Iomega Storecenter Ix2 .. The Rise, the Fall and?

 Names 

So a long time ago I picked up a Iomega Storecenter IX2. Most often when you search for this you will come up with the more modern versions. There is an IX2 'two hundred', and now a modern IX2, along with a IX2 "slash D L".

Already in that first paragraph we have come across a problem with this device. I put those last items in quotes to help those looking for the original item and not the new improved ones. It is extremely hard to find posts and articles regarding this generation because of the names that they chose for the newer ones. Conspiracy theories exist that Iomega is trying to make up for a failed release by burying it as much as possible online. I however remain unsure that it isn't just the forgetful nature of big companies once the newer set of tech arrives.

Compatibility

You might be curious about the newer IX2 and if things like software apps and firmware are compatible. They are not.

	Iomega ix2 (new)	Iomega ix2-"two hundred"	Iomega ix2 (original)
CPU	Marvell 88F62E2A1C160 Kirkwood @ 1.6 GHz	Marvell 88F6281 Kirkwood @ 1 GHz	Marvell 88F5182 "Orion"
Ethernet	Marvell 88E81318	Marvell 88E81116R	Marvell 88E1118
RAM	256 MB	256 MB	128 MB
Flash	128 MB	128 MB	4 MB

Table 1: Component summary and comparison

The above table was originally posted here http://www.smallnetbuilder.com/.
As you can see the hardware varies significantly from the original to the newer versions, so any firmware wouldn't be transferable between them. Missing from the above table is the "d l" version, that would go in the "IX2 (new)" column, since it is merely the diskless version of that unit.

Rise 

As is true of many of my tech gadgets I originally got this NAS on Craigslist. It was an excellent price, and I had need of a large amount of storage due to an upcoming bout of video editing. At the time I was running a coffee shop that also doubled as a music venue. Many of the bands there liked coming around and using our compute resources to whip up a quick song or video.

It came originally with 2 500gb drives that you can configure either in Raid 1 or RAID 0. In theory Raid 1 mirrors both drives giving you some additional security, Raid 0 stripes both drives giving you added speed at the cost of security.  

I had need of copious volumes of speed and storage space so I set it up as RAID 0. I have rarely had a problem with reliability, as modern hard drives are a fairly hardy sort.

This worked for nearly a year. Its hard to tell how old the unit was originally, but it was at least a year or two old when I got it.

The NAS itself worked great. It has an easy web setup, the ability to setup different user accounts, you can password protect storage, its easily sharable on a network (windows, mac, or linux). It has a gigabit LAN connection so its fairly fast, though it seems limited by drive speed.

Fall

Then came the crash. I attempted to look up one of saved files to find that it wasn't available on the network. Repeated attempts led to nothing. Looking up the web interface indicated a failed drive. Restarting the unit led to a blinking light and no further availability of the web interface. It was essentially dead.

The very next day one of our friends, who worked with bands extensively came in after a show. His computer had been blown up by bad electrics at a local bar, he was hoping for the backups he had stored on our IX2. :( This was not happy times.

Looking online led to the previously mentioned frustration with the names. ..also the prevailing but useless wisdom that once a RAID 0 array fails you are screwed.

And...

I know that there often tricks to be used, so I didnt trash the drives, or overwrite the data as some might do. What I did do was go out and buy some new drives since the old ones had failed. I was now going to come up with 2 2TB drives in RAID 1, not wanting to repeat old mistakes, but wanting more space at the same time.

Various discoveries were made in this process:

1. The fan setup in this case is noticeably bad. Just by looking it seemed ineffective. Others have some to the same conclusion.
   
   This post here shows how someone else fixed a similar situation including the fan. Planetkris post.
2.  The raid is broken performance wise. Switching to RAID 1 is supposed to be slower then useing RAID 0. The striping version is supposed to nearly double your output, while showing the space of both drives as available. The size was there, but the speed was actually SLOWER then in the mirrored version. So no point in using that at all!
3. Its fairly hard, and takes a lot of messing about with the file systems on these drives to get the new drives to work. Ideally you would just pull 1 out, have the device update it, then add the second. This is old tech though, and luck is just not with us here.
   
   This post here deals with that problem BFtech blog post 

 Between all the information in those posts, and the searching I did using them I revived the old IX2 and have been using it ever since. Just about 2 years have gone by now, and it still runs strong.

I didn't do everything exactly as posted in those other pieces. At the time I thought the fan wasn't SOooo bad, so I merely cut out some of the grills on the front and the back. They were literally just tiny slits originally, and barely any air could move, so that clearly had to change.  Moving to the larger drives took some experimentation and multiple tries at copying the partitions, but eventually it did take. That experience mirrors that of many commenters so don't give up if your first try through the directions doesn't do it.

Blowing off the dust...

Now that it been a while I decided to dig out those old drives that I had stored so long ago. No reason to let some decently large hard drives just gather dust forever. Why not see if anything of use can be done with the info on them? By now Ive been living without all the data that was there, I wouldn't really feel it if it was indeed lost forever. Then I could reclaim the drives and use them for something in my computers before they became completely obsolete.

What a surprise.. I was actually able to rescue the original data! It took learning some obscure methods of linux partioning that Iomega decided to use, but it was manageable. Ill outline the gist of it below, so you can do the same if the need arises.

On the IX2 you have 2 drives hooked to a small motherboard. Each of the drives has 2 partitions. Here, Ill show you.

That first partition that you see in blue is quite important. That is the firmware for the entire device. That gets loaded into the motherboard on startup, and this is how the whole thing gets going. Often theres an electronic storage area on any given piece of electronics that holds its firmware, Iomega decided to use the hard drives instead.

This has some drawbacks, in that if the drive loses some of that info the whole device goes down. But aren't both drives the same? Yes, however the IX2 is not intelligent enough to switch over to the working version if 1 messes up. You can do this externally, and attempt to revive it, but this often is too little too late. This does work for updating to new larger hard drives though. Just take out the drives, use your favorite partitioning software to setup a similar structure and copy the data from that first partition over. ....and repeat. 

That 2nd partition is where Iomega got tricky though. Its not just a simple partition like you and I would normally mess with in lets say Partition Magic or the windows disk manager. No they decided to use something fancier here called LVM (logical volume management) partitions. This is a whole different, obscure, and very complex bit of doing a similar thing. ..and almost no O.S. is going to understand it by default. With some love and understanding, a bit of reading, and some additional packages you can get Linux to deal with it.

That isn't a step by step guide, obviously. But there should be enough information there to help you rescue any information you may have lost. .. I hope.

Shrink that file! ..yes but by how much? 

I was moving a rather large virtual disk today, which led to an entire odyssey of problems, as these minor tasks often seem to do. Towards the end I decided to compress the virtual disk for archiving.

I have a few different tried and true compressors on my computer, why not test them out to see who does better on this particular task. Compressing 1 monolothic huge, presumably compressible file is the challenge.

Using the MS Windows tools I already brought the disk down from a hefty 9.5 gigs to 5.4. Could we go much further than that, or does MS provide compression with their compacting tool?

So we have 4 contenders, and 3 using two different compression levels:

• Winrar 2.4
• Hamster Freezip 3.0
• Winzip 17
• 7zip 9.2
• 7zip 17.01

Winrar 2.4 is a venerable choice. However over the ages its proven its worth with a good balance of speed and the ability to out compress zipfile versions on the normal settings. More recent versions of zip compressors have begun making the compression process about the same. However the need to preload even large files has made the programs as a whole generally slower to use.

7zip and Hamster have the benefit of being free. Hamster on the other hand, despite having a name Im not fond of, seems to be fairly well optimized and speedy. Sadly it also preloads the contents to view an archive which can make it slow, although not as slow as Winzip.  7zip has its own special branch of zip compression which is often much smaller, but is quite slow at compressing. Conversely it doesn't prescan contents, making viewing an archive a handy and speedy process.

Winzip is arguably the heavy guns. They are the longtime standard, and quite pricey. So they'll have to perform above the rest to make continued investment worthwhile. As stated they already stumble with their long load times (which they didn't used to have in older versions btw).

So how did they do? Compressing a 5.4GB virtual disk In order of file size:
(done with a core I7 Ivy, on an Intel nvme SSD)

• 7zip 17.01 7z ultra settings - 3,246,744 KB time 5:01
• winzip zipx (option named 'best method smallest size') - 3,432,207 KB time 11:01
• 7zip 7z slow but often quite small - 3,433,919 KB time 10:22 
• 7zip 17.01 7z - 3,454,266 KB time 4:01 
• Winrar - 3,643,304 KB time 3:08
• 7zip zip format - 4,056,192 KB time 4:12
• 7zip 17.01 zip format - 4,056,192 KB time 4:21
• winzip zip format (option named 'legacy compression') - 4,077,092 KB time 0:55
• Hamster full ('*** compression' in their terms) -4,081,097 KB time 1:01
• Hamster normal ('** compression' in their terms) - 4,085,040 KB time 0:49

The above was using the automatic and easy to change settings of each program. Much more in depth changes are possible with 7zip.
-Changing 7zip 17.01 to use 'large memory pages' reduced the 7z time to 3:59 for identical file size.

Speed wise 7zip 9.2 was slower than all the others with both types. The others varied per how much work they did, though hamster came out last in absolute size, its 2 star compression does most of the work and faster than anything else. Modern 7zip has come along nicely for speed.

I had 2 versions of 7zip handy. Version 9.2 is in the range of 8 years old now vs 17.1 that came out this year. The results between versions were not what I would have guessed. They've gotten bigger, and in the case of normal zip format, slower as well. However the 7z version has gotten massively faster, apparently by making use of all available cores, which they don't do in the normal zip algorithm or the older version.

Given the file sizes and times I think Ive got to hand it to the free programs. This affirmed my normal use which is Hamster for day to day compression, with 7zip for those special big jobs. ..and winrar, well its just nice to have around. That is until I put 7zip 17.01 in the mix. Now with the additional speed, file size, and customization options I may at last be able to cut down to 1 archiving program. ...its been a long time coming.

Copy / Paste broken in win10 ... a venerable problem.

So here I was working away on things as normal when I go to paste a link into my browser. No go.

Again and again no text showing up anywhere. At first I thought it was limited to the browser, then discover its computer wide. No pasting into notepad, a command prompt, anywhere.

Looking around the web I find that its not an uncommon error. The few times its mentioned in official blogs there are normally dozens of comments from people looking for solutions, with little offered.

Heres a page that lists most of the common solutions in various posts.
https://social.technet.microsoft.com/Forums/windows/en-US/5bbc11e8-ca2d-41ac-b640-d66ce971f58f/copy-paste-clipboard-issues-not-working?forum=w7itprogeneral

Of course.. for me none of that worked. I imagine, since you're still reading it hasn't worked for you either. I did finally come across an obscure programming tool that got me through without having to reboot. That was the goal.
Look for:
getopenclipboardwindow.zip

I got mine from here:
http://windowsxp.mvps.org/temp/GetOpenClipboardWindow.zip

Of course you should try to make sure that your source is as reputable as possible. Try to avoid the links from pastebin and others unless you have some reason to trust the original poster. No need to sneak in a whole new problem while we solve this one.

Run the program therein. What does it do? It may show an error initially, but then it will come up with a window showing you the 'pid' of the process that is the problem. P.I.D. is the 'process id' number, a unique number that every process running on your computer gets.

If this solution is going to help the program will start a window with this line:
"PID of process with clipboard opened is"  Followed by a number. That number is what you want.

Apparently the issue can be that a process (what most of us call a program) keeps the clipboard open and locks it. From there nothing else on the computer can make use of it. In my case it was soffice.exe from libreoffice that was the culprit. In my case I had closed the program days before, but it had never actually gone away.

What do you do with a pid? Well it differs with different operating systems. This is for win 10, where I experienced the problem, but similar solutions can be found for all versions.

Right click on the task bar and hit 'Task Manager' to start that up. Go to the 'Details' tab, and from there hit the top of the 'Pid' column. That sorts all the processes that you have running by the process id. Scroll down to the id number that you found above.

Highlight that process in the list. Then hit the 'End Task' button. That will kill off the offending application. In my case this ended the problem. It may well come back if that program is run again, so be aware, and decide if you want to keep that particular program.

converting pronto codes to cde files for the Nirvis Slink-e

Hello all.. been a long time. :)

Background

Theres a bit of vintage gear that I use that takes some work to get anything out of. Its called a Nirvis Slink-e. No longer made, as the original maker got bought out by M$ years ago. It was the only piece of hardware able to do some nifty things with minidisc (what I usually use it for) and home automation (what most people used it for).

Nicely the original maker has left their website up for reference here:
You'll note the extremely basic HTML on the page. Kinda adds to the vintage feel I guess. :\
http://www.nirvis.com/

Problem is, the format of the files this thing used on a computer were always a bit of witchcraft to me. I got one to work a little once through experimentation, but never reliably. Why? I didn't take the time to learn the format and actually know what I was doing. Well I finally did it right.






This project is about understanding infrared remote signals, and how to convert them between a few standard formats. The most commonly available format as given on Remotecentral.com are known as Pronto codes. Pronto being a brand of programmable touchscreen remotes released by Philips, before they pretty much completely gave up on home electronics and automation.

There are information pages there on understanding the pronto file format I read many times. It never seemed to help that much since the general format is explained pretty well, but how to take that and move it into the cde file I was wanting to work with for the slink-e was never explained. So there I sat with a bit of understanding, but few working files.

This weekend I decided to take it the rest of the way and provide help to those that wanted it. ..and I succeeded.

Here Ill provide an example code for a technics sa-ax7 receiver taken from remote central in pronto, and exactly how its represented in my cde file to successfully send out a working code through the slink-e. My hope is that with this note and the standard tutorial below you will be able to convert your own codes.
http://www.remotecentral.com/features/irdisp1.htm

What can be done with all this information? Well, once understood you can send infrared remote codes out from a computer, through the slink-e, and control any electronics you want. I've wanted to do this for ages, since the remote to my receiver died, and replacing it is in the range of $50. That is far too much to be paying on an amp I've had since forever. Yet it requires a remote control to use some functions, like setting up the surround sound. I've been stuck, until now.

The code

If you cut and paste into a cde file from between the separation lines you should have 1 working command. Read the comments (words after the pound signs) to understand the meaning, or where the numbers come from.

In the case of that big old pronto code, just put the # signs at the start of a line to make it work as a comment. I wont bother trying to format it better with all the changes that can happen during posting.
--------------------------------------------------------------
desc=technics saax7 and cd eur646496
type=ir
carrier=40000
sleep=-300000
zero=400 -400
one=400 -1200
start=3200 -1600
stop=400 -50000
repeat=1

#first 2 sets of 12 numbers indicate the device in this
#case a technics saax7 receiver or a slmc7 cd player
#2nd 2 mark the command, other tutorials break that
#down further but this works for id.

##an example pronto code for the saax7-vcr1
#0000 0070 0000 0064 0080 0040 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 #0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 #0010 0010 0010 0010 0010 0030 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 #0010 0010 0010 0010 0010 0010 0010 0030 0010 0030 0010 0030 0010 0030 0010 0010 0010 0010 0010 0030 0010 0010 #0010 0030 0010 0030 0010 0030 0010 0030 0010 0030 0010 0010 0010 0010 0010 0ab3 0080 0040 0010 0010 0010 0030 #0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 #0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0010 0010 0030 #0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0030 #0010 0030 0010 0030 0010 0010 0010 0010 0010 0030 0010 0010 0010 0030 0010 0030 0010 0030 0010 0030 0010 0030 #0010 0010 0010 0010 0010 09ec
##when breaking this down note the repeats and the sections as such
#0000 0070 - learned code, carrier frequency
#0000 0064 - one time burst lenght (none), length of information
#section (64 bits)
#above goes into the header along with the stop start codes, though
#they are taken in multiples of each other
#the 400 comes from 1/frequency in ms.
#0080 indicates 8 400ms pulses or 3200
#0040 indicates 4 400 ms pulses or 1600
#
#
#in this case the info section has a repeat so it includes
#24 bits * 2, plus 16 for the additional start stop codes
#0080 0040 - start code (8 bites)
#take all the 0010 0010, and 0010 0030 codes and interpret them
#as 0 and 1 respectively
#0010 0ab3 - end code, indicates the end of copy 1 (8 bits)
#0080 0040 - start code (8 bits)
#0010... copy2 section
#0010 09ec - final end code (8 bits finally makes 64 when
#added to info sections and other 8 bit sets)
#note how the end codes differ. In IR 10% difference can be
#seen as noise. So these could be made the same.
#white space appears to be ignored in slinke files
#numbers divided below as I found the number sequence..
#remember binary is written
#'backwards' here so least sig bit is to the left.

010000000000 010000000101 000000000111 100101111100:saax7-vcr1
--------------------------------------------------------------

Byte: G+ gonna take your kidz

I wrote this in response to an article on wired here:
http://www.wired.com/epicenter/2012/01/google-facebook-teens-safety/

Google+ is late to the party, and they're taking it slowly. Doesnt mean they are worse off, FB was pretty slow starting off in the beginning, and pitiful compared to myspace. When Myspace kept shooting its functionality in the foot, only then did FB start to pick up steam.

G+ needs to find something that makes it atractive to the young people before they will flock to it. Still won't be as fast as FBs rise because at the moment they aren't self destructing as fast. FB may have had privacy issues, and ad issues, and hacking issues, but none of that messes with their core functionality, people canstill get on there and do what they like doing.

As much as the nerd in all of us wants to proclaim the g+ feature-set superior, it doesnt really matter because its not the nerd in all of us that they need to appeal to in order to get a full share of the mass market. They need something 'neat', music, or cool decorations on your page, or something that every little teeny bopper will love as the new shiny thing. . ...not just something technically superior.

They may still win out, but its just blatant guessing to predict it right now.

Kodak and the US

Why is Kodak in trouble? Because they're not Kodak anymore. Really that
says a lot about the US in general as well.

As noted in my last post I'm doing a lot of scanning of old photos right now, mostly from the 50s to the 70s for the todays crop. So of course that means a lot of exposure (haha) to Kodak slides and all the promo material surrounding it.

Of course in the business news the last few days are headlines of Kodak's bankruptcy. They want to sell and sue their way out of their financial problems, but that's no long term strategy, that's desperately grasping at straws.

Dealing with old Kodak products you really get a feeling for what that company was. It cared about a good product. ..no that's not even true, they cared about having the absolute best product, always, every time even as a sprawling international company.


Back in the days when you payed for long distance calls, paper mail was the norm, and it took a week to get a letter to Europe they maintained an epic network of the best products. The pictures I'm dealing with were were supplied, taken, and developed from Europe, to Texas, Kansas, and Alaska. Yet they maintain the same quality from place to place.

We can do that nowadays no problem, but I was just thinking that was quite a feat back then. It must have been incredibly difficult to just get a memo out talking about the new graphic to put on the slides. ..just that kind simple problem that they innovated solutions for back then.

What made them really the best though, wasn't consistency. It was quality, they had it, and they invested in it, and they promoted it.

Reading posts from archivist nowadays the old timers praise endlessly their 30, 40, sometimes 50 year Kodachrome slides and prints for not fading. At the same time they show a stream of competitors products losing all the information they were entrusted with. This didn't happen by luck either, there were many formulas used even by Kodak alone, but they invested in the research to know what would last.

Still we didn't know that for sure until today. So you couldn't base your success on that alone. In fact that kind of focus on the long term would be unheard of nowadays. You don't get something built to last unless you pay an especially high premium for archival quality. Pay any less and you're assumed to be losing your data and merchandise in about a 5 year period. Though you will most likely get lucky and outpace that by a fair amount.

So if even long term quality wasn't the secret what was it? Detail I think, craftsman level detail, even on a mass produced, heavily franchised product. No matter where you bought a Kodak product you knew you were getting the best available, and you were happy to pay a little bit more for it. Note I say a little, because even though it was the best it was still priced at a usable point, not out of reach even for the everyday user.

I notice things like a special little note put into the boxes announcing how every slide would now be dated. Even such a small improvement in the process is brought to your attention to appreciate, when it could have been ignored just as easily. At the same time this wasn't just marketing hype, showing you an obvious feature in the hopes you wouldn't notice its flaws, because again it was Kodak and assumed to be among the best.

Currently when you go to purchase Kodak you don't get anything near that. You get a jumble of digital cameras mixed in with the rest. To be fair they have decent if not ground breaking features. Sadly they have lackluster software that is decent, but nothing more then that. You may look around and see rebadged accessories like SD cards, nothing special other then the packaging sealed with the otherwise generic item in China was printed to go with your new camera.

Why would you buy that? ..and of course no one does. The thing is they, and we as the US need to start making our own products, and caring whether or not they suck. ..then amazingly people will start coming back to our brands.

Unfortunately its a bit of work.. like communicating that graphic was in the 50s. We can do it though, its just a matter of whether we want to. Whether we want to continue renting our country from the Chinese, or would rather start affording our lifestyle.

HP Scanjet 4670 - the beauty and the.. well..

The HP Scanjet 4670

Starting off my 2nd post with my first piece of antiquated tech. Its a pretty scanner actually, and you don't hear that in a sentence too often.

It's fast, quiet, and doesn't take up much space on the desktop. They could have gone a long way with this one, unfortunately its a dead series. :(

They stopped making this scanner back in the days of XP. Which really isn't all that far off in the real world. In compuland however thats ages ago! Worse then even that however, they stopped making drivers for it before Vista even came out. Apparently they teased some people with the idea of drivers the first few years Vista was out in the wild, but then cancelled the whole thing.

This is really one of the reasons that Corporations suck! They dont have to, and they don't alway, but quite often, its just a pile of suckage after the glitzy ad campaign wears off. They could have a legacy department that slowly works on updating drivers for things, supports older epuipment, puts up documentation for when they finally figure its no longer worth it. All things that would be cheap for them. ..maybe even make the drivers part an unofficial thing with rewards to their coders for legacy drivers. ..or even opensource the old ones when they are clearly no longer interested in taking care of the brand they nurtured all they way into our pocket books.

I'm not that mad about it really. I bought this particular one used. Though I do remember being impressed when I saw the thing new in at Fry's. Never saw one again until this one popped up on Craigslist for $25.

Its great because it doesn't have any lid. The holder doubles as a lid if you want it to, but you can lift it out of the cradle and put the glass up against anything you want to scan! This is very cool, and I've found it to be handy a number of times, especially when trying to get to certain parts of larger books or magazines.

The reason I'm so deeply involved with it right now is that I'm working on scanning in the piles of slides that my family have generated over the eons. I'd dearly love to afford one of those dedicated slide scanners with the "digital ice" technology and all that magic. As you can probably guess from the $25 price point though a $1500 dedicated film scanner just isnt in the budget right now. So I make do.

Unfortunately that brings me to the weak part of this scanner, driver quality. :( HP does have a generic driver that will run this scanner in Vista and win 7, "yay"! But it won't run the "transparent materials adapter", the part that scans film and slides. "Boo!".One awesome thing I've discovered though is Vuescan. One of the top few sets of scanning software out there. Why is it so awesome you might ask? ..because the developer has a similar sense for older tech to myself. Vuescan supports just about every scanner ever made, out of date ones, even SCSI ones.

So I can now run my scanner, and even the TMA adapter that I need for the slides. (So many slides!) So this is where the speed comes in handy.

Unfortunately theres still a problem.. the quality of the scans for these slides is somewhat questionable. Its clearly a software issue, but one that will never be resolved since HP gave up the ghost on this one. Any very dark slide will come up with vertical banding across the entire length of the picture. :( This is very noticeable in any dark picture that you have to lighten up to see the subject matter.

Good enough for my current needs, and Ill be working long hours into the night with photoshop repairing the damage to some of these. But with just a little tweak to the firmware HP could have made me a believer in their products. Maybe even enough to go out and by some more of them. Come at the tech world the way I do however it gives me a good scope to see just how well a company is going to treat a customer after they buy that shiny bauble at the store. ..and in this case it screams, "Not very well."

Good things Ive discovered - Vuescan
Bad Things Ive learned - HP doesnt like supporting legacy products, the HP scanjet 4670 had real potential but will just make most people angry.
Things broken for this article - none. yay!