Get Time
Spotlight Sessions: Specification Chat Transcript

Chat transcript from Specification Chat: Spotlight Session with AssertionsOn
Handle Comment
jmpld40Welcome to the chat everyone
RaulEnriqueSilHi all word
jmpld40Welcome to AssertionsOn as well
yoitsfrancisis it starting?
jmpld40We'd like to keep this chat focused on Specification competitions and what they are all about
chownikhilcan any one explain what this is all about
AssertionsOnThanks everyone for joining us in todays chat.
[hitcher]so.. what is it all about?
yoitsfrancishi all!
yoitsfrancisits midnight in the philippines
zhch12121its midnight in china
yoitsfrancisFirst of all, I saw your posts in the forums. Congratulations for representing the Specifications Spotlight Sessions. So here are my questions:
[hitcher]yoitsfrancis: you from the philippines?
yoitsfrancisQuestion #1: Lets say that I'm new to this type of track. I have a feeling that in order for someone to win in this contest, you must have some experience in doing Specifications or similar tasks in Software Engineering. Let's say that I don't have that
AssertionsOn[hitcher]: Maininly, specification (and concept) competitions deal with looking at a very high level at a clients idea and coming up with a technical solution.
yoitsfrancisI notice that there are no good Specifications deliverables for us to study unlike Design and Development that is available in the TC Catalog because of the obvious nature that this type of competition is client specific.
yoitsfrancisQuestion #1a: Are there good books that you can train yourself to be good at this? Question #1b: Are there good websites or links that are worth studying for this type of competitions?
yoitsfrancisQuestion #1c: Is there more good resource that you might share that I missed?
[hitcher]AssertionsOn: oh. got nosebleed.
yoitsfrancis[hitcher]: yup
[hitcher]yoitsfrancis: kabayan :)
AssertionsOnyoitsfrancis: Not necessarily. Any exp. in building applications will be helpful, especially web applications.
AssertionsOnyoitsfrancis: I mainly rely on past experiences for guiding my submissions, as well as my creativity. It helps a lot to think outside the box.
AssertionsOnyoitsfrancis: Google and TopCoder are great references
AssertionsOnyoitsfrancis: Also, by looking at what past winners have produced greatly helps.
AdamSeleneyoitsfrancis: Also, we have some resources on our wiki. They are not 100% up to date to the current specification challenges, but the content should still be helpful for the basics of what we are looking for.
AdamSeleneyoitsfrancis: Check this link:
yoitsfrancisoh i see!
yoitsfrancisQuestion #2: Let's say that I am equally good in Specifications, Architecture, Design, Development, Studio, Assembly and Testing. Is Specifications a track is really really really worthwhile pursuing?
AdamSeleneyoitsfrancis: For conceptualization, see here:
yoitsfrancisI know that this track is relavitely young compared to others but do you think money has already been made in Design and Development based on the vast number of components in the Catalog and you would recommend on shifting to Specifications?
yoitsfrancisAnd may I know the reason why?
AssertionsOnPlus, all, yo u have to remember that these challenges are not as technical as design and developement challenges, so you don't have to be too techincal. If you can explain your ideas in a straightforward way, you have a great chance of winning.
AdamSeleneyoitsfrancis: The specification track will have less trips to the TCO, but for every project we do we expect to have conceptualization, specification, and architecture work.
AdamSeleneyoitsfrancis: Realistically, though we do have a large catalog, some amount of component work may be necessary for most projects... so I wouldn't write those off yet.
AssertionsOnyoitsfrancis: Yes. Just look at the last 2 months of competitions. There has been a ton of Concept/Spec challenges been put out.
AssertionsOnyoitsfrancis: Some people don't have the tech background for those types of competitions, and these would allow them to be involved in the TC process, as well as gain experience, feedback, and potentially some cash.
AdamSeleneyoitsfrancis: Yes, indeed - we don't expect everyone from component competitions to be great at specification, and vice versa.
AdamSelenethat was @AssertionsOn, been a while since I messed with the arena chat interface =)
yoitsfrancisAdamSelene: Yup, I understand
yoitsfrancisQuestion #3: I saw fellow member's post in the forums and feel them that its hard to earn prize money at TopCoder. Let's say that I have a full time job, work 8-12 hours a day earning $400-$900 a month.
AssertionsOnExactly, I am not great at design and development competitions, but I love these competitions since I don't have to dive into all the technical aspects of an appllication. I can come up with some great ideas and present them in a meaningful way.
yoitsfrancis(I love Software Engineering so if ever I land in Specifications that would be totally fine with me). Is devoting full time in Specifications the answer? If so, how would I make that shift from a full time job to a full time TopCoder competing member?
yoitsfrancis(Well, if you personally ask me on the shift thing, that is really really tough, especially that you don't get enough rest so if you have a better technique in shifting, I am all ears).
AdamSeleneyoitsfrancis: Competing full time at TopCoder is an important personal choice - you need to weigh the factors involved. I don't think we can guarantee that any single type of competition will always have enough volume...
AssertionsOnyoitsfrancis: Remember there are some full time design/devs at TopCoder who rely on it for their only income. For us 40hr/week guys, it can be tough to keep up with them.
yoitsfrancisAssertionsOn: good for you then!
yoitsfrancisAssertionsOn: you do the things you love!
AdamSeleneyoitsfrancis: we would love for that to be the case, but it is dependent on a lot of factors - client work volume, the overall economy, etc.
AssertionsOnyoitsfrancis: I personally work 40+ a week and have children, so it is semi-stressful. But when you win, it is so worth it.
yoitsfrancisAssertionsOn: yeah i feel you on that one
AssertionsOnAdamSelene: I agree. Don't expect to put 1-2 hours into these compeititons. Usually, I put in about 10-20 per contest.
AssertionsOnPlus, these competitions (as all of them) are very addictive. After my first one, I wanted to beat the other competitors so bad it kept me up at night trying to figure out how to beat them.
AssertionsOnCause with members like Mig-29, argolite, Ghostar, AleaActaEst, and others, it is both intimidating yet inspirational when you go against them and win.
yoitsfrancisAssertionsOn: so, how does that feel?
AssertionsOnyoitsfrancis: What feeling are you referencing ?
AssertionsOnyoitsfrancis: Well, it feels good to think you can be in the same sentence with them (wrt Spec/Concept competitions). They are all great at what they do, and you hope to achieve their level of work.
yoitsfrancisQuestion #4: Let's say that my happiness can be simply defined as me having $100,000 every year in my pocket (Hey, I'm totally happy with SRMs, but as long as we humans have a civilization and society, I need the money!), is Specifications the track to
yoitsfrancisgo through compared to others? And may I know the reason why?
AssertionsOnSo, let me ask the members a question: What keeps you from participating in these competitions?
yoitsfrancisguys, thats all the questions i have on my side
ivernyoitsfrancis: have you considered trying a competition or two and seeing how it works for you?
GhostarThe lack of a real straightforward scorecard keeps me from competing.
GhostarI like a little bit of structure to the review process
AssertionsOnyoitsfrancis: Are you saying it is it worth it, to me personally?
AdamSeleneyoitsfrancis: We would like all our competition tracks to be very profitable for our members. Given volume of work, we hope any track will be able to keep people happy.
AssertionsOnyoitsfrancis: I think competiting in these is both fun and gives you great experience, even if you don't win.
AdamSeleneyoitsfrancis: I would recommend picking a track based on your aptitude more than anything else
AdamSeleneyoitsfrancis: I think that will give you the most money in the long run, all external factors being equal.
AssertionsOnAdamSelene: I agree with you here, about picking the right track for your personality and exp.
AdamSeleneGhostar: In your experience are problems more in the scorecard or in inexperienced reviewers? We'd like to fix both problems obviously
AssertionsOnAdamSelene: For me, I can develop and design components, but not at the speed some of the current ones do it. But I know I can push out some specifications really quick.
yoitsfrancisAssertionsOn: answering your question, i have an affinity for a good design
GhostarI haven't actually gone through the review process in a spec. contest, but from what I saw in the scorecard, it was fairly subjective as to what the client liked
GhostarThis is kind of intimidating, since it's hard to judge what the client is exactly looking for
GhostarThe questions help, but it still can be confusing
AssertionsOnAdamSelene: (wrt Ghostar issues) These are being worked actively by TC . It just takes a little time to get those issues worked out. But I have also had some frustrations along the way myself.
AdamSeleneGhostar: Understood - so as a barrier to entry you expect clear scorecards and guidelines.
jmpld40Ghostar: Just so you are all aware, COncept and SPec comps are in the spotlight right now at TC and visit the threads or post here if you have ideas that should be on our radar
jmpld40Ghostar: and we have a team dedicated to reviewing the challenges so far, and the feedback, and will implement changes
GhostarThat would help. I'm not sure I like the idea of having the client actively involved in the review process
AdamSeleneGhostar: Have you had a chance to look at the architecture scorecards that were posted? Would posting scorecards in that format help for these challenges? We didn't get much feedback on the arch cards.
GhostarI did look them over, but didn't have time to get too in depth with them
GhostarI think that posting the scorecards will help, and I know that is being worked on, from what I have seen in the forums
AdamSeleneGhostar: But, would a presentation like that help? Updating to have all guidelines, including scoring
GhostarI think so
AssertionsOnGhostar: "client in the review process" -> Unfortunately, this is going to keep happening (I think). They need to be involved to make sure their business needs are met at this high of a level.
GhostarI understand that, but it definitely makes the process more initimidating, and a little more chaotic
AssertionsOnGhostar: But they are getting members more involved in the review, so the comments/appeals/etc are getting extremely better.
AdamSeleneI think what we are moving towards is client review only at a high level during conceptualization. I think everyone agrees with Ghostar that we need accountability and regularity in more detailed review
AssertionsOnAdamSelene: Agreed
AssertionsOnGhostar: (Don't mess up my handle or you might get flagged for profanity ;) )
GhostarAssertionsOn: It seems like Mig-29 is winning most of these. Do you have a strategy to beat him?
GhostarAssertionsOn: I'm new to the chat room :)
AssertionsOnGhostar: Yes I have a strategy to beat them, but I can't give that away unfortunately. Then I would have everyone else beating me, and that would not be fun.
GhostarAssertionsOn: Do you find that a lot of these competitions are similar enough in nature to pull together different pieces of past competitions? I do that quite a bit with architecture competitions, and it saves a bunch of time, and cuts down on mistakes
GhostarAssertionsOn: I figured :)
AdamSeleneGhostar: I would expect to see that sooner than later - especially with conceptualization - many clients for instance need social networking features
AssertionsOnGhostar: There are small instances of reusing past competition content, but not really. I would say about 90% of my submission is brand new information. That is one area I need to improve in.
GhostarAssertionsOn: That's kind of another barrier to entry, since there isn't a way (that I know of) to download past submissions
barbarabAre there any good examples of past specification winners available? There are some great algo tutorials, but none for specifications or architecture that I can find.
AssertionsOnGhostar: I do think TC has posted sample competitions up in the forums and included them in the distrobutions, so if you follow that you should be fine.
AdamSelenebarbarab: We are working on providing more examples, but it is a tough issue as the output for these competitions is often very client specific
GhostarAssertionsOn: I like to be able to see all of them, not just a single sample. That way I can find ones that have similar goals to a new one posted.
AssertionsOnbarbarab: Not that I know of, to date. There is a sample set of deliverables, but I don't know if they are pubically available or not.
GhostarAdamSelene: I know we posted a couple of architecture submissions in the wiki awhile back. I'm not sure if they are still there. They were for internal TC systems
AdamSeleneGhostar: I think that is fairly unlikely, again because of client concerns.
GhostarAdamSelene: That's true :(
AdamSeleneGhostar: yes, for internal projects I'm sure we will post them; I do not know that there has been more than one completed so far though
AssertionsOnGhostar: Agreed. I would like to see personally the differences between the Round 1 and Round 2 submissions, just to see how much difference there is.
AdamSeleneYou brought up the round concept so I will jump on that...
AdamSeleneThe round 1 submissions are NOT intended to be full submissions
AdamSeleneHas that been at all clear in recent competitions?
AdamSeleneThe idea is that round 1 will flesh out use cases, and round 2 will provide the complete document.
GhostarAdamSelene: Aren't only the top 3 allowed to go on to round 2?
AdamSeleneI don't believe so, round 1 is not intended to winnow competition.
AdamSeleneIt may have been tried that way in the past.
AssertionsOnGhostar: Nope, only those who submit for R1 can go to R2
AdamSeleneThe reasoning behind rounds is just that we have often had this situation:
AdamSeleneClient likes submission #1, and #2, and #3
AdamSelenenot because of styling, or quality
AdamSeleneBut because #2 has an idea that the other 2 missed
AdamSeleneand #3 integrates the uses cases more clearly
AdamSelenebut say #1 is the best written and won the contest.
AssertionsOnAdamSelene: But, remember that R1 submissions get screened so if you don't have everything filled out, you may not go to R2
AdamSeleneTo avoid this situation, round 1 should unify the core content of the submissions
AdamSeleneAssertionsOn: Yes, I believe that is just to try to ensure that all competitors in round 1 are actively participating
AdamSeleneAssertionsOn: and not just waiting for round 2 to decide whether to compete
AssertionsOnAdamSelene: Agreed.
GhostarAdamSelene: I guess this is another concern I have with the competition. Doesn't that cut down on the competitive advantage I may have by coming up with good ideas?
GhostarAdamSelene: Since those ideas may be shared with the rest of the competitors?
AdamSeleneGhostar: yes and no
AssertionsOnGhostar: Yes. This is one issue I have with the new round formats.
AdamSeleneGhostar: yes - it may slightly, but the intention is by paying round 1, you are compensated for the best ideas
AdamSeleneGhostar: the net intended effect of the challenge is to deliver a complete, quality business requirements document to the client
AdamSeleneGhostar: the advantage should be the ability to do that, in the end
AssertionsOnAdamSelene: But, these prizes are sometimes too low to warrant putting out all the good ideas for others to latch onto.
GhostarAssertionsOn: I agree with that
AdamSeleneI am at a bit of a loss to answer that, thinking.
GhostarAdamSelene: I know that TC's main concern is getting the client what it needs, but in this particular case, that goal and how it is achieved seems to go against the competition concept
AssertionsOnGhostar: And therefore results in the construction of your winning strategy.....
AdamSeleneOK, so we need two things
AdamSelene1) all the good ideas
AdamSelene2) a document that is written well and clearly that contains all of #1
AdamSelenewe are certainly open to any format that arrives at that
AdamSeleneIf we shift the balance of money, does that solve your concern with the round format?
AdamSelenee.g. 50% round 1, 50% roudn 2?
GhostarAdamSelene: In my mind, why not do it like Studio? The client picks a winner, and if they want ideas from other submissions, they pay for those as well, and then the overall winner goes through final fixes, adding in the things the client wants
AdamSelenewe can't go any further than that, because then people won't take the time to complete #2, will they?
AssertionsOnAdamSelene: Interesting idea.....
AdamSeleneBut that's what we're doing with the two round format, basically
AssertionsOnAdamSelene: People like me and Mig-29 will still compete ;)
AdamSelenein theoretically, a lot less time
cyGNUsAdamSelene: then make payouts only after competing in round #2
AdamSeleneI guess in your model you have potetntially more money in round 1
GhostarAdamSelene: True, but at the expense of not having a true "competition"
AdamSeleneno, it's still competitive imo, you're still conflating the two competitions
AdamSelene#1 idea generation, #2 idea documentation
AdamSeleneI may come up with a great idea and not be able to write it down concisely
GhostarAdamSelene: Why not break it up into two competitions then?
AdamSelenethat is the intention with the round format =)
AssertionsOnIs there anyone new to specification challenges that need more information about how to compete, etc.? I know we are diving into challenge specifics, so if anyone has any more general questions, speak up.
GhostarAdamSelene: Yes, but you *have* to compete in the idea generation to compete in the documentation portion
AdamSeleneSo you would be OK with open registration for rd 2?
AssertionsOnAdamSelene: Not really.
GhostarAdamSelene: I guess so. I still am uncomfortable with the idea of rounds. I think it should be two distinct competitions, with separate reviews and payments for each
AdamSeleneWell, they are.
AdamSeleneRound 1: reviewed by client, paid
AdamSeleneRound 2: reviewed by members hopefully, paid
AdamSeleneThis is exactly the intent behind the roudn format, with the minimum of overhead.
GhostarAdamSelene: I still don't like it :)
GhostarAdamSelene: I want to be the winner of a full contest, not just 1/2 of one
GhostarAdamSelene: I want my competition history to say that I won this or that
AssertionsOnGhostar: They are working on this, I believe.
AdamSeleneI see. So if we presented it as clearly two separate competitions, you would like it better.
AdamSeleneAnyone else agree? AssertionsOn?
GhostarAdamSelene: I think so, but don't hold me to that :)
gt494AdamSelene: would reg of round 2 depend on participation in round 1?
barbarabAdamSelene: Yes, two separate challenges will encourage and reward teamwork - especially if all good ideas are rewarded appropriately in the ideas competition.
AssertionsOnAdamSelene: I personally don't like that idea.
AdamSelenegt494: I don't have any preference, Ghostar seems to like no, AssertionsOn, yes.
AssertionsOnAdamSelene: But, I would have to think about it some more.
Ghostargt494: I think that idea generation and documentation are two separate skills
AssertionsOnAdamSelene: I am just starting to like the new format myself.
Ghostargt494: Look at how many submissions were received for the AOL contest, which was purely idea generation
AssertionsOnGhostar: The senstations contest?
AdamSeleneThat was much lower resolution than conceptualization would need to be though.
Ghostargt494: Yes
AdamSelene(Believe me, I've read almost every submission)
AdamSeleneThat was just a call for projects.
GhostarAdamSelene: True.
AssertionsOnGhostar: I think they are totally different, and you would not get 100+ submissions.
AdamSeleneConceptualization needs to identify all core use cases.
AdamSelenePersonally I don't think that they are particularly disjoint because the same skill that lets you analyze what the client says and figure out what they really want will also guide you in the details of your submission.
GhostarAdamSelene: I think it would definitely need to be not quite so free form, for obvious reasons, but I think that splitting the two couldn't hurt
AdamSeleneI think if you are the only one who thought of Use Case XXX, you've been thinking about it for that much longer and have a basic understanding of why the client wants it
AdamSelenethat your competitors lack.
AdamSeleneIt certainly helps the competitors to be given your idea
AdamSeleneBut I think you still have an edge over them.
AdamSeleneBut I am not the one actually doing these competitions so I will tend to believe you guys more than what I dream up =)
barbarabAdamSelene: It would probably improve round 2 submissions if you had an example or two of well written specs - possibly from TopCoder internal projects
AssertionsOnAdamSelene: I don't think there is any advantage once competitors have you solution. It is better for the client though.
GhostarAdamSelene: True. It's hard for me to have a different perspective on this, since I am pretty good at documenting ideas, but not so good at actual idea generation
AdamSelenebarbarab: Yes, we understand and will get examples clearly posted as soon as we are able.
GhostarAdamSelene: I guess I just feel that there is much more straight creativity in round 1 which may suite certain members more than round 2
barbarabAdamSelene: Thank you
AdamSeleneAssertionsOn: they won't have your solution though, all that should be shared after round 1 is the core use case.
AdamSeleneAssertionsOn: If I tell everyone that "Vote for comment" is a core use case
AdamSeleneAssertionsOn: Do I end up with TC's +/-? Slashdot's karma system? what?
AdamSeleneAssertionsOn: there is still a lot of room to distinguish yourself.
AssertionsOnAdamSelene: Disagreed. Past comments have been extremely detailed, and stated details that one document had and the others did not.
AdamSeleneAssertionsOn: OK, that is a problem and not the intent of the round system.
AssertionsOnAdamSelene: All for the price of $50.
jmpld40As we start winding down this chat, does anyone else have any questions about how to compete or about the actual competition for AssertionsOn or the TC Admins?
jmpld40Specification Competition will be part of the TCO and all the details will be anounced next week..
AssertionsOnI want to personally thank everyone for joining the chat and asking questions. I hope to see you all in these competitions.
GhostarAssertionsOn: I have a feeling I will see you in Vegas :)
jmpld40however...the AOL/TC Sensations Developer Challenge Idea Generation Competition is over now so apps will need to be built!
AssertionsOnEven if you don't win, they give you great experience for future endeavors.
AssertionsOnGhostar: I hope so. It would be an honor.
jmpld40These competitions are starting sooner than you think so i'm hoping you'll all give them a try
kalcany one here???
barbarabAdamSelene: Good luck finding a system that rewards teamwork!
AssertionsOnbarbarab: LOL!
jmpld40A big thank you to AssertionsOn, AdamSelene and all others for participating in the chat.
AssertionsOnThank you TopCoder for asking me to join this chat. I had a great time discussing these competitions.
jmpld40I'll be posting the entire transcript as soon as I can and we can discuss things further in the accompnaying forum.
jmpld40Remember we will have 2 other chats, one next week with Ghostar about Architecture so don't miss them.
jmpld40Any other questions about Spec competitions? anyone? :)
AssertionsOnGhostar: I'll be at that chat, so get ready for some tough questions.
GhostarAssertionsOn: Bring it on :)
AssertionsOnGhostar: Consider yourself served ;)
lancordhi !
AssertionsOnlancord: Hello
AdamSelenelancord: Hi there, you're coming in on the tail end of the chat =) Do you have any questions about Concept/Spec challenges?
lancordjus in short plzz tell me wat in specific i nee dto know for design and ev.
lancordi mean a;lgo is not making me money
lancordthr ?
AssertionsOnlancord: High level system use cases and good documentation/idea skills
AdamSeleneThere are some tutorials and other information on our wiki:
AssertionsOnlancord: I would chekc out the wiki page AdamSelene put up
AssertionsOnlancord: it is very helpful in giving a brief intro to these new competition types
AdamSeleneIf you are interested in Conceptualization, see
lancordokay.......i and how much time will it take to get good in these things
lancordlike if u i devote 5-6 hrs daily
AssertionsOnlancord: I personally put in about 10 hrs a day on challenges, if not more.
AdamSeleneIf you can spend that much time per day, I would expect you to start seeing results within 3 to 4 competitions. Every competition should give you feedback on how well you are doing, and where you need to improve.
jmpld40we'll also have a chat about design and dev right here on wed. feb 11 at 11:00 a.m. EST
lancordokay hmmm......thanks
lancordthanks for the advices
AssertionsOnlancord: No prob. also check out the forums for future discussions about these comp.
jmpld40read more here:
jmpld40thanks again everyone and hope to see you next week!
AssertionsOnThanks everyone!
AdamSeleneThanks everyone, feel free to email me at if you have more questions.