The Future of SDA

2015-01-25 04:11:05 pm

Might be magic...

I generally don't wade into these debates much, but I have a few random ideas that I thought I would put out there. As I was typing this I noticed that it looks a lot like what
Mystery wrote, so I think other people could be on the same wavelength.

* Completely agree with the suggestion that moooh raised around making runs available with an "unverified" tag or some such descriptor that people can filter on

* It sounds like the "glitchless" category is something that a lot of people would like, but rather than having to make official decisions on what constitutes a glitch or not, possibly what could be useful is a reputation-based rating system for listed runs on a variety of categories, of which glitchless would be one category (something like stackoverflow.com). This would remove the responsibility from the staff and leave it up to the community to decide what's a glitch and what's not. Users with a low rep would not be able to vote (to prevent spam users from voting) and games with a high ratio of glitch votes would display a message next to the run "This speedrun relies on game glitches to achieve the optimal time". For controversial ones where the community votes a fairly even split "This speedrun contains some parts which could be considered game glitches"

* I was trying to think of a way that the video encoding responsibilities could be reduced for staff as well - possibly some sort of distributed transcoding system? People upload their runs to the site, a server farms pieces of encoding work out to volunteers who run a client on their home PC and the video gets encoded in chunks in a set format and style (basically a distributed Yua). Whether something like this exists already is something to research.

* Could some of these projects (especially the distributed video transcoding) be floated as university projects? I feel that part of the difficulty of running SDA is that a lot of people want to be speedrunners but not necessarily want to contribute much outside of running their favourite games and having them posted up. I hate to admit that I generally fall into this category too, though I have made a few contributions here and there when I get motivated.

* If you read this far you've probably guessed I'm heavily in favour of site automation. Making things easier for the staff is one good way to keep things moving - when tasks become a chore, things grind to a halt. Thanks to UraniumAnchor for all the efforts here. I guess part of this thread is helping to decide where SDA wants to go so those areas can be automated.

I generally don't wade into these debates much, but I have a few random ideas that I thought I would put out there. As I was typing this I noticed that it looks a lot like what 
Mystery wrote, so I think other people could be on the same wavelength.

* Completely agree with the suggestion that moooh raised around making runs available with an "unverified" tag or some such descriptor that people can filter on

* It sounds like the "glitchless" category is something that a lot of people would like, but rather than having to make official decisions on what constitutes a glitch or not, possibly what could be useful is a reputation-based rating system for listed runs on a variety of categories, of which glitchless would be one category (something like stackoverflow.com). This would remove the responsibility from the staff and leave it up to the community to decide what's a glitch and what's not. Users with a low rep would not be able to vote (to prevent spam users from voting) and games with a high ratio of glitch votes would display a message next to the run "This speedrun relies on game glitches to achieve the optimal time". For controversial ones where the community votes a fairly even split "This speedrun contains some parts which could be considered game glitches"

* I was trying to think of a way that the video encoding responsibilities could be reduced for staff as well - possibly some sort of distributed transcoding system? People upload their runs to the site, a server farms pieces of encoding work out to volunteers who run a client on their home PC and the video gets encoded in chunks in a set format and style (basically a distributed Yua). Whether something like this exists already is something to research.

* Could some of these projects (especially the distributed video transcoding) be floated as university projects? I feel that part of the difficulty of running SDA is that a lot of people want to be speedrunners but not necessarily want to contribute much outside of running their favourite games and having them posted up. I hate to admit that I generally fall into this category too, though I have made a few contributions here and there when I get motivated.

* If you read this far you've probably guessed I'm heavily in favour of site automation. Making things easier for the staff is one good way to keep things moving - when tasks become a chore, things grind to a halt. Thanks to UraniumAnchor for all the efforts here. I guess part of this thread is helping to decide where SDA wants to go so those areas can be automated.

ballofsnow

2015-01-25 04:43:41 pm

Heh these threads come and go. Guess I've been here too long.

Here's what I think should happen:

Step 1. SDA community fundraises enough cash to pay a professional company to implement all these "SDA 2.0" ideas.

I mean who else is going to do it?

Alko

Edit history:

Alko: 2015-01-25 04:57:03 pm

Alko: 2015-01-25 04:46:28 pm

2015-01-25 04:45:41 pm

Caution: This user contains Kana ^_^

Why do these topics always happen when I'm away from the internet and come back to read 5 whole pages? ~.~ The fact that this thread appeared proves that SDA isn't dead or dying.

Quote from AlecK47:

Quote from Crow!:

Quote from moooh:

An idea the staff has had is to keep around the current publication process but make runs available right away in some form. Think of it like a tier system where a bronze tier run might be not yet officially timed and might only exist for watch using a link to YT. A run could move up to the silver tier when it has published comments and an official time and finally be promoted to the gold tier when the run is also available in high quality encodes for download.
Visitors who'd want SDA old style could just filter the game list for "gold only" or something like that. Thoughts on this?

I like the idea but I don't like that nomenclature - the tiers makes it sound like the runs themselves are better or worse, when it's the site's treatment of the run which is complete or incomplete. More descriptive category names like "Untimed", "No Download" and "Complete" might be better.

Seconding this.

Thirding this. (and fixed name)

Quote from Moooh:

For any of this to be possible, as well as doing simple things like adding a KB-link on each game page (which has always been a fantastic idea), we are in the need of migrating a lot of static html pages into the database. This is a long and tough process that we hope to be able to get community assistance for.

Erm, I would have thought that some Perl, Python or whatever script should be capable of doing it? I mean, the game pages are pretty much organised all in the same way: Overview, information for run 1, comments for run 1, repeat the final two for any further runs. Should be automateable?

I don't know why people keep assuming things like Nate having completely disappeared? Or the forum not being maintained? Nate only posted something about enhancing the security less than a few months ago (my memory about when something happens is always shady …) and he usually has been quick to answer any posts in the taigaforum thread … but w/e.

Glitchless is an extre~mly slippery slope … Oh wait, you knew that. But let me elaborate:

Is backwalking in OoT a 'bannable' glitch? (For those unfamiliar: Walking backwards is faster than walking forwards. No idea if the developpers conciously made that so.)
Is supersliding a 'bannable' glitch? (You are pushed back from a bomb explosion and use some spacy mechanics to glide 'forever'.)
I mean, the latter is definitely unintended, but it still is entirely within the game world, and would have been accepted to SDA in the previous no-leaving-the-game-world times. A LOT of people advocating glitchless runs during the German AGDQ restream would have been fine with supersliding, but where to draw the line?

Let's see, what else:
+ redesigning the front page, maybe to turn portal-like, as long as it's not too overbuilt
+ getting more people to add content to the KB (I plead guilty)
+ making the KB visible
+ archive-style game pages with past runs (would volunteer to help dig up links/comments)
+ define clear quality standards for runs. (I know we have only game A/V, but are there any resolutions/bitrates/framerates (guessing native?)/codecs/whatever minimally required?) I've still heard rumours of quality standards being on the moon.

– no to immediate public verification (the 'big games' find enough verifiers within days. No need imho).
– no to a change of forums. Tweaks yes, but keep it taiga.

Re new categories: One could always add the question of 'accept this new category?' to private verifications (and auto-reject new categories if the verification went public due to too small a community). Note that that wouldn't apply to low/100 % categories which would/should be allowed by default, as long as they are consistent enough. Not sure about no major skips (as is currently described) but not opposed.

Re different timing methods: Hum. In cases, where community timing is well-known and different, it could just be added to the run information?

Edit: I totally forgot about allowing Japanese turbo controllers for RPGs … I'm leaning against this because it can lead to a lot faster time (eliminating human error when mashing through text plus potentially faster fighting plus easier, more consisten stutter stepping in FFIX. However, with a clear separation it could work.

Also, a minor final spoiler:

Quote from Mystery:

Side note, but...

Quote from Nyasupan:

...I agree with Croonikeys when he says:

Cronikeys is female. Please don't assume gender.

I see where you're coming from. However, we don't see people face to face; we see the text that they write plus a more or less random avatar next to it. Other than checking each user's profile before answering (could also be a lie or missing … not happening. Sorry, mate), having seen them on stream or met them in person, there's no way to guess. Spelling out names all the time, the way Japanese does it, isn't an option and neither is resorting to languages such as Finnish (hän meaning he/she/it). S/he is damn awkward to read and write so a thoughtless 'he' remains the easiest option.
Also, people tend to forget, even if they should know. I should now know Cronikeys is female, but I'll likely have forgotten by the time I respond to a post of her's the next time ^^' No offence ment Wink

Why do these topics always happen when I'm away from the internet and come back to read 5 whole pages? ~.~ The fact that this thread appeared proves that SDA isn't dead or dying.

[quote="AlecK47"][quote="Crow!"][quote=moooh]An idea the staff has had is to keep around the current publication process but make runs available right away in some form. Think of it like a tier system where a bronze tier run might be not yet officially timed and might only exist for watch using a link to YT. A run could move up to the silver tier when it has published comments and an official time and finally be promoted to the gold tier when the run is also available in high quality encodes for download.
Visitors who'd want SDA old style could just filter the game list for "gold only" or something like that. Thoughts on this?[/quote]
I like the idea but I don't like that nomenclature - the tiers makes it sound like the runs themselves are better or worse, when it's the site's treatment of the run which is complete or incomplete.  More descriptive category names like "Un[b]tim[/b]ed", "No Download" and "Complete" might be better.[/quote]
Seconding this.[/quote]
Thirding this. (and fixed name)

[quote="Moooh"]For any of this to be possible, as well as doing simple things like adding a KB-link on each game page (which has always been a fantastic idea), we are in the need of migrating a lot of static html pages into the database. This is a long and tough process that we hope to be able to get community assistance for.[/quote]
Erm, I would have thought that some Perl, Python or whatever script should be capable of doing it? I mean, the game pages are pretty much organised all in the same way: Overview, information for run 1, comments for run 1, repeat the final two for any further runs. Should be automateable?

I don't know why people keep assuming things like Nate having completely disappeared? Or the forum not being maintained? Nate only posted something about enhancing the security less than a few months ago (my memory about when something happens is always shady …) and he usually has been quick to answer any posts in the taigaforum thread … but w/e.

Glitchless is an extre~mly slippery slope … Oh wait, you knew that. But let me elaborate:
[spoiler]Is backwalking in OoT a 'bannable' glitch? (For those unfamiliar: Walking backwards is faster than walking forwards. No idea if the developpers conciously made that so.)
Is supersliding a 'bannable' glitch? (You are pushed back from a bomb explosion and use some spacy mechanics to glide 'forever'.)
I mean, the latter is definitely unintended, but it still is entirely within the game world, and would have been accepted to SDA in the previous no-leaving-the-game-world times. A LOT of people advocating glitchless runs during the German AGDQ restream would have been fine with supersliding, but where to draw the line?[/spoiler]

Let's see, what else:
+ redesigning the front page, maybe to turn portal-like, as long as it's not too overbuilt
+ getting more people to add content to the KB (I plead guilty)
+ making the KB visible
+ archive-style game pages with past runs (would volunteer to help dig up links/comments)
+ define [i]clear[/i] quality standards for runs. (I know we have only game A/V, but are there any resolutions/bitrates/framerates (guessing native?)/codecs/whatever minimally required?) I've still heard rumours of quality standards being on the moon.

– no to immediate public verification (the 'big games' find enough verifiers within days. No need imho).
– no to a change of forums. Tweaks yes, but keep it taiga.

Re new categories: One could always add the question of 'accept this new category?' to private verifications (and auto-reject new categories if the verification went public due to too small a community). Note that that wouldn't apply to low/100 % categories which would/should be allowed by default, as long as they are consistent enough. Not sure about no major skips (as is currently described) but not opposed.

Re different timing methods: Hum. In cases, where community timing is well-known and different, it could just be added to the run information?

Edit: I totally forgot about allowing Japanese turbo controllers for RPGs … I'm leaning against this because it can lead to a lot faster time (eliminating human error when mashing through text plus potentially faster fighting plus easier, more consisten stutter stepping in FFIX. However, with a clear separation it could work.

Also, a minor final spoiler:
[spoiler][quote="Mystery"]Side note, but...

[quote="Nyasupan"]...I agree with Croonikeys when [b]he [/b]says:[/quote]
Cronikeys is female. Please don't assume gender.[/quote]
I see where you're coming from. However, we don't see people face to face; we see the text that they write plus a more or less random avatar next to it. Other than checking each user's profile before answering (could also be a lie or missing … not happening. Sorry, mate), having seen them on stream or met them in person, there's no way to guess. Spelling out names all the time, the way Japanese does it, isn't an option and neither is resorting to languages such as Finnish (hän meaning he/she/it). S/he is damn awkward to read and write so a thoughtless 'he' remains the easiest option.
Also, people tend to forget, even if they should know. I should now know Cronikeys is female, but I'll likely have forgotten by the time I respond to a post of her's the next time ^^' No offence ment ;) [/spoiler]

Cronikeys

Edit history:

Cronikeys: 2015-01-25 09:12:00 pm

2015-01-25 07:37:34 pm

Quote from Alko:

Quote from Moooh:

For any of this to be possible, as well as doing simple things like adding a KB-link on each game page (which has always been a fantastic idea), we are in the need of migrating a lot of static html pages into the database. This is a long and tough process that we hope to be able to get community assistance for.

Erm, I would have thought that some Perl, Python or whatever script should be capable of doing it? I mean, the game pages are pretty much organised all in the same way: Overview, information for run 1, comments for run 1, repeat the final two for any further runs. Should be automateable?

I could do that in like 15 lines of code. I'm a little bit shocked that the pages are static html files. Moooh - if you want this done I could literally do it in a day. Just tell me what format you want the data in (json, csv, sql, xml, whatever). That shouldn't be the hold up in forward progress.

EDIT: Made a simple proof of concept (http://puu.sh/f2qaX/0eee91473a.txt) - obvious errors, but this was produced automatically with about 20 lines of code. Let me know if you're interested in my help @ whomever

[quote="Alko"][quote="Moooh"]For any of this to be possible, as well as doing simple things like adding a KB-link on each game page (which has always been a fantastic idea), we are in the need of migrating a lot of static html pages into the database. This is a long and tough process that we hope to be able to get community assistance for.[/quote]
Erm, I would have thought that some Perl, Python or whatever script should be capable of doing it? I mean, the game pages are pretty much organised all in the same way: Overview, information for run 1, comments for run 1, repeat the final two for any further runs. Should be automateable?[/quote]
I could do that in like 15 lines of code. I'm a little bit shocked that the pages are static html files. Moooh - if you want this done I could literally do it in a day. Just tell me what format you want the data in (json, csv, sql, xml, whatever). That shouldn't be the hold up in forward progress.

EDIT: Made a simple proof of concept (http://puu.sh/f2qaX/0eee91473a.txt) - obvious errors, but this was produced automatically with about 20 lines of code. Let me know if you're interested in my help @ whomever

UraniumAnchor

2015-01-25 11:43:31 pm

Not a walrus

Quote from Cronikeys:

Quote from Alko:

Quote from Moooh:

For any of this to be possible, as well as doing simple things like adding a KB-link on each game page (which has always been a fantastic idea), we are in the need of migrating a lot of static html pages into the database. This is a long and tough process that we hope to be able to get community assistance for.

Erm, I would have thought that some Perl, Python or whatever script should be capable of doing it? I mean, the game pages are pretty much organised all in the same way: Overview, information for run 1, comments for run 1, repeat the final two for any further runs. Should be automateable?

I could do that in like 15 lines of code. I'm a little bit shocked that the pages are static html files. Moooh - if you want this done I could literally do it in a day. Just tell me what format you want the data in (json, csv, sql, xml, whatever). That shouldn't be the hold up in forward progress.

EDIT: Made a simple proof of concept (http://puu.sh/f2qaX/0eee91473a.txt) - obvious errors, but this was produced automatically with about 20 lines of code. Let me know if you're interested in my help @ whomever

I scanned for a couple of pages that I know are going to be problems and it didn't handle them correctly at all. (Any IL table, for one.)

Most game pages are indeed trivial and can be handled with a simple dynamic page generator, but there's a still TON of pages that have weird quirks that need to be worked around in some way, either with extra support from the page generator or just deciding that the special case isn't worth it.

We started putting runs into a proper database some time in 2012. Just to give you a rough idea of how much still needs to be migrated to the database:

There are 790 published runs according to the queue. Some of these are actually obsolete but not marked as such.

After searching through every static html file I could find on the server (including some that aren't linked from anywhere), I found:

2298 unique run IDs (many of them obsolete, but still <10%).
2953 unique level run IDs.

Of those:

779 run IDs matched a run in the queue database.
1312 level run IDs matched a level run in the queue database.

This also means that a huge portion of the site is more than three years old. And not just because "oh nobody submits any more", but there's just a ton of random games that nobody's bothered to improve since then. Throwing all of that out would be a mistake, and I'm pretty sure most people reading this thread would agree with that.

I will say, though, that I've made quite a bit of progress in recent days in improving a lot of back-end things that have been holding us up.

On a somewhat related note the number of people who drop a submission in the form and then never bother to send us ANYTHING, not even a youtube link, let alone a recorded file, is quite high. I don't have a percentage handy but as of right now there are 35 runs in the queue that were submitted at least a month ago that are still waiting on ANY sort of file, and of those 21 have been waiting at least 3 months.

[quote="Cronikeys"][quote="Alko"][quote="Moooh"]For any of this to be possible, as well as doing simple things like adding a KB-link on each game page (which has always been a fantastic idea), we are in the need of migrating a lot of static html pages into the database. This is a long and tough process that we hope to be able to get community assistance for.[/quote]
Erm, I would have thought that some Perl, Python or whatever script should be capable of doing it? I mean, the game pages are pretty much organised all in the same way: Overview, information for run 1, comments for run 1, repeat the final two for any further runs. Should be automateable?[/quote]
I could do that in like 15 lines of code. I'm a little bit shocked that the pages are static html files. Moooh - if you want this done I could literally do it in a day. Just tell me what format you want the data in (json, csv, sql, xml, whatever). That shouldn't be the hold up in forward progress.

EDIT: Made a simple proof of concept (http://puu.sh/f2qaX/0eee91473a.txt) - obvious errors, but this was produced automatically with about 20 lines of code. Let me know if you're interested in my help @ whomever[/quote]

I scanned for a couple of pages that I know are going to be problems and it didn't handle them correctly at all. (Any IL table, for one.)

Most game pages are indeed trivial and can be handled with a simple dynamic page generator, but there's a still TON of pages that have weird quirks that need to be worked around in some way, either with extra support from the page generator or just deciding that the special case isn't worth it.

We started putting runs into a proper database some time in 2012. Just to give you a rough idea of how much still needs to be migrated to the database:

There are 790 published runs according to the queue. Some of these are actually obsolete but not marked as such.

After searching through every static html file I could find on the server (including some that aren't linked from anywhere), I found:

2298 unique run IDs (many of them obsolete, but still <10%).
2953 unique level run IDs.

Of those:

779 run IDs matched a run in the queue database.
1312 level run IDs matched a level run in the queue database.

This also means that a huge portion of the site is more than three years old. And not just because "oh nobody submits any more", but there's just a ton of random games that nobody's bothered to improve since then. Throwing all of that out would be a mistake, and I'm pretty sure most people reading this thread would agree with that.

I will say, though, that I've made quite a bit of progress in recent days in improving a lot of back-end things that have been holding us up.

On a somewhat related note the number of people who drop a submission in the form and then never bother to send us ANYTHING, not even a youtube link, let alone a recorded file, is quite high. I don't have a percentage handy but as of right now there are 35 runs in the queue that were submitted at least a month ago that are still waiting on ANY sort of file, and of those 21 have been waiting at least 3 months.

BlueGlass

2015-01-25 11:48:04 pm

Yep yep yep!

Just a preface, I'm pretty tired at the moment, so I apologize for any incoherency in my thoughts.

Wow, this has been a very interesting read. Personally, I find myself agreeing to all the main points and ideas brought up.

I know I'm not a frequent poster or too active in the community, but I'd like to take a minute to state my viewpoint as someone who has been following speedrunning and SDA since ~2004.

I saw some good discussion earlier about how one route for the site's future would be to provide some useful service for the speedrunning community that doesn't currently exist. In my opinion, one thing it would be good at addressing is the complete disorganization of information for new speedrunners looking to get into a game.

To explain, I mean that there isn't a good, obvious, centralized location to learn up-to-date tricks and strategies for a game. Instead, one has to scour the SDA forums, TASvideos, Twitch, Youtube, and the KB for the needed information. And even then, they'll still miss stuff.

I think, as several others have already suggested, making the KB more prominent (resulting in it being both more extensive and more noticeable) is probably the best solution to this. I've seen a good number of comments that SDA is old, irrelevant, and outdated, yet a large percentage of speedrunners still use it to one degree or another despite that. Having a clear, centralized source of information would help to fix these problems.

Well, that was my off-the-top-of-my-head thoughts. I would also like to say that I love SDA and am very excited to see what comes out of all this.

On a side note, was very excited to see a post from Yoshi348. He used to be very active on M2K2 back when I first started lurking, so it's a nice feeling to know that he still follows the community.

Just a preface, I'm pretty tired at the moment, so I apologize for any incoherency in my thoughts.

Wow, this has been a very interesting read.  Personally, I find myself agreeing to all the main points and ideas brought up.

I know I'm not a frequent poster or too active in the community, but I'd like to take a minute to state my viewpoint as someone who has been following speedrunning and  SDA since ~2004.

I saw some good discussion earlier about how one route for the site's future would be to provide some useful service for the speedrunning community that doesn't currently exist.  In my opinion, one thing it would be good at addressing is the complete disorganization of information for new speedrunners looking to get into a game.

To explain, I mean that there isn't a good, obvious, centralized location to learn up-to-date tricks and strategies for a game.  Instead, one has to scour the SDA forums, TASvideos, Twitch, Youtube, and the KB for the needed information.  And even then, they'll still miss stuff.

I think, as several others have already suggested, making the KB more prominent (resulting in it being both more extensive and more noticeable) is probably the best solution to this.  I've seen a good number of comments that SDA is old, irrelevant, and outdated, yet a large percentage of speedrunners still use it to one degree or another despite that.  Having a clear, centralized source of information would help to fix these problems.

Well, that was my off-the-top-of-my-head thoughts.  I would also like to say that I love SDA and am very excited to see what comes out of all this.

On a side note, was very excited to see a post from Yoshi348.  He used to be very active on M2K2 back when I first started lurking, so it's a nice feeling to know that he still follows the community.

bpcookie

2015-01-26 12:15:54 am

Quote from TheMG2:

Here is the thing, there needs to be a want for a glitchless category in the specific communities. The metroid prime community has 0 desire for any sort of glitchless category. Therefore it would make 0 sense to have one for that game.

Is there a site where I can fill out an application to be a part of the Metroid Prime community?

Quote from Judgy:

This is also true for Borderlands 2 / Pre-sequal.

Borderlands 2 for example with glitches is around 2:20:00 for NG W/DLC, without glitches it's (as far as I can remember) 3.5 - 4 hours (if not more) for NG W/DLC.

Hence why exactly NO-ONE runs without glitches.

Then how do you know how long a glitchless run is? I don't understand that logic.

I will be more than happy to respect you guys' opinions on their own merit, you don't need to throw around these emphatic absolute zeroes and capitals.

But you know, we're on the same side here. You guys are saying no, that doesn't fit these two specific games. Ok, great. It doesn't have to fit all games. I'm saying be open to in cases where there is interest. If that's zero cases, then that fact will reveal itself organically and there will be no need for protest.

[quote="TheMG2"]
Here is the thing, there needs to be a want for a glitchless category in the specific communities. The metroid prime community has 0 desire for any sort of glitchless category. Therefore it would make 0 sense to have one for that game.[/quote]

Is there a site where I can fill out an application to be a part of the Metroid Prime community?

[quote="Judgy"]

This is also true for Borderlands 2 / Pre-sequal.

Borderlands 2 for example with glitches is around 2:20:00 for NG W/DLC, without glitches it's (as far as I can remember) 3.5 - 4 hours (if not more) for NG W/DLC.

Hence why exactly NO-ONE runs without glitches.   [/quote]

Then how do you know how long a glitchless run is? I don't understand that logic.

I will be more than happy to respect you guys' opinions on their own merit, you don't need to throw around these emphatic absolute zeroes and capitals.

But you know, we're on the same side here. You guys are saying no, that doesn't fit these two specific games. Ok, great. It doesn't have to fit all games. I'm saying be open to in cases where there is interest. If that's zero cases, then that fact will reveal itself organically and there will be no need for protest.

Judgy

2015-01-26 12:25:17 am

Borderlands 2 Glitch Hunter/ router.

Quote from bpcookie:

Then how do you know how long a glitchless run is? I don't understand that logic.

I will be more than happy to respect you guys' opinions on their own merit, you don't need to throw around these emphatic absolute zeroes and capitals.

Because I've been finding glitches and shortcuts on Borderlands 2 since it was a 5 hour slog (originally) and the best we got it to without glitches was around 3 and a half hours.

and Yes we are both on the same team here I was just confirming the notion that Metroid wasn't the only community to have basically shunned Glitchless because, quiet frankly, its is BORING to watch.

As stated by yourself it should be on a per game basis, for example I would run Borderlands 2 Glitched but I also run Rainbow Six: Vegas 2 without glitches.

I wasn't trying to show anyone up or make them feel stupid, just wanted to back up what was a relevant point.

UraniumAnchor

2015-01-26 12:25:33 am

Not a walrus

Quote from bpcookie:

I'm saying be open to in cases where there is interest. If that's zero cases, then that fact will reveal itself organically and there will be no need for protest.

The issue with "glitchless" it that's it's far too vague. Is abusing some level geometry that lets you skip a significant portion of the game considered a glitch if all you have to do is jump in a clever fashion? Does it count as a glitch if it's just some special combo that's not listed in the manual? (Maybe shield dashing in SOTN would be a good example here, or basically any game with animation cancelling.)

moooh

2015-01-26 12:28:05 am

Exoray

Quote from Cronikeys:

I could do that in like 15 lines of code. I'm a little bit shocked that the pages are static html files. Moooh - if you want this done I could literally do it in a day. Just tell me what format you want the data in (json, csv, sql, xml, whatever). That shouldn't be the hold up in forward progress.

EDIT: Made a simple proof of concept (http://puu.sh/f2qaX/0eee91473a.txt) - obvious errors, but this was produced automatically with about 20 lines of code. Let me know if you're interested in my help @ whomever

Thanks, but I'm afraid it's a bit more complex than that. A lot of pages for games with releases on multi-consoles don't mention what system a run was actually made on so that has to be manually researched at those times. Category is not a simple string but made up of distinct tags (to allow for filtering capabilities), but I suppose you could implement some kind of fuzzy-matching for those.
There's also quite a few quirks like UA mentioned. IL tables that exist in multiple variants, different naming conventions for single-run pages and for multi-run pages, runs with extra files provided, runs with translated comments and the list goes on up to 20 identified quirks (and probably a couple of unidentified).
I'm not saying that parts of this cannot be scripted, I'm just letting you know that if you want to help with this using a scripted solution, you'll have to be prepared to devote quite a bit more than just a day Smiley

One thing that should be a bit easier to script that would save me loads of time is if someone can get me a list of the IL level names (in Max Payne "Woden's Manor" is a level name. In some other game, "Level 2" is a level name).
A list of all those, grouped by the game page name, would be great.

If anyone wants to know more about the migration challenges we face or have suggestions that could help with it, we should take that to a separate thread.

[quote="Cronikeys"]
I could do that in like 15 lines of code. I'm a little bit shocked that the pages are static html files. Moooh - if you want this done I could literally do it in a day. Just tell me what format you want the data in (json, csv, sql, xml, whatever). That shouldn't be the hold up in forward progress.

EDIT: Made a simple proof of concept (http://puu.sh/f2qaX/0eee91473a.txt) - obvious errors, but this was produced automatically with about 20 lines of code. Let me know if you're interested in my help @ whomever[/quote]
Thanks, but I'm afraid it's a bit more complex than that. A lot of pages for games with releases on multi-consoles don't mention what system a run was actually made on so that has to be manually researched at those times. Category is not a simple string but made up of distinct tags (to allow for filtering capabilities), but I suppose you could implement some kind of fuzzy-matching for those.
There's also quite a few quirks like UA mentioned. IL tables that exist in multiple variants, different naming conventions for single-run pages and for multi-run pages, runs with extra files provided, runs with translated comments and the list goes on up to 20 identified quirks (and probably a couple of unidentified).
I'm not saying that parts of this cannot be scripted, I'm just letting you know that if you want to help with this using a scripted solution, you'll have to be prepared to devote quite a bit more than just a day :)

One thing that should be a bit easier to script that would save me loads of time is if someone can get me a list of the IL level names (in Max Payne "Woden's Manor" is a level name. In some other game, "Level 2" is a level name).
A list of all those, grouped by the game page name, would be great.

If anyone wants to know more about the migration challenges we face or have suggestions that could help with it, we should take that to a separate thread.

Cronikeys

Edit history:

Cronikeys: 2015-01-26 01:48:56 am

Cronikeys: 2015-01-26 01:14:00 am

2015-01-26 01:13:27 am

Quote from moooh:

If anyone wants to know more about the migration challenges we face or have suggestions that could help with it, we should take that to a separate thread.

good plan.

fwiw i could use giantbomb for game meta (that's not already on the FullList.html or whatever) and i already know how to deal w/ il tables, variants, etc. I'm very experienced at data migration. I know about most of the quirks you and ua mentioned.

For more control, I could even make a semi-automated system where someone (i don't mind doing it) just goes through one by one w/ my script pre-populating the fields and then whoever could either change a quirky weird thing or deal with some of the related issues UA mentioned. whatever works, but yes - different thread would probably be best.

totally unrelated edit: please no glitchless. not only is it arbitrary, but that's what let's plays are for.

bpcookie

2015-01-26 01:21:31 am

Quote from Alko:

Glitchless is an extre~mly slippery slope … Oh wait, you knew that. But let me elaborate:

Is backwalking in OoT a 'bannable' glitch? (For those unfamiliar: Walking backwards is faster than walking forwards. No idea if the developpers conciously made that so.)
Is supersliding a 'bannable' glitch? (You are pushed back from a bomb explosion and use some spacy mechanics to glide 'forever'.)
I mean, the latter is definitely unintended, but it still is entirely within the game world, and would have been accepted to SDA in the previous no-leaving-the-game-world times. A LOT of people advocating glitchless runs during the German AGDQ restream would have been fine with supersliding, but where to draw the line?

To me, those two examples are very clear. Backwalking is not a glitch. Bomb slide is a glitch. I could be wrong, but I imagine that that would be a pretty strong consensus. But that's not even the point. If they decided that bomb slides are allowed in glitchless, while I wouldn't agree with that specific decision personally, I'd happily concede that point to the consensus and enjoy watching the glitchless-as-defined-by-committee run. And whichever way that decision fell, that would be the standard that all future glitchless runs would adhere to.

I think the point that no one seems to want to accept is that everyone doesn't have to 100% agree with every decision about every glitch in every game. You just say we're going to appoint 3 knowledgeable people to once and for all decide on these borderline cases, and then whatever the decision is, that's what we go with as our set of rules for that category of that game forever more. It's subjective. There's no wrong answer. It's like arguing balls and strikes in baseball. Yeah, people are going to argue about it. But they appoint a guy to make those decisions, and that's that.

[quote="Alko"]Glitchless is an extre~mly slippery slope … Oh wait, you knew that. But let me elaborate:
[spoiler]Is backwalking in OoT a 'bannable' glitch? (For those unfamiliar: Walking backwards is faster than walking forwards. No idea if the developpers conciously made that so.)
Is supersliding a 'bannable' glitch? (You are pushed back from a bomb explosion and use some spacy mechanics to glide 'forever'.)
I mean, the latter is definitely unintended, but it still is entirely within the game world, and would have been accepted to SDA in the previous no-leaving-the-game-world times. A LOT of people advocating glitchless runs during the German AGDQ restream would have been fine with supersliding, but where to draw the line?[/spoiler]

[/quote]

To me, those two examples are very clear. Backwalking is not a glitch. Bomb slide is a glitch. I could be wrong, but I imagine that that would be a pretty strong consensus. But that's not even the point. If they decided that bomb slides are allowed in glitchless, while I wouldn't agree with that specific decision personally, I'd happily concede that point to the consensus and enjoy watching the glitchless-as-defined-by-committee run. And whichever way that decision fell, that would be the standard that all future glitchless runs would adhere to.

I think the point that no one seems to want to accept is that everyone doesn't have to 100% agree with every decision about every glitch in every game. You just say we're going to appoint 3 knowledgeable people to once and for all decide on these borderline cases, and then whatever the decision is, that's what we go with as our set of rules for that category of that game forever more. It's subjective. There's no wrong answer. It's like arguing balls and strikes in baseball. Yeah, people are going to argue about it. But they appoint a guy to make those decisions, and that's that.

Naegleria

(user is banned)

2015-01-26 01:23:38 am

anytime you want to ban glitches, the category should describe that

sonic 2 "no level wraps" or no horizontal underflow, not sonic 2 glitchless

a link to the past no exploration glitch, not alttp glitchless

etc

bpcookie

2015-01-26 02:02:28 am

Quote from UraniumAnchor:

The issue with "glitchless" it that's it's far too vague. Is abusing some level geometry that lets you skip a significant portion of the game considered a glitch if all you have to do is jump in a clever fashion? Does it count as a glitch if it's just some special combo that's not listed in the manual? (Maybe shield dashing in SOTN would be a good example here, or basically any game with animation cancelling.)

Pick 5 people knowledgeable about SOTN, and we'll ask them. If 3 or more say shield dashing is allowed, then it's allowed. And just like that we have our rule sheet for all future SOTN runs.

Quote from Cronikeys:

please no glitchless. not only is it arbitrary, but that's what let's plays are for.

It's perfectly valid for you to be against glitchless and argue that it's arbitrary. But this part about let's plays is not only clearly wrong, but needlessly condescending. What you're saying implies that only the fastest category is relevant, and that all others should be written off as casual.

Anyway, sorry to everyone. I think this is a great thread with many important ideas, and I don't mean to monopolize it. I've spoken my peace about the idea of a glitchless category and I won't keep bringing it back up.

[quote="UraniumAnchor"]
The issue with "glitchless" it that's it's far too vague. Is abusing some level geometry that lets you skip a significant portion of the game considered a glitch if all you have to do is jump in a clever fashion? Does it count as a glitch if it's just some special combo that's not listed in the manual? (Maybe shield dashing in SOTN would be a good example here, or basically any game with animation cancelling.)
[/quote]

Pick 5 people knowledgeable about SOTN, and we'll ask them. If 3 or more say shield dashing is allowed, then it's allowed. And just like that we have our rule sheet for all future SOTN runs.

[quote="Cronikeys"]please no glitchless. not only is it arbitrary, but that's what let's plays are for.[/quote]

It's perfectly valid for you to be against glitchless and argue that it's arbitrary. But this part about let's plays is not only clearly wrong, but needlessly condescending. What you're saying implies that only the fastest category is relevant, and that all others should be written off as casual.

Anyway, sorry to everyone. I think this is a great thread with many important ideas, and I don't mean to monopolize it. I've spoken my peace about the idea of a glitchless category and I won't keep bringing it back up.

Reiska

2015-01-26 03:41:16 am

1) Naegleria makes a pretty good point, can't deny
2) personal pet peeve but I hate seeing "let's play" thrown around like an insult, there is still a very definite gap between most _actual_ let's plays and a run that is still aiming for the fastest time possible within certain constraints (e.g. playing the game as "intended" by the developers). admittedly, I'm a little biased in this regard because I do semi-casually speedrun glitchless any% in an extremely broken game (that used to be a popular speedgame until it was broken so thoroughly). I'm capable of *doing* the glitched routes, but they just aren't fun to play in my opinion as they skip either what I consider to be some of the most fun parts of the game, or are an actual credits warp after 50 seconds of play.

that said, like bpcookie, I don't want to monopolize this thread with this topic either, as there are other important topics being discussed as well, so if people want to discuss further in this direction we probably should take it to another thread.

1) Naegleria makes a pretty good point, can't deny
2) personal pet peeve but I hate seeing "let's play" thrown around like an insult, there is still a very definite gap between most _actual_ let's plays and a run that is still aiming for the fastest time possible within certain constraints (e.g. playing the game as "intended" by the developers).  admittedly, I'm a little biased in this regard because I do semi-casually speedrun glitchless any% in an extremely broken game (that used to be a popular speedgame until it was broken so thoroughly).  I'm capable of *doing* the glitched routes, but they just aren't fun to play in my opinion as they skip either what I consider to be some of the most fun parts of the game, or are an actual credits warp after 50 seconds of play.

that said, like bpcookie, I don't want to monopolize this thread with this topic either, as there are other important topics being discussed as well, so if people want to discuss further in this direction we probably should take it to another thread.

Mystery

Edit history:

Mystery: 2015-01-26 04:08:34 am

Mystery: 2015-01-26 04:06:51 am

Mystery: 2015-01-26 04:06:04 am

Mystery: 2015-01-26 03:55:56 am

Mystery: 2015-01-26 03:45:39 am

2015-01-26 03:45:14 am

Quote from Melodia:

Again, maybe people like me who would rely on the feed for updates is in the minority but it's a point...

I use feed too (That's what they're there for! Facebook/twitter was NOT meant to be a source of news), but I haven't encountered any problems as far as I know.

Quote from Alko:

Edit: I totally forgot about allowing Japanese turbo controllers for RPGs … I'm leaning against this because it can lead to a lot faster time (eliminating human error when mashing through text plus potentially faster fighting plus easier, more consisten stutter stepping in FFIX. However, with a clear separation it could work.

I honestly think it's unfair to ban turbo controllers. Some people just hate mashing. Are you going to ban them from submitting a run because of that?

Quote from Alko:

I see where you're coming from. However, we don't see people face to face; we see the text that they write plus a more or less random avatar next to it. Other than checking each user's profile before answering (could also be a lie or missing … not happening. Sorry, mate), having seen them on stream or met them in person, there's no way to guess. Spelling out names all the time, the way Japanese does it, isn't an option and neither is resorting to languages such as Finnish (hän meaning he/she/it). S/he is damn awkward to read and write so a thoughtless 'he' remains the easiest option.
Also, people tend to forget, even if they should know. I should now know Cronikeys is female, but I'll likely have forgotten by the time I respond to a post of her's the next time ^^' No offence ment Wink

So you would rather knowingly possibly be rude than take a little extra time to avoid gender specific words? I don't get you. Is it really going to kill you to avoid specifically writing he or she? If you avoid it, it won't matter if you don't know the gender, or if the gender is wrong, or if you've forgotten. Just saying. You don't assume people's gender in real life if you forget it.

Quote from ballofsnow:

I mean who else is going to do it?

Why don't we just ask the community for some help? I mean, seriously, there are a lot of people here with a lot of talent. Use it!

[quote="Melodia"]Again, maybe people like me who would rely on the feed for updates is in the minority but it's a point...[/quote]
I use feed too (That's what they're there for! Facebook/twitter was NOT meant to be a source of news), but I haven't encountered any problems as far as I know.

[quote="Alko"]Edit: I totally forgot about allowing Japanese turbo controllers for RPGs … I'm leaning against this because it can lead to a lot faster time (eliminating human error when mashing through text plus potentially faster fighting plus easier, more consisten stutter stepping in FFIX. However, with a clear separation it could work.[/quote]
I honestly think it's unfair to ban turbo controllers. Some people just hate mashing. Are you going to ban them from submitting a run because of that?

[quote="Alko"]I see where you're coming from. However, we don't see people face to face; we see the text that they write plus a more or less random avatar next to it. Other than checking each user's profile before answering (could also be a lie or missing … not happening. Sorry, mate), having seen them on stream or met them in person, there's no way to guess. Spelling out names all the time, the way Japanese does it, isn't an option and neither is resorting to languages such as Finnish (hän meaning he/she/it). S/he is damn awkward to read and write so a thoughtless 'he' remains the easiest option.
Also, people tend to forget, even if they should know. I should now know Cronikeys is female, but I'll likely have forgotten by the time I respond to a post of her's the next time ^^' No offence ment ;) [/quote]
So you would rather knowingly possibly be rude than take a little extra time to avoid gender specific words? I don't get you. Is it really going to kill you to avoid specifically writing he or she? If you avoid it, it won't matter if you don't know the gender, or if the gender is wrong, or if you've forgotten. Just saying. You don't assume people's gender in real life if you forget it.

[quote="ballofsnow"]I mean who else is going to do it?[/quote]
Why don't we just ask the community for some help? I mean, seriously, there are a lot of people here with a lot of talent. Use it!

Reiska

2015-01-26 03:53:06 am

The problem is that English is kind of poor at actually being gender neutral (using "they" is the best way English can handle it, really, but it comes off as somewhat awkward and impersonal IMO), for all that I otherwise agree with your point.

moooh

Edit history:

moooh: 2015-01-26 04:02:18 am

2015-01-26 04:00:33 am

Exoray

uuuh, Mystery: that's definitely not a quote from me you have there last in your post. Not sure why you labeled it as me by hand..

Mystery

Edit history:

Mystery: 2015-01-26 04:28:29 am

Mystery: 2015-01-26 04:07:37 am

2015-01-26 04:05:12 am

Oops, sorry, sorry! So sorry! The quoting got a bit messed up. Let me try to fix that.

EDIT: Fixed

I believe quotes got inserted into the middle of others, which happens if the cursor isn't at the end of a post when quoting. I had to try to fix it by hand, so mistakes slipped by.

Again, sorry for that.

Nyasupan

Edit history:

Nyasupan: 2015-01-26 05:09:51 am

2015-01-26 04:18:13 am

Quote from Mystery:

So you would rather knowingly possibly be rude than take a little extra time to avoid gender specific words? I don't get you. Is it really going to kill you to avoid specifically writing he or she? If you avoid it, it won't matter if you don't know the gender, or if the gender is wrong, or if you've forgotten. Just saying. You don't assume people's gender in real life if you forget it.

Can we please stop right here and not make a drama out of this? I wrote the initial message in my spare time between classes at university, and I assumed a gender. You pointed out my mistake and I fixed that barely 3 hours later.

Quote from Nyasupan:

Quote from Mystery:

Cronikeys is female. Please don't assume gender.

I am sorry, I will edit asap.

I've not received any sort of message from Cronikeys reporting I've disrespected her with that message, I too have a neutral nickname with a girl in the avatar and someone could assume I am a girl. I wouldn't mind to be referred as a she and then correct it. It happens. We're talking of something else in this thread, let's keep the discussion focused on that.

[quote="Mystery"]So you would rather knowingly possibly be rude than take a little extra time to avoid gender specific words? I don't get you. Is it really going to kill you to avoid specifically writing he or she? If you avoid it, it won't matter if you don't know the gender, or if the gender is wrong, or if you've forgotten. Just saying. You don't assume people's gender in real life if you forget it.[/quote]
Can we please stop right here and not make a drama out of this? I wrote the initial message in my spare time between classes at university, and I assumed a gender. You pointed out my mistake and I fixed that barely 3 hours later.

[quote="Nyasupan"][quote="Mystery"]Cronikeys is female. Please don't assume gender.[/quote]
I am sorry, I will edit asap.[/quote]

I've not received any sort of message from Cronikeys reporting I've disrespected her with that message, I too have a neutral nickname with a girl in the avatar and someone could assume I am a girl. I wouldn't mind to be referred as a she and then correct it. It happens. We're talking of something else in this thread, let's keep the discussion focused on that.

Yagamoth

2015-01-26 05:58:17 am

INTJ

To sum up what I'd wish to have from SDA:

1) An archive (~museum), where I can go back in time to see how a game developed in the speedrunning community
-> Almost in place already, only need to see the runs that are already obsoleted on the site

2) A more easily accessible knowledge base per game linked to the discussion thread. Essentially what a wiki is, but easier to use
-> Both adding and editing information should be simple
-> Having to learn markup is actually quite the barrier

3) A platform for discussion of speedrunning related things
-> Already exists as the forums

============================

The main reason why I go to SDA runs once in a while is because I love the detailed comments and commentary. But by now, having to download high quality runs is more of a hindrance than anything else to me personally. Something like Youtube is so much simpler to start watching. Further, many runs don't really have commentary to them, making a live-run-vod more appealing where I at least get the runners reaction to various events in the run.

Personally I feel a lot like Cyghfer in most aspects. If I want to watch a speedrun of a specific game, I never go to SDA anymore. I look things up on Youtube, Twitch and sometimes Nicovideo. I have no idea how to say this in a way, that no one currently involved in the process will feel hurt, but I really have to agree with Cyghfers stance:

I wouldn't really care if SDA ceased its speedrun-publishing operations

It feels ungrateful of me to say that, considering how many runs on SDA I've enjoyed over the course of the years, considering how many people are involved in the verification process and it's improvement. But that's just how I feel - Sorry about that

To sum up what I'd wish to have from SDA:

1) An archive (~museum), where I can go back in time to see how a game developed in the speedrunning community
-> Almost in place already, only need to see the runs that are already obsoleted on the site

2) A more easily accessible knowledge base per game linked to the discussion thread. Essentially what a wiki is, but easier to use 
-> Both adding and editing information should be simple
-> Having to learn markup is actually quite the barrier

3) A platform for discussion of speedrunning related things
-> Already exists as the forums

============================

The main reason why I go to SDA runs once in a while is because I love the detailed comments and commentary. But by now, having to download high quality runs is more of a hindrance than anything else to me personally. Something like Youtube is so much simpler to start watching. Further, many runs don't really have commentary to them, making a live-run-vod more appealing where I at least get the runners reaction to various events in the run.

Personally I feel a lot like Cyghfer in most aspects. If I want to watch a speedrun of a specific game, I never go to SDA anymore. I look things up on Youtube, Twitch and sometimes Nicovideo. I have no idea how to say this in a way, that no one currently involved in the process will feel hurt, but I really have to agree with Cyghfers stance:

I wouldn't really care if SDA ceased its speedrun-publishing operations

It feels ungrateful of me to say that, considering how many runs on SDA I've enjoyed over the course of the years, considering how many people are involved in the verification process and it's improvement. But that's just how I feel - Sorry about that

ais523

2015-01-26 07:47:58 am

I'm actually really happy with what SDA is doing at the moment, and wouldn't mind much if nothing changed. There is scope for improvement, of course, but where we are right now is still a pretty good place.

It would be nice to see the knowledge base more used. I've already started updating it for Neverwinter Nights (I made the equivalent page on TASvideos for NetHack). One thing that would be nice would be more centralised cross-linking between various strategy archives. (For instance, for the Metroid Prime games, metroid2002 has a lot of information but much of it is outdated; and much of the more recent Metroid information is in its forums. You wouldn't necessarily discover either fact just reading SDA, although you'd quickly be told if you asked in any of the relevant threads.) It would be nice if it were easy to find all the strategic information for a game.

I guess the delay in hitting the front page is less of a problem for me because a) the run is typically on YouTube while I'm waiting, and b) a really good speedrun takes years to produce anyway. I appreciate that it can be a problem for some people, though. (For what it's worth, TASvideos also has a long delay, but the entire process is public; the most time-consuming part is normally video encoding, followed by verification.)

I'm actually really happy with what SDA is doing at the moment, and wouldn't mind much if nothing changed. There is scope for improvement, of course, but where we are right now is still a pretty good place.

It would be nice to see the knowledge base more used. I've already started updating it for Neverwinter Nights (I made the equivalent page on TASvideos for NetHack). One thing that would be nice would be more centralised cross-linking between various strategy archives. (For instance, for the Metroid Prime games, metroid2002 has a lot of information but much of it is outdated; and much of the more recent Metroid information is in its forums. You wouldn't necessarily discover either fact just reading SDA, although you'd quickly be told if you asked in any of the relevant threads.) It would be nice if it were easy to find all the strategic information for a game.

I guess the delay in hitting the front page is less of a problem for me because a) the run is typically on YouTube while I'm waiting, and b) a really good speedrun takes years to produce anyway. I appreciate that it can be a problem for some people, though. (For what it's worth, TASvideos also has a long delay, but the entire process is public; the most time-consuming part is normally video encoding, followed by verification.)

z1mb0bw4y

2015-01-26 10:39:49 am

Gets the cake.

I think the biggest thing TASvideos has going for its verification process is that someone usually has a temp encode up the same day the run is submitted. Something like that would be very nice for everyone to be able to see a run that has reached verification, even if they're not verifying the run. Perhaps just a low quality youtube encode or something along those lines.

IsraeliRD

Edit history:

IsraeliRD: 2015-01-26 11:37:56 am

2015-01-26 11:33:48 am

Dragon Power Supreme

I'd like to throw in a few cents in as well.

Like a few others, I don't visit other speedrunning sites frequently, so the speedruns I watch come primarily from SDA. While I can check YT/Twitch, I rarely do unless I want to see the latest record or WIP pretty badly, and this happens maybe once a month. In addition to that, I watch TASes from TASVideos when they are published.
As of currently I'm at the army, and while I get to go home daily, some days or weekends I'm staying in the base. When I'm at home, I have 4 hours to catch up on the day's events, prepare for tomorrow and do whatever else I want in the remainder of the time. While at the army, I feed off my mobile internet. As one can imagine, no work happens towards SDA during these overnights on base. During these times I prefer to watch a TAS I downloaded as it passes the time just great.

My current roles on SDA are these: timing runs, taking gamepage screenshots and posting updates. I also help with PRC if necessary, but most often this is not needed.
- Timing runs: I love it. I haven't got bored of this, and pretend to enjoy swearing at the brutality of PC games, heavily segmented runs or some ILs. I've been told by others that timing runs is "black magic" they don't want to get near, and I completely understand them. I got burned out only once and that was due to two Abe games that gave me a headache.
- Gamepage screenshots: I already time the runs, and watch them, so why not do that as well? This actually helped previous updaters way more than you can imagine since they didn't have to download the games, take screenshots and sometimes even had to search the nets for a title screen which may at times become pretty difficult. There used to be a script from nate that randomly took screenshots, but I never used it. It probably works. I don't want a "loading" screen as a screenshot though Tongue

I know there are some gamepages with exactly that too lol.
- PRC (Prerelease Check): Simple and easy 99% of the time. Last one I did was Dune 2000 which had a bunch of problems from verification so it's a good thing I checked it as well... It's also the run no one would dare try and PRC it in our galaxy but I was crazy enough to do it. Dune 2000 this Saturday's update hype.
- Updating. It used to be hard. The "theme" some people dislike is actually something I like, but I never purposely hold up one or more runs just so a themed update will happen in a later date when the "missing run" will come up to the end of the queue. In the past you'd make the update and write it probably a day or so before you'd post (and in some cases, same day), which was stupid imo.
Nowadays I have updates ready weeks, even more than a month ahead of time. All written out and just need to go the gamelist and gamepage intros, shift files and done. My research into the games in the update usually mean I can do a quick blurb on the game, do the gamepage intro quietly and even have some knowledge about the game so it doesn't turn out to be something too terrible.
Regarding those who will complain about having updates ready a month ahead: timing speedruns can be brutal, and PC games can hold things up. With the little time I have (weekends mostly free), I can spend time doing timings without holding up the whole damn queue because of one particular game. I also don't want to write updates daily... again, it's all about not burning out.

Please don't go into how you can reduce my workload. I doubt you can. Timing runs is the big one and I've handled it well; I can time lots of runs in a weekend, so I generally don't tend to be a problem in the queue. If my weekend goes into a PC run or two, I will immediately make up for the remainder of the runs by eating my free time during the week.

I started seeing the SDA front page updates from Enhasa's era. I haven't played 99.99% of games on SDA, and watching runs for me is usually composed of the game page's intro text, the screenshots and runner's comments. If my interest is sparked by all those, I'd download the run and watch it.
I liked Enhasa's updates and I further enjoyed the updates LLCoolDave and ShadowWraith did; the themes and jokes imo worked well if they didn't go overboard, and a quick joke isn't too hard to do and can bring a smile to some people. Keeping my own updates small means I don't burn myself out, can add in a quick joke (now a title-subtitle combo) and maybe add stuff further down the update. My "I have no regrets" is where I get to have some fun and is the extent where my updates will go in terms of "going overboard".
I also enjoy giving runners respect for their runs, and thus giving them some time on the front page rather than have it thrown out the next day by another speedrunner's run. It gives them a nod for their hard work and wait while their runs goes through the site. I know that for those who want the run "now" this matters not, but I feel it does for the runners who submit.

TASVideos got mentioned here a few times. I agree that their verification is easier than SDA, but most people forget why: they don't have to worry about A/V, missing footage, possible cheating and other things. They only need to launch the replay file into an emulator and verify it syncs. No cheating, no missing footage etc. Much harder here unfortunately.

The thing I dislike about with TASVideos is that while they may have a faster turnaround, I am completely turned off by their "TODO: Describe this movie here." in their description. As someone who TASed and knows how stupid difficult this can get and the patience needed, I would be turned off completely and never want to submit again if my hard work resulted in a publisher's laziness to take 5 minutes to Google for a game description. Some TODOs of these even stay up there for days!, even if the run is in the Moon Tier. Some of the game descriptions are provided by the TASers and are copied nearly word for word by the publishers... this happened to my runs as well. I wonder if I would've gotten a TODO if I didn't submit a description.

Someone mentioned all the hoops to get through SDA. What about TASVideos? It is simple enough for a quick and dirty hack, but when I wanted to do my own Hocus Pocus TASes I read through their requirements and steps and instantly got turned off. I needed to have an approved image that had some blurbs about TASVideos. I needed to download several programs, do code stuff and a whole bunch of things I have no understanding or knowledge of. I don't have the time, patience or programs and resources to do that, so I never did that. Compare with SDA's yua. I dump the run in, check on StatID, put in whatever I want, mark in the qualities that are in red color and off it goes making perfect SDA submittable runs. Their requirements might be easier today, I dunno.

Also hats of for nate for doing so many runs encoding even to this day. TASVideos has a problem IMO where their standards are so high (4k YouTube, anyone?) that their encoders have a huge turnaround... don't know how many of them stay for a year or longer than that. Probably most lose interest after 6-12 months. It also causes problems in having runs published because it takes their encoders time to do the required encodes. I can't claim to know their current standards or how long they take to encode, but I do generally remember it taking damn long. The upside is I can watch the quick 'n dirty encode on YouTube that someone randomly does (if at all) rather than wait weeks for the run to be published.

Not saying I don't like TASVideos. I do very much and fully respect them. These guys provide the runs I watch at the army overnight and keep me entertained when I'm bored at home. I also vote for the TASes I watch, and you'll see I have 36.4% of their published movies as me voting on them. The number would be easily 50%+ because it took me a couple of years before I registered and then voted. I don't have the time to be involved in their community, but these guys have top respect from me and they were my first VOD to watch after AGDQ 2014 & 2015 (my backlog is from SGDQ 2013).

Lastly... volunteers. It's so easy to say things, but by god actually doing them is a different world altogether and unfortunately most don't realise that. Kudos to the past and current SDA and TASVideos staff for all their hard work, both websites wouldn't have gotten as far as they have if it wasn't for them.

Anyway these are my ramblings. I hope I provided some insight into what I do.
Shit that took longer to write than I wanted, better get ready for tomorrow.

I'd like to throw in a few cents in as well.

Like a few others, I don't visit other speedrunning sites frequently, so the speedruns I watch come primarily from SDA. While I can check YT/Twitch, I rarely do unless I want to see the latest record or WIP pretty badly, and this happens maybe once a month. In addition to that, I watch TASes from TASVideos when they are published.
As of currently I'm at the army, and while I get to go home daily, some days or weekends I'm staying in the base. When I'm at home, I have 4 hours to catch up on the day's events, prepare for tomorrow and do whatever else I want in the remainder of the time. While at the army, I feed off my mobile internet. As one can imagine, no work happens towards SDA during these overnights on base. During these times I prefer to watch a TAS I downloaded as it passes the time just great.

My current roles on SDA are these: timing runs, taking gamepage screenshots and posting updates. I also help with PRC if necessary, but most often this is not needed.
- Timing runs: I love it. I haven't got bored of this, and pretend to enjoy swearing at the brutality of PC games, heavily segmented runs or some ILs. I've been told by others that timing runs is "black magic" they don't want to get near, and I completely understand them. I got burned out only once and that was due to two Abe games that gave me a headache.
- Gamepage screenshots: I already time the runs, and watch them, so why not do that as well? This actually helped previous updaters way more than you can imagine since they didn't have to download the games, take screenshots and sometimes even had to search the nets for a title screen which may at times become pretty difficult. There used to be a script from nate that randomly took screenshots, but I never used it. It probably works. I don't want a "loading" screen as a screenshot though :P [size=7]I know there are some gamepages with exactly that too lol.[/size]
- PRC (Prerelease Check): Simple and easy 99% of the time. Last one I did was Dune 2000 which had a bunch of problems from verification so it's a good thing I checked it as well... It's also the run no one would dare try and PRC it in our galaxy but I was crazy enough to do it. Dune 2000 this Saturday's update hype.
- Updating. It used to be hard. The "theme" some people dislike is actually something I like, but I never purposely hold up one or more runs just so a themed update will happen in a later date when the "missing run" will come up to the end of the queue. In the past you'd make the update and write it probably a day or so before you'd post (and in some cases, same day), which was stupid imo.
Nowadays I have updates ready weeks, even more than a month ahead of time. All written out and just need to go the gamelist and gamepage intros, shift files and done. My research into the games in the update usually mean I can do a quick blurb on the game, do the gamepage intro quietly and even have some knowledge about the game so it doesn't turn out to be something too terrible.
Regarding those who will complain about having updates ready a month ahead: timing speedruns can be brutal, and PC games can hold things up. With the little time I have (weekends mostly free), I can spend time doing timings without holding up the whole damn queue because of one particular game. I also don't want to write updates daily... again, it's all about not burning out.

Please don't go into how you can reduce my workload. I doubt you can. Timing runs is the big one and I've handled it well; I can time lots of runs in a weekend, so I generally don't tend to be a problem in the queue. If my weekend goes into a PC run or two, I will immediately make up for the remainder of the runs by eating my free time during the week.

I started seeing the SDA front page updates from Enhasa's era. I haven't played 99.99% of games on SDA, and watching runs for me is usually composed of the game page's intro text, the screenshots and runner's comments. If my interest is sparked by all those, I'd download the run and watch it.
I liked Enhasa's updates and I further enjoyed the updates LLCoolDave and ShadowWraith did; the themes and jokes imo worked well if they didn't go overboard, and a quick joke isn't too hard to do and can bring a smile to some people. Keeping my own updates small means I don't burn myself out, can add in a quick joke (now a title-subtitle combo) and maybe add stuff further down the update. My "I have no regrets" is where I get to have some fun and is the extent where my updates will go in terms of "going overboard".
I also enjoy giving runners respect for their runs, and thus giving them some time on the front page rather than have it thrown out the next day by another speedrunner's run. It gives them a nod for their hard work and wait while their runs goes through the site. I know that for those who want the run "now" this matters not, but I feel it does for the runners who submit.

TASVideos got mentioned here a few times. I agree that their verification is easier than SDA, but most people forget why: they don't have to worry about A/V, missing footage, possible cheating and other things. They only need to launch the replay file into an emulator and verify it syncs. No cheating, no missing footage etc. Much harder here unfortunately.

The thing I dislike about with TASVideos is that while they may have a faster turnaround, I am completely turned off by their "TODO: Describe this movie here." in their description. As someone who TASed and knows how stupid difficult this can get and the patience needed, I would be turned off completely and never want to submit again if my hard work resulted in a publisher's laziness to take 5 minutes to Google for a game description. Some TODOs of these even stay up there for days!, even if the run is in the Moon Tier. Some of the game descriptions are provided by the TASers and are copied nearly word for word by the publishers... this happened to my runs as well. I wonder if I would've gotten a TODO if I didn't submit a description.

Someone mentioned all the hoops to get through SDA. What about TASVideos? It is simple enough for a quick and dirty hack, but when I wanted to do my own Hocus Pocus TASes I read through their requirements and steps and instantly got turned off. I needed to have an approved image that had some blurbs about TASVideos. I needed to download several programs, do code stuff and a whole bunch of things I have no understanding or knowledge of. I don't have the time, patience or programs and resources to do that, so I never did that. Compare with SDA's yua. I dump the run in, check on StatID, put in whatever I want, mark in the qualities that are in red color and off it goes making perfect SDA submittable runs. Their requirements might be easier today, I dunno.

Also hats of for nate for doing so many runs encoding even to this day. TASVideos has a problem IMO where their standards are so high (4k YouTube, anyone?) that their encoders have a huge turnaround... don't know how many of them stay for a year or longer than that. Probably most lose interest after 6-12 months. It also causes problems in having runs published because it takes their encoders time to do the required encodes. I can't claim to know their current standards or how long they take to encode, but I do generally remember it taking damn long. The upside is I can watch the quick 'n dirty encode on YouTube that someone randomly does (if at all) rather than wait weeks for the run to be published.

Not saying I don't like TASVideos. I do very much and fully respect them. These guys provide the runs I watch at the army overnight and keep me entertained when I'm bored at home. I also vote for the TASes I watch, and you'll see I have 36.4% of their published movies as me voting on them. The number would be easily 50%+ because it took me a couple of years before I registered and then voted. I don't have the time to be involved in their community, but these guys have top respect from me and they were my first VOD to watch after AGDQ 2014 & 2015 (my backlog is from SGDQ 2013).

Lastly... volunteers. It's so easy to say things, but by god actually doing them is a different world altogether and unfortunately most don't realise that. Kudos to the past and current SDA and TASVideos staff for all their hard work, both websites wouldn't have gotten as far as they have if it wasn't for them.

Anyway these are my ramblings. I hope I provided some insight into what I do.
[size=4]Shit that took longer to write than I wanted, better get ready for tomorrow.[/size]

Alko

2015-01-26 03:47:23 pm

Caution: This user contains Kana ^_^

Ah, damnit, I completely forgot about IL tables and other stuff. Yeah, that makes coding it a lot harder.

Quote from Naegleria:

anytime you want to ban glitches, the category should describe that

sonic 2 "no level wraps" or no horizontal underflow, not sonic 2 glitchless

a link to the past no exploration glitch, not alttp glitchless

etc

This.

Quote from Mystery:

Quote from Alko:

Edit: I totally forgot about allowing Japanese turbo controllers for RPGs … I'm leaning against this because it can lead to a lot faster time (eliminating human error when mashing through text plus potentially faster fighting plus easier, more consisten stutter stepping in FFIX. However, with a clear separation it could work.

I honestly think it's unfair to ban turbo controllers. Some people just hate mashing. Are you going to ban them from submitting a run because of that?

Yes. Because it removes an element of skill from the run. See all the praise 0xwas got behind the scenes at AGDQ when he finally beat the mashing game. Note: I hate mashing and I'm not good at it. Wink

Taking the rest private because irrelevant to the topic.

Ah, damnit, I completely forgot about IL tables and other stuff. Yeah, that makes coding it a lot harder.

[quote="Naegleria"]anytime you want to ban glitches, the category should describe that

sonic 2  "no level wraps"  or no horizontal underflow,  not sonic 2 glitchless

a link to the past no exploration glitch, not alttp glitchless

etc[/quote]
This.

[quote="Mystery"][quote="Alko"]Edit: I totally forgot about allowing Japanese turbo controllers for RPGs … I'm leaning against this because it can lead to a lot faster time (eliminating human error when mashing through text plus potentially faster fighting plus easier, more consisten stutter stepping in FFIX. However, with a clear separation it could work.[/quote]
I honestly think it's unfair to ban turbo controllers. Some people just hate mashing. Are you going to ban them from submitting a run because of that?[/quote]
Yes. Because it removes an element of skill from the run. See all the praise 0xwas got behind the scenes at AGDQ when he finally beat the mashing game. Note: I hate mashing and I'm not good at it. ;)

Taking the rest private because irrelevant to the topic.