Board ArchivesSite FeaturesActiveworlds SupportHistoric Archives |
grimble grimble2000@btinternet:com // User Search
grimble grimble2000@btinternet:com // User SearchWondering how his question about sombody making a 3.4 VB SDK got to such a long boring string of postsFeb 18, 2003, 1:05am
Programming LanguagesFeb 10, 2003, 8:27pm
I'd say its "concept" time, rather than a new language. If you are to make
the most of either C++ or Java you'll need to get to grips with OO to a level where C++ makes a difference (i.e. not just writing C code and using a C++ compiler). Otherwise its just syntax and you don't learn anything new and rewarding. If it was a straight choice, I'd say C++ is more straight forward, especially for a windows developer, as it without all the additional confusion of packages, runtimes et al. Not enough "get up and go" about Java to feel like your making that early breakthrough to keep the interest. My gut feeling would be a bit of reading into high-level OO concepts, then C++, then ONLY OO in C++, then Java. Java makes much more sense when you have a solid grasp of the underlying premises, which are too different to procedural languages to make a simple transition using typical get-stuck-in learning methods employed by the casual programmer. If its used right (classes, classes, classes), C++ is a very productive language and quick to learn with the right grounding. My thoughts ... feel free to contradict. Grims [View Quote] Programming LanguagesFeb 18, 2003, 1:00am
Oh bloody hell! Well all I can say is that if you haven't even grasped the
basic requirements of C memory management, then you shouldn't be touching it or C++ anyway. Having no garbage collection in C++, in my opinion, makes for better understanding of his/her application by the developer as you're forced to respect the context of an object or store. Its not an "evil C++ memory error", its code written by someone who hasn't made he effort to lean the fundamentals of a language. Of course, there are so few free resources around for learning C++ on the web that its hardly surprising someone would fall into that trap. Blimey Guv'nor! Grims. [View Quote] Programming LanguagesFeb 18, 2003, 12:01pm
heh ... Wasn't an attack, was pure, unadulterated frustration at reading
complete trash. What utter crap, geesh! Surely you have to be making this up. [View Quote] Bot HelpFeb 18, 2003, 9:22pm
The main AW SDK documentation is available at www.activeworlds.com/sdk. This
documentation is for the C interface, but the COM wrapper commands are pretty much the same with a few idiosyncracies (which I believe are highlighted in the readme file in the zip). If you have any further questions, best to ask in the SDK newsgroup. Grims [View Quote] Xela and Avatar functionFeb 26, 2003, 10:49pm
Need Bot ScriptMar 9, 2003, 10:32pm
"Thoudands of lines of code" can be replaced by the creative application of
simple maths, binary operators and specific valued enums (a basic example being the message box constants for buttons, default button and icon). Grims [View Quote] Need Bot ScriptMar 31, 2003, 10:53am
Classes and arrays still need handling. If you're looking at cumulative
values (like those you described), careful selection of binary enum values removes a lot of the work in identifying state. Its a basic programming technique that, when used effectively, provides powerful shortcuts within an applicaton. [View Quote] Suggestions for bot site featuresMar 8, 2003, 12:43pm
I hope this doesn't sound like out-and-out criticism, its just my thoughts
dumped into a post. My expectations when I visited the site was for a site that it would provide a hub for all things "Bots", but I was ultimately dissapointed. Things I would like to find (and feel would be of help to development-type people) are: 1. Information on SDKs for all platforms (VB, delphi, C/C++, Java, etc.) where available, including news on alternative development methods (like the java development environment package mentioned a while ago in the SDK newsgroup, can't remember the link so sorry to whoever it was), download locations, help sites, etc. 2. Links to development related sites (www.programmersheaven.com, www.vbcity.com, various msdn.microsoft.com/library root pages for language references, etc.). This might slow the tide of "how do I trim a string in VB" type questions. 3. Unravelling the apparent mysteries of the SDK (i.e. understandable tutorials on property surveys and terrain surveys/maintenance being obvious candidates here based on the number of queries that come through). 4. Forums - Not discussion/"please help me wipe my nose" forums like the AW newgroups seem to have become, but more akin to those available on sites like www.perlmonks.org, where ideas and code are volunteered with moderated comments/alternatives/enhancements posted against them. I think this would prove very popular (IF you are prepared/have the time to review the "posts" before adding them to the thread). Nothing wrong with having interim discussion threads on each item where anyone can post comments, but the main item content would be strictly maintained by yourself. 5. Source Code libraries (both whole functional areas as well as code snippets) ... I would see this being closely related to the previous item. 6. Full Bots (with good documentation - augmented by yourselves if possible/necessary). Obvious examples of omissions would be Xelagot, Preston, Magsbot, etc. (with the appropriate owners approval of course). 7. Generic "helpers" available for download (like ActiveX DLLs or C DLLs callable from C, C++ and VB) for common functions such as Session Table management, object management, etc. that allow these things to me handled by a defined interface rather than including code in an application to support them (lots of possibilities here - good for newbies especially where things like session management will be a mystery and a simple call to a DLL in the Avatar Add/Change/Delete event handlers would maintain the list and a few methods to query the list). 8. Another nice idea (I think) would be something along the lines of www.asynchrony.com where AW related projects can be proposed, appropriately skilled teams formed, etc. and the deliverables made available (if possible) through the site. This would encourage developers interested in AW to interact more, sharing skills and knowledge, the end result (hopefully) being less disperate "releases" of bots by individuals and more (sorry about this to those that "release" bots) useful and well thought out offerings. Experience could be gained by all those taking part in a project improving the group as a whole. Maybe even have the concept of one or two "tag-along" newbies who can watch from the sidelines but still be involved in the project. Some projects can be "open source" to act as a reference for newbies (or anyone) after completion. The "helpers" mentioned above could be good test cases for this perhaps? No offence, but the site currently seems like "KAH and Bowen's Bot Stuff" which I don't think is your intention. I would think you'd want to get to a point where AW were gagging to include a prominent link to your site on their SDK help pages (and on other sites too) - where your site not only augments the information available on the AW site, but becomes an essential part of learning and improving skills related to AW SDK development. My feeling is that, if you want to succeed bigtime with this and actually make a difference to AW bot developers, there needs to be something that makes the site interesting as a "home" rather than just a reference or download resource site. In my opinion, some of the more interactive items above would go towards achieving this. I think there is a need for a dedicated "bot" site, but it should offer more than a few downloads and news items. I'd lose the "the rest ..." section too, and keep it focussed entirely on what you're trying to achieve. I'd even say focus directly on the development side, with just a download page for the finished articles. Everyone has their own views on sites like this, but these are my thoughts and I hope they're helpful, Grims. [View Quote] Suggestions for bot site featuresMar 8, 2003, 1:04pm
Why was the followup for this set to "bots" only? Surely this is relevant to
SDK also. Grims [View Quote] Suggestions for bot site featuresMar 8, 2003, 1:10pm
Suggestions for bot site featuresMar 8, 2003, 2:21pm
I didn't mean any of it as disparaging/critical in any way.
I'd really like to see SOME resource appear that pulls this group of developers together. Everyone has something to offer in here, but there's too much "look what I've done, take it or leave it" going on, driven by their own needs (which often vary from others needs). Things like Asynchrony encourage people to be instigate projects that they cannot complete on their own. How many "I need someone to make me a bot that ..." posts do we see? If these people are left to at least be involved in an organised project to meet their needs, then there'll be less annoying "Give me stuff" posts around. Most people involved in things like AW SDK development are there because they enjoy learning and being involved in things. Unless I've totally misjudged the developers in this group, there will be interest in joint projects. The SDK newsgroup has phases of being a general chat forum and for the content that IS relevant it would be great to have a single resource of quality information to point to as a useful reference. It will need some commitment obviously but if you can introduce a site that gets this group of eternal individuals to work together, it can only be a benefit and maybe even encourage AW to progress the SDK more rather than concentrating on the browser. Currently, SDK updates seem to be purely as a consequence of new features in the browser rather than having new non-visual interaction possibilities as the driver in their own right. This group could be so much more productive and enjoyable, its such a shame to see it whimper in and out of life over and over again. Grims [View Quote] I wish I wish I wish...Feb 10, 2003, 1:38pm
seed object functionFeb 10, 2003, 8:09am
yeah, but some people just insist on not applying any common sense when
reading a post just so they can post a smart-ass remark. [View Quote] seed object functionFeb 10, 2003, 8:14pm
seed object functionFeb 10, 2003, 8:43pm
What makes it really annoying (to me) is that its not even one of the stupid
wishlist items that get posted here. Its short, simple, and makes sense as a wish for AW and yet three people have chosen to jump on a single mis-placed word. This world is a fecking mess, it really is. Grims [View Quote] seed object functionFeb 11, 2003, 6:25am
Do you read? Can you understand acknowledgement if you can read it? One of
these two answers is "No". Hint: "Its short, simple, and makes sense as a wish for AW" Grims [View Quote] Imbedded RobotsFeb 20, 2003, 8:51am
Well thankyou for that clarification TZ ... although I am still
none-the-wiser regarding your reasoning. I didn't mean "you stupid dope, can't you f**king read?", I was merely indicating, in a short and sweet manner, that you were explaining that something can't currently be done in a wishlist newsgroup, where existing limitations are not an issue (but likely development/implementation considerations and costs are). Grims [View Quote] I wish for a reflection obejct actionFeb 12, 2003, 9:36pm
One step further on the idea (same implementation basically) is remote
cameras. They've been mentioned before but on the back of a lame "wouldn't it be cool if we could put a camera in one place and be able to view what it sees on a picture object somewhere else" ticket (Ugh!). Remote cameras really need to be native to the 3D engine (like WildTangent provides through portals) but they're better suited for closed environments like gaming levels, and are usually strategically placed to avoid too much additional rendering work. Nothing wrong with putting a reflection distance parameter on the command though (with a max distance enforced by the browser) ... that would encourage builders to use them intelligently to decent effect. Since I'm blabbing about WildTangent WebDriver (I use this as an example of something I've tried out, its not a plug which would be pointless anyway) .... There are a few things in it that would be great to be able to use in a decent world in AW (especially surface shading which I think would add some impressive effects to scenes). To be honest, I have no idea what RenderWare supports or is capable of as its too darned expensive to justify playing with. I've strayed somewhat, but you get my drift. Grims [View Quote] terrain editing rightsFeb 14, 2003, 12:55am
At one object per cell with terrain data (not blindly one object per cell),
I don't believe this represents a significant impact on either, and any impact that it does represent is far outweighed by the benefit it brings to a building world that wants to share terrain rights. I think the bot is more intelligent than its given credit for in terms of managing its own memory and querying the terrain objects. Experienced developers generally have more than sufficient foresight in such matters. Grims. [View Quote] EncroachingFeb 20, 2003, 8:32am
Now you're getting into areas that will get messy (just from a requirements
point of view, not even from coding and storage) as time goes on - I think you're pushing the envelope of cost-benefit here. I would expect such a feature to have a significant impact on the current storage and registry aspects, as well as attempting to identify what is "shared property", when it becomes "non-shared property", when that cell becomes "non-shared", changes in object positioning, orientation or even shape/size, etc. I mean would any object overlapping that cell become "shared property"? Who has the right to delete/change "shared property"? When does someone become unable to make property "shared" or change existing "shared property". If person A deletes their "shared property" in a cell to replace it, does the cel become the owner of the other participant? Too many questions and too much effort/cost for a minor feature that only worldowners who open their world to potentially untrustworthy builders would make use of. Otherwise, you have privs which can be set on a temporary basis to alow someone to overlap your work, the option of having no registry at all plus the obvious bot functionality (Hmmm that reminds me ...). My opinion of course, but I think its a minor issue. Grims. [View Quote] EncroachingFeb 20, 2003, 8:47am
Ummm ... I'm commenting on the feasilbility of the item ... not trying to
explain that AW doesn't support it now *boggle*. Perhaps you should work on that resistance ;O). Grims [View Quote] EncroachingFeb 20, 2003, 9:06am
As I said, there is a mire of requirements nightmares before you even get to
the code. If someone can identify all the ins and out, document them, and still believe that the item is worth the effort involved in the work they have just done plus the identification of suitable solution options plus the development and testing costs, then fine. I just rattled off a few complexities off the top of my head, but you can see that the scope of such a change would cover several core building related areas (hence higher risk in the change) and leave questionmarks over the ownership of property in the future (as one potential drawback). Its all about identifying all the requirements and assessing the risk and potential benefit of the change. If you can say 100% that you can identify all the issues involved and still come up with prudent solution that won't take 2 over man-weeks to implement (including your "thinking time") then I'm happy to accept that I have an over cynical view of it. A solution with a ball-park development estimate of anything more costly than 2 man-weeks, to me, doesn't justify it being looked at for a fringe item. The later in the development that the errors are found, the more expensive it becomes to fix them - just work out how much the 3.4 Beta has costs in real terms - Ouch!! Grims. [View Quote] Actually, I think it could be done pretty efficiently, but it would all depend on how it's coded. AW 3.1Feb 20, 2003, 9:08am
SDK ...Feb 20, 2003, 8:43am
Its been a while since I put a new wish in here, so I thought I'd drag up an
old one. We have a new version of AW on the horizon, with a few changes to the SDK - some of which are useful, some of which are necessary to support the new browser functionality (i.e. approximately 1 million new world attributes), but nothing on the scale of the "Object click events now include the Object X, Y, Z, Description, etc....." revelation. One item that contiunues to raise its head from developers is the rather lacking "AW_EVENT_OBJECT_DELETE" event attributes, currently residing in the "it would be so much more useful if" category for most developers. The wish is, of course, that the object attributes be included in the event like the "Click" event now has ... so that a deleted object can be restored should this be necessary without the need to query the cell in case someone does delete it. I'm not trying to second guess any reasons its not been addressed previously (like at the same time as the click event got the benefit of these useful attributes) or the difficulty associated in implementing it. I would like to know, though, whether this one item pencilled in for a future realease, or whether its technically a big issue because of legacy code. This request gets brought up a lot by SDK users, so a timely appearance would be welcome. I was disappointed to see that it wasn't included in build 29 as it gives world management bots the power they need to rectify "issues" with deletions. While I'm here ... if possible, identifying the object number/cell of the deleted object in the AW_EVENT_OBJECT_ADD event (for an object- move) would also make a lot of people's lifes much much easier. Ta, Grims. Global ChatMar 1, 2003, 7:24am
|