D&D 5E Artificer Class, Revised: Rip Me A New One

[RealAlHazred]

Originally posted by Tempest_Stormwind:

I'm back from over a week away from the keyboard. Finally getting a chance to work on the updates:
 

• I've cleaned up the editing. There's no reference to infusions as anything other than spells except in the description of the artificer's spellcasting (and the name of the potion / arcane device abilities).
• The Homunculists' Guild finally has rules attached to it (but I'm not satisfied with its last ability). I dropped the idea of upgrading warforged in favor of bringing back the specialist homunculi from 3e (all of them except the persistent harrier, which was kind of forgettable).
• The notes are cleaned up quite a lot - this now has very little extra text beyond what would appear in the book.
• I altered the starting package to include a shield option for melee-centric artificers.
• Power Surge is now a one-action spell with a Concentration duration, and the acceleration clause in the magitechnicians' guild is dropped.
• The 14th level ability is now an expanded augmentation list which includes necrotic, radiant, psychic, and poison damage, since that's pretty straightforward, but I haven't yet had a chance to check MM resistances / vulnerabilities to those damage types as CR increases.
• Considering how potent tunable resistances can be, I've adjusted Armor Augmentation such that magic weapons bypass its resistance if applied to armor. The 14th level ability removes this limitation. It's still a useful spell, definitely, but it isn't a super powerhouse for any tank at all levels, and this change reduces the impact that introducing the artificer has on Lore bards. (To get the most out of these, you need artificer levels)

 
I'm also playing around with wording that allows an artificer's "extra component" to work even if he's got his hands full, so he can infuse a weapon he's wielding if his other hand holds a shield, for instance. This isn't implemented yet, and might prove a bit complex to do.
 
Otherwise, I think this is nearing completion. Early playtests (including at level 1, where the only magic you have is through arcane devices (scrolls), potions, Magecraft, and rituals) have proven satisfying, and I'm liking how the text has turned out after all of these revisions. I might stitch together a document with artwork and formal typesetting still.
 
 
A side note from early testing: I've seen the following multiclass interactions:
-Splashing Rogue to get expertise in Arcana and a bonus-action Dodge appears to be pretty damn useful, particularly for the spellforgers' guild. Alchemists giving their bombs to Dex-happy rogues is potentially exploitable as well, but I don't like the idea of limiting bombs to the alchemist alone. All told, rogue is a natural partner for the artificer.
-A bard dip - again, for expertise, but also for bardic inspiration on Prototype's check - appears to be exactly as useful as it appears to be, in that it's good but the delayed advancement of artificer spells known, spell slots / reserve limits, and the split spellcasting scores slow you down somewhat compared to single-classing either route. It also takes five bard levels to get inspiration on a short-rest refresh, which is less of a factor early on but as your spell slots grow to outstrip your Charisma modifier, it becomes a bigger factor. Incidentally, using the book of schema instead of spells known also prevents this kind of multiclassing from fueling Prototype's library to disturbingly large proportions.
-A warlock dip remains useful here for fueling augmentation spells and Prototype with recurring spell slots. However, Eldritch Blast does not combine well with any real artificer abilities (unless you're making prototypes of spells like Hunter's Mark), and the casting stats don't line up.
-Wizard and artificer could get out of hand quickly, but the inability to copy from one book to the other seems to be enough to keep a sanity brake in place. This kind of "sage" character is interesting when he poaches low-level spells that by most accounts he shouldn't be able to use (like Hex, Hunter's Mark, Entangle, and so on) and builds arcane devices (scrolls) to use them as supplements to his existing spells, or when he uses Arcane Recovery as a ghetto craft reserve. Combining Find Familiar with the golemists' homunculus for two little helpers is also an interesting twist (and though I haven't personally tested it yet, he can also build a device for Find Steed and the assorted Conjure spells - the Hordeificer rides again.)
-On the reverse of this, dipping artificer as most of the above classes provides a decent chunk of utility. While you'd dip warlock for one or two specific spells (typically Hex and Eldritch Blast) available rather frequently, you'd dip Artificer for a small number of scrolls, potions, or prototypes / augmented weapons to round out a strategy. It's a good dip, and one I haven't tested as thoroughly as I'd like - but it does appear to be somewhat hard to break.
-The only class that appears to be a good mix for artificer that I haven't checked yet is Fighter, and that's mostly because of their durability and Action Surge (though the improved weapons and armor to enhance certainly factors in as well!).
 
Additionally, these test characters were run using variant human, hill dwarf, rock gnome, and warforged (both official and my personal variant, using the soldier subrace). All of these had distinct advantages as artificers, bringing out particular playstyles somewhat naturally (I was very pleasantly surprised by the spellforgers' dwarf, for instance). There's still two +Int races out there to test that I haven't looked into, but both are unconventional yet understandable choices for artificers - high elves and tieflings. There's even some mechanical synergies between them that I didn't expect
 
A huge factor that's shown up so far is concentration - that dramatically limits the kind of stunts the artificer can pull off, particularly with arcane devices / prototypes. You find yourself focusing instead on out-of-combat rituals and in-combat oddball instantaneous effects as a result, which is kind of exactly what the artificer was supposed to be doing as a mad inventor or troubleshooter anyway. The ability of an artificer to build a buffstack onto himself and his team that's so powerful it may as well be mecha armor is gone, and with it most of the arty's biggest exploits.
 
There's still some high-level problems that show up when combined with Simulacrum (as written, craft reserve recovers even if spell slots don't), but Simulacrum is a very problematic spell to begin with, and this is only possible with Artificer 20 builds or some multiclasses with bard or wizard. I'm still toying around with other high-level builds, in particular those that get 9th level spells from other sources and dip Artificer afterwards.

 

[RealAlHazred]

Originally posted by rampant:

Mage hand, heavily modified, and eldritch blast work as class features because they are solid, reliable, and limited. SSI up til your latest re-write was none of these. Now it's solid and reliable, an improvement mind you, but it's still pretty unbounded.
 
Sustain augmentation is a little too much, what if it allows you to sustain one at a time for free, instead of making all instances of the spells not require concentration?
 
For the artificer spells known I'm a little leery of letting them trade up old spells for higher level ones. 
 
You know my feelings regarding allowing them to learn new schema wizard style, especially with the fact that they aren't restricted to a specific list.

 

[RealAlHazred]

Originally posted by Tempest_Stormwind:

rampant wrote:Mage hand, heavily modified, and eldritch blast work as class features because they are solid, reliable, and limited. SSI up til your latest re-write was none of these. Now it's solid and reliable, an improvement mind you, but it's still pretty unbounded.

I'm actually geniunely curious as to why you're changing your mind on this, because Prototype and the old SSI are virtually identical, especially for the magitechnicians' guild. If it wasn't solid or reliable before, why is it solid and reliable now?
 
As for unbounded, see below.
 

Sustain augmentation is a little too much, what if it allows you to sustain one at a time for free, instead of making all instances of the spells not require concentration?

Considering how damage types and vulnerabilities are pretty broadly distributed, I'm not sure about why. I do "get" that this requires testing, though.
 
EDIT: I should note there are two consequences to removing concentration instead of just saying it can be meaintained. In this case, if I've cast Enhance Ability on an ally's belt, and I try to enhance my weapon, it would cancel the Enhance Ability under your wording, but not under mine.
 

For the artificer spells known I'm a little leery of letitng them trade up old spells for higher level ones.

It's the same as bards, sorcerers, rangers, eldritch knights, arcane tricksters, and warlocks.
 

You know my feelings regarding allowing them to learn new schema wizard style, especially with the fact that they aren't restricted to a specific list.

Again, how many spells did your last wizard have in his spellbook? Compare that number to the minimum (twice the wizard's level, plus four). Note that the artificer knows a number of schema equal to his level plus three (basically, half as many to start) and can't copy schema out of defeated wizard spellbooks. Unless the DM hands them out (say, by finding a schema or spell etched on ancient ruins) or you find scrolls (which aren't purchaseable by default!), you won't really get any more than what the class itself provides you. (Go and re-read the downtime activities - it's not like you can just research up new ones in this edition.)

 

[RealAlHazred]

Originally posted by rampant:

Prototype was garbage, Infuse arcane device is what I'm talking about.
 
Basically I just don't want Artificers to be maintaining a tons of buffs at a time, the ability to maintain 1 or two extra buffs, especially specific buffs, isn't a huge deal but I want to avoid letting it get out of hand. Nuking the number of buffs a mage could have running at once is one of the things 5e did right.
 
My last wizard? I assume you don't mean 4e because they had sensibly bounded spell books, so it's been a while (was not a fan of 3e caster classes prior to the warlock to be honest), hmmmmm that's probably the rogue-wizard MC, He started off with a level in or two in rogue and had so much fun with use magic device that he decided to be a real wizard. So he picked up his first level in wizard and began filching spells and scrolls left and right, by the time he reached 5th level effective wizard (he eventually went arcane trickster) he was getting close to needing another spell book, besides the extra copy he kept in case someone jacked his main tome, it would have been more but back in 3e you had other stuff  to spend gold on. To say nothing of how in parties with multiple wizards they always choose different spells on level up and traded, thus doubling their spells known.
 
Simply put it doesn't matter that not every wizard can get a huge number of extra spells known this way. The fact that the class allows them to gain additional powers known without paying an opportunity cost is broken as hell. Not as bad as most of your divine casters mind you, but still busted and requires DM enforcement to balance, which is the same as being unbalanced. The Artificer is multiple times worse since he's not just picking off a specific list but any spell in the game.

 

[RealAlHazred]

Originally posted by Tempest_Stormwind:

rampant wroterototype was garbage, Infuse arcane device is what I'm talking about.

That's interesting, actually. Infuse Arcane Device has been in this artificer since the get-go: it's Imbue Scrolls with a new name. But that new name midway through does enough to change it so that you can view it through the same SSI lens that you view Prototype.
 

Basically I just don't want Artificers to be maintaining a tons of buffs at a time, the ability to maintain 1 or two extra buffs, especially specific buffs, isn't a huge deal but I want to avoid letting it get out of hand. Nuking the number of buffs a mage could have running at once is one of the things 5e did right.

And that's a fair point, which I haven't tested as fully as I need to yet. But considering how these are limited to a very small list (just the basic augmentations, which are NOT as strong offensively as they are defensively, and, for spellforgers, the basic ones that upgrade equipment), I wouldn't be super-paranoid about it. Caution is still prudent, though.
 

My last wizard? I assume you don't mean 4e because they had sensibly bounded spell books, so it's been a while (was not a fan of 3e caster classes prior to the warlock to be honest), hmmmmm that's probably the rogue-wizard MC, He started off with a level in or two in rogue and had so much fun with use magic device that he decided to be a real wizard. So he picked up his first level in wizard and began filching spells and scrolls left and right, by the time he reached 5th level effective wizard (he eventually went arcane trickster) he was getting close to needing another spell book, besides the extra copy he kept in case someone jacked his main tome, it would have been more but back in 3e you had other stuff  to spend gold on. To say nothing of how in parties with multiple wizards they always choose different spells on level up and traded, thus doubling their spells known. 
Simply put it doesn't matter that not every wizard can get a huge number of extra spells known this way. The fact that the class allows them to gain additional powers known without paying an opportunity cost is broken as hell. Not as bad as most of your divine casters mind you, but still busted and requires DM enforcement to balance, which is the same as being unbalanced. The Artificer is multiple times worse since he's not just picking off a specific list but any spell in the game.

So, you admit you haven't played a 5e wizard, and are judging this off of 3e with its common scrolls, purchasable scrolls, and downtime research to expand books further. Judging characters based on how certain classes played in entirely separate games is sort of like saying that 5e War or Tempest clerics can't be good in melee, because they have a prohibition against edged weapons and poor THAC0.

 

[RealAlHazred]

Originally posted by rampant:

Yes scrolls would LIKELY be harder, however it doesn't matter if it's 1 or 100 the fact that they get new powers at no opportunity cost is broken, and no I never researched a single spell with that character, that what the thief skills were for, do you have any idea how stingy those NPC wizards are? Even at an offer 3-to-1 the elf bastard wouldn't even so much as let me look at the general location of his spare books. Well guess whose laughing now, racial bonus to perception doesn't do jack if you don't put any ranks in it you Ivory Tower Journal Jockey! Prior wizards (or similar classes, there were a couple other book casters in the old days) however did certainly use the research option to pick up important spells that the base spells learned didn't cover or if I needed something specific and couldn't buy or steal it faster than researching it on my own would allow. Not recommended for volume spell learning but a great way to get specific spells or the occasional custom job. Back before energy substitution I liked to have a couple of sonic spells on hand, and for some reason they never published a spell that deployed exploding mice capable of independent movement and running on walls, so I had to make my own.
 
Admittedly I DM more than I get to play these days, which has it's rewards mind you, I'd just like to spend some more time back on the player's side of the screen. I've got this really nifty idea for a Paladin-Sorc I've been meaning to try since '09.
 
As for it being an entirely separate game, ... not really. I mean it's different, but it's enough like 3e that it makes a lot of the same mistakes, and will thus have a lot of the same issues, it's sidestepped some of them nicely, concentration buffs, and the end of the never to be sufficiently damned full-attack action stand out as excellent improvements. However other issues remain, clerics for example, several divine casters really with their automatic access to every spell on their lists as soon as they get the level requirements including any new spells the dm allows into the world. Wizards are slightly more hindered certainly, and the fact that every class has it's own spell list and we don't have nearly as much of this nonsense with each spell being on 5-10 class lists is definitely helping keep them in line. However the new edition also works for them in certain ways. Without the need to spend huge amounts of loot on gear wizards can dumped their extra cash into spell knowledge, and are still as able to trade spells as ever. 
 
This **** is broken, clerics, wizards, and similar casters get more powers either for doing jack all, or for shelling out the coin and a day tops to scribe a new spell they find, with no further opportunity cost, and doing it without even bounding it to a certain list of spells is even more so. Look back on the tier system from the 3e and pathfinder days, you'll find that the most broken classes almost invariably shared this trait, the ability to learn new powers with no permanent opportunity costs, basically the ability to be good at everything.
 
As for Infuse arcane device it's got two important advantages over SSI/Prototype, one it's reliable, it doesn't require extra checks to use, two Anyone can use it if I'm reading the DMG rules on spell scrolls right. Casters still have a massive advantage when casting the stuff that actually involves a DC, but that's not a huge issue, and even if it was it's a simple matter to fix honestly that doesn't even have to be placed in the class description, just tweak how non-casters work with items that cast spells, which is frankly something that needs to be done anyway.

 

[RealAlHazred]

Originally posted by Tempest_Stormwind:

Revisions:

• Realizing a major goof on my part, I switched the magitechnician's Inventor to recover after a long rest, not a short rest, as otherwise it's basically taking the warlock and outdoing it, particularly once their level 11 ability kicked in.
• With that change, all the main abilities that used craft reserve were on a long rest, rather than a short rest, so I switched its default recovery to a long rest and clarified the exceptions (magecraft, Augmentation Savant, Infuse Bomb, and Power Surge). This leaves all the other uses for craft reserve - Infuse Arcane Device, Infuse Potion, Inventor, and Spell Flask - on long-rest recharge. Now, if you're using craft reserve to get the effect of a spell or something close enough to a spell, you can't recover the spent reserve on a short rest. The exception to this is the Spellforgers' Guild Augmentation Savant, which recovers reserve spent on augmenting his own equipment, since that's so central to the guild - but I took the opportunity to clarify that it won't give him this recovery if he's casting stronger spells to augment his allies' gear. (This does mean he's really efficient on the reserve, but his actions are spoken for, since he's going to be fighting most of the time once battle starts.). 

 

rampant wrote:Yes scrolls would LIKELY be harder, however it doens't matter if it's 1 or 100 the fact that they get new powers at no opportunity cost is broken, and no I never researched a single spell with that character, that what the thief skills were for, do you have any idea how stingy those NPC wizards are? Even at an offer 3-to-1 the elf bastard wouldn't even so much as let me look at the general location of his spare books. Well guess whose laughing now, racial bonus to perception doesn't do jack if you don't put any ranks in it you Ivory Tower Journal Jockey! Prior wizards (or similar classes, there were a couple other book casters in the old days) however did certainly use the research option to pick up important spells that the base spells learned didn't cover or if I needed something specific and couldn't buy or steal it faster than researching it on my own would allow. Not recommended for volume spell learning but a great way to get specific spells or the occasional custom job. Back before energy substitution I liked to have a couple of sonic spells on hand, and for some reason they never published a spell that deployed exploding mice capable of independent movement and running on walls, so I had to make my own.

1) Readily available scrolls at market.
2) Readily available scrolls in treasure hoards. 
3) Downtime activities that allowed research.
4) Readily available other spellcasters to crib from (note: artificers and wizards can't use each others' books here)
 
Yeah, you're still operating on the 3e paradigm. I'm not going to consider this criticism until you've had some experience operating without easy access to spellbook expansion methods in 5e. 
 

As for it being an entirely separate game, ... not really. I mean it's different, but it's enough like 3e that it makes a lot of the same mistakes, and will thus have a lot of the same issues, it's sidestepped some of them nicely, concentration buffs, and the end of the never to be sufficiently damned full-attack action stand out as excellent improvements. However other issues remain, clerics for example, several divine casters really with their automatic access to every spell on their lists as soon as they get the level requirements including any new spells the dm allows into the world. Wizards are slightly more hindered certainly, and the fact that every class has it's own spell list and we don't have nearly as much of this nonsense with each spell being on 5-10 class lists is definitely helping keep them in line. However the new edition also works for them in certain ways. Without the need to spend huge amounts of loot on gear wizards can dumped their extra cash into spell knowledge, and are still as able to trade spells as ever.  
This **** is broken, clerics, wizards, and similar casters get more powers either for doing jack all, or for shelling out the coin and a day tops to scribe a new spell they find, with no further opportunity cost, and doing it without even bounding it to a certain list of spells is even more so. Look back on the tier system from the 3e and pathfinder days, you'll find that the most broken classes almost invariably shared this trait, the ability to learn new powers with no permanent opportunity costs, basically the ability to be good at everything.

It's similar to 3e in many ways, but it is not 3e. I'm a longtime 3e veteran and love that system to death, but the similarities between it and 5e are really more superficial than substantial. Spellcasting is one of the bigger overall differences, including the size of wizard spellbooks. Functionally the book is a little bigger due to upcasting (i.e. a cleric doesn't need to know six different healing spells - Cure Wounds is all six, similar to how psionic powers work). However, in terms of actual different spell effects, it's remarkably lower, since spell scrolls and enemy spellbooks are much rarer and not purchasable at all, and downtime research is not able to produce spells.
 
I would be utterly shocked if a wizard found more than, say, his character level in wizard scrolls over the course of his career, unless the DM specifically plants scrolls for some reason (I've been known to do that) or if the campaign has a bazillion wizards who are willing to trade spells instead of sending the PCs out for quests to pay for the privilege of copying spells from their books. In no case is the spell a player wants going to be freely available, either - I can't just buy Simulacrum on the open market, and just because I know a particular area is more likely to have 2nd level spells doesn't have any guarantee that it'll have Enhance Ability. I'll run a more rigorous test on this when I get home and have access to my loot tables.
 

As for Infuse arcane device it's got two important advantages over SSI/Prototype, one it's reliable, it doesn't require extra checks to use, two Anyone can use it if I'm reading the DMG rules on spell scrolls right. Casters still have a massive advantage when casting the stuff that actually involves a DC, but that's not a huge issue, and even if it was it's a simple matter to fix honestly that doesn't even have to be placed in the class description, just tweak how non-casters work with items that cast spells, which is frankly something that needs to be done anyway.

No, that's wrong. Scrolls (as a class of item) can be used by anyone, but spell scrolls can only be used by casters with the spell in question on their list (the specific rules on spell scrolls trumps the general rules on scrolls - for an example of a scroll that isn't a spell scroll, look at the Scroll of Protection, which is similar to Magic Circle but usable by anyone.)
 
For me, the interesting bit was that, back when this was still Imbue Scrolls, you were judging it separately. Now that it's been cosmetically changed to Infuse Arcane Device, you're viewing it the same way I intended it, as a more reliable (but slow to build) version of Spell Storing Item. Prototype allows you to build this stuff faster and more frequently (limited by spell slots (and possibly reserve if you're a magitechnician)), but isn't anywhere near as reliable as if you take the time to do it right (as an arcane device / spell scroll). And that's the core point here - they're different mechanics, but functionally the same ability. 

 

[RealAlHazred]

Originally posted by Tempest_Stormwind:

DarkSphinx wrote:Wow!  Very nice!  This, I can actually use! (Couldn't stomach the watered down WotC wizard subclass).

Thanks. FWIW, the original WotC one has grown on me to a point, but only if one changes the scrolls to arcane devices (see here for why(x); it's more than just the cosmetic change), and while I grudgingly admit it's an artificer, it's not the Eberron artificer. That's where this steps in.
 
I'm still trying to come up with a better level 17 ability for the Golemists guild, and that's the last bit of new mechanics the class needs before it's finished, unless testing reveals something weird - say, with the consequences of removing Concentration from a small set of equipment buffs. (A potential fallback if this is problematic instead has those spells behave at a higher level than the spell slot actually used on them, tuned to make up for the artificer's slow spell level progression - i.e. an artificer's 2/3s casting would still produce Magic Weapon, Magic Armor, and Elemental Weapon at the same bonus as a full-casting wizard, but only those spells.) Regarding tweaks to existing mechanics, I'm also contemplating removing the heavy crossbow, as while crossbows are some of the best fits, I don't want it to be too shoehorned into using them.
 
EDIT: Turns out a simple "those spells behave as if they were cast in a slot one higher than they actually were" gets you pretty darn close to what a wizard can do. Just a matter of putting it in the right place in the progression, then.
 

rampant ripped you a nice one, but you've sewed up everything beautifully!

I wouldn't say he ripped me a new one, but he was appropriately brutal, and I'm very appreciative of that - it made my bad ideas die and a better class emerged from it. That's half of why I will always ask for people to Rip Me A New One, rather than PEACH. (The other half is that every single word in that term is something that goes without saying in a conversation (which is voluntary, based upon reading what the others write, but also soliciting feedback and assuming the action is done in good faith), so about the only good thing it has going for it is a catchy acronym.) Survival of the fittest always beats "not in the face" when designing rules.

 

[RealAlHazred]

Originally posted by rampant:

OK I found the SPELL scroll item as a specific item and yes you're right. So wow yeah lot less impressed with the class now.
 
Look I already explained how the spell book thing is still an issue, the fact that it happens once is broken, your example of extra spells equal to their character level over the course of the game? Yeah that's completely borked up! The fact that the artificer doesn't have a specific spell list for their schema makes it worse because at least the wizards can't do everything, yet. I want you to imagine two artificers in a party, assuming the DM isn't running them in a constant state of terror and poverty such that they spend every day fighting for survival with little more than the shirts on their backs, what reason would they have not to trade schema? Boom those artificers now know twice as many schema as a similar artificer in a party that got a paladin instead. 5e make sit harder than 3e I admit that, my complaint is that it still happens at all, because it is inherently unbalanced.
 
Furthermore because the spell scrolls only work for artificers, casters, and thief-rogues of sufficient level to have use magic device, the artificer gains massive enabling powers or not, based on party composition.  So the power and influence of the artificer swings greatly based on whether he's got casters backing him up or not.
 
That's not good for the same reason 'warlordesque' classes and subclasses that don't work well with casters (I'm looking at you every non-casting leader-type class from 3e) aren't good. Every class has to be able to function, and be balanced, in a party with any potential combination of other classes. 

 

[RealAlHazred]

Originally posted by Tempest_Stormwind:

rampant wrote:OK I found the SPELL scroll item as a specific item and yes you're right. So wow yeah lot less impressed with the class now.

On the one hand, you keep seeming to expect the artificer to turn everyone else on your team into spellcasters - you've asked for prototypes to work for other characters, and you're disappointed that nonspellcasters can't make use of spell scrolls / arcane devices.
 
On the other hand, you constantly raise the specter of the artificer having access to too many spells (and, as I said above, I will not consider further criticism on this point from you until you've played or run a 5e wizard - it's nowhere near as easy to expand the spellbook here).
 
You do know that implementing the first of these suggestions makes the second a much larger issue, but ignoring the first (i.e. following the rules as they stand) also makes the second much less significant, right?
 
If it ain't broke...

I want you to imagine two artificers ina  party, assuming the DM isn't running them in a constant state of terror and poverty such that they spend every day fighting for survival with little more than the shirts on their backs, what reaosn woudl they have not to trade schema? Boom those artificers now know twice as many schema as a similar artificer in a party that got a paladin instead. 5e make sit harder than 3e I admit that, my complaint is that it still happens at all, because it is inherently unbalanced.

This is, yes, potentially an issue.
 
However, I think you're overlooking two critical points.
 
First, the relible arcane devices (which are slow to build and replace, and unlike the 5e spellcasters are literally fire-and-forget, since you need to build two Sleep devices to cast Sleep twice) are entirely limited by your craft reserve. A level 9 artificer has at most 13 points of craft reserve, and may have far less if they didn't max out Int. He will have two 3rd level schemas by default; if there are two artificers in the party, he might have four. (This is four levels after the wizards picked up 3rd level spells, and the wizards have just picked up 5th level spells.) If he wants to build third-level arcane devices, he can hold at most four of them in a day, and that also ties up all of his craft reserve so he can't have any other arcane devices, any active potions, any uses of magecraft, or any of his subclass' abilities (three of the four subclasses increase the demand on craft reserve). Even with all of this, he's forced to pick which spells he creates when he builds the devices, and is unable to pick better ones from a list of prepared spells or somesuch when battle starts. Oh, and these devices don't use your Intelligence for DC or spell attack - they're locked to DC 15 / +7, which for a 9th level artificer would be equivalent to a 16 Int no matter how high yours is. (This is deliberate, actually - part of the cost of reliability is that you don't supercharge your abilities.)
 
Second, if he wants to exceed this amount, or have a wider set of versatility, or to use his Intelligence directly in combat with a schema, he can use his spell slots on them - through Prototype, which you take every opportunity to insult as useless and unreliable. These are also not usable in combat (except for the Magitechnicians, who have to spend a Hit Die to use them - and on multi-day adventures, those start falling pretty fast since you don't recover all spent hit dice on a long rest), have a timed shelf life, and (for the strong ones) there's a pretty high chance of failure. In this case, a 9th level artificer has at best +9 Arcana, and a 3rd level prototype (of which he can create two, before he runs out of spell slots) needs a DC 16 check to trigger - that's a 30% failure rate. In 3.5 terms, this is like trying to cast spells while wearing chainmail, except here it takes a minute to set up (or two rounds and a Hit Die for Magitechnicians).
 
And all of this for a maximum of 4 3rd level spells across two artificers, each of whom (if they focus on absolutely nothing but this) can deliver at most six such effects (four weaker than normal, and two with a rather high risk of failure, and all of whom require additional set up time and aren't chosen on the fly); if two wizards worked together to do the same they'd have access to up to twenty such spells, and could deliver seven such effects each (and they'd be stronger, since more than half of those are upcast in higher-level slots, all of them are cast with perfect reliability, and all of them use the wizard's full Intelligence score; furthermore, they're picked appropriately on the fly from a list, and it's easy to have fourteen of those 20 spells prepared. None of this even considers what the spells are - artificers have no specific proficiency with any particular form of schema (except, perhaps, rituals, and tomelocks have already shown that's nowhere near as bad as people think), while wizards have ways of making one specific school completely badass).
 
If anything, this is telling me that I should increase the reserve cost for arcane devices (something that my tests have suggested as well - it's not a serious problem but it is exploitable) rather than decrease the versatility.
 

Furthermore because the spell scrolls only work for artificers, casters, and thief-rogues of sufficient level to have use magic device, the artificer gains massive enabling powers or not, based on party composition.  So the power and influence of the artificer swings greatly based on whether he's got casters backing him up or not. 
That's not good for the same reason 'warlordesque' classes and subclasses that don't work well with casters (I'm lookign at you every non-casting leader-type class from 3e) aren't good. Every class has to be able to function, and be balanced, in a party with any potential combination of other classes. 

Right, and the power and influence of a short-range radius paladin aura is stronger the more people fight in melee with him instead of hanging back while he tanks, and the power and influence of a rogue's sneak attack improves if he's got a meleeist on the team instead of a team of archers, and the rogue's ambushing ability (exemplified on the assassin) becomes much less useful if anyone on the team isn't also a stealth expert, and the power and influence of a fighter's Protection combat style or Rally maneuver increase the more people he has to protect from hit point damage, and the power and influence of a wizard increases the more other wizards he's fighting alongside, and....
 
And the artificer's weapon-enhancing buffs increase in utility when there's more people who fight with weapons (especially once he starts upcasting or sustaining them) and decrease in utility when there's more people who fight with cantrips. And his arcane devices increase in utility if he's got more spellcasters (of the right sort - a wizard can't use a device of Entangle, but a druid can) on the team, but decrease in utility if there's more fighters (but, again, this isn't always true, as the alchemist does exactly what you want at the high levels).
 
Also, if you think none of the 3e leader-type noncasters didn't play well with noncasters, you never read White Raven Tactics nor the Determined Caster aura nor anything that worked off of debuffs (I'm currently playing a character who's basically a walking kill zone for our spellcasters, for instance).
 
 
 
EDIT: Revisions made:

• Heavy Crossbow removed; spellforgers can still get it, and everyone else can still use the light crossbow.
• Arcana training is now optional.
• Starting equipment now clearly provides a book of schema.
• The overall progression has been adjusted.  The old ability timing was set back when it was a 3/4 caster, not a 2/3 caster; the adjusted timing evens out dead levels. The closest to "dead" (in terms of not getting new spells, new spell levels, or new abilities) are 13 (which coincides with a proficiency bonus increase) and 18 (I added a new spell known at this level to cover for that).
• Arcane Devices now decay if you don't maintain them for 24 hours (this is automatic just by keeping them with you, and I'm leaving the description up to the player). This puts them in the same category as every other item you create through craft reserve(temporary), and it further reinforces why they're not sold. Also, I broadened it to function off of any artisan tools, not just tinker tools (it makes sense that a mason artificer would etch runes into tablets and an alchemist would brew some chemical reaction instead of constructing a Rube Goldberg machine), but the name "Arcane Device" still harkens to the tinker tools.
• Potions of Resistance now clearly let you set which type of resistance you're granting - provided it's one of the ones from your weapon augmentation spell (acid, cold, fire, lightning, thunder; after 15th that expands to include necrotic, radiant, poison, and psychic). Note that they give a chosen resistance to anyone, using *their* action, for 1 hour without relying on concentration, three levels earlier than you can do the same with Armor Augmentation, and these potions are the only way to get resistance to Force (though you have to gamble to make Force resistance potions).
• Spellforgers' Guild: Sustained Infusion is replaced with Augmented Infusion, which gives +1 spell level to the three gear-enhancement spells. These spells still require concentration (the only spells that don't now are the augmentations from high-level artificers).
• Golemists' Guild: Construct Dominance now has a different effect. While technically I see a few different checks and balances connected to this ability, a look at the constructs in the game already makes any of the reasonable ones purely symbolic (i.e. it's a level 17 ability, and Iron Golems are CR 16), so I omitted them. This might change as constructs are added to the game. (It's modelled off of the high level Necromancer ability, for reference.)
• I changed Reconstruction (raise dead) to retain the standard ordeal, but removed its time limit. It's Construct-only (and they lack a soul), so there's no reason to wait for Resurrection to be able to re-activate ancient relics if you can find them.
• Weapon and Armor Augmentation now no longer require somatic components. Well, technically, they do, but it's whatever manipulation you need to get the material components going, and removing the S component allows you to augment your own weapons and armor once your hands are full (which really matters for spellforgers and any artificer using a shield).
• Jumpstart now clearly can't be used to keep prototypes working. (If that were allowed, you'd basically be casting 6th level spells out of 3rd level slots.) The wording was ambiguous earlier.
• Power Surge's overload explosion damage is now on par with a basic Fireball.
• Added object interactions on Repair Damage, Inflict Damage, Energy Ward, and Stone Construct. 
• I added two 7th level spells (Mordenkainen's Sword and Symbol, similar to Animate Objects and Glyph of Warding), since it was kind of bugging me that they got up to 7th level slots but had no spells for them. (I liked how this allowed upcasting, but no other class has slots but no spells. Even the warlock gets spells for levels he lacks slots (for use with Mystic Arcanum and, presumably, spell scrolls)). I also added Blade Ward as a cantrip, as I've changed my "no combat cantrip" philosophy into "no offensive cantrips" instead.
• General grammar editing throughout. It's still five pages printed.

Also, a reminder about arcane device strength: It functions as a Spell Scroll, which doesn't use your own Intelligence. Instead, it works out to the same DC / Spell Attack as a spell cast by the lowest-level wizard capable of doing it, with 16 Intelligence who puts +2 Int in at 4th and 8th level, except this never continues to scale (i.e. the stats for a 3rd level scroll will always match those of a 5th level wizard with an Intelligence of 18, even if you're higher level).
 
The arcane device DCs and spell attacks, taken from the basic rules, are as follows (with a column added for a comparable wizard's Int to get this result; the effective proficiency bonus scaling is clearly based on the full-caster's rate of spell slot progression):

Lvl	Rar	DC	Atk	Wiz Lvl/Int
Cantrip	C	13	+5	1 / 16
1	C	13	+5	1 / 16
2	U	13	+5	3 / 16
3	U	15	+7	5 / 18
4	R	15	+7	7 / 18
5	R	17	+9	9 / 20
6	V	17	+9	11 / 20
7	V	18	+10	13 / 20
8	V	18	+10	15 / 20
9	L	19	+11	17 / 20

Cantrip and 8th-9th level scrolls included just for curiosity. Cantrips can't be entered into the book of schema, and 8ths and 9ths are only open to wizard/artificer multiclasses who won't have enough craft reserve to create arcane devices of them. (They can build prototypes, but this'll cost them the spell slot and might fail, unlike just casting an 8th or 9th-level spell normally.)
 
This means it's actually quite doable to make an artificer who doesn't have a high Intelligence, by focusing on arcane devices and augmentation instead of Arcana / Prototype. They'll just choose to use their natural ability scores under Personal Weapon Augmentation. A high Intelligence is still desirable for artificers who focus on Prototype or DC/Attack-driven spells, and throwing your own bombs is a ranged spell attack and not a ranged weapon attack, so Int is still favored exclusively by three-quarters of the path choices (and it's still viable for the spellforgers).