I've poked around and can't find a way to produce higher resolution printed cards.

I'll assume that it is a bandwidth issue - higher resolution = more cost. If so, a suggestion: Is there any chance that either
A. as part of higher levels of membership a higher resolution card is generated (incoming cards, that is?)
B. that eQSL.cc might pull images from, say, flikr, to generate cards without additional bandwidth?

I love the service, please don't get me wrong. The only downside seems to be that the cards are printed from jpg/gif images instead of higher resolution formats that look better and look more like, well, real QSLs. I just feel like...for the sake of a little bandwidth, this could be improved.

73 and thanks for the awesome service,
Jay KA3JLW

---

KA3JLW Jay Horman

That would be great Jay. I don't know if it's the bandwidth so much as the storage capacity for everyone's eQSL design. An 64kb or even an 584kb file takes up alot less storage than say a 1.4mb jpeg or 4mb tiff file. Maybe with Gold, you could upload a large file, but what about all the others (AG and not AG) that are not gold? They would still only have small file eQSLs in storage. Maybe the Admins of this site have some ideas?
73!

---

N1ORK Orest 'Andy' Zajac

I don't know how many eqsl users there are, but let's assume 1 million for the sake of ease.

And let's assume they all upload a 300 dpi image (jpg - no need to use TIFF format at 300 dpi, and that's plenty for a photo quality print). For a standard QSL size of 3.5 x 5.5 inches, that's a jpg of 1050 x 1650 pixels and will work out to a file size around 500 kB at 90% quality... but let's call it 1 MB for the sake of a round number.

AND, just to make it really cool, the system can generate a small version for dial-up folks at 96 dpi (the current size) which will add another 100 kB to the size.

1 million users times 1.1 MB per user = 1100 GB of data.

With current technology and costs, using eight 500 GB drives for RAID reliability and speed with approximately double estimated capacity for future growth, that will cost $650 CDN to procure the equipment:

Example (there are others, this was just easy - no endorsement is implied): http://www.tigerdirect.ca/applications/SearchTools/item-details.asp?EdpNo=3167558&CatId=525

So let's not say how much it will cost to store all this data. The cost to store the data is essentially nothing. RAID controllers are on-board almost every modern computer purchased in the last 5 years, so there's no cost associated even with that... but let's assume that's an issue too and throw in another $600 for a new motherboard, CPU, and power supply. And multiply it by three so you have a load-balanced database server farm. Still well under $4000 for a load-balanced, reliable, database server farm with 2000 GB of RAID storage for 1 million eQSL users with photo quality QSLs stored.

As far as I can tell, the eQSL application does not store a separate picture for each eQSL you receive or send, it merely references the picture the user has uploaded and applies the QSO data when you ask to see it, so the storage requirements for the pictures are quite modest, even if the pictures are large.

If 50000 eQSLs are retrieved per day, and they're all the big ones, that works out to bandwidth of 50 GB per day - about the same as downloading 10 movies per day. At 40% network load, that's about 15 Mbps of upload required from eQSL to the Internet - a healthy internet connection to be sure. Likely if there's a cost problem, this is where it might be, although Verizon (selected at random, I am not endorsing this company) seems to offer a business service in this range for about $250 US per month, so presumably this is available from numerous vendors in a similar price range.

I also don't think there are 1 million eQSL users.

While I agree that 10 years ago, storage and bandwidth would have been a huge issue, I think time and technology has rendered that argument obsolete.

bleah, a yellow sac spider just crawled on me... now I have the willies. ugh.

---

VE3OIJ P. Darin Cowan
Edited by VE3OIJ P. Darin Cowan on May 7, 2008 at 03:56 AM

Darin,
I understand all your numbers and agree with your reasoning. I'm just wondering why the Administrators of this site don't go forward with this idea? I know that they just upgraded the servers, so maybe they are progressing in this direction. Don't know what the total donations are per month, but with the figures you bring up for both storage and bandwidth, I know I couldn't afford it.
Hope you got that spider!
73!

---

N1ORK Orest 'Andy' Zajac

I know they just did server upgrades, that's why I don't think it would require new servers to augment the space, and hard drives are REALLY REALLY inexpensive.

The real question is how much bandwidth costs compared to incoming revenue. I *HOPE* that there's enough interest in eQSL to generate the few hundred dollars a month it must cost to run this... I'd hate to think that the admins are out-of-pocket on this stuff. They don't sell advertising, so they have to survive on donations.

If the admins are using out-of-pocket money to drive eQSL, eQSL is in trouble. One illness or unexpected expense and the whole system could be hosed. I've never had the impression that things are that precarious. A quick look down the donors list reveals over $80,000 in donations just from the gold members. That's $40,000 in servers and software, and enough money to pay $400 a month in internet for 8 years, so money is unlikely the problem. eQSL is remarkably transparent with it's funding - a real breath of fresh air for an organization that takes donations. Kudos to the admins for that.

I suspect it has more to do with just time and desire. It was mentioned in a previous thread that big pictures are a hassle for people on dial-up. That's true, although it could be overcome by allowing people with slow connections to set that as a profile option and receive small pictures. That would mean changing the eQSL application which could be a hassle (and cost money) if that work has to be contracted out. I would like to think that it isn't in eQSL's plan to march to the beat of the least technologically advanced radio amateur in any case, but the download for dial-ups thing is a pretty valid concern that would have to be addressed in any implementation of photo-quality QSL pictures.

---

VE3OIJ P. Darin Cowan

Thanks for the info and insight Darin. This is a great site and I hope it continues for a long time. Maybe the High-res eQSLs will come some day, but for now, I think the Admins have done a great job and perform a great service. It's too bad the LoTW folks can't come on board. Can you imagine if they combined and used TQSL for verification of QSOs and eQSL for generation of electronic QSL cards? All award systems could merge (eQSL, ARRL, CQ and others) into a global award system. But I guess I'm just a dreamer............
73 Darin and thanks again for your info!

---

N1ORK Orest 'Andy' Zajac

If you haven't guessed, implementing projects like this is a big part of what I do for a living, although I am not involved with eQSL and am only providing my opinion as a user.

I think the admins do a great job as well, and I can live without the better quality qsl pics, but I do hope that improving the print quality is somethign that is on the radar.

---

VE3OIJ P. Darin Cowan

Thanks for all the replies - way back when :)

Some other ideas on how to do this:

1. Have the pictures stored elsewhere (like flickr) and find a way to "marry" the picture upload with the QSO data. It is done now, the only difference is the file will pull from a different source. That may solve any bandwidth issues but probably creates new ones.

2. Have a throttle for those with dial-up. One dial up ISP I had provided a browser plug-in that would reduce picture file sizes by reducing the resolution on the fly. Maybe the site serves those with higher bandwidth and then provides a link or download to that kind of throttle software for those with dial-up. But then again, I don't know how many users are on dial-up and we certainly want to welcome/encourage "3rd world" participants who may not have access to broadband.

3. Provide the card in relatively lo-rez as it is now, with a "click here" to get it in high-rez, and maybe the "click here" is only active for silver/gold members. (can you guess I'm silver?) This may help drive sales.

4. Bigger picture idea, I posted this on QRZ.com: Paper QSL printers must feel a little threatened by LotW and eQSL. So how about joining with them a bit? Here's how - offer these options:
- Have UX5UO design your card: $25
- Have UX5UO host your card to provide others with a higher quality QSL: $25/yr
That way the QSL maker becomes a partner with eQSL, people can use a 'unified' design with their paper QSLs, and higher bandwidth / resolution cards are offered.

As earlier in the thread, I LOVE the idea of LoTW officially connnecting. And I love unicorns and hitting the lottery... :)

---

KA3JLW Jay Horman