Are conferences good value for money (as a use for free software donations/funding)? Like I get that they're good, I like attending conferences. But if I could choose between employing someone full-time for a month (especially somewhere with low cost of living like India or Eastern Europe) and buying flights and a hotel room for a weekend, I think I'd choose the former.
Conferences are essential in building a healthy community - and seeing who you actually "work for", especially as an open source developer. Sure, you can sit on Zoom for an hour or two and hold a talk, but you really can't beat the face-to-face experience.
I think it's something you need to experience to see the value in.
In my experience, conferences are a very nice to have, unique experiences, interesting people. But productivity? No. When it comes to being productive, I get far more done sitting at my desk at home. I totally understand that conferences are a really nice thing to go to for contributors. But its more a sort of compensation for unpaied work. At some point, you are invited to go to a foreign country, and you get a sort of geek-holiday. Thats nice. But I never considered it essential for my contributions.
In fact, my impression actually turned around when the "bring your children" thing started to be normal. I vividly remember a few talks at DebConf where I was not able to follow the speaker because of verbal small children in the talk room. Since then, I really really prefer to listen to talk recordings at home.
I always find conferences to be a real boost for productivity personally. You always come out full of ideas and papers to read or implement. The motivation and passion of others can be contaminating as well.
We don't go to face-to-face meeting for papers. We come for live interaction, major release level decision-making etc. We tried video conferencing a few times. It's fun but hard to organize (timezones), too messy, and time-consuming. So are chats.
> productivity? No. When it comes to being productive, I get far more done sitting at my desk at home.
It depends. For creating production-quality code certainly true.
But for design discussions, feature priotisation etc. an in person meeting often is better. Also such events /can/ lead to interesting results from hacking sessions where people collaboratively spin out ideas and code stuff. That code typically needs cleanup, but can spark ideas. Way home from a conference for me often was the time where I had creative new ideas :)
I don't know about conferences per-say, but something like LibreGraphicsMeeting is, as people developing lots of open source graphics get together and swap ideas.
Luckily there is one there has been some funding, at least for travel.
> But if I could choose between employing someone full-time for a month (especially somewhere with low cost of living like India or Eastern Europe) and buying flights and a hotel room for a weekend, I think I'd choose the former.
A month's worth of a bad programmer's time might be worth a lot less than meeting motivated good software engineers at a conference and potentially turning one or two into contributors.
Definitely fund these guys. I really hope they can get to the point where they are paid as well as those of us that work in commercial software development.
As nice as Photopea is, it's browser only. Furthermore, GIMP provides a variety of different plugins, extensibility options, hardware support for graphics tablets etc.
So the answer is yes, GIMP is still a lovely piece of free software that's utilized by numerous people out there to great effect.
It's gratis, but not libre. Some people value this attribute of a software, choosing their toolkit exclusively or primarily so. And for these people, a nonfree software is a not a replacement for a free software.
Also yes, because it's only distributed as a web app, one day it might vanish and that's it. If it were AGPL licenced, for example, anyone could fire up another instance or embed it via Electron, and the software would live on.
Indeed. I really want to like GIMP, but it doesn't hold a candle to my old copy of Photoshop CS4 from ten years ago. From what I can recall, the GIMP UI has hardly changed since then too.
I suppose they're probably constrained by what is possible with GTK? Basically it seems like Linux needs a better standard for user interfaces -- I can't think of a GTK app that doesn't have awkard position/spacing/layout issues.
> From what I can recall, the GIMP UI has hardly changed since then too.
There have actually been changes, and even if they aren't major, someone like me who has never used anything but GIMP was annoyed/slightly confused by them. I don't know what the issue with GIMP is, besides tools that Photoshop appears to have or has better implementations for. My impressions is that it is mostly Photoshop users, that are expecting something else, or am I mistaken?
> My impressions is that it is mostly Photoshop users, that are expecting something else, or am I mistaken?
I use GIMP for my (modest) editing needs, and I like it. But some of the behaviour is.... weird.
For example, if you've got a region selected and you use the flood fill tool, it will only fill pixels within that region. OK, that's logical enough, a tool only applies to selected pixels, right?
But if you've got a region selected and you use the move tool, it'll move the entire layer. If you want to move only the pixels in the selected region, you gotta convert it to a floating selection.
And after you've moved your floating selection, you want to select a different region to move, so you drag out a new selection. Not so fast, your floating region has now vanished. You gotta defloat the selection by selecting nothing. But you can't do that by choosing 'None' in the 'Select' menu, that's greyed out for some reason.
I'm sure these behaviours all make sense to a developer or expert user - but nobody's going to be telling their friends that GIMP is easy to pick up :)
I think I’ve seen GIMP’s “move tool moves the frame” behavior on somewhere else, perhaps in ancient versions of Photoshop. Back in the time GIMP UI was close enough to those, like in the legendary CS2 or v5.5 days.
Krita has a different focus though. What photoshop is to Illustrator, is GIMP to Krita.
I use both and I think GIMP is really usable and the UI isn't that bad. There were some changes though, so people should take a look at the latest version.
I'm guessing that Blender has more funding available to it and/or has retained more of its senior developers, compared to GIMP.
Expecting that a piece of software undertakes an ambitious UI overhaul (which is no small feat) without one or both of those resources shows, to me, a lack of understanding.
It should not be underestimated how much Blender is helped by the project management by Ton.
As far as I know GIMP does not have a face for the press, does not organize conferences, has no one contacting businesses for funding. And it has an unfortunate joke name.
I agree. Two things come to mind regarding the interface of GIMP.
One is what you wrote - that it needs a product manager and tons of manhours just to do what's essentially a refactor. Not to mention that GIMP devs said that many UI changes would imply backend changes too, because of limitations. And they are working on those limitations, among other things.
Second part of the point is that Adobe must have like 50 people work on Photoshop, and I guess even more when you consider the overlaps with their other products. Compare this to the active contributor number to GIMP.
The second point is that many people compare it to Photoshop like Photoshop is a gold standard of usability. It's just as a nightmare as GIMP if you're not used to it. So, it's not really GIMP's fault that people are used to something else.
I think the fact that Blender has had same person steering its development since the days when it was an in-house tool for a small animation studio makes a tremendous difference.
The other factor is that Blender was always a tool by and for actual companies doing real commercial work and understands those needs. GIMP on the other started as a university project that got picked up the GNU project, and has always been more of a 'research' project.
From my historic experience with both, the GIMP project always seemed actively hostile to UI change suggestions, and even keeping their name is a thumb in the eye to going mainstream. Blender has been more about being a good product from the very start, I've been following it since before it was open sourced. It's current success is a path it started walking way before it had lots of devs and resources.
Blender was extremely hostile to UI change suggestions as well until only recently when they changed things drastically with the insanely successful Blender 2.8 project. From then on donations (including corporate sponsors) grew massively. Blender has made massive strides in the past few years, and much of it can be attributed to them changing their stance on UI/UX changes.
Almost not joking: I wonder if it wouldn't be better if Blender grew image editing features. After all, it now has video editing features, and I hear they're decent.
I just noticed they don't even have screenshots on their website. That's horrible in my opinion. I'm not using a program where I have to actively search for 3rd party websites to find some pictures of the UI.
This. This is the biggest reason why I don't use GIMP more often. The tooling is great, but the usability is atrocious. I would give them all the money for a UI overhaul.
Photoshop and Illustrator's UIs are so illogical, arbitrary and tooltip-free they give skilled users Stockholm syndrome. GIMP's UI is logical and good, and you have 10x the odds of finding something you've never looked for before quickly than you do with Photoshop.
I've never used Photoshop so can't comment on that, but I've always found GIMP's UI to be simple and straightforward.. but maybe that's because I've used it for so long. I really don't have a problem with it.
I'd rather GIMP focused on adding features that Photoshop and other competitors have.
There's also some low-hanging fruit if someone wants to contribute. E.g. the developer docs are rather outdated. I recently implemented AMD FSR as a Gimp plugin but all the docs led to code that resulted in a slew of deprecation warnings. Also poor performance and some thing just not working right at all (tile cache). I kinda got frustrated with it and dropped the gimp parts altogether, but if I had more time, it shouldn't be hard to figure out and maybe improve the docs or write an up to date tutorial.
I think you're not the first person facing this issue. If you feel like it, organizing a hackathon around it, so that you and others can update the docs and polish your respective plugins would be a huge contribution! Speaking as someone who is not a GIMP contributor myself.
As an occasional user of GIMP I'll gladly donate to the developers. This is a piece of software that has made me money, so it is only fair I give back.
Biggest mistake of free software: It's too much hassle to donate. Show me a PayPal yearly subscription confirmation dialog on first startup for something less than 10 dollars, and if it takes 2 clicks at most and less than 10 seconds, you will get my money.
The expression itself is fine, but the original (something like "Free software isn't worth the hassle of forking over") was a confusing collision of terms. It took me a while to work out what meaning they intended. I'm generally relaxed about alternative grammar.
This is something that can easily be argued. Not every project wants a donation, for one. Management of donations is also not straightforward - in case of GIMP, the project donations are managed by Gnome.
Second point: why show it on the first start? People don't trust or like the software at first, so they would just click X and begin using the software. Some would even think it's intrusive.
Rather, I've seen solutions where the software would ask for money right after it's been useful to the user. Duplicati for example would tell you to donate, suggested amounts, right after it successfully restored a backup.
> in case of GIMP, the project donations are managed by Gnome.
I'm sorry, I believe the article directly contradicts your statement since its mentioned they receive donations outside of GNOME, in platforms such as Liberapay, and it's the only way to pay for actual development.
It's complicated. The GIMP donation page lists multiple donation avenues, but if you wish to donate to the GIMP project in general, that's the one that's managed by Gnome. I was referencing this paragraph: https://www.gimp.org/donating/#donate-to-the-project
UI and UX are two very different things. UI is any interface an user might use, be it a app on a phone screen, voice commands etc.
User experience is the study in how to optimize an users' experience, so that they can solve issues or tasks faster, easier or better. UX might affect how an app behaves or how it looks but it isn't tightly tied into the visuals of an application. Aesthetics are just a small part of UX but usually a rather big part in UI.
> Second point: why show it on the first start? People don't trust or like the software at first
No one donates to software they don't trust. By the time they read the donation banner and consider it, its because they use the software and trust it.
This should be as direct and hassle free as possible, I agree with OP.
I found it really easy to pay them 10$ per month. Clicked the paypal "P" in the top right corner and set it up in a few clicks. The main busy work was logging in to Paypal since I don't stay logged in there and use 2FA.
Well, it's not exactly THAT simple, because first as a project you need to have a registered non-profit organization as well as a bank account. Then, you need to be registered with Paypal, and then once everything is setup and you've chosen Paypal as your main avenue of donations, they're going to shut you down because everyone doing something remotely good gets shut down from that capitalist cauldron of evil.
So in fact you need to be on Liberapay, Patreon, Tippee, OpenCollective, Paypal, HelloAsso and more in order to be sure you're not going to lose your source of donations from one day to the next. I mean if only every finance services provider was as transparent and benevolent as Liberapay, all would be fine. But we live in a grim world, and even Liberapay was almost shut down due to not finding a suitable finance service provider after Mangopay threw them out.
That's why i truly appreciate when some established non-profits help gather funds for smaller projects. For example, the disroot.org hosting coop redirects some funds to free-software projects they either use directly or encourage their users to use. What i appreciate even more, is when i can donate cash in person to volunteers during free-software conferences, and/or buy a souvenir like a nice t-shirt... That's truly the only way i know where my money is going, cryptocurrencies aside.
How is this any different? Some people don't have PayPal accounts. Is one proprietary, closed-source platform better than the other in this scenario? I'm not a Patreon user (but have an old account), but at least they don't require a "real" phone number, and don't have a terrible CAPTCHA system to log in.
Well if you want donations from the most people possible, then yes PayPal is a way better option just on popularity alone - 6 versus 377 million users, or 491mil for GoFundMe, who doesn't require an account to donate.
Some people have had PayPal accounts for a decade plus, like me. I'm really not about to make a transaction with a random company simply cause I don't need to enter my phone number.
1. Paypal
2. Patreon
3. Stripe
4. Bitcoin, XLM, etc. Just do all the cryptos.
5. Contacts so larger donors can work out large donations or recurring donation plans via other methods.
I actually get this sentiment. I don't find too much use in GIMP specifically anymore. darktable/RawTherapee can more than pull their weight in photo editing and can handle RAW non-destructively. Krita is better optimized for drawing. Hugin stitches and stacks images better. You may as well use ImageMagick for resizing. GIMP being a jack of all trades, has left it the master of none for me. Maybe had they had more funds 6-8 years ago, it would be in a better spot but right now, I'd rather throw donations at these other projects.
I’ve donated to GNOME before, but like this new website says, that goes to GIMP as a way to spend on community needs like hardware and conferences. I’ll have to set up Patreon, because directly contributing to Øyvind Kolås’s actual ability to live and contribute sounds like a much better goal.
I'm confused by what appears to be a contradiction at the end.
These are consecutive sentences:
> Note that it is also possible to fund several contributors through GIMP Liberapay account as an interesting alternative.
> What these donations through GNOME still cannot do is funding paid development, so if that’s what you want, please fund the developers directly as explained above.
I would rather put money into a single pot for GIMP contributors than have to chose and donate to individual contributors. Can I do that through the liberapay link (https://liberapay.com/GIMP/)? Most things suggest that I can, but the text makes it a little confusing.
Has anyone reached out to the two creators of the GIMP?
They're worth many hundreds of millions of dollars from their time at Google, a successful exit to Square, and their eventual exit with Cockroachdb (and they deserve every penny of it)
As someone else noted, there is a no-code startup that just raised 50M and GIMP struggles to fund two developers.
I have often thought that one of the most effective ways, if I was say a Texan oil billionaire, most effective ways of spreading my world view would be to pay people to code useful free government and non-gov tools - all software is opinionated- as texan billionaire i would want it to have my opinions
Re: Texan Billionaire I just remembered the second Harry Palmer film.
I feel like the distributions have completely dropped the ball in this regard.
They're positioned ideally to at least have some measure of what packages are most frequently installed. Even with the use of volunteer public mirrors that share no access logs with the distributor, the package management architecture could be structured to consult distro-controlled metadata servers in locating the mirrors and preserving some visibility into which packages are most installed. They control the package tooling, they should have considered this.
With even vague knowledge of which packages are most used, they could be distributing donations to those upstream projects. They obviously already have the dependency graph as well, so even without popularity metrics, can already know which packages are cornerstones (kernel, libc, openssl, systemd, mesa immediately come to mind).
The distro installers and package management tools should be providing convenient ways for end-users to donate via whatever mechanisms, paypal, patreon, bitcoin, etc. Just give me one mechanism to throw money at my distro of choice with the clear indication that it will be dispersed to upstream developers of important software. Also give me an easy way to explicitly allocate my funds to specific packages if I am so inclined. Make it all transparent on a web site where the money goes on a high level, how much is coming in, and which projects are getting what %age of that, what % is administration overhead, etc.
99 comments
[ 3.6 ms ] story [ 155 ms ] threadI think it's something you need to experience to see the value in.
In fact, my impression actually turned around when the "bring your children" thing started to be normal. I vividly remember a few talks at DebConf where I was not able to follow the speaker because of verbal small children in the talk room. Since then, I really really prefer to listen to talk recordings at home.
is it really what GIMP needs though ?
It depends. For creating production-quality code certainly true.
But for design discussions, feature priotisation etc. an in person meeting often is better. Also such events /can/ lead to interesting results from hacking sessions where people collaboratively spin out ideas and code stuff. That code typically needs cleanup, but can spark ideas. Way home from a conference for me often was the time where I had creative new ideas :)
Luckily there is one there has been some funding, at least for travel.
But is the article about conferences? I thought it was about directly supporting some of the developers of GIMP, perhaps I misread
A month's worth of a bad programmer's time might be worth a lot less than meeting motivated good software engineers at a conference and potentially turning one or two into contributors.
So the answer is yes, GIMP is still a lovely piece of free software that's utilized by numerous people out there to great effect.
Also yes, because it's only distributed as a web app, one day it might vanish and that's it. If it were AGPL licenced, for example, anyone could fire up another instance or embed it via Electron, and the software would live on.
I suppose they're probably constrained by what is possible with GTK? Basically it seems like Linux needs a better standard for user interfaces -- I can't think of a GTK app that doesn't have awkard position/spacing/layout issues.
There have actually been changes, and even if they aren't major, someone like me who has never used anything but GIMP was annoyed/slightly confused by them. I don't know what the issue with GIMP is, besides tools that Photoshop appears to have or has better implementations for. My impressions is that it is mostly Photoshop users, that are expecting something else, or am I mistaken?
I use GIMP for my (modest) editing needs, and I like it. But some of the behaviour is.... weird.
For example, if you've got a region selected and you use the flood fill tool, it will only fill pixels within that region. OK, that's logical enough, a tool only applies to selected pixels, right?
But if you've got a region selected and you use the move tool, it'll move the entire layer. If you want to move only the pixels in the selected region, you gotta convert it to a floating selection.
And after you've moved your floating selection, you want to select a different region to move, so you drag out a new selection. Not so fast, your floating region has now vanished. You gotta defloat the selection by selecting nothing. But you can't do that by choosing 'None' in the 'Select' menu, that's greyed out for some reason.
I'm sure these behaviours all make sense to a developer or expert user - but nobody's going to be telling their friends that GIMP is easy to pick up :)
I use both and I think GIMP is really usable and the UI isn't that bad. There were some changes though, so people should take a look at the latest version.
I mean, that's the theory, but I just find photo editing to be all around much more comfortable in Krita despite using GIMP for years.
Also one of Photoshop's main domain is digital painting (https://www.google.com/search?q=making+digital+painting+with...) which is what Krita focuses on. You don't really draw with Illustrator.
I meant Fresco instead of Illustrator, I forget the names of Adobe apps constantly.
Expecting that a piece of software undertakes an ambitious UI overhaul (which is no small feat) without one or both of those resources shows, to me, a lack of understanding.
As far as I know GIMP does not have a face for the press, does not organize conferences, has no one contacting businesses for funding. And it has an unfortunate joke name.
One is what you wrote - that it needs a product manager and tons of manhours just to do what's essentially a refactor. Not to mention that GIMP devs said that many UI changes would imply backend changes too, because of limitations. And they are working on those limitations, among other things. Second part of the point is that Adobe must have like 50 people work on Photoshop, and I guess even more when you consider the overlaps with their other products. Compare this to the active contributor number to GIMP.
The second point is that many people compare it to Photoshop like Photoshop is a gold standard of usability. It's just as a nightmare as GIMP if you're not used to it. So, it's not really GIMP's fault that people are used to something else.
The other factor is that Blender was always a tool by and for actual companies doing real commercial work and understands those needs. GIMP on the other started as a university project that got picked up the GNU project, and has always been more of a 'research' project.
We ask for rationalization of suggested changes. For some reason people think it's hostile.
The GIMP's weakness is color management.
I'd rather GIMP focused on adding features that Photoshop and other competitors have.
https://www.dropbox.com/s/tll8m18gyh1o1is/gimp-contributor-o...
I'd appreciate actionable feedback on that plan.
:)
Second point: why show it on the first start? People don't trust or like the software at first, so they would just click X and begin using the software. Some would even think it's intrusive.
Rather, I've seen solutions where the software would ask for money right after it's been useful to the user. Duplicati for example would tell you to donate, suggested amounts, right after it successfully restored a backup.
I'm sorry, I believe the article directly contradicts your statement since its mentioned they receive donations outside of GNOME, in platforms such as Liberapay, and it's the only way to pay for actual development.
A perennial issue for GIMP, at least their branding is consistent though :)
User experience is the study in how to optimize an users' experience, so that they can solve issues or tasks faster, easier or better. UX might affect how an app behaves or how it looks but it isn't tightly tied into the visuals of an application. Aesthetics are just a small part of UX but usually a rather big part in UI.
No one donates to software they don't trust. By the time they read the donation banner and consider it, its because they use the software and trust it.
This should be as direct and hassle free as possible, I agree with OP.
So in fact you need to be on Liberapay, Patreon, Tippee, OpenCollective, Paypal, HelloAsso and more in order to be sure you're not going to lose your source of donations from one day to the next. I mean if only every finance services provider was as transparent and benevolent as Liberapay, all would be fine. But we live in a grim world, and even Liberapay was almost shut down due to not finding a suitable finance service provider after Mangopay threw them out.
That's why i truly appreciate when some established non-profits help gather funds for smaller projects. For example, the disroot.org hosting coop redirects some funds to free-software projects they either use directly or encourage their users to use. What i appreciate even more, is when i can donate cash in person to volunteers during free-software conferences, and/or buy a souvenir like a nice t-shirt... That's truly the only way i know where my money is going, cryptocurrencies aside.
Some people have had PayPal accounts for a decade plus, like me. I'm really not about to make a transaction with a random company simply cause I don't need to enter my phone number.
These are consecutive sentences:
> Note that it is also possible to fund several contributors through GIMP Liberapay account as an interesting alternative.
> What these donations through GNOME still cannot do is funding paid development, so if that’s what you want, please fund the developers directly as explained above.
I would rather put money into a single pot for GIMP contributors than have to chose and donate to individual contributors. Can I do that through the liberapay link (https://liberapay.com/GIMP/)? Most things suggest that I can, but the text makes it a little confusing.
They're worth many hundreds of millions of dollars from their time at Google, a successful exit to Square, and their eventual exit with Cockroachdb (and they deserve every penny of it)
I have often thought that one of the most effective ways, if I was say a Texan oil billionaire, most effective ways of spreading my world view would be to pay people to code useful free government and non-gov tools - all software is opinionated- as texan billionaire i would want it to have my opinions
Re: Texan Billionaire I just remembered the second Harry Palmer film.
They're positioned ideally to at least have some measure of what packages are most frequently installed. Even with the use of volunteer public mirrors that share no access logs with the distributor, the package management architecture could be structured to consult distro-controlled metadata servers in locating the mirrors and preserving some visibility into which packages are most installed. They control the package tooling, they should have considered this.
With even vague knowledge of which packages are most used, they could be distributing donations to those upstream projects. They obviously already have the dependency graph as well, so even without popularity metrics, can already know which packages are cornerstones (kernel, libc, openssl, systemd, mesa immediately come to mind).
The distro installers and package management tools should be providing convenient ways for end-users to donate via whatever mechanisms, paypal, patreon, bitcoin, etc. Just give me one mechanism to throw money at my distro of choice with the clear indication that it will be dispersed to upstream developers of important software. Also give me an easy way to explicitly allocate my funds to specific packages if I am so inclined. Make it all transparent on a web site where the money goes on a high level, how much is coming in, and which projects are getting what %age of that, what % is administration overhead, etc.