Calibre : minor suggestions

[carpii]

Just an idea for calibre..

Id like to randomly download ebooks from various sources and just save them, into a single directory.

As it stands, I then have to load them into Calibre, convert them and then upload them to the device.

Id like to see a few features to streamline all of this...

1) Ability to set a default 'single directory' path, so that clicking the Add Books button will use that path instead of prompting each time

2) Ability to watch a directory for changes, and automatically convert any books which appear here, into a set format. Ie.. watch /home/carpii/ebooks and if it finds any new books, convert them to Sony format, and tell calibre to automatically upload them to the device the next time its connected

This would make a great addition I think

Finally, a lot of open source software have a 'pledge' area, whereby you can say how much you would be willing to donate if this feature was added. This would be great, and a nice way to help encourage people to contribute to the softwares development.

Maybe I could set the ball rolling by pledging $100 for the above features?

Thanks

[kovidgoyal]

A pledge area is a good idea as there are a lot of such minor feature suggestions and I only tend to act on the ones I think are cool/important. Do you know of any webservice I can use to set it up?

[carpii]

Originally Posted by kovidgoyal

A pledge area is a good idea as there are a lot of such minor feature suggestions and I only tend to act on the ones I think are cool/important. Do you know of any webservice I can use to set it up?

I dont Im afraid

Most of the ones Ive seen seem to be just bespoke PHP scripts integrated with the site.

Heres an example..

http://www.codeweavers.com/compatibility/toplists/

[JGB]

That sounds like haxial.net's system. They have a list of all the upcoming features people want in their software, they work on the ones they want, or put them in the list where they please, if you want your feature moved up the list, you can post a bribe, if it's high enough for the amount of work, one of the developers is going to get your feature done quick

[kovidgoyal]

What service do they use to keep the payments on hold until the feature is complete?

[frandavid100]

Originally Posted by carpii

1) Ability to set a default 'single directory' path, so that clicking the Add Books button will use that path instead of prompting each time

2) Ability to watch a directory for changes, and automatically convert any books which appear here, into a set format. Ie.. watch /home/carpii/ebooks and if it finds any new books, convert them to Sony format, and tell calibre to automatically upload them to the device the next time its connected

That's a pledge I would contribute to.

[kovidgoyal]

Hmm after some googling I couldn't find any simple way to set up a bounty system. The problem is, you need an escrow service to hold the pledges and some 3rd party to decide when the feature is "complete".

[moggie]

Originally Posted by kovidgoyal

Hmm after some googling I couldn't find any simple way to set up a bounty system. The problem is, you need an escrow service to hold the pledges and some 3rd party to decide when the feature is "complete".

Given that you are the person who does most/all of the coding for calibre we could just trust you . And we could use the wiki page to hold the pledges and you could edit and sign when someone sends you cash .

Another way of putting it would be how much cash would be needed to increase the chance of implementing a feature ? I wouldn't expect it to replace the day job just yet...

For example I would pay something to allow one instance of calibre to be able to talk to another via the web service ( manually configured ) and have the servers library show up on the client and have a send to device button on the client. That way my main store would be on my server but I would connect the device up to my portable....

[kovidgoyal]

Originally Posted by moggie

Given that you are the person who does most/all of the coding for calibre we could just trust you . And we could use the wiki page to hold the pledges and you could edit and sign when someone sends you cash .

Another way of putting it would be how much cash would be needed to increase the chance of implementing a feature ? I wouldn't expect it to replace the day job just yet...

For example I would pay something to allow one instance of calibre to be able to talk to another via the web service ( manually configured ) and have the servers library show up on the client and have a send to device button on the client. That way my main store would be on my server but I would connect the device up to my portable....

You'd still need some way to hold the pledged payments and return them if it looks like a feature is not going to be implemented.

As for the client-server, why not just use the content server? Provided your device appears as mass storage device, you can just save whatever books you want directly to the device.

[moggie]

Originally Posted by kovidgoyal

You'd still need some way to hold the pledged payments and return them if it looks like a feature is not going to be implemented.

Thats where the trust bit would come from , if the sums were not too high people would be happy I suspect that you held the money and promised to return it .... Especially if it was very clear this was how it worked.

Originally Posted by kovidgoyal

As for the client-server, why not just use the content server? Provided your device appears as mass storage device, you can just save whatever books you want directly to the device.

But then it wouldn't get collection and tag information ?

[carpii]

I think the pledge system is really honor-based, in that Id promise to donate a set amount if a feature I liked was added. Then once the feature becomes available I would go ahead and donate.

Alternatively you could let people make pledges, and then once you've decided its a feature you are able to develop, you send an automated mailshot asking them to fulfill their pledge. It gets a bit complicated at this point though. What if only 3 of the 20 people who pledged actually follow through.

I dont know, the honor system is clearly open to abuse, and maybe its not that workable for Calibre. I do think it might be worth exploring though

But my pledge still stands, and Id be happy to donate for it upfront.

[kovidgoyal]

Well a trust based system is easy enough to set up. How's this:

• Open a ticket describing the wanted feature
• I'll reply saying how much it would cost to get that feature implemented
• People who want that feature can add comments with the amount they are willing to pledge
• If/when the amount reaches the target, I'll add a comment saying I'm ready to implement the feature
• All the pledgers can then send donations with a comment indicating they're for the feature and once the total is achieved, I'll start working on the feature

The two sources of risk are:

1) All the pledgers don't actually donate once the total is reached. Then those that did will have their money in limbo
2) I don't implement a feature even after receiving all pledges.

This has the benefit that you dont risk your money until enough people have pledged.

[JGB]

The way the Haxial one works is quite simple.
http://www.haxial.com/bribe/

and here is there current bribe list

http://www.haxial.com/bribe/brlist.html

I think it seems like a good idea, basically Haxial retains control over the main development, and they implement main stuff as their priority, but for features that users want, people chip in, whichever has the most money gets started on first, when that is complete the next one is started. They also have Bounty Hunters, basically people who are not Haxial but who will choose to write the code for a specific feature when the money is good enough.
This way more then one person can chip in, the list is very clear and easy to understand and all in all a lot of fun

EDIT: they collect the money in advance, no fancy software or pledging involved

[daviddem]

Since this thread is called minor suggestions, here is one: it took me three days to realize that I could convert to epub via the GUI. That is because the settings button is tiny. I think it belongs with the big buttons at the top.

[kovidgoyal]

Originally Posted by daviddem

Since this thread is called minor suggestions, here is one: it took me three days to realize that I could convert to epub via the GUI. That is because the settings button is tiny. I think it belongs with the big buttons at the top.

The reason the config button is tink is that it should not be used frequently. I'll probably implement a different UI for setting the output format soon that will not require the use of the config dialog.