredux-devel mailing list > > Fof...@li... > > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > -- Andrew Turner ajt...@hi... 42.4266N x 83.4931W http://highearthorbit.com Northville, Michigan, USA idealistic technocrat Photos - http://flickr.com/photos/ajturner Travel - http://highearthorbit.com/projects/location/ |
|
From: Benjamin S. <bs...@cr...> - 2005-11-28 18:53:23
|
Whatever we do, I would like to reiterate the desire to maintain backwards compatability for the first release, which means we can't get rid of 'read' yet. Ultimately, I think I like the idea of using tags for all status, and we might have some "reserved" tags (read, unread, etc) that don't show up as normal tags in the UI, but act just like tags. --Benjy On Mon, Nov 28, 2005 at 01:50:42PM -0500, Andrew Turner wrote: > Do we want to support some enumerated "Flag" that is separate from > tags/categories for things like read, unread, "flagged", "delete" ? >=20 > I guess incorporating this into tags is the most flexible means. Would > require less coding to do a search on "all items with tags windows, > programming, unread" and completely flexible "flags" >=20 > What say anyone else? >=20 > Andy >=20 >=20 > On 11/28/05, Kevin <ke...@dr...> wrote: > > Benjamin Stewart wrote: > > > > <snip> > > > > >(Oh, this does bring up the question of whether or not we want to chan= ge > > >the name of the 'read' column at some point, so we can stop confusing > > >mysql) > > > > > > > > +1. Using reserved words as column names is a bad idea. The backtick > > workaround is not cross DB compatible. This column could even be > > eliminated by using a "read" or "unread" tag. > > > > --Kevin > > > > > > ------------------------------------------------------- > > This SF.net email is sponsored by: Splunk Inc. Do you grep through log = files > > for problems? Stop! Download the new AJAX search engine that makes > > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > > http://ads.osdn.com/?ad_id=3D7637&alloc_id=3D16865&op=3Dclick > > _______________________________________________ > > Fofredux-devel mailing list > > Fof...@li... > > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > > >=20 >=20 > -- > Andrew Turner > ajt...@hi... 42.4266N x 83.4931W > http://highearthorbit.com Northville, Michigan, USA > idealistic technocrat >=20 > Photos - http://flickr.com/photos/ajturner > Travel - http://highearthorbit.com/projects/location/ >=20 >=20 > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://ads.osdn.com/?ad_idv37&alloc_id=16865&op=3Dclick > _______________________________________________ > Fofredux-devel mailing list > Fof...@li... > https://lists.sourceforge.net/lists/listinfo/fofredux-devel |
|
From: Andrew T. <ajt...@hi...> - 2005-11-28 18:50:59
|
Do we want to support some enumerated "Flag" that is separate from tags/categories for things like read, unread, "flagged", "delete" ? I guess incorporating this into tags is the most flexible means. Would require less coding to do a search on "all items with tags windows, programming, unread" and completely flexible "flags" What say anyone else? Andy On 11/28/05, Kevin <ke...@dr...> wrote: > Benjamin Stewart wrote: > > <snip> > > >(Oh, this does bring up the question of whether or not we want to change > >the name of the 'read' column at some point, so we can stop confusing > >mysql) > > > > > +1. Using reserved words as column names is a bad idea. The backtick > workaround is not cross DB compatible. This column could even be > eliminated by using a "read" or "unread" tag. > > --Kevin > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://ads.osdn.com/?ad_id=3D7637&alloc_id=3D16865&op=3Dclick > _______________________________________________ > Fofredux-devel mailing list > Fof...@li... > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > -- Andrew Turner ajt...@hi... 42.4266N x 83.4931W http://highearthorbit.com Northville, Michigan, USA idealistic technocrat Photos - http://flickr.com/photos/ajturner Travel - http://highearthorbit.com/projects/location/ |
|
From: Andrew T. (nilspace) <nil...@us...> - 2005-11-28 18:44:53
|
I think a tag is like a "categorization" of feeds and items. This has been the 'traditional' meaning and the one that most users would come in understanding from Flickr, del.icio.us, etc. A "category" would mean a genre and is more general. - tags are specific classifications (php, toread, code, political, republican, article) - categories are broad classificiations (programming, politics, local) One of the discussion lines awhile ago was to not do categories, and just do tags, since tags should be able to encapsulate anything a category is. However, perhaps they are "too specific", and a hierachy in an OPML file would be difficult, since feeds have multiple tags. Now an OPML file can contain tag markups of feeds and items, but this wouldn't be an outline form, just more metadata within the feed tags. Categories are pretty much done, just some things to clean up in the code (initial install when no categories exist or imported feeds w/o a category - I guess just add an initial "None" category) But do you think tags vs. categories as two organization schemes will confuse users? You brought up a good point about "Tagging" searches. I definitely think saved searches ("smart folders") are great. These searches would be available for reading or for providing an aggregated feed. Tagging searches is interesting, can you provide an example of how it would work? Andy On 11/28/05, Kevin <ke...@dr...> wrote: > > Carlos Kozuszko wrote: > >And what about tagging the feeds instead of individual items ? Or > > maybe tags for both could be implemented. > > I thought that's what categories were for. (grouping of feeds) We could > support both, but a clear definition of what a tag is and what a category > is should be made. > > Here is my crazy brainstorming of what a tag could be. > > 1. tag attach to item > 2. tag attach to feed > 3. tag attach to tag > 4. tag attach to category > 5. tag attach to search result > > So, I've partially implemented #1. Which of the others listed really > represent what a tag *IS*? I thought about implementing tags as only #5= . > Then, the tag is stored as a list of search criteria. The nice thing > about doing this is it allows the user create a tag using any of the abov= e > criteria, at the same time. Tags could be made this flexible. Do we nee= d > them to be? Am I completely off base on what a tag is supposed to > accomplish? > > > > I've already found a use for categories, lets say im importing a > > opml file (as I actually do), the tree-like structure for the categorie= s > > could/should be preserved when adding those feeds to our database. > > Likewise, when we export a opml file the categories should be > > reflected in that file. > > How many levels should the category tree support? Can a category contain > feeds and other categories at the same time? > > > > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://ads.osdn.com/?ad_idv37&alloc_id=16865&opclick > _______________________________________________ > Fofredux-devel mailing list > Fof...@li... > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > -- Andrew Turner ajt...@hi... 42.4266N x 83.4931W http://highearthorbit.com Northville, Michigan, USA idealistic technocrat Photos - http://flickr.com/photos/ajturner Travel - http://highearthorbit.com/projects/location/ ( |
|
From: Benjamin S. <bs...@cr...> - 2005-11-28 18:40:27
|
> btw, Benjy was the one who was experimenting with adodb. Uh, no, it was Kevin. ;-) --Benjy |
|
From: Kevin <ke...@dr...> - 2005-11-28 17:47:53
|
> I think what I'm trying to say is tags are great - but are often too > explicit/small. For example, in my del.icio.us tags, I tend to have 3 > or 4 bookmarks with any 1 tag. This makes them not that useful for > broad readings. I want more than just "cocoa" or "auto" tags. I want > "Politics" or "Technology". > > I guess I'm just being too picky. Using tags in general would let me > go ahead and mark these broad generalizations for feeds and then read > a group of feeds based on "Technology" etc. So I think just supporting > tags is fine (albeit much harder). > > What about the problem of "tag soup". People are very liberal with > tagging - and not always consistent either. Any suggestions? The tag consistency problem could be handled using tag aliases. Multiple names could be attached to tags and are only used for RSS item category auto-tagging. Example: I have an existing tag of "windows" I have a feed that uses the rss item category of "Windows XP" I create a tag alias "Windows XP" which points to the real tag "window= s" When the item is inserted and auto-tagged using the rss item category,= =20 it will follow the alias and tag the item using "windows", instead of "Windows XP" What do you think? Most of my rss feeds don't use the optional category feature, but I can see the value in this. --=20 Kevin |
|
From: Kevin <ke...@dr...> - 2005-11-28 17:36:16
|
Andrew said: > > I can see the benefits of having a user-defined taxonomy of tags. For > one thing, many feeds supply the blog's categories for the item - so > we could optionally apply these tags by default. I think this is a good idea to have the item category(s) carry over to tags. This would require having the add or update functions auto-create tags as needed to match the item categories. --=20 Kevin |
|
From: Kevin <ke...@dr...> - 2005-11-28 17:23:04
|
Carlos Kozuszko wrote: >And what about tagging the feeds instead of individual items ? Or > maybe tags for both could be implemented. I thought that's what categories were for. (grouping of feeds) We could support both, but a clear definition of what a tag is and what a category is should be made. Here is my crazy brainstorming of what a tag could be. 1. tag attach to item 2. tag attach to feed 3. tag attach to tag 4. tag attach to category 5. tag attach to search result So, I've partially implemented #1. Which of the others listed really represent what a tag *IS*? I thought about implementing tags as only #5= . Then, the tag is stored as a list of search criteria. The nice thing about doing this is it allows the user create a tag using any of the abov= e criteria, at the same time. Tags could be made this flexible. Do we nee= d them to be? Am I completely off base on what a tag is supposed to accomplish? > I've already found a use for categories, lets say im importing a > opml file (as I actually do), the tree-like structure for the categorie= s > could/should be preserved when adding those feeds to our database. > Likewise, when we export a opml file the categories should be > reflected in that file. How many levels should the category tree support? Can a category contain feeds and other categories at the same time? |
|
From: Carlos K. <ck...@gm...> - 2005-11-28 13:43:23
|
Kev said: >>>Here is my plan for the tag schema. It adds 2 new tables. One to holdthe tag names, and the other to cross reference tags to items. And what about tagging the feeds instead of individual items ? Or maybe tags for both could be implemented. I've already found a use for categories, lets say im importing a opml file (as I actually do), the tree-like structure for the categories could/should be preserved when adding those feeds to our database. Likewise, when we export a opml file the categories should be reflected in that file. btw, Benjy was the one who was experimenting with adodb. Carlos. |
|
From: Kevin <ke...@dr...> - 2005-11-28 07:10:06
|
Andrew Turner wrote: >Thanks for the quick succession/stream of development msgs. Actually >brought up a bunch of good points. The primary issue being I am >responsible for having checked in incomplete/bad code. I'm very sorry >about that. I was rushing to get a baseline in for other people to >fool around with and didn't do appropriate testing. I'll not do this >in the future. > >Anyways, you're correct that we need to be able to handle >updates/inserts of new tables. I believe Carlos was working on >wrapping up the mysql calls to work with wither MySQL or PostgreSQL. >Perhaps we can handle wrapping up clean select/inserts there? The >other, easier, alternative is to add an "upgrade.php" that does tests >and updates to tables as necessary. This could be a set of tests and >inserts, or an actual intelligent upgrade depending on version number. > > > ADODB supports updating a table schema to match a template. It will add columns and change column types as needed. It's pretty slick. I prefer this to hand coding an intelligent upgrade script. -Kevin |
|
From: Kevin <ke...@dr...> - 2005-11-28 06:59:35
|
Benjamin Stewart wrote: >Update, 2 seconds later: yes, the existing HEAD does not like not >having any categories defined. As soon as I defined one, panel started >working (though view still has the problem defined below), and a second >problem is that the panel immediately listed the feeds I had added >without a category as belonging to the new category (obviously wrong). > >Anyway, I'm just going to wait until tags land to worry about this too >much. > >Kevin, what database schema are you planning on using for tags? > > > Here is my plan for the tag schema. It adds 2 new tables. One to hold the tag names, and the other to cross reference tags to items. I am about 25% finished. I probably won't have something worthy of a checkin until next week. CREATE TABLE `$FOF_TAG_TABLE` ( `id` int(11) NOT NULL auto_increment, `name` varchar(250), PRIMARY KEY (`id`) ) TYPE=MyISAM; CREATE TABLE `$FOF_X_TAG_ITEM_TABLE` ( `id` int(11) NOT NULL auto_increment, `tag_id` int(11) NOT NULL default '0', `item_id` int(11) NOT NULL default '0', PRIMARY KEY (`id`) ) TYPE=MyISAM; --Kevin |
|
From: Kevin <ke...@dr...> - 2005-11-28 06:57:04
|
Benjamin Stewart wrote: <snip> >(Oh, this does bring up the question of whether or not we want to change >the name of the 'read' column at some point, so we can stop confusing >mysql) > > +1. Using reserved words as column names is a bad idea. The backtick workaround is not cross DB compatible. This column could even be eliminated by using a "read" or "unread" tag. --Kevin |
|
From: Andrew T. <ajt...@hi...> - 2005-11-28 01:39:45
|
Thanks for the quick succession/stream of development msgs. Actually brought up a bunch of good points. The primary issue being I am responsible for having checked in incomplete/bad code. I'm very sorry about that. I was rushing to get a baseline in for other people to fool around with and didn't do appropriate testing. I'll not do this in the future. Anyways, you're correct that we need to be able to handle updates/inserts of new tables. I believe Carlos was working on wrapping up the mysql calls to work with wither MySQL or PostgreSQL. Perhaps we can handle wrapping up clean select/inserts there? The other, easier, alternative is to add an "upgrade.php" that does tests and updates to tables as necessary. This could be a set of tests and inserts, or an actual intelligent upgrade depending on version number. Benjy, can you add a feature request to the SF site for FoFRedux? I see Kevin's already added one. ;) Ok, so let's wait to hear from Kevin regarding the plan for tags. It probably is a good idea to have those implemented rather than fixing the category problems. Andy On 11/27/05, Benjamin Stewart <bs...@cr...> wrote: > (the broken view appears to be the known timestamp issue. I'll shut up > and just integrate my changes now.) > > On Sun, Nov 27, 2005 at 03:47:20PM -0800, Benjamin Stewart wrote: > > Update, 2 seconds later: yes, the existing HEAD does not like not > > having any categories defined. As soon as I defined one, panel started > > working (though view still has the problem defined below), and a second > > problem is that the panel immediately listed the feeds I had added > > without a category as belonging to the new category (obviously wrong). > > > > Anyway, I'm just going to wait until tags land to worry about this too > > much. > > > > Kevin, what database schema are you planning on using for tags? > > > > --Benjy > > > > On Sun, Nov 27, 2005 at 03:44:24PM -0800, Benjamin Stewart wrote: > > > * on a vanilla install, add feeds does not seem to be working. I > > > suspect this might have something to do with php settings of registe= r > > > globals vs. .. whatever the other options are. Switching add.php to = use > > > _REQUEST instead of _GET and _POST fixed it, and I checked in the fix= to > > > head. > > > > > > * Even on the vanilla install, I'm not getting a working installation= . > > > the panel only lists one feed (even though I've subscribed to many), > > > does not list the feed's name, and view does not list any results. Am > > > investigating. Possibly related to not having any categories defined= ? > > > If so, will hack around this pending categories via tags. > > > > > > * I guess we should handle feed: urls (thanks a billion, apple.). > > > Anyone interested, or should we add it to wherever "the" todo list > > > lives? > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi= les > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://ads.osdn.com/?ad_id=3D7637&alloc_id=3D16865&op=3Dclick > _______________________________________________ > Fofredux-devel mailing list > Fof...@li... > https://lists.sourceforge.net/lists/listinfo/fofredux-devel > -- Andrew Turner ajt...@hi... 42.4266N x 83.4931W http://highearthorbit.com Northville, Michigan, USA idealistic technocrat Photos - http://flickr.com/photos/ajturner Travel - http://highearthorbit.com/projects/location/ |
|
From: Benjamin S. <bs...@cr...> - 2005-11-27 23:58:50
|
(the broken view appears to be the known timestamp issue. I'll shut up and just integrate my changes now.) On Sun, Nov 27, 2005 at 03:47:20PM -0800, Benjamin Stewart wrote: > Update, 2 seconds later: yes, the existing HEAD does not like not > having any categories defined. As soon as I defined one, panel started > working (though view still has the problem defined below), and a second > problem is that the panel immediately listed the feeds I had added > without a category as belonging to the new category (obviously wrong). > > Anyway, I'm just going to wait until tags land to worry about this too > much. > > Kevin, what database schema are you planning on using for tags? > > --Benjy > > On Sun, Nov 27, 2005 at 03:44:24PM -0800, Benjamin Stewart wrote: > > * on a vanilla install, add feeds does not seem to be working. I > > suspect this might have something to do with php settings of register > > globals vs. .. whatever the other options are. Switching add.php to use > > _REQUEST instead of _GET and _POST fixed it, and I checked in the fix to > > head. > > > > * Even on the vanilla install, I'm not getting a working installation. > > the panel only lists one feed (even though I've subscribed to many), > > does not list the feed's name, and view does not list any results. Am > > investigating. Possibly related to not having any categories defined? > > If so, will hack around this pending categories via tags. > > > > * I guess we should handle feed: urls (thanks a billion, apple.). > > Anyone interested, or should we add it to wherever "the" todo list > > lives? |
5 messages has been excluded from this view by a project administrator.
