AGS tips.

[passer-by]

You spend years creating the first 90% of the game and eleven hours for the rest of it. Mistake.
People *will* notice.

[SSH]

Don't spend all your time responding to posts in the forums instead of making your game

Damn!

[cat]

Switch tasks

If you don't want to make backgrounds anymore switch to animations or coding. There is always enough to do.

[olafmoriarty]

If your stories suck, grab a great story from someone else.
Seriously. Look to the classics. There are tons of great books that
a) belong to the great classics so that pretty much everybody has heard or them
b) are so old that the copyright has expired on them, so they're absolutely free to use
c) have great potential to be turned into awesome games

I recommend Gutenberg.org as a starting point -- tons of great stories, most of them with no video game adaptation whatsoever. So if stories are your weak spot, why not instead of struggling for months with coming up with a decent story, just find an awesome story and make a game out of it?

Not all stories are perfect for making games out of, but surprisingly many of them do. Imagine an adventure game about Oliver Twist, a whodunit game starring Sherlock Holmes, or maybe a young Hercule Poirot. I'm playing with the thought of making a game out of "The Wonderful Wizard of Oz" myself.

[Ultra Magnus]

If you get disenchanted, play an existing game that's similar to the one you're making.
Example:-
If your graphics suck and you start to think "nobody will want to play this", find and play a game that's popular despite having crap graphics.
If that game works, why can't yours?

[Laukku]

I'll update this list if I can come up with more stuff...

Remember to turn debugging off when you release your game.

[radiowaves]

For the love of the gods, use .png
Seriously. Trying to post .bmp files will have dire (diiiiiiire) consequences, and most of the time, .jpg will make a sloppy mess out of your art.

Just for reminder. PNG is only good for final conversion. Don't save your wips in PNG, various programs tend to mess it up sometimes, I have had experience it going darker after too many oversaves. I make all my wips in bmp and I just don't care about your slow connection.

Final product is what matters.

My tip:

Don't cramp yourself too many different projects, wether AGS or not.
It is most likely they will begin to interfere with each other and take your time away, you lose motivation. It is also important to have some rest time when working on a project.

[Gord10]

Plug off your modem
Unless you have to talk with your teammates or find resources/help from net, plug your modem off. Stay offline.
Because Internet has a big potential to prevent you from working. You can waste your hours that you should had spent on your game's workings in chat rooms, forums, messengers, etc.

[evenwolf]

Don't let talking about your game become a substitution for making your game.

If you're the type of person that loses interest in an idea just because someone bad mouths it or claims that you stole the idea,  STOP talking about your ideas OR begin to take comments with a grain of salt.  Yes, there was a similar joke in The Simpsons, big deal.    Use the joke anyway as long as you came up with it. 

Your first project will undoubtedly have many mistakes but no mistake is worse than losing focus and quitting.    Don't be the guy who starts and quits to the point of absurdity and never releases a finished game. ahem..

Perfection can wait.  Focus on completion.   Continue to take notes for all your games but work toward a deadline for your first smaller project.    Scale is your friend.   Start small with either a RON game, the MAGS competition, or simply give yourself several small goals such as a two room game with three puzzles.   You will make mistakes, and those mistakes will be ones that you learn from for the next, bigger project.    Always repeat this mantra: 

"Big game later. I'm not sacrificing anything but beginner's mistakes."

[AJA]

Work in small chunks. Deadlines are your friends.

Set achievable and relatively small goals for yourself and have a deadline for them. If there's something I've learned from the One Room One Week competitions, it's the fact that it's easier to get your game finished if you have a deadline. What I mean is that you should, for example, give yourself two weeks to finish a specific room in your game and try to get it done in that time. When you've managed that, set a new goal and tweak the deadline if it didn't work that well with the previous goal. Don't try to finish your game in one large chunk.

[paolo]

This is an excellent thread and there have been some excellent tips posted so far.

Zor has already said this, but I think it is so important that I'm going to repeat it:

Plan out the whole game before starting to create it.

Taking a half-completed idea and starting to work on it in AGS is an almost guaranteed way of ensuring you'll never finish your game. So here's my related tip:

When planning out your game, work out early on how it is going to end.

Once you know how it is going to end (and how it's going to start), you'll have a good idea of what needs to happen in the middle of the game in order for it to get from the beginning to the end.

Write a to-do list and keep it up to date.

Even small games benefit from this approach. You know you (or your co-creators) need to draw the graphics, work out the plot, do the scripting, compose the music, put in the sound effects, do the beta-testing, fix the bugs and find somewhere to host the game, so there is a lot to keep track of. All of these tasks will have subtasks, and new subtasks will come up all the time. If you're writing one particular scene and find something that isn't working in another scene, write it down so that you can work on it when you've finished writing the current scene and don't forget about it.

A little and often

CodeJunkie has already mentioned the "often" bit ("Don't stop working on a project for more than a few days") but similarly, don't go to the opposite extreme and code/draw/whatever for 18 hours non-stop - you'll fry your brain and become thoroughly sick of your game.

Get someone to proofread the text in your game.

This is important* whether you suck at spelling and grammar or have a PhD in it them.** If you suck at spelling and grammar, be honest with yourself - your game is probably going to have a lot of spelling and grammar mistakes in it. For some people playing your game, they won't be bothered by this, but for other people, it will be an annoying distraction. If there are a lot of mistakes, some people might think your game's not worth playing and give up on it. If, on the other hand, you're a spelling bee champion five years running, still get someone to check your spelling and grammar. It's all too easy to make typos or use homophones ("their" instead of "they're", for example) accidentally.

* Yeah, yeah, call me a grammar Nazi if you want, but I think this is important. And any typos or grammatical errors in this post will be because I didn't re-read the message closely enough before posting. So there  

** Oops, just proved my point, really.

[aussie]

Don't post a GiP thread about your game until you've nearly finished it

[olafmoriarty]

Don't be afraid to ask for help.
If you think that you're an expert on drawing, animation, music, story design, coding, dialogue, etcetera, etcetera, it is an interesting challenge to create a game from scratch all by yourself. But most often you'll realize that there is at least one field of the gamemaking progress you're not that good at. If, for instance, you have an excellent story and you code like Gandhi, but you really hate to draw, players will notice that your passion is not in it, if you even finish the game instead of giving it up the second you have to sit down and animate a walkcycle. Instead, look through the "Offer your services" thread and find that it's full of people eager to help out on games, and if you find someone who offers to help out with what you struggle with (and you like their references), send them a PM. If you recruit a team, you won't have to do the parts you hate all by yourself.

Look through available modules.
Chances are that if a GUI or game technique is very popular, there exists a module for it. Skim through the modules forum and see if you find the things you need. There's no need to reinvent the wheel.

[Makeout Patrol]

Translation files are an excellent proofreading tool.

One thing I wish I'd thought of earlier is creating a new .trs file, copy-pasting the content into Word or something similar, and hitting 'spell check.' You'll have to find the text in the actual game files and replace it manually, but you'll have to do that no matter what method of proofreading you use. You get the advantage of a machine-based spell check, and it only takes into account the text that actually appears to the player. Once you're done, delete the translation from the editor.