i payed for above offer,
but find < 512 MBram10 GBdisk 512 GBtraffic> in cp
test only?
id=718905
~~~
Offer #98
SKU: RYZEN SKU# V00-D3
1 cores
384MB RAM
15GB NVMe Storage
1 IPv4
1792GB Bandwidth
LOS ANGELES, CA
$7.60 per YEAR
All orders that bugged out should be logged. I'll be going through and catching up on fixing them now. After I'm done I'll make another post with more information and then we can go through any problems still left over with the provisioning. Give me about an hour or so and I'll provide another update.
There's one order stuck on provisioning which I'm investigating further now, and two orders stuck on vserverid not found.
I'm going to be attempting to rework the provisioning system from scratch again, if that doesn't work I'll revert back to the old version. By old version I mean about a week or two older. That version was tested thoroughly for the 512MB/10GB bug, but has some bugs of its own that would slow down the provisioning process (it doesn't use the new logic of clustering provisioning.) I'll let you guys know what version we end up on after I'm done. But for now we're still on the current version that has a high chance of provisioning as 512MB/10GB which will be automatically logged and corrected (no need to report that problem here) and a small chance of having the vserverid bug as well. I'll try to reintroduce the vserverid bug catcher (since it wasn't happening frequently, it was removed, but we used to have logic that reverted the entire process if it caught a vserverid bug and retried. I disabled it since if it gets stuck and retried over and over and over it could potentially cause other problems.)
@tototo said:
Offer 96 is listed as $7.60 but it seems to be actually lower than that?
This is another bug. It's not a major bug, it's a bug with price reduction handling. It makes sure it gets the important part done correctly which is actually updating the price on the service that you order, but pays less attention to updating the displayed price. This isn't a big deal as long as prices are dropping, but if prices are going up it can be problematic. Basically, there's sometimes a delay on it updating what is displayed. Since that's not prioritized in the process. It usually fixes itself in a couple minutes. Let me know in this case if it doesn't catch itself soon and correct it. Actually I'm not sure what you're talking about specifically unless I'm looking at something wrong, but we're currently at offers 100 through 102, right? That might be another bug where your page needs to be reloaded. Let me know if that's the case as I'll have to look into that.
There used to be a bug that caused it to do that (stop updating) but I reworked the frontend to make that not a possibility so I'm wondering if it's from your backend notifications or how long you've had the page open.
Right now I'm seeing offer 100 listed as $8.28 but being actually $7.04.
I've also caught another bug, where the system completely errors out on the ordering page, related to the price change warning (most likely.) Yay, bugs! It goes away after a refresh.
@msatt said: @VirMach - Same bug, unfortunately Amsterdam Offer 82 not applied correctly to Invoice 1661531
Thanks,
Are the Amsterdam ones the only ones that are broken? I guess I need to wait and not try for another ams offer...
No, Amsterdam offers are just more popular, so more orders, so more reports. Amsterdam is filled differently than other locations so it's popping out offers people like more I think. They also have different multipliers per location so perhaps Amsterdam is giving out those type of offers (that match preferences.) Or maybe people just like that location.
We just got a big batch of new orders, going through those as well now to fix the above bugs for them. These orders were placed after my message above.
Provisioning bug is related to another WHMCS bug, where while paying the invoice/checking out, customer creates a second user account. We've never been able to figure this bug out, as usually we don't receive any feedback from customers that experience it. So if you're order #611199, or invoice #1661547, please reach out here so we can work out what could have caused that bug to occur.
This is for "MassivePrivate-VM"
I thought the name was funny so I added that in as well for clarification.
@VirMach Yeah everything you said is correct. I was late to post because I was doing some checks, but it’s certain that it wasn’t Offer #100-102. In other words, it wasn’t a bug where the offer page failed to update.
I rushed to place an order to check if the actual price being lower than the offer page price was a bug, but I should’ve waited for Offer #100 instead of jumping on #98 lol
Yes, just confirming there will be a good amount of those, after I add them into the pool. They'll be on the chunkier side, I think, because we're almost out of IP addresses in Miami.
@animeiscool said:
Update: Seems like offer #54 is going through the same thing. $11.52 on flash sale site, but cheaper price on the main website.
Update 2: Offer #53 is now $53.01 on the main site, compared to the $10.01 on the flash sale site.
Idk what exactly is happening, but I believe in VirMach.
This was a massive bug related to the system halting per the post above yours. I think it worked itself out. It was just displaying the older offer mixed with a newer offer. This shouldn't happen ever again, unless it happens again due to the reason the bug occurred which was me editing the code in a really bad way.
Speaking of which, I had to revert the code to fix it. The bug I was trying to fix which was not reported is that sometimes when plan gets sold out, it just gets stuck.
@mucstudio said:
It shows that it has been opened normally, but when I click in, I can't find the server!!!
You're one of the two vserverid bugs. I'm still working on that, sorry for the wait.
Not happy about this one. But at least it's only for two of them. But that will make it so much harder to catch.
@sh97 said:
Feeling the RAM is kinda imbalanced for the offers
I noticed this issue as well, I don't have a proper fix it for it other than making it reduce the other resources, which could be a fix since it would also reduce price. I was actually working on a better fix where in those situations it gives you the extra disk and whatever else but just charges less for it, and uses RAM as the primary price indicator.
I'll see what we can do about it though but the "problem" is basically that it's too accurate now. It just means we have Ryzen nodes that are limited on memory as opposed to Epyc or E5. And a lot of the crazy high disk offers in the past (from filling nodes that had a lot of disk before we switched over have expired, and a lot of the crazy low disk offers have been renewed at a higher rate, that's left nodes with kind of with a lot of disk space. And since they're Ryzen, and CPU usage is now low as a result, for offers mostly sold on E5's in the past... yeah.)
Oh really important regarding RAM: the higher RAM plans are currently less likely to show up because they'd be on nodes with a lot of memory left over, which happen to also be nodes that are locked off/have been locked off for some time. We went with the nodes being unlocked that are the most stable which happen to be the nodes with the highest amount of memory used since they're usually actively being refilled and just higher memory usage in general from being online for so long. So once these get entered back into the pool you'll start seeing your higher memory offers.
@tototo said: @VirMach Yeah everything you said is correct. I was late to post because I was doing some checks, but it’s certain that it wasn’t Offer #100-102. In other words, it wasn’t a bug where the offer page failed to update.
I rushed to place an order to check if the actual price being lower than the offer page price was a bug, but I should’ve waited for Offer #100 instead of jumping on #98 lol
Let me know if you notice any specific way this works out, as in let me know if it appears to be after a price drop, it's lower on the order page, and then corrects itself or if it stays broken the entire time. No rush, just whenever it's noticed if you have more information for me on it.
It's already on my (growing) list of things to fix but lower on the priority list for now.
After I'm done with fixing all the important bugs, I want some feedback on the pricing/resource generator as I'll likely want to rework that system to some degree. There's already a system that compares the plan to a few "model" plans. So I want your feedback on some of these "model" plans and how they should be, otherwise I'll use the order data to tune the existing "model" plans.
The "model" plans are plans we know are popular and how they should be and it uses them to compare and make some last minute adjustments to generated plans.
So basically let me know the type of plans you're looking for and would like to see. Give me the RAM, CPU, Disk, Data as well as the price range (like preferred price to maximum price) you'd be willing to pay for it and in which locations. I'll compile all these together and see if I can work it into a better modeling system.
On another note:
Sometimes I'll come across a plan and wish it was a bit more balanced in my opinion. Like if it had less bandwidth (personally I'd want at least 1-2 ish TB) and/or just one core compared to having two cores, and a bit higher ram like around 1-1.5 GB, instead of ranging from 700ish MB and below, but more realistically I wouldn't really want anything 500 MB and below. As for disk space, I don't see myself needing much of it. And maybe a price adjustment on the lower side , I could see myself getting some more flash sales. I don't know if there's any others who are in the same boat.
@VirMach said: let me know the type of plans you're looking for and would like to see. Give me the RAM, CPU, Disk, Data as well as the price range (like preferred price to maximum price) you'd be willing to pay for it and in which locations. I'll compile all these together and see if I can work it into a better modeling system.
Another bug, it seems like the TTS is more likely to pierce through this issue, but not the sound alert. If you have the tab in the background, then at some point in time I think the browser puts it to sleep and doesn't play the sound alert (even if you have TTS+sound enabled.) The current fix is to just open the tab back up and you should be good again for some amount of time. Maybe in practice this is a non-issue since it could just be a "me" problem with having like 300 tabs open.
There's 2 or 3 Epyc servers sitting behind me and one of them already deployed in Tokyo since January 2024, but just not being used.
Sometimes I'll come across a plan and wish it was a bit more balanced in my opinion. Like if it had less bandwidth (personally I'd want at least 1-2 ish TB) and/or just one core compared to having two cores, and a bit higher ram like around 1-1.5 GB, instead of ranging from 700ish MB and below, but more realistically I wouldn't really want anything 500 MB and below. As for disk space, I don't see myself needing much of it. And maybe a price adjustment on the lower side , I could see myself getting some more flash sales. I don't know if there's any others who are in the same boat.
There was a system I was working on that got scrapped which would have basically resolved your issue. I was then going to work on a system afterward where there's a flash sale within your service details page after you order, but obviously after we're done with the sale, it would only benefit people that bought a plan they didn't really want and they'll be able to finally tune it/customize it to what they want so that won't work out either.
The system was going to basically generate a range of specifications for each flash special to let you decrease some things to lower the price or increase some things and pay more. It was mostly aimed (initially) toward additional IP addresses and disk, IIRC.
I'll see if it's feasible to work this in, or try to come up with some other solution as that does seem to be a big problem (at least the part where you're getting resources that are just lopsided and you don't get to opt out and it gets calculated in the price.) One of the solutions would be not being charged as much for the amount that's way over a reasonable range, for example if there's a plan with 384MB RAM, 3 Cores, then we should try to weigh it down to where either the cores are reduced and/or we charge less per core for it. The system already does technically kind of keep this in mind a little but not to the degree I'd like for it to do.
One other idea that was explored at some point in time was having sub-plans for each plan. You'd click order now or another button, and you'd be presented with 3 plans for the one of 3 plans, for a total of 9 plans. One of them would be the plan as advertised. One would be a variant that's smaller, one would be a variant that's larger. That was decided against because while it could technically utilize the system as it currently is then we'd have more plans to track and at that point we might as well just present 9 different plans and the likelihood of one of them matching what you want is technically statistically the same since we do also randomize plans already (it decides whether to make it larger, smaller, and so on, it gets a dice roll.)
I guess the issue we need to fix is making sure that the plan that closely resembles what you want has the potential to ever exist.
@VirMach said: let me know the type of plans you're looking for and would like to see. Give me the RAM, CPU, Disk, Data as well as the price range (like preferred price to maximum price) you'd be willing to pay for it and in which locations. I'll compile all these together and see if I can work it into a better modeling system.
@animeiscool said: On a serious note: How much ram is usually installed on those EPYC servers?
I don't remember. We were somewhere around 384GB to 1024GB though. They have something like 8 to 16 slots and 32GB to 64GB per stick but could go all the way up to 256GB sticks (which we definitely didn't use.) I think the 16 slots were dual Epyc ones, we might have had one of those.
So either 384GB, 512GB, or 768GB most likely.
@animeiscool said: Imagine if there was an anime waifu body pillow chilling, next to the server rack. Or some special anime stickers on the Tokyo servers that are data center approved/server rack approved .That would be cool, but I don't think any data center would want that :P
The body pillow would definitely be a fire hazard.
Bug found, the bug where no welcome email was sent has been fixed.
New bug is that now up to 3 emails are sent. Well actually, it's even worse, it's where up to 3 services are generated. This might also be causing the vserverid bug.
IMPORTANT UPDATE: only use the service that is actively being displayed on the service details page/billing area. The others will be terminated (we'll likely suspend them gracefully first, but we're not going to honor any support requests where you decided to use all of them and then want to rotate out to some other generated one.)
Will be working on redoing the creation system completely.
CREATIONS ARE BEING PAUSED: The system will still keep track of which services are ordered and need to be created, but it won't be provisioned immediately. I'll provide an update on when it's reworked and unpaused.
Plan #103
$9.44 PER YEAR
512MB DDR4 RAM
1 CPU vCORE
30GB SSD (NVMe)
3456GB BANDWIDTH
1 DEDICATED IPv4
(As for the discount rate, I don't know since the plan is being switched out.)
As an example, I feel like this plan is close to being an ideal plan, at least in my mind. I don't what the minimum disk space is for these plans. Although, I wonder how much more RAM you could get if you balanced the plan by lowering the bandwidth to something like 2000 GB. However, we need to keep in mind how much can be balanced out, while still being feasible for VirMach. Anyone else got any feedback or thoughts?
@animeiscool said: As an example, I feel like this plan is close to being an ideal plan, at least in my mind. I don't what the minimum disk space is for these plans. Although, I wonder how much more RAM you could get if you balanced the plan by lowering the bandwidth to something like 2000 GB. However, we need to keep in mind how much can be balanced out, while still being feasible for VirMach. Anyone else got any feedback or thoughts?
I actually do not believe we charge for bandwidth in the current logic. It just decides how much bandwidth to give you based on all the other specifications, pricing, and location. So in that case, if I remember correctly (I'm not sure, we had so many different versions) then it would mean it would have zero effect on the price and other resource levels.
For 33%* off one-time. Remember, one-time means non-recurring. One-time discount. Not one time use for the coupon. And remember this means since it's 33% off, it equals the first 1/3rd~ of the service being rendered as free and therefore the first 4 months of service being 0% SLA, and no support per the disclaimers. (That means the first coupon was 1/4 or 3 months, second coupon was 1/2 or 6 months.)
Comments
Virmach is alive! I'll go check my servers after a year and see if they are idling peacefully.
Are the Amsterdam ones the only ones that are broken? I guess I need to wait and not try for another ams offer...
i payed for above offer,
but find < 512 MBram10 GBdisk 512 GBtraffic> in cp
test only?
id=718905
~~~
Offer #98
SKU: RYZEN SKU# V00-D3
1 cores
384MB RAM
15GB NVMe Storage
1 IPv4
1792GB Bandwidth
LOS ANGELES, CA
$7.60 per YEAR
All orders that bugged out should be logged. I'll be going through and catching up on fixing them now. After I'm done I'll make another post with more information and then we can go through any problems still left over with the provisioning. Give me about an hour or so and I'll provide another update.
Offer 96 is listed as $7.60 but it seems to be actually lower than that?
512MB/10GB disk issues have ben fixed.
There's one order stuck on provisioning which I'm investigating further now, and two orders stuck on vserverid not found.
I'm going to be attempting to rework the provisioning system from scratch again, if that doesn't work I'll revert back to the old version. By old version I mean about a week or two older. That version was tested thoroughly for the 512MB/10GB bug, but has some bugs of its own that would slow down the provisioning process (it doesn't use the new logic of clustering provisioning.) I'll let you guys know what version we end up on after I'm done. But for now we're still on the current version that has a high chance of provisioning as 512MB/10GB which will be automatically logged and corrected (no need to report that problem here) and a small chance of having the vserverid bug as well. I'll try to reintroduce the vserverid bug catcher (since it wasn't happening frequently, it was removed, but we used to have logic that reverted the entire process if it caught a vserverid bug and retried. I disabled it since if it gets stuck and retried over and over and over it could potentially cause other problems.)
This is another bug. It's not a major bug, it's a bug with price reduction handling. It makes sure it gets the important part done correctly which is actually updating the price on the service that you order, but pays less attention to updating the displayed price. This isn't a big deal as long as prices are dropping, but if prices are going up it can be problematic. Basically, there's sometimes a delay on it updating what is displayed. Since that's not prioritized in the process. It usually fixes itself in a couple minutes. Let me know in this case if it doesn't catch itself soon and correct it. Actually I'm not sure what you're talking about specifically unless I'm looking at something wrong, but we're currently at offers 100 through 102, right? That might be another bug where your page needs to be reloaded. Let me know if that's the case as I'll have to look into that.
There used to be a bug that caused it to do that (stop updating) but I reworked the frontend to make that not a possibility so I'm wondering if it's from your backend notifications or how long you've had the page open.
Right now I'm seeing offer 100 listed as $8.28 but being actually $7.04.
I've also caught another bug, where the system completely errors out on the ordering page, related to the price change warning (most likely.) Yay, bugs! It goes away after a refresh.
No, Amsterdam offers are just more popular, so more orders, so more reports. Amsterdam is filled differently than other locations so it's popping out offers people like more I think. They also have different multipliers per location so perhaps Amsterdam is giving out those type of offers (that match preferences.) Or maybe people just like that location.
We just got a big batch of new orders, going through those as well now to fix the above bugs for them. These orders were placed after my message above.
Provisioning bug is related to another WHMCS bug, where while paying the invoice/checking out, customer creates a second user account. We've never been able to figure this bug out, as usually we don't receive any feedback from customers that experience it. So if you're order #611199, or invoice #1661547, please reach out here so we can work out what could have caused that bug to occur.
This is for "MassivePrivate-VM"
I thought the name was funny so I added that in as well for clarification.
@VirMach Yeah everything you said is correct. I was late to post because I was doing some checks, but it’s certain that it wasn’t Offer #100-102. In other words, it wasn’t a bug where the offer page failed to update.
I rushed to place an order to check if the actual price being lower than the offer page price was a bug, but I should’ve waited for Offer #100 instead of jumping on #98 lol
Yes, just confirming there will be a good amount of those, after I add them into the pool. They'll be on the chunkier side, I think, because we're almost out of IP addresses in Miami.
This was a massive bug related to the system halting per the post above yours. I think it worked itself out. It was just displaying the older offer mixed with a newer offer. This shouldn't happen ever again, unless it happens again due to the reason the bug occurred which was me editing the code in a really bad way.
Speaking of which, I had to revert the code to fix it. The bug I was trying to fix which was not reported is that sometimes when plan gets sold out, it just gets stuck.
You're one of the two vserverid bugs. I'm still working on that, sorry for the wait.
Not happy about this one. But at least it's only for two of them. But that will make it so much harder to catch.
I noticed this issue as well, I don't have a proper fix it for it other than making it reduce the other resources, which could be a fix since it would also reduce price. I was actually working on a better fix where in those situations it gives you the extra disk and whatever else but just charges less for it, and uses RAM as the primary price indicator.
I'll see what we can do about it though but the "problem" is basically that it's too accurate now. It just means we have Ryzen nodes that are limited on memory as opposed to Epyc or E5. And a lot of the crazy high disk offers in the past (from filling nodes that had a lot of disk before we switched over have expired, and a lot of the crazy low disk offers have been renewed at a higher rate, that's left nodes with kind of with a lot of disk space. And since they're Ryzen, and CPU usage is now low as a result, for offers mostly sold on E5's in the past... yeah.)
Oh really important regarding RAM: the higher RAM plans are currently less likely to show up because they'd be on nodes with a lot of memory left over, which happen to also be nodes that are locked off/have been locked off for some time. We went with the nodes being unlocked that are the most stable which happen to be the nodes with the highest amount of memory used since they're usually actively being refilled and just higher memory usage in general from being online for so long. So once these get entered back into the pool you'll start seeing your higher memory offers.
Let me know if you notice any specific way this works out, as in let me know if it appears to be after a price drop, it's lower on the order page, and then corrects itself or if it stays broken the entire time. No rush, just whenever it's noticed if you have more information for me on it.
It's already on my (growing) list of things to fix but lower on the priority list for now.
After I'm done with fixing all the important bugs, I want some feedback on the pricing/resource generator as I'll likely want to rework that system to some degree. There's already a system that compares the plan to a few "model" plans. So I want your feedback on some of these "model" plans and how they should be, otherwise I'll use the order data to tune the existing "model" plans.
The "model" plans are plans we know are popular and how they should be and it uses them to compare and make some last minute adjustments to generated plans.
So basically let me know the type of plans you're looking for and would like to see. Give me the RAM, CPU, Disk, Data as well as the price range (like preferred price to maximum price) you'd be willing to pay for it and in which locations. I'll compile all these together and see if I can work it into a better modeling system.
@VirMach Do you still have spare epyc servers?
On another note:
, I could see myself getting some more flash sales. I don't know if there's any others who are in the same boat.
Sometimes I'll come across a plan and wish it was a bit more balanced in my opinion. Like if it had less bandwidth (personally I'd want at least 1-2 ish TB) and/or just one core compared to having two cores, and a bit higher ram like around 1-1.5 GB, instead of ranging from 700ish MB and below, but more realistically I wouldn't really want anything 500 MB and below. As for disk space, I don't see myself needing much of it. And maybe a price adjustment on the lower side
1C/2G/30G 500GB/W $12/yr. in L.A.
Another bug, it seems like the TTS is more likely to pierce through this issue, but not the sound alert. If you have the tab in the background, then at some point in time I think the browser puts it to sleep and doesn't play the sound alert (even if you have TTS+sound enabled.) The current fix is to just open the tab back up and you should be good again for some amount of time. Maybe in practice this is a non-issue since it could just be a "me" problem with having like 300 tabs open.
There's 2 or 3 Epyc servers sitting behind me and one of them already deployed in Tokyo since January 2024, but just not being used.
There was a system I was working on that got scrapped which would have basically resolved your issue. I was then going to work on a system afterward where there's a flash sale within your service details page after you order, but obviously after we're done with the sale, it would only benefit people that bought a plan they didn't really want and they'll be able to finally tune it/customize it to what they want so that won't work out either.
The system was going to basically generate a range of specifications for each flash special to let you decrease some things to lower the price or increase some things and pay more. It was mostly aimed (initially) toward additional IP addresses and disk, IIRC.
I'll see if it's feasible to work this in, or try to come up with some other solution as that does seem to be a big problem (at least the part where you're getting resources that are just lopsided and you don't get to opt out and it gets calculated in the price.) One of the solutions would be not being charged as much for the amount that's way over a reasonable range, for example if there's a plan with 384MB RAM, 3 Cores, then we should try to weigh it down to where either the cores are reduced and/or we charge less per core for it. The system already does technically kind of keep this in mind a little but not to the degree I'd like for it to do.
One other idea that was explored at some point in time was having sub-plans for each plan. You'd click order now or another button, and you'd be presented with 3 plans for the one of 3 plans, for a total of 9 plans. One of them would be the plan as advertised. One would be a variant that's smaller, one would be a variant that's larger. That was decided against because while it could technically utilize the system as it currently is then we'd have more plans to track and at that point we might as well just present 9 different plans and the likelihood of one of them matching what you want is technically statistically the same since we do also randomize plans already (it decides whether to make it larger, smaller, and so on, it gets a dice roll.)
I guess the issue we need to fix is making sure that the plan that closely resembles what you want has the potential to ever exist.
Added to spreadsheet of requests.
How much ram is usually installed on those EPYC servers?
I don't remember. We were somewhere around 384GB to 1024GB though. They have something like 8 to 16 slots and 32GB to 64GB per stick but could go all the way up to 256GB sticks (which we definitely didn't use.) I think the 16 slots were dual Epyc ones, we might have had one of those.
So either 384GB, 512GB, or 768GB most likely.
The body pillow would definitely be a fire hazard.
Bug found, the bug where no welcome email was sent has been fixed.
New bug is that now up to 3 emails are sent. Well actually, it's even worse, it's where up to 3 services are generated. This might also be causing the vserverid bug.
IMPORTANT UPDATE: only use the service that is actively being displayed on the service details page/billing area. The others will be terminated (we'll likely suspend them gracefully first, but we're not going to honor any support requests where you decided to use all of them and then want to rotate out to some other generated one.)
Will be working on redoing the creation system completely.
CREATIONS ARE BEING PAUSED: The system will still keep track of which services are ordered and need to be created, but it won't be provisioned immediately. I'll provide an update on when it's reworked and unpaused.
edit
(As for the discount rate, I don't know since the plan is being switched out.)
As an example, I feel like this plan is close to being an ideal plan, at least in my mind. I don't what the minimum disk space is for these plans. Although, I wonder how much more RAM you could get if you balanced the plan by lowering the bandwidth to something like 2000 GB. However, we need to keep in mind how much can be balanced out, while still being feasible for VirMach. Anyone else got any feedback or thoughts?
$9.44 PER YEAR
1512MB DDR4 RAM
1 CPU vCORE
30GB SSD (NVMe)
300GB BANDWIDTH
1 DEDICATED IPv4
I'll take it if it's in L.A
4 cores
4GB RAM
150GB NVMe Storage
1 IPv4
5000GB Bandwidth
MIAMI, FL
I actually do not believe we charge for bandwidth in the current logic. It just decides how much bandwidth to give you based on all the other specifications, pricing, and location. So in that case, if I remember correctly (I'm not sure, we had so many different versions) then it would mean it would have zero effect on the price and other resource levels.
Coupon "tenpera50" has been expired.
New code: BF33Third
For 33%* off one-time. Remember, one-time means non-recurring. One-time discount. Not one time use for the coupon. And remember this means since it's 33% off, it equals the first 1/3rd~ of the service being rendered as free and therefore the first 4 months of service being 0% SLA, and no support per the disclaimers. (That means the first coupon was 1/4 or 3 months, second coupon was 1/2 or 6 months.)
👀
Was originally going to be 38% then switched to 33% after I realized 4/12 months is 33.3% -- math is hard.
Delayed bus:
On time:
No Paypal now?
You can use PayPal to add funds. IIRC it has been this way for a while.
Is NYC Metro really Buffalo?
/s