Mobile devices today are coming increasingly powerful but constrains in computing power and battery capacity make it prohibitive to convert in real time media information upon downloading or streaming on a mobile client. With limited codec support in most smartphone devices with the exception of Asian PMP devices catering to torrent-loving media consumers, Internet downloaded media in DivX, MKV, or Blu-Ray MPEG-4 have to be converted to H.264 prior loading to device or streaming. Transcoding is a time-consuming and often manual activity that needs to be eliminated and automated. Vuze/Azureus have changed that with automatic transcoding to iPhone including import in iTunes, PSP support with thumbnail support, probably soon to support Android devices as well. Yet, Vuze or similar solutions need to be accessed on the transcoding station (powerful Desktop PC) and synced with the device. On the go, often the user may find out he wants to watch a video that he has at home, but it is not transcoded yet. Carrying all transcoded media on the device sometimes is not viable – user may have terabytes of videos at home, which convert to hundreds of gigabytes of transcoded data. He may not want to have transcoded copy to every media file. Finally, not every device requires the same transcoded copy – user may have a HTC Tattoo budget smartphone and an Archos5 PMP. While codec support is the same, resolutions are totally different and bitstream support differs as well or the user may not require same quality for each device he has. The following combination of server and mobile client solves this problem.
- Remote-control 24/7 Server PC/Mac/Linux box that serves as torrenting client as well as transcoding server. Every downloaded movie or TV show is compared using RegEx with Scraper information using similar methods to XBMC to gather full movie or TV show information and episode list. This builds Web Server which the Video Player can access. The user, via special Video Player with Web interface access integrated in it, can list all the movies and TV Shows on his home computer in a rich User Experience full of Fan Art, DVD Box covers, screen thumbnails and more, and can check which one is watched or not. While Web Based, the whole HTML/image pack of the remote server is cached in the background to allow the user to access a snapshot of the server’s offerings even if offline. The server also should support merging subtitles into videofiles via FFMPEG, and also must receive additional options for video/audio/subtitle track choice for MKV packagers (i.e. you can request transcoding for an anime series, by choosing default or specifying exact audio (Eng/Jap) and subtitles (Yes/No/which ones).
- Android OS Video Player. No need for fancy codec support as we will be streaming or downloading media with transcoding on demand. The user chooses what Movies or TV Shows he wants to watch. The Video Player demands these shows from the remote server, informing him about the specifics of the device – in our case, it’s an Archos PMP, so the server knows that he has to serve 800×480 MPEG-4 to the device using specific kbps setting (Archos has 500GB hard drive so it can take 1.5Mbps, but if it was a HTC Tattoo it would be 384 kbps video). The remote server starts transcoding all requested videos. It sends an estimated transcoding time to the Video Player. Using Interface bar or Home Screen Widget, the user may see how much time is required before the videos are completely transcoded. Once the videos are transcoded, or if they were transcoded previously and transcoded copy for that device exists already, it it sent to the device via background downloaded service that supports resuming if the user drops connection from time to time due to mobility. Once the video is uploaded to the device, the user may start watching it. Or, if streaming is possible, he can start watching it once certain buffering is made possible and downstream can support buffering – the user is informed in real time whether streaming will go perfectly or there may be hiccups.
- The Video Player may also support additional automation such as: (1) deleting transcoded cached and/or original copy files on local mobile device and/or remote server once the video is being completely watched; (2) merging video player UI with data from various torrent client web clients, such as mTorrent, Vuze, Transmission – see while watching what is the remote up/down ratio for the video so you know whether it is safe to delete both as data and as torrent from used torrent client.
What do you guys think is the most intuitive way to post to social networks our condition, thoughts and memories from the day without having to spend more than couple seconds in expressing it; or how can we remember to post about things that we do not have time or condition to write at the exact real time moment it happens?
One way to post to social network is through predefined messages. We could design an UI that detects series of swipes and taps in gestures, each gesture triggering a predefined message for sending. We could even chain several gestures to connect several short sentences or portions of a sentence. If the sentence has ad-lib locations, Voice-To-Text could fill in the blanks. Predefined Twitter messages can have variants based on Locations. (i.e. “I am still at work” may be “I am at home” if at home location) Location is based on GPS, CellID, WiFi ID Presence.
- John is running TwitterTap app. TwitterTap has a black screen interface or a Home Screen replacement which mnimics turned off phone so his boss doesn’t know he’s actually running a Twitter app. The phone screen acts as a touch sensitive Twitter/Facebook interface for entering predefined Twitter messages.
- John is at work, he was told a great joke. This made him happy.
- John taps three times on the screen. This means “I am still at work” from the predefined list of phrases. John swipes twice down up on the phone screen. Two up swipes means Positive Rank 2, which means “I am happy, because $1!”.
- The screen shows a minimalistic icon of Microphone or keyboard. Johh can add $1 = “I heard a joke” either via microphone icon voice-to-text or keyboard. Adding all ad-lib blanks shows a small, minimalistic message and approval button, or directly sends the message to all added services.
- Sometimes we don’t want to send immediately, but to remind ourselves about something that happened today when later we have time to message it. John’s taps and sweeps are enough to set a draft format that he can later at home resume (in sequence if multiple drafts) with missing information. For example, John has no time to sweet the joke today, so all he does is press a draft button, tap few letters if he wants on a virtual keyboard – “jk” for a joke – and save it as a draft. At home or while walking back to home, he pulls his phone and starts sending all the Twits. Each message comes in timeline sequence, offering him to add blanks, rewrite or remove. Each pre-defined message may have current and past mode. If John sends the message later, “I am still at work” becomes “While I was at work”, “I am happy, because $1!” becomes “I LOLed a lot, because $1!”. He decides to use voice-to-text to speak “I heard a killer joke, i’ll tweet it later” and the message is now complete. Message is auto-shortened by removing vowels, switching words for sms shortspeak etc. if longer than 140 characters, then it’s sent.
Paranoid fits aside, Google is demonstrating incredible technology that will transform Google Earth from geolocation database to a virtual world and indispensable news-monitoring tool. Just imagine that in couple years, reading news will be nothing compared to zooming and staring at little manequins crowded at important events. Such as zooming over union strikes will show you how people on the strike move along the streets, accompanied with real-time news-flow from news sites and the mandatory censoring and manipulation of data by switching to simulated peaceful mode when the police squads are sent tasering people in sight.
If you had this technology today provided via Google Maps API, how would you use it?
The Mandatory Google Wave thread, so we all enjoy the real-time Web innovation. If anyone gets an invite or has invites to spare, please share in this thread. My email is…
…and I still haven’t got one.
Microsoft has something steaming hot at his hands, something to make even the most fervent Apple-cart fanbois cringe in envy. I think we’re all aware of Microsoft Courier dual-screen tablet.
Why I believe Microsoft Courier can succeed?
- Self-protective design. Similar to Nintendo DS/DSi, both screens protect each other in a closed state. Apple Tablet does not promise any protective design, making caring for 7-10 inch screen an issue to consider.
- LCD Tablet for the masses. Wacom Cintiq plays Adobe’s game and holds a high price tag of $999 barring amateurs and enthusiasts from access to LCD drawing. Even without pressure sense of Wacom tablets, Courier could prove to be the best way to learn people to draw using reference material, positioning art into panels for comics or storyboards.
- Usability-minded product focus with real-life applications. Apple Tablet’s expected marketing will be too much focused on multimedia and gaming. Rumored support for iPhone apps will provide initial productivity boost but iPhone apps are known to provide constrained or pseudo-usability. It will take time, or might prove impossible, to provide the level of integration that Courier could provide out of the box.
Why I believe Microsoft Courier can fail?
- Exorbitant price. Given that demonstrations has shown Courier being targetted at designers, engineers, and such. first generation hardware can make this device unreachable for the general population at prices over $800. EeeReader from ASUS may be the alternative, stealing the basic design and cheap-ifying it.
- Closed Platform. Without SDK and third-party support, tossing information in a spiffy OneNote or drawing with MS Paint inspired Corel Painter wannabe can get old pretty fast. The worst thing that could happen to Courier is to end up a closed platform, or Apple-like cripped hardware held back by business policy not to make it too open to developers. Once again, Microsoft will have one chance to play fair ot lose the game to the taiwanese companies. ASUS would love if M$ plays El Jobso with Microosft Courier, as it’s a chance for EeeReader.
What are you going to use a Microsoft Courier in your life?
Gizmodo can be a real spirit-drowner sometimes when they tell you Palm Pre business is bound to drag worse than dealing with App Store. I shudden remembering my business operation with Apple and App Store, so if Palm is supposed to be worse, I am that close to giving up. When you hear of 27 weeks in review limbo, you know it’s time to get all those printed app ideas and shred them to tiny bits.
On the brighter side, Android Open Alliance is a group of Droid coders who believe Google Experience apps are not made of heavenly light and shouldn’t put tears in Cyanogen’s eyes, as they can be exchanged with fully open-source set of apps that provide the same usability. So far they are looking for complete solutions to prepackage and build up some cheap short-term hype, but I am hoping they listen to my issue-writing wisdom and develop everything from scratch in order to implement more advanced, more capable and plugin-friendly solutions that support even more services and features out of the core.
Engadget have posted today the announcement of Lighthouse SQ7 by AdelaVoice – Internet Tablet intended for posting on Twitter and Facebook via voice recognition. The tablet has bare minimum of features, does not use any worth noting operating system, has almost no multimedia capabilities and is limited to Twitter, Facebook widgets, Image gallery, and Webkit based browsing. Yet, it demonstrates near-perfect Voice-casting to Twitter and Facebook. Impressive.
Alas, the device requires Push-to-Talk feature that needs manual reach to the device. This makes it hard to use anywhere, such as lying on the bed, standing across the room, or being anywhere you don’t have the device with you. If I were to design such a product, I would have done it a screenless computer worn around the neck, around the wrist. The device would then communicate with just about any screen-based device such as computers, mobile phones, LCD picture frames for text confirmation, or would allow direct no-confirmation posting, or screenless confirmation by text-to-voice repeating and voice command agreement to post as text, agreement to post as audio attachment (via hosting audio file) or canceling.