*BLF LT1 Lantern Project) (updated Nov,17,2020)

[DBSAR]

@sbslider

@dekozn

Any updates so far? I’m hoping to have one in my hands beginning july or is that just wishfull thinking?

just wishfull thinking IMO, but keep hope alive.

Were just hoping & waiting on Barry & Sofirn for updates.

[evil_andy]

I say let it take however long it needs so that it gets done right the first time.

[SIGShooter]

Here are the additions since last Sunday night. Spreadsheets linked in post #1 have been updated and reflect exactly what is in my database including of course the BLF user name

Thanks

1714 Aviationtech
1715 Behyot
1716 SamHaLeKe
1717 jericho0o
1718 lightblows
1719 lightblows
1720 bonfire
1721 hoabegrubn
1722 email
1723 email
1724 email
1725 email
1726 Sparky.
1727 Sparky.
1728 Sparky.
1729 everwest
1730 InitiatePenguin
1731 Flying Luminosity
1732 Cajones32
1733 D4rka2nx
1734 ludolf

[illuminos]

I think I was missed, interested in 2, post 4811.

[SIGShooter]

@illuminos

I think I was missed, interested in 2, post 4811.

Sorry about that. I added you for 2 but the results won’t be updated until next Sunday. However you’re in my database.

!!

[illuminos]

@SIGShooter

@illuminos

I think I was missed, interested in 2, post 4811.

Sorry about that. I added you for 2 but the results won’t be updated until next Sunday. However you’re in my database.

No problem, I’m sure it’s no easy task keeping track of these posts.

[ToyKeeper]

@SIGShooter

I added you for 2 … you’re in my database.

Thanks, that’s an interesting look at how the interest list works. Maybe I’m weird, but I think it’s fascinating to see how different people solve similar problems.

My solution was a bit different… Instead of a database, it used a plain text file… but that’s relatively immaterial. The main differences seem to be in the rest of the system. When I wanted to update the list, my process was:

• Run “grab.sh” to download new posts.
• Run “tkbot.py” to process the new posts. This involved a bit of manual input. Basically, it’d go through a loop like this:
  • Display the next post’s text content, minus any quoted parts.
  • Analyse the text and display the script’s best guess about what the post means (comment or request).
  • Give a text prompt, for manual categorization.
  • The user can then press enter (accept the script’s guess), or type a value. Like “c” for a comment, or a number for a request. “1” for one light, “2” for 2 lights, etc. Also negative numbers for cancellations.
• When that’s done, run “interest.py” to generate a new human-friendly copy of the interest list.
• Copy/paste that into BLF.

So it mostly just required one or two keystrokes per post. And the text analyzer wasn’t very accurate, so eventually I turned it off and just made the default action “c” for “comment”. That way, it was just Enter for regular comments, or a number followed by Enter for requests.

If I ever do something like that again though, I want to make it fully automatic, using explicit bot control commands that people can do on their own. Managing the list manually was a lot of work even with the extra tools I made for it.

[SIGShooter]

@ToyKeeper

@SIGShooter

I added you for 2 … you’re in my database.

Thanks, that’s an interesting look at how the interest list works. Maybe I’m weird, but I think it’s fascinating to see how different people solve similar problems.

My solution was a bit different… Instead of a database, it used a plain text file… but that’s relatively immaterial. The main differences seem to be in the rest of the system. When I wanted to update the list, my process was:

• Run “grab.sh” to download new posts.
• Run “tkbot.py” to process the new posts. This involved a bit of manual input. Basically, it’d go through a loop like this:
  • Display the next post’s text content, minus any quoted parts.
  • Analyse the text and display the script’s best guess about what the post means (comment or request).
  • Give a text prompt, for manual categorization.
  • The user can then press enter (accept the script’s guess), or type a value. Like “c” for a comment, or a number for a request. “1” for one light, “2” for 2 lights, etc. Also negative numbers for cancellations.
• When that’s done, run “interest.py” to generate a new human-friendly copy of the interest list.
• Copy/paste that into BLF.

So it mostly just required one or two keystrokes per post. And the text analyzer wasn’t very accurate, so eventually I turned it off and just made the default action “c” for “comment”. That way, it was just Enter for regular comments, or a number followed by Enter for requests.

If I ever do something like that again though, I want to make it fully automatic, using explicit bot control commands that people can do on their own. Managing the list manually was a lot of work even with the extra tools I made for it.

Since I already had Oracle up on a VM it was easy enough to create a schema and a few tables & other objects for the lantern project. It only takes me a few seconds to enter the BLF user name and post number into my insert script which then writes the request into a detail table and creates an entry in a master user name table if needed. I probably spend less than 5 minutes per day entering the requests so it’s not a big effort on my part. The same tables/objects are reusable for other projects if the need ever arises so that’s a plus I guess.

I couldn’t figure out a way to automatically and accurately analyze the postings so I thought about writing a front-end to let people enter them but I didn’t want to expose my computer to the outside world

[DBSAR]

@SIGShooter

Since I already had Oracle up on a VM it was easy enough to create a schema and a few tables & other objects for the lantern project. It only takes me a few seconds to enter the BLF user name and post number into my insert script which then writes the request into a detail table and creates an entry in a master user name table if needed. I probably spend less than 5 minutes per day entering the requests so it’s not a big effort on my part. The same tables/objects are reusable for other projects if the need ever arises so that’s a plus I guess.

I couldn’t figure out a way to automatically and accurately analyze the postings so I thought about writing a front-end to let people enter them but I didn’t want to expose my computer to the outside world

[nick779]

Im so freaking excited for this lantern. Ive had a Q8 with a warm flash gel and a diffuser running as a bedside candle for a good 8 months now. I cant wait to have something I dont have to mess with that has an adjustable tint.

This might be the first light I dont mess with at all.

[sbslider]

@SIGShooter

Since I already had Oracle up on a VM it was easy enough to create a schema and a few tables & other objects for the lantern project. It only takes me a few seconds to enter the BLF user name and post number into my insert script which then writes the request into a detail table and creates an entry in a master user name table if needed. I probably spend less than 5 minutes per day entering the requests so it’s not a big effort on my part. The same tables/objects are reusable for other projects if the need ever arises so that’s a plus I guess.

I couldn’t figure out a way to automatically and accurately analyze the postings so I thought about writing a front-end to let people enter them but I didn’t want to expose my computer to the outside world

There was someone that approached me while I was keeping this list, and he created a bot that would search the thread of folks who wanted to be added to the interest list. It seemed like he was close to making it work, then he disappeared.

huey18

last post in this thread

[SIGShooter]

@sbslider

@SIGShooter

Since I already had Oracle up on a VM it was easy enough to create a schema and a few tables & other objects for the lantern project. It only takes me a few seconds to enter the BLF user name and post number into my insert script which then writes the request into a detail table and creates an entry in a master user name table if needed. I probably spend less than 5 minutes per day entering the requests so it’s not a big effort on my part. The same tables/objects are reusable for other projects if the need ever arises so that’s a plus I guess.

I couldn’t figure out a way to automatically and accurately analyze the postings so I thought about writing a front-end to let people enter them but I didn’t want to expose my computer to the outside world

There was someone that approached me while I was keeping this list, and he created a bot that would search the thread of folks who wanted to be added to the interest list. It seemed like he was close to making it work, then he disappeared.

huey18

last post in this thread

I remember seeing that when he posted it. The problem as I recall was that it depended on the request being in the exact format required in order to scrub it from the post. Since it only takes me a few minutes each day to manually scan new postings it’s not a big deal for me to copy and paste the user name and post number into my table. Post number isn’t really needed since each insert tags the row with an incrementing sequence number (in case it matters for sending out the PM’s) but I like to have it just as another data point.

[sycomonkeys]

I would like to get two more of these. That would make 3 in total for me.
I really hope the production of these goes in full swing soon with hurricane season starting out here in FL.
No matter when it does, I know they’ll totally be worth the wait with the dedication everyone has put in to it.

[sbslider]

@SIGShooter

Since it only takes me a few minutes each day to manually scan new postings it’s not a big deal for me to copy and paste the user name and post number into my table. Post number isn’t really needed since each insert tags the row with an incrementing sequence number (in case it matters for sending out the PM’s) but I like to have it just as another data point.

I’m so glad you were able to pick this up when you did, as my work got crazy busy for a few months, and I would have struggled to keep up.

Keep up the good work! :+1:

[SIGShooter]

@sbslider

@SIGShooter

Since it only takes me a few minutes each day to manually scan new postings it’s not a big deal for me to copy and paste the user name and post number into my table. Post number isn’t really needed since each insert tags the row with an incrementing sequence number (in case it matters for sending out the PM’s) but I like to have it just as another data point.

I’m so glad you were able to pick this up when you did, as my work got crazy busy for a few months, and I would have struggled to keep up.

Keep up the good work! :+1:

Not working gives me lots of time to spend on BLF :laughing:

[el_duderino]

Can you please put me down for one more, for a total of 3 (pending final price)?

P.S. Thank you to the whole team for making this a reality, and tracking the staggering amount of interest.

[zra]

I’m interested in one, please

Thank you

[LiteBriteKnight]

Interested

[ToyKeeper]

@sbslider

There was someone that approached me while I was keeping this list, and he created a bot that would search the thread of folks who wanted to be added to the interest list. It seemed like he was close to making it work, then he disappeared.

I was disappointed that he disappeared. It seemed like a really good start on something which could help a lot of people.

I’ve been tempted to build something too, but I’ve been way too busy. The only things I’ve automated here lately are things I needed directly for projects, like sending out a couple thousand private messages.

[Ramblings]

interested for 2!