Upcoming Quicksilver Sprint

[Rob McBroom]

I’ve worked out some time with my employer to dedicate exclusively to Quicksilver. It looks like I’ll be able to take off the last week of July for this purpose.

Thing is, this will be unpaid leave, so Patrick and I have also discussed me getting paid with donation money for my time during the sprint. Since money’s involved we want to have this discussion in public. The main points are the hourly rate, and the expectations.

First, what does everyone think of $50/hour?

As for what I’ll work on, here’s the list I’ve had building up in Remember the Milk. These descriptions are enough to remind me to work on something, but could be meaningless to others, so let me know if you want details on a certain point.

• add regex features to the Text Manipulation plug-in
• Allow triggers to override the default action
• Fix collection icons
• blog post about Event Triggers
• investigate new iTunes API
• Log to Console action
• look into wi-fi power management
• blog about the “smart” spacebar
• Add “sort by date”
• enhance TextMate current selection
• Unit Tests

But to be up-front, I don’t want people to expect a new feature bonanza. I would also like to spend some time on the things that always get pushed aside, like learning some fundamental AppKit and Xcode things I never got around to. I’d also like to look into how we can fund and do sprints like this more often.

Comments? Objections?

Thanks.

--
Rob McBroom
http://www.skurfer.com/

[Howard Melman]

I think this is great.

I don't mind donation money going to this. I don't have a real sense of it but $50/hr sounds like a reasonable rate. For a full week of 40 hours that's $2000. I have no idea what's in the donation pool, but I wouldn't want it depleted just for this.

As for the work items:
- I was just going to repost that I *really* miss Safari history support. It's been put off for a couple of years now and I'd think a lot of our users use Safari. Perhaps generalizing existing sqlite stuff so it could be easily applied to other things is an item for this sprint?
- iTunes stuff strikes me as very used features so new API stuff sounds good (though I don't know what it does). Is iTunes stuff that's skipped over?
- power management sounds like a good thing to be able to manipulate w/ Quicksilver, not sure what wi-fi power management is
- I don't use the Text Manipulation plugin so I'm not sure how many will use new regex features with it
- Unit tests are good and do get skipped over
- I'm not sure of the utility of a log to Console action. Seems more useful to be able to read the console via Quicksilver and then use Text Manipulation on it. Also Event Triggers :)
- I think adding Siri support for Sierra would be pretty cool. "Siri run my trigger for ..." "Siri take that last thing we spoke about and bring it into a QS window as selection" Make a collection or results list out of the latest results from Siri. Exposing all actions to Siri? A "Send to Siri" action so that I could use Siri to do something with the object?

I don't know what these are:
- Allow triggers to override the default action
- sort what by date?
- enhance TextMate current selection

I'm not sure of the value of using this extraordinary opportunity for blog posts. Particularly at $50/hr.

Howard

> --
> You received this message because you are subscribed to the Google Groups "Quicksilver - Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to quicksilver---deve...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Mike Petonic]

I've donated in the past.  Love Quicksilver. I just signed up for an hour of your time.  

Great employer, BTW!

[Patrick Robertson]

To fill in some blanks:

> I have no idea what's in the donation pool, but I wouldn't want it depleted just for this.

Right now we have $3000 USD in PayPal, €374 in Flattr (= $420 USD) and I think around £800 in the UK bank account. In total we have around $4500.

I am in agreement that we should still keep some donation money kept back for a rainy day, but am happy for most of this money to go towards paying Rob to work on QS.

We could also set up a Kickstarter or Fundraiser to try and get awareness for this and hopefully get more funds (thanks Mike for already donating $50!).

I am also not sure what $50/hr is like in real terms, but I trust that Rob’s time is worth it, and that Rob is not likely to be screwing over the QS Org, so am happy to accept whatever rate he considers acceptable for living costs.

As for what Rob should do with his time. I think it’s a good idea not to go crazy with new features, but I do think that since this money is coming from QS donations, we should use the money responsibly for what the majority of people want to see done. Perhaps we can look at the most commented issues on GH and use that as a system for ranking things to fix/implement? Rob should definitely have a say though on what he works on - it’s his time after all! - but I do think we should come up with a more methodical and accountable method for choosing what to work on.

For example, I feel like regex features for the text manipulation plugin is a "nice to have" new features, but there are perhaps other areas we could develop that will benefit a larger majority of our users.

I don’t know a whole lot about the ‘Django Foundation’ (the foundation behind the Python Django project), but they have a Django fellowship[1] that from what I’ve seen seems to work well. Perhaps we can explore this and see how they deal with the following:

* The pay during the fellowship period

* How to decide on what tasks to undertake

* Accountability and how to measure the success/outcome of the fellowship

* Funding the fellowship - how do they raise money?

[1] https://www.djangoproject.com/fundraising/#fellowship-program

[Rob McBroom]

On 24 Jun 2016, at 0:41, Mike Petonic wrote:

> I've donated in the past. Love Quicksilver. I just signed up for an
> hour
> of your time.

Thanks!

> Great employer, BTW!

Indeed. They’re also hosting qsapp.com for free.

A couple of additional comments on what to work on…

I’ll definitely be looking at these.

https://github.com/quicksilver/Quicksilver/issues/assigned/skurfer

Some are very small.

I should probably spend as much time as I can on things you’ll
actually notice. I’m at a loss as to how we could figure out what
people are most interested in, but since you have the power Paddy, feel
free to skim the list of open issues and assign more to me if you see
something obvious.

The to-do list I shared was just the whole list with no priority
attached. I won’t necessarily be looking at all of it.

[Mike Petonic]

On Tuesday, June 28, 2016 at 12:53:07 PM UTC-7, Rob McBroom wrote:

> Great employer, BTW!

Indeed. They’re also hosting qsapp.com for free.

Great stuff!  

 

A couple of additional comments on what to work on…

I’ll definitely be looking at these.

https://github.com/quicksilver/Quicksilver/issues/assigned/skurfer

Some are very small.

I should probably spend as much time as I can on things you’ll
actually notice. I’m at a loss as to how we could figure out what
people are most interested in, but since you have the power Paddy, feel
free to skim the list of open issues and assign more to me if you see
something obvious.

Power Paddy :-)  Heh.

No particular preference, use your best judgement, as you do.  Well, on second thought, I wouldn't mind having a mechanism that changes the sort order of files and directories in the most elegant, extensible way possible.  I'm not sure what I mean, but not have a particular mechanism hard-coded in.  I'm sure that goes without saying (since it's the approach and philosophy of QS anyway).

When I'm looking at MindMaps or other files, I want to sort by last-modified-time because that's my stack.  

It'd be great, while the list of files/dirs are still open to be able to switch the sort-order in case you go into a directory that you usually traverse alphabetically, for instance.  Having that preference remembered on a per-directory basis would be great, too.

Not a biggie, but not something that I could even think about coding because it gets into the heard of the QS util itself.

Maybe when QS is extensible via Swift (is it yet?), I'll take a crack.  Just looking at the SmallTalk syntax of ObjectiveC gives me cancer.

Thanks for the ongoing development, Rob (and rest of the QS dev team).

[Patrick Robertson]

> you have the power Paddy, feel  free to skim the list of open issues and assign more to me if you see something obvious. 

Hehe, power! I’m going through and assigning a few issues to you (and also closing a load in the process which is great).

There’s nothing glaringly obvious that I’ve seen so far, you  an obviously ignore the ones I’ve assigned if you don’t agree with the priority.

I think the list of assigned issues you have is already pretty good.

Other things:

* I created one more issue on deprecated methods - I think it’s been a while since we’ve done some ‘internal cleanup’ and removed warnings etc., but I think now would be a good opportunity to go through and look for deprecated methods so they don’t bite us when a new OS release drops support for them.

* I will definitely try and be more available during the sprint period, and I think it would be great if the other core QS Devs were also available - it’d be great if we can shift some of the open pull requests as well!

* Shall I start a Kickstarter campaign or similar, then try and get the word out on this?

[Howard Melman]

> On Jul 3, 2016, at 7:34 PM, Patrick Robertson <robertso...@gmail.com> wrote:
>
> * I will definitely try and be more available during the sprint period, and I think it would be great if the other core QS Devs were also available - it’d be great if we can shift some of the open pull requests as well!

IRC during the sprint or is there a better way? (Slack?) I can' help with code, but will do whatever I can.

Howard

[Rob McBroom]

On 3 Jul 2016, at 23:05, Howard Melman wrote:

> IRC during the sprint or is there a better way? (Slack?) I can' help
> with code, but will do whatever I can.

I was thinking we should set up Slack anyway. Might encourage more
people to participate. Any objections?

[Patrick Robertson]

> I was thinking we should set up Slack anyway. Might encourage more people to participate. Any objections?

I’m using it for work. Seems useful, I say go for it.

I’ve now got a bit more time on my hands, so do people think setting up a kickstarter is a good idea? We can set a target of $2000 - for 40 hrs of Rob’s work - and take it from there.
If we go ahead then we’d need some kind of blurb and ‘rewards’ for donating. I was thinking for the rewards we could have something along the lines of:

* $20: A coffee break for Rob. Eternal thanks from the QS community
* $50: 1hr of Rob’s time: Your name on the Quicksilver supporters' list (distributed with every release of QS) - it’d go under the ‘credits’ section of our ‘Credits.html’ page
* $100: 2hrs of Rob’s time: A free Quicksilver t-shirt AND your name on the Quicksilver supporters' list (this might be the final kick in the pants I need to execute this!
* $200: 4hrs of Rob’s time: A say on what Rob spends these 4hrs on. We already have a list of things the Quicksilver development team considers top-priority, have your say on what you think should be done from this list (I’m not sure on this one, it’s starting to sound scarily like a bounty, but restricted to what we’ve already screened).

Rob: This week sometime I’ll go ahead and look at what you asked me to look at before the sprint… although I can’t seem to find the email. Was there a priority of which PR you wanted checking over?

[Rob McBroom]

On 11 Jul 2016, at 21:00, Patrick Robertson wrote:

I was thinking we should set up Slack anyway. Might encourage more people to participate. Any objections?

I’m using it for work. Seems useful, I say go for it.

OK. I can work on that tomorrow.

I’ve now got a bit more time on my hands, so do people think setting up a kickstarter is a good idea?

We have enough money now, but it might be good to keep it that way for the next sprint. It might also draw more attention to the cause, but I’m a little scared by that. I don’t want expectations to get out of hand.

I have no idea how much work is involved in such a thing, or if two weeks is enough time to accomplish anything, so I don’t want to make the decision for you. I suppose, if we’re trying to be ready for future sprints, the campaign can go beyond the start of this one so maybe the short notice isn’t a concern.

We can set a target of $2000 - for 40 hrs of Rob’s work - and take it from there.

If we go ahead then we’d need some kind of blurb and ‘rewards’ for donating. I was thinking for the rewards we could have something along the lines of:

* $20: A coffee break for Rob. Eternal thanks from the QS community

* $50: 1hr of Rob’s time: Your name on the Quicksilver supporters' list (distributed with every release of QS) - it’d go under the ‘credits’ section of our ‘Credits.html’ page

* $100: 2hrs of Rob’s time: A free Quicksilver t-shirt AND your name on the Quicksilver supporters' list (this might be the final kick in the pants I need to execute this!

* $200: 4hrs of Rob’s time: A say on what Rob spends these 4hrs on. We already have a list of things the Quicksilver development team considers top-priority, have your say on what you think should be done from this list (I’m not sure on this one, it’s starting to sound scarily like a bounty, but restricted to what we’ve already screened).

I think those are all good ideas. Just two points:

• Can it be less about me personally? :-) I know I’m dedicating time for this round, but hopefully the money we raise today will end up going to others for similar reasons.
• If we let donations influence the work, I think we need to limit the options more. But then I think, the things that make that list are probably things we want to work on anyway, so maybe there’s no point?

Rob: This week sometime I’ll go ahead and look at what you asked me to look at before the sprint… although I can’t seem to find the email. Was there a priority of which PR you wanted checking over?

I think you’re referring to this comment:

https://github.com/quicksilver/Quicksilver/issues/1277#issuecomment-230484839

Mainly just reviewing the points in the pull request. It’ll need to be redone, but the basic idea hasn’t changed.

[Howard Melman]

> On Jul 11, 2016, at 11:46 PM, Rob McBroom <mailin...@skurfer.com> wrote:
>
> On 11 Jul 2016, at 21:00, Patrick Robertson wrote:
>
>> I’ve now got a bit more time on my hands, so do people think setting up a kickstarter is a good idea?
>
> We have enough money now, but it might be good to keep it that way for the next sprint. It might also draw more attention to the cause, but I’m a little scared by that. I don’t want expectations to get out of hand.
>
> I have no idea how much work is involved in such a thing, or if two weeks is enough time to accomplish anything, so I don’t want to make the decision for you. I suppose, if we’re trying to be ready for future sprints, the campaign can go beyond the start of this one so maybe the short notice isn’t a concern.
>
>
>> We can set a target of $2000 - for 40 hrs of Rob’s work - and take it from there.
>
>> If we go ahead then we’d need some kind of blurb and ‘rewards’ for donating. I was thinking for the rewards we could have something along the lines of:
>
>> * $20: A coffee break for Rob. Eternal thanks from the QS community
>> * $50: 1hr of Rob’s time: Your name on the Quicksilver supporters' list (distributed with every release of QS) - it’d go under the ‘credits’ section of our ‘Credits.html’ page
>
>> * $100: 2hrs of Rob’s time: A free Quicksilver t-shirt AND your name on the Quicksilver supporters' list (this might be the final kick in the pants I need to execute this!
>
>> * $200: 4hrs of Rob’s time: A say on what Rob spends these 4hrs on. We already have a list of things the Quicksilver development team considers top-priority, have your say on what you think should be done from this list (I’m not sure on this one, it’s starting to sound scarily like a bounty, but restricted to what we’ve already screened).
>
> I think those are all good ideas. Just two points:
>
> • Can it be less about me personally? :-) I know I’m dedicating time for this round, but hopefully the money we raise today will end up going to others for similar reasons.
> • If we let donations influence the work, I think we need to limit the options more. But then I think, the things that make that list are probably things we want to work on anyway, so maybe there’s no point?

I've funded several but not setup any so FWIW... Maybe gofundme.com is a better option for this round. You can set any timeframe you want, I haven't seen a Kickstarter for under 28 days. Also it's more about donations rather than reward levels and it's not all or nothing on reaching a goal. Also I think Kickstart donators typically want a specific "thing" at the end and not just "hey we fixed these 10 bugs...". But I don't know anything about the fees or setup stuff for any of these sites.

Howard

[Rob McBroom]

On 12 Jul 2016, at 10:14, Howard Melman wrote:

> I've funded several but not setup any so FWIW... Maybe gofundme.com is
> a better option for this round. You can set any timeframe you want, I
> haven't seen a Kickstarter for under 28 days. Also it's more about
> donations rather than reward levels and it's not all or nothing on
> reaching a goal. Also I think Kickstart donators typically want a
> specific "thing" at the end and not just "hey we fixed these 10
> bugs...". But I don't know anything about the fees or setup stuff for
> any of these sites.

There’s also IndieGoGo, which MailMate used recently for a very
similar purpose.

[Mike Petonic]

On Wednesday, June 22, 2016 at 8:40:12 PM UTC-7, Rob McBroom wrote:

But to be up-front, I don’t want people to expect a new feature bonanza. I would also like to spend some time on the things that always get pushed aside, like learning some fundamental AppKit and Xcode things I never got around to. I’d also like to look into how we can fund and do sprints like this more often.

Comments? Objections?

Comment: Agreed.  As a contributor of a whole hour (!!!) of Rob's time, I agree.  I think that we need to reinvest in the internals of QS so that new features that ultimately *do* get added on don't need to be constructed on to of spaghetti code.

For instance, I was writing an Applescript action to *finally* help me solve my "sort by last modification date" issue.  Turns out that the capability to do this was present all along. I can write an Applescript action that takes a directory in the first-pane and then returns the contents of that directory in last-modified sort order.  And I wanted an optional third pane that allowed me to specify how many most-recently-modified files to return (default of entire directory).  Well, turns out there's a bug from about a year ago where optional third-pane arguments are pretty random if one isn't supplied ( Problems with three-pane AppleScript actions #2071 ).

QS has got a great modular architecture, but some internal things must be patched up first.  And it's this modular architecture that would allow us to extend QS w/o having to recompile QS.  That's nirvana.

My recommendation: Fix the stuff that would empower users to extend QS without having to recompile QS.  Or add features in (if they're lacking).  But we should worry about superfluous stuff that are one-offs that probably should be fixed by solving for the general case instead of the specific.

 

[Mike Petonic]

On Sunday, July 17, 2016 at 5:09:32 PM UTC-7, Mike Petonic wrote:

For instance, I was writing an Applescript action to *finally* help me solve my "sort by last modification date" issue.  Turns out that the capability to do this was present all along. I can write an Applescript action that takes a directory in the first-pane and then returns the contents of that directory in last-modified sort order.  And I wanted an optional third pane that allowed me to specify how many most-recently-modified files to return (default of entire directory).  Well, turns out there's a bug from about a year ago where optional third-pane arguments are pretty random if one isn't supplied ( Problems with three-pane AppleScript actions #2071 ).

QS has got a great modular architecture, but some internal things must be patched up first.  And it's this modular architecture that would allow us to extend QS w/o having to recompile QS.  That's nirvana.

Dang, Guys.  Rob, did you start your sprint yet?  Because I was browsing Github and noticed that you fixed this.  **impressed**.  Thank you.

Thank you!

[Patrick Robertson]

Just to make a final word on crowdfunding: I think it’s a very good idea for the future, but agree with Rob that this is perhaps not a time when we’d need to use it.

We have $2,000+ in our PayPal account, so we are OK for this round. I still think it’d be worth creating a blog post and sharing it around on Twitter etc. I’ll get on that (Rob - I’ll make sure to keep it less about you personally)

[Rob McBroom]

On 19 Jul 2016, at 20:15, Mike Petonic wrote:

> Rob, did you start your sprint yet?  Because I was browsing Github
> and noticed that you fixed this.

Not yet. And I didn’t fix that, just reviewed and merged it. 😀

[Rob McBroom]

On 22 Jun 2016, at 23:40, Rob McBroom wrote:

> Thing is, this will be unpaid leave, so Patrick and I have also
> discussed me getting paid with donation money for my time during the
> sprint. Since money’s involved we want to have this discussion in
> public. The main points are the hourly rate, and the expectations.

My total for last week (M-F only) was 30.5 hours. Not the 70 hour
marathon I was hoping for, but better than 30 minutes here and there.

I say cap it at 24 hours. I wasn’t doing it for the money, I just need
to be able to justify a week away from work. Sound good?

[Patrick Robertson]

A meagre 30.5 hours eh?

I’m good with paying for the full 30.5 hours if that is what you worked, but whatever you feel is justifiable for yourself personally, then that should also work as well.
The more we have in the bank now the sooner we could potentially do another sprint… with more of us involved!

So 24 hours x $50 would be US$1200. If we calculated for the whole 30.5 hours it’d be $1525. Both are fine I think.
Is sending the money via PayPal the easiest thing for you? Quicksilver will have to pay about a 5% fee, but that’s the pain of using PayPal in the first place…
I’ve just done some searching, and it seems you *can* add a US bank account to a UK PayPal account by calling them up. I will try and get on to this so that it can directly be deposited into your account, avoiding the fee.
Rob: send me your bank details when you get a chance and I’ll look into this.

[Rob McBroom]

On 11 Aug 2016, at 23:04, Patrick Robertson wrote:

> I’m good with paying for the full 30.5 hours if that is what you
> worked, but whatever you feel is justifiable for yourself personally,
> then that should also work as well.
> The more we have in the bank now the sooner we could potentially do
> another sprint… with more of us involved!

Agreed. Cap it at 24.

I contacted you in Slack about the PayPal stuff. Thanks.

[Mike Petonic]

Attaboy, Rob!  Thanks for the extra time into the project.  It's just a token compared to what you just donated on this sprint, but I just bought another hour of your (or someone else's) time for the next sprint!

[Rob McBroom]

On 16 Aug 2016, at 23:28, Mike Petonic wrote:

> It's just a token compared to what you just donated on this sprint,
> but I just bought another hour of your (or someone else's) time for
> the next sprint!

Thanks (again)!

[Patrick Robertson]

Thanks for the support Mike!

A UK PayPal account cannot add a US bank account (unfortunately), so I have transferred the money directly via PayPal to Rob. There was a cross-border fee of 1%, so the total money out was $1,212 ($1,200 + the $12 fee)

Rob: this should all be in your PayPal account now.

As things stand we have US$1,987.24 in PayPal. That should be enough for another sprint! ;-)

[Mike Petonic]

On Friday, August 19, 2016 at 5:44:26 AM UTC-7, Patrick wrote:

Thanks for the support Mike!

You and Rob (and the entire active dev team for QS) are most welcome.  I'll contribute in other ways when I get a little more time on my hands.  I'm kind of itching to do a plugin for LastPass.  Took a quick look at the build instructions (thanks, Rob) and when I have a bit of time, I'll try to contribute to the code base.